middotbull i
IPA Report lfullber August 1988
OllMSElT SUPPORT OF THE OFFSHORE BOOM illgt SKIMMER TRIALSbull
ST JOHI Is IEWPOUlIDLAllIgt
G L McCovn M J Bo111t and J H llash Ro7 F Weston Inc Leonardo BJ 07737
Contract llo 68-03-3li50
ProJect Otticer Robert W Rillger
RISK REDUCTIOI DGIBBRIJG LABORATORY SUPERFUllD TECRlOLOGY DEMOBSTRATIOJ DIVISIOJ
RELEASES CONTROL BRABCR US EllVlIONMEBTAL PROTECTION AGENCY
EDIS01 1 JJ 08837
Thie Stu47 vae conducted in Cooperation vith
Minerals Management Senice US Department ot the Interior
Reeton VA and
Environment Canada Ottawa Canada
RISK REDUCTIOJ EllGIBEERING LABORATORY OFFICE OF RESEARCH illgt DEVELOPMEBT
U S EllVlIONMEBTAL PROTECTIOI AGENCY CIBCilINATI ORIO li5268
24521
bull
______
DISCLADllR
The information in this document hu been funded vh~ or in part b7 the United States Environmental Protection Agency under Contract lo 68-03-3450 to Roy F Weston Inc It has been subjected to the Agencys peer and adshyministrative review and it has been approved tor publication as an EPA docushyment
Approval does not signity that the contents neces11arUyen renect the vieVll and policies ot the US Environmental Protection Agency nor does mention ot trade names or commercial products constitute endorsement or recOJ1111endation tor use by the Agency
ii
PORIWORD
[l-2 paragraphbull ~ Standard IPA Foreword)
Thill report contains a bullllllllllampr7 ~ actirltiebull performed and data obtained b7 Ro7 1 Weston Inc the operating contractor for IPAe OHMSBTT tebullt facility in support of a eeriebull of interagenc7 cooperative exercises renown aa the 1987 OftShore Boom and Skimmer Trialbull These ezercieebull were conducted ill Canadian ocean vaten near St Johnbullbull llevtoundland iD September 1987 and iirrolved Joint sponsorship planning and performampDCe b7 the US Enrlronmenshytal Protection Ageney Environment Canadamp the CampDadim Cout Guard the US Minerals Management Service and numeroubull other pamprticiPBntbull The work pershyformed by the OHMSEIl operating contractor conetitutea oal7 a eampll portion of tbe overall proJect and Do attempt hu been made in this report to describe the complete operations that were conducted or the results that were obtained except u they impacted OllMSEll support operations The report focuses on the planning implementation md data collection ampctiTitiebull provided by Ro7 r Weston Inc md i t1 bullubcontracton IC1 lechDOloa Inc and EnvireepoDle Inc in bullupport of the overall effort1
iii
ABSTRACT
1olloYill8 near~ two years ot planning and coordination a Joint Cbullnadfan - United States oil spill control and cleanup exercise vu conshyducted ott the shore ot lfevtoundland in September 1987 The BPA Oil and Hazshyardous Materials Silllulated lDTironmental lest 1ank (OJIMSBll) tacllity operatshying contractor Rar 1 Weston Inc supported this exercise by proTiding an instlllllented oil containment boom tor eTaluation ot boca pertoriaance methods oil recoTery instrumentation analysis ot the recoTered product and direct support ot the ottshore operations with ten OBMSBll personnel
To evaluate relative boom and water motions preHure transducers and digital data loggers were procured installed on an aTailable oil containment boca and calibrated at the OHMSEll test facility light channels ot such depth-measurill8 instrumentation with one-hour recording capability Vere proTided on the boom that was shipped to St Johns Revtoundland tor the test sequences that occurred during the period ot September 2o-21i 1987
Instrumentation to measure tlov rates and gross Tolumes ot recoTered oil during the sltilllmer-eTaluation phase ot the offshore trials vaa also designed fabricated calibrated and operated by OHMSEll personnel lhis srstem conshysisted ot an in-line Tenturi tlov meter arrangement tank depth-sounding ~es and both in-line and stratified-liquid tank saapling deTicH Samples collected during the recoTery operations were subsequent~ analrsed by ASDf aetbods to determine the percent ot water in the recoTered product
The ottahore teats that were conducted consisted ot a practice run with oa4 simulated oil release on September 21 and a tul1-ecale exercise with releaee ot oil on September 24 Although operational ditticultin were enshycoUDtered in maintaining an acceptable configuration ot the instrumented boom deployed during the ottahore tests titty-eix minutes ot data were obtained OD all eight channelbull during the practice run
Following repair ot instllllllentation cabling that had been damaged during the practice run torty-six minutes ot data from tour channels were recorded during the actual exercise amid continuing problem related to the seaworthishyness ot the instrumented boom In addition tlov rates recoTered TOlUllleB and oilwater ratios were determined tor the various skimmers deployed during the oil recovery phase ot the operations
The OHMSETT actiTitiee were ~ a small portion ot the overall effort during the offshore tests and this report focuses on the experimental proceshydures and findings related to the instrumented boom and the oil recoTery measurements The overall test operations are discussed ~ u they relate to the portions involving OHMSETT participation
iT
COhElllS
Forevordbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 111 A~act bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1T bhibita bullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbull bullbullbullbullbullbullbullbullbullbullbullbull - bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbullbull bull bullbullbullbullbullbullbullbullbullbullbullbullrl AbbreTiationa and S)mbolbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullixActnovledgment bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullx
1 Introduction and Background bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1 2 Conclusionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 3 Recommendationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 4 Materials and Methodbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Boom Selectionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boca Performance Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Flov and RecoTbulllJ Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull PhotoVideo Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Laboratory Equipaent and Materiamplbullmiddotmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other lquipnent and Materials bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
5 Experimental Proceduresbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Booa Illllersion Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boom Tension Measurementbull bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 1lov and RecoTbullrJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull LaboratOlJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
6 Results and Discussionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Rererencebullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Appendices
A PlampIIIled OBMSETT Personnel Duties and Stations During the Ottabore Trialbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
B Computer Programs used to Analyze Teat Datbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Glos aampl) bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull
T
EXHIBITS
1 Transducer Mounting Locations AloDg the Boom
2 Detail Of Transducer MountiDg
3 Cable Harness Arrangement
4 Strain Load Link Arrangement
5 Layout Of Venturi Tube On RecoverT Ve1111el
6 Ex8111ple Of Rav Data output From The Calibration
1 Calibration Curve Of Transducer lo 1
8 Calibration Curve Of Transducer lo 2
9 Calibration Curve Of Transducer lo 3
10 Calibration ot Transducer lo 4
11 Calibration Curve Of Transducer lo 5
12 Calibration Curve For Transducer lo 6
13 Calibration Curve For Transducer lo 1
14 Calibration Curve For Transducer lo 8
15 Example of Rav Data Plot tor the Second Calibration Series
16 Wave Record (Plot Of Immersion Data) From Transducer 1 During The Fixed-Sensor Tests
17 Magnitude-Frequency Spectrum For Transducer 4 Fixed Sensor With Waves
18 Magnitude-Frequency Spectrum tor Transducer 4 Moving Sensor With Waves
19 Tank Testing Of Both Boom Sections
20 Wave Records For Transducers During Initial BoOlll Tests With Waves and Applied Tension
21 Magnitude-Frequency Spectra For Transducers During Tank Tests
vi
22 Measured Tendon On The Boom Por (a) Pirat Ralf Of Teat And (b) Second Bamplf Of Teat During The Practice Run
23 (a) Calibration Of A Rosemount Differentiampl Pressure Gaugebull and (b) Venturi lube
24 Demonstrated Linearit7 Of Venturi Plow Meter At 05 Meters of Water lull Scample Deflection
25 Telog Data Output For 5-second Averaging At 32 gpa Wominampl Plow Rate
26 Wave Record For Sensor 1 During The Practice Run
27 Wave Record Por Sensor 2 During The Practice Run
28 Magnitude Spectrum
29 Wave Record Por Unit 2 Sensor 4 During Offshore Teats
30 Magnitude Sprectrua
31 One-Third Significant Wave Heights Obtained During The Offshore Teat And During The Practice Run
32 Fluid Recoverr Rages For The Jramo Skimmer From Flow Rate Measurements
33 Fluid Recoverr Rates For The Jramo Skimmer From Tank Depth Measurements
34 Fluid Recoverr Rates For The GT185 Skimmer From Flov Rate Measurements
35 Fluid Recover) Rates For The GT185 Skimmer From Tank Depth Measurements
36 Results Of Laboratorr Anal7sia Of In-Line Samples For Percent Water
31 Results Of Laboratorr Analysis Of Stratified Tank Samples For Percent Water
38 Operational Data Logs From Practice Run On 92187
39 Operational Data Logs From Offshore Tests On 92487
vii
LIST OF ABBRBVIATIOIS AID SYMBOLS
ABBREVIATIOfS
OIDCSElr -- Oil and Hazardous Materials Simulated Enrtromientampl Teat Tank
CCG -- Canadian Coast Guard
SIXBOLS
ACKIOWLEDGMDT
Thill vorlr vabull supported b7 Tamprious aember aaenciebull of the US EPAbull OU amplld Hazardous Materialbull Simulated BnTiromental Teat faalr (OHMSEll) Intershyaa-cr Technical COllllittee (OITC)
The OITC currentl7 includes the US EPA US Coast Guard the Mineralbull Management Senice the Canadian Consenation and Protection SerTice and the US lfaT)
The support and encourage-nt ot n11111eroua personnel troa OITC aaencies are aclrnovledged In particular ve Yish to express appreciation to the tolshylaving sponsors and clients tor their continuing support
o US EnTironmental Protection Agenc7 shyRobert Billger Project Otticer and Richard Griffiths
o U S Minerals Manage-nt SerTice shyEdvard TeDD7110D
o EnTironmental Canada - Barrr Whittalrer
o Canadian Coast Guard - William Ryan and Warne Balle7
We are grateful tor the supreme ettorts that the OHllSElf statt proTided during the vorlr at the OBMSETT tacilit7 and in Canada Tralr the project YU
a team ettort the manr and Taried ideas cOIDllleDts and extra ettorts ot the entire statt contributed in no small measure to the succes1 ot the program
ix
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
DISCLADllR
The information in this document hu been funded vh~ or in part b7 the United States Environmental Protection Agency under Contract lo 68-03-3450 to Roy F Weston Inc It has been subjected to the Agencys peer and adshyministrative review and it has been approved tor publication as an EPA docushyment
Approval does not signity that the contents neces11arUyen renect the vieVll and policies ot the US Environmental Protection Agency nor does mention ot trade names or commercial products constitute endorsement or recOJ1111endation tor use by the Agency
ii
PORIWORD
[l-2 paragraphbull ~ Standard IPA Foreword)
Thill report contains a bullllllllllampr7 ~ actirltiebull performed and data obtained b7 Ro7 1 Weston Inc the operating contractor for IPAe OHMSBTT tebullt facility in support of a eeriebull of interagenc7 cooperative exercises renown aa the 1987 OftShore Boom and Skimmer Trialbull These ezercieebull were conducted ill Canadian ocean vaten near St Johnbullbull llevtoundland iD September 1987 and iirrolved Joint sponsorship planning and performampDCe b7 the US Enrlronmenshytal Protection Ageney Environment Canadamp the CampDadim Cout Guard the US Minerals Management Service and numeroubull other pamprticiPBntbull The work pershyformed by the OHMSEIl operating contractor conetitutea oal7 a eampll portion of tbe overall proJect and Do attempt hu been made in this report to describe the complete operations that were conducted or the results that were obtained except u they impacted OllMSEll support operations The report focuses on the planning implementation md data collection ampctiTitiebull provided by Ro7 r Weston Inc md i t1 bullubcontracton IC1 lechDOloa Inc and EnvireepoDle Inc in bullupport of the overall effort1
iii
ABSTRACT
1olloYill8 near~ two years ot planning and coordination a Joint Cbullnadfan - United States oil spill control and cleanup exercise vu conshyducted ott the shore ot lfevtoundland in September 1987 The BPA Oil and Hazshyardous Materials Silllulated lDTironmental lest 1ank (OJIMSBll) tacllity operatshying contractor Rar 1 Weston Inc supported this exercise by proTiding an instlllllented oil containment boom tor eTaluation ot boca pertoriaance methods oil recoTery instrumentation analysis ot the recoTered product and direct support ot the ottshore operations with ten OBMSBll personnel
To evaluate relative boom and water motions preHure transducers and digital data loggers were procured installed on an aTailable oil containment boca and calibrated at the OHMSEll test facility light channels ot such depth-measurill8 instrumentation with one-hour recording capability Vere proTided on the boom that was shipped to St Johns Revtoundland tor the test sequences that occurred during the period ot September 2o-21i 1987
Instrumentation to measure tlov rates and gross Tolumes ot recoTered oil during the sltilllmer-eTaluation phase ot the offshore trials vaa also designed fabricated calibrated and operated by OHMSEll personnel lhis srstem conshysisted ot an in-line Tenturi tlov meter arrangement tank depth-sounding ~es and both in-line and stratified-liquid tank saapling deTicH Samples collected during the recoTery operations were subsequent~ analrsed by ASDf aetbods to determine the percent ot water in the recoTered product
The ottahore teats that were conducted consisted ot a practice run with oa4 simulated oil release on September 21 and a tul1-ecale exercise with releaee ot oil on September 24 Although operational ditticultin were enshycoUDtered in maintaining an acceptable configuration ot the instrumented boom deployed during the ottahore tests titty-eix minutes ot data were obtained OD all eight channelbull during the practice run
Following repair ot instllllllentation cabling that had been damaged during the practice run torty-six minutes ot data from tour channels were recorded during the actual exercise amid continuing problem related to the seaworthishyness ot the instrumented boom In addition tlov rates recoTered TOlUllleB and oilwater ratios were determined tor the various skimmers deployed during the oil recovery phase ot the operations
The OHMSETT actiTitiee were ~ a small portion ot the overall effort during the offshore tests and this report focuses on the experimental proceshydures and findings related to the instrumented boom and the oil recoTery measurements The overall test operations are discussed ~ u they relate to the portions involving OHMSETT participation
iT
COhElllS
Forevordbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 111 A~act bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1T bhibita bullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbull bullbullbullbullbullbullbullbullbullbullbullbull - bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbullbull bull bullbullbullbullbullbullbullbullbullbullbullbullrl AbbreTiationa and S)mbolbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullixActnovledgment bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullx
1 Introduction and Background bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1 2 Conclusionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 3 Recommendationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 4 Materials and Methodbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Boom Selectionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boca Performance Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Flov and RecoTbulllJ Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull PhotoVideo Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Laboratory Equipaent and Materiamplbullmiddotmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other lquipnent and Materials bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
5 Experimental Proceduresbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Booa Illllersion Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boom Tension Measurementbull bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 1lov and RecoTbullrJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull LaboratOlJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
6 Results and Discussionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Rererencebullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Appendices
A PlampIIIled OBMSETT Personnel Duties and Stations During the Ottabore Trialbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
B Computer Programs used to Analyze Teat Datbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Glos aampl) bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull
T
EXHIBITS
1 Transducer Mounting Locations AloDg the Boom
2 Detail Of Transducer MountiDg
3 Cable Harness Arrangement
4 Strain Load Link Arrangement
5 Layout Of Venturi Tube On RecoverT Ve1111el
6 Ex8111ple Of Rav Data output From The Calibration
1 Calibration Curve Of Transducer lo 1
8 Calibration Curve Of Transducer lo 2
9 Calibration Curve Of Transducer lo 3
10 Calibration ot Transducer lo 4
11 Calibration Curve Of Transducer lo 5
12 Calibration Curve For Transducer lo 6
13 Calibration Curve For Transducer lo 1
14 Calibration Curve For Transducer lo 8
15 Example of Rav Data Plot tor the Second Calibration Series
16 Wave Record (Plot Of Immersion Data) From Transducer 1 During The Fixed-Sensor Tests
17 Magnitude-Frequency Spectrum For Transducer 4 Fixed Sensor With Waves
18 Magnitude-Frequency Spectrum tor Transducer 4 Moving Sensor With Waves
19 Tank Testing Of Both Boom Sections
20 Wave Records For Transducers During Initial BoOlll Tests With Waves and Applied Tension
21 Magnitude-Frequency Spectra For Transducers During Tank Tests
vi
22 Measured Tendon On The Boom Por (a) Pirat Ralf Of Teat And (b) Second Bamplf Of Teat During The Practice Run
23 (a) Calibration Of A Rosemount Differentiampl Pressure Gaugebull and (b) Venturi lube
24 Demonstrated Linearit7 Of Venturi Plow Meter At 05 Meters of Water lull Scample Deflection
25 Telog Data Output For 5-second Averaging At 32 gpa Wominampl Plow Rate
26 Wave Record For Sensor 1 During The Practice Run
27 Wave Record Por Sensor 2 During The Practice Run
28 Magnitude Spectrum
29 Wave Record Por Unit 2 Sensor 4 During Offshore Teats
30 Magnitude Sprectrua
31 One-Third Significant Wave Heights Obtained During The Offshore Teat And During The Practice Run
32 Fluid Recoverr Rages For The Jramo Skimmer From Flow Rate Measurements
33 Fluid Recoverr Rates For The Jramo Skimmer From Tank Depth Measurements
34 Fluid Recoverr Rates For The GT185 Skimmer From Flov Rate Measurements
35 Fluid Recover) Rates For The GT185 Skimmer From Tank Depth Measurements
36 Results Of Laboratorr Anal7sia Of In-Line Samples For Percent Water
31 Results Of Laboratorr Analysis Of Stratified Tank Samples For Percent Water
38 Operational Data Logs From Practice Run On 92187
39 Operational Data Logs From Offshore Tests On 92487
vii
LIST OF ABBRBVIATIOIS AID SYMBOLS
ABBREVIATIOfS
OIDCSElr -- Oil and Hazardous Materials Simulated Enrtromientampl Teat Tank
CCG -- Canadian Coast Guard
SIXBOLS
ACKIOWLEDGMDT
Thill vorlr vabull supported b7 Tamprious aember aaenciebull of the US EPAbull OU amplld Hazardous Materialbull Simulated BnTiromental Teat faalr (OHMSEll) Intershyaa-cr Technical COllllittee (OITC)
The OITC currentl7 includes the US EPA US Coast Guard the Mineralbull Management Senice the Canadian Consenation and Protection SerTice and the US lfaT)
The support and encourage-nt ot n11111eroua personnel troa OITC aaencies are aclrnovledged In particular ve Yish to express appreciation to the tolshylaving sponsors and clients tor their continuing support
o US EnTironmental Protection Agenc7 shyRobert Billger Project Otticer and Richard Griffiths
o U S Minerals Manage-nt SerTice shyEdvard TeDD7110D
o EnTironmental Canada - Barrr Whittalrer
o Canadian Coast Guard - William Ryan and Warne Balle7
We are grateful tor the supreme ettorts that the OHllSElf statt proTided during the vorlr at the OBMSETT tacilit7 and in Canada Tralr the project YU
a team ettort the manr and Taried ideas cOIDllleDts and extra ettorts ot the entire statt contributed in no small measure to the succes1 ot the program
ix
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
PORIWORD
[l-2 paragraphbull ~ Standard IPA Foreword)
Thill report contains a bullllllllllampr7 ~ actirltiebull performed and data obtained b7 Ro7 1 Weston Inc the operating contractor for IPAe OHMSBTT tebullt facility in support of a eeriebull of interagenc7 cooperative exercises renown aa the 1987 OftShore Boom and Skimmer Trialbull These ezercieebull were conducted ill Canadian ocean vaten near St Johnbullbull llevtoundland iD September 1987 and iirrolved Joint sponsorship planning and performampDCe b7 the US Enrlronmenshytal Protection Ageney Environment Canadamp the CampDadim Cout Guard the US Minerals Management Service and numeroubull other pamprticiPBntbull The work pershyformed by the OHMSEIl operating contractor conetitutea oal7 a eampll portion of tbe overall proJect and Do attempt hu been made in this report to describe the complete operations that were conducted or the results that were obtained except u they impacted OllMSEll support operations The report focuses on the planning implementation md data collection ampctiTitiebull provided by Ro7 r Weston Inc md i t1 bullubcontracton IC1 lechDOloa Inc and EnvireepoDle Inc in bullupport of the overall effort1
iii
ABSTRACT
1olloYill8 near~ two years ot planning and coordination a Joint Cbullnadfan - United States oil spill control and cleanup exercise vu conshyducted ott the shore ot lfevtoundland in September 1987 The BPA Oil and Hazshyardous Materials Silllulated lDTironmental lest 1ank (OJIMSBll) tacllity operatshying contractor Rar 1 Weston Inc supported this exercise by proTiding an instlllllented oil containment boom tor eTaluation ot boca pertoriaance methods oil recoTery instrumentation analysis ot the recoTered product and direct support ot the ottshore operations with ten OBMSBll personnel
To evaluate relative boom and water motions preHure transducers and digital data loggers were procured installed on an aTailable oil containment boca and calibrated at the OHMSEll test facility light channels ot such depth-measurill8 instrumentation with one-hour recording capability Vere proTided on the boom that was shipped to St Johns Revtoundland tor the test sequences that occurred during the period ot September 2o-21i 1987
Instrumentation to measure tlov rates and gross Tolumes ot recoTered oil during the sltilllmer-eTaluation phase ot the offshore trials vaa also designed fabricated calibrated and operated by OHMSEll personnel lhis srstem conshysisted ot an in-line Tenturi tlov meter arrangement tank depth-sounding ~es and both in-line and stratified-liquid tank saapling deTicH Samples collected during the recoTery operations were subsequent~ analrsed by ASDf aetbods to determine the percent ot water in the recoTered product
The ottahore teats that were conducted consisted ot a practice run with oa4 simulated oil release on September 21 and a tul1-ecale exercise with releaee ot oil on September 24 Although operational ditticultin were enshycoUDtered in maintaining an acceptable configuration ot the instrumented boom deployed during the ottahore tests titty-eix minutes ot data were obtained OD all eight channelbull during the practice run
Following repair ot instllllllentation cabling that had been damaged during the practice run torty-six minutes ot data from tour channels were recorded during the actual exercise amid continuing problem related to the seaworthishyness ot the instrumented boom In addition tlov rates recoTered TOlUllleB and oilwater ratios were determined tor the various skimmers deployed during the oil recovery phase ot the operations
The OHMSETT actiTitiee were ~ a small portion ot the overall effort during the offshore tests and this report focuses on the experimental proceshydures and findings related to the instrumented boom and the oil recoTery measurements The overall test operations are discussed ~ u they relate to the portions involving OHMSETT participation
iT
COhElllS
Forevordbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 111 A~act bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1T bhibita bullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbull bullbullbullbullbullbullbullbullbullbullbullbull - bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbullbull bull bullbullbullbullbullbullbullbullbullbullbullbullrl AbbreTiationa and S)mbolbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullixActnovledgment bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullx
1 Introduction and Background bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1 2 Conclusionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 3 Recommendationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 4 Materials and Methodbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Boom Selectionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boca Performance Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Flov and RecoTbulllJ Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull PhotoVideo Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Laboratory Equipaent and Materiamplbullmiddotmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other lquipnent and Materials bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
5 Experimental Proceduresbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Booa Illllersion Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boom Tension Measurementbull bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 1lov and RecoTbullrJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull LaboratOlJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
6 Results and Discussionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Rererencebullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Appendices
A PlampIIIled OBMSETT Personnel Duties and Stations During the Ottabore Trialbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
B Computer Programs used to Analyze Teat Datbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Glos aampl) bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull
T
EXHIBITS
1 Transducer Mounting Locations AloDg the Boom
2 Detail Of Transducer MountiDg
3 Cable Harness Arrangement
4 Strain Load Link Arrangement
5 Layout Of Venturi Tube On RecoverT Ve1111el
6 Ex8111ple Of Rav Data output From The Calibration
1 Calibration Curve Of Transducer lo 1
8 Calibration Curve Of Transducer lo 2
9 Calibration Curve Of Transducer lo 3
10 Calibration ot Transducer lo 4
11 Calibration Curve Of Transducer lo 5
12 Calibration Curve For Transducer lo 6
13 Calibration Curve For Transducer lo 1
14 Calibration Curve For Transducer lo 8
15 Example of Rav Data Plot tor the Second Calibration Series
16 Wave Record (Plot Of Immersion Data) From Transducer 1 During The Fixed-Sensor Tests
17 Magnitude-Frequency Spectrum For Transducer 4 Fixed Sensor With Waves
18 Magnitude-Frequency Spectrum tor Transducer 4 Moving Sensor With Waves
19 Tank Testing Of Both Boom Sections
20 Wave Records For Transducers During Initial BoOlll Tests With Waves and Applied Tension
21 Magnitude-Frequency Spectra For Transducers During Tank Tests
vi
22 Measured Tendon On The Boom Por (a) Pirat Ralf Of Teat And (b) Second Bamplf Of Teat During The Practice Run
23 (a) Calibration Of A Rosemount Differentiampl Pressure Gaugebull and (b) Venturi lube
24 Demonstrated Linearit7 Of Venturi Plow Meter At 05 Meters of Water lull Scample Deflection
25 Telog Data Output For 5-second Averaging At 32 gpa Wominampl Plow Rate
26 Wave Record For Sensor 1 During The Practice Run
27 Wave Record Por Sensor 2 During The Practice Run
28 Magnitude Spectrum
29 Wave Record Por Unit 2 Sensor 4 During Offshore Teats
30 Magnitude Sprectrua
31 One-Third Significant Wave Heights Obtained During The Offshore Teat And During The Practice Run
32 Fluid Recoverr Rages For The Jramo Skimmer From Flow Rate Measurements
33 Fluid Recoverr Rates For The Jramo Skimmer From Tank Depth Measurements
34 Fluid Recoverr Rates For The GT185 Skimmer From Flov Rate Measurements
35 Fluid Recover) Rates For The GT185 Skimmer From Tank Depth Measurements
36 Results Of Laboratorr Anal7sia Of In-Line Samples For Percent Water
31 Results Of Laboratorr Analysis Of Stratified Tank Samples For Percent Water
38 Operational Data Logs From Practice Run On 92187
39 Operational Data Logs From Offshore Tests On 92487
vii
LIST OF ABBRBVIATIOIS AID SYMBOLS
ABBREVIATIOfS
OIDCSElr -- Oil and Hazardous Materials Simulated Enrtromientampl Teat Tank
CCG -- Canadian Coast Guard
SIXBOLS
ACKIOWLEDGMDT
Thill vorlr vabull supported b7 Tamprious aember aaenciebull of the US EPAbull OU amplld Hazardous Materialbull Simulated BnTiromental Teat faalr (OHMSEll) Intershyaa-cr Technical COllllittee (OITC)
The OITC currentl7 includes the US EPA US Coast Guard the Mineralbull Management Senice the Canadian Consenation and Protection SerTice and the US lfaT)
The support and encourage-nt ot n11111eroua personnel troa OITC aaencies are aclrnovledged In particular ve Yish to express appreciation to the tolshylaving sponsors and clients tor their continuing support
o US EnTironmental Protection Agenc7 shyRobert Billger Project Otticer and Richard Griffiths
o U S Minerals Manage-nt SerTice shyEdvard TeDD7110D
o EnTironmental Canada - Barrr Whittalrer
o Canadian Coast Guard - William Ryan and Warne Balle7
We are grateful tor the supreme ettorts that the OHllSElf statt proTided during the vorlr at the OBMSETT tacilit7 and in Canada Tralr the project YU
a team ettort the manr and Taried ideas cOIDllleDts and extra ettorts ot the entire statt contributed in no small measure to the succes1 ot the program
ix
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
ABSTRACT
1olloYill8 near~ two years ot planning and coordination a Joint Cbullnadfan - United States oil spill control and cleanup exercise vu conshyducted ott the shore ot lfevtoundland in September 1987 The BPA Oil and Hazshyardous Materials Silllulated lDTironmental lest 1ank (OJIMSBll) tacllity operatshying contractor Rar 1 Weston Inc supported this exercise by proTiding an instlllllented oil containment boom tor eTaluation ot boca pertoriaance methods oil recoTery instrumentation analysis ot the recoTered product and direct support ot the ottshore operations with ten OBMSBll personnel
To evaluate relative boom and water motions preHure transducers and digital data loggers were procured installed on an aTailable oil containment boca and calibrated at the OHMSEll test facility light channels ot such depth-measurill8 instrumentation with one-hour recording capability Vere proTided on the boom that was shipped to St Johns Revtoundland tor the test sequences that occurred during the period ot September 2o-21i 1987
Instrumentation to measure tlov rates and gross Tolumes ot recoTered oil during the sltilllmer-eTaluation phase ot the offshore trials vaa also designed fabricated calibrated and operated by OHMSEll personnel lhis srstem conshysisted ot an in-line Tenturi tlov meter arrangement tank depth-sounding ~es and both in-line and stratified-liquid tank saapling deTicH Samples collected during the recoTery operations were subsequent~ analrsed by ASDf aetbods to determine the percent ot water in the recoTered product
The ottahore teats that were conducted consisted ot a practice run with oa4 simulated oil release on September 21 and a tul1-ecale exercise with releaee ot oil on September 24 Although operational ditticultin were enshycoUDtered in maintaining an acceptable configuration ot the instrumented boom deployed during the ottahore tests titty-eix minutes ot data were obtained OD all eight channelbull during the practice run
Following repair ot instllllllentation cabling that had been damaged during the practice run torty-six minutes ot data from tour channels were recorded during the actual exercise amid continuing problem related to the seaworthishyness ot the instrumented boom In addition tlov rates recoTered TOlUllleB and oilwater ratios were determined tor the various skimmers deployed during the oil recovery phase ot the operations
The OHMSETT actiTitiee were ~ a small portion ot the overall effort during the offshore tests and this report focuses on the experimental proceshydures and findings related to the instrumented boom and the oil recoTery measurements The overall test operations are discussed ~ u they relate to the portions involving OHMSETT participation
iT
COhElllS
Forevordbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 111 A~act bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1T bhibita bullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbull bullbullbullbullbullbullbullbullbullbullbullbull - bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbullbull bull bullbullbullbullbullbullbullbullbullbullbullbullrl AbbreTiationa and S)mbolbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullixActnovledgment bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullx
1 Introduction and Background bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1 2 Conclusionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 3 Recommendationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 4 Materials and Methodbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Boom Selectionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boca Performance Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Flov and RecoTbulllJ Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull PhotoVideo Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Laboratory Equipaent and Materiamplbullmiddotmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other lquipnent and Materials bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
5 Experimental Proceduresbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Booa Illllersion Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boom Tension Measurementbull bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 1lov and RecoTbullrJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull LaboratOlJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
6 Results and Discussionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Rererencebullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Appendices
A PlampIIIled OBMSETT Personnel Duties and Stations During the Ottabore Trialbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
B Computer Programs used to Analyze Teat Datbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Glos aampl) bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull
T
EXHIBITS
1 Transducer Mounting Locations AloDg the Boom
2 Detail Of Transducer MountiDg
3 Cable Harness Arrangement
4 Strain Load Link Arrangement
5 Layout Of Venturi Tube On RecoverT Ve1111el
6 Ex8111ple Of Rav Data output From The Calibration
1 Calibration Curve Of Transducer lo 1
8 Calibration Curve Of Transducer lo 2
9 Calibration Curve Of Transducer lo 3
10 Calibration ot Transducer lo 4
11 Calibration Curve Of Transducer lo 5
12 Calibration Curve For Transducer lo 6
13 Calibration Curve For Transducer lo 1
14 Calibration Curve For Transducer lo 8
15 Example of Rav Data Plot tor the Second Calibration Series
16 Wave Record (Plot Of Immersion Data) From Transducer 1 During The Fixed-Sensor Tests
17 Magnitude-Frequency Spectrum For Transducer 4 Fixed Sensor With Waves
18 Magnitude-Frequency Spectrum tor Transducer 4 Moving Sensor With Waves
19 Tank Testing Of Both Boom Sections
20 Wave Records For Transducers During Initial BoOlll Tests With Waves and Applied Tension
21 Magnitude-Frequency Spectra For Transducers During Tank Tests
vi
22 Measured Tendon On The Boom Por (a) Pirat Ralf Of Teat And (b) Second Bamplf Of Teat During The Practice Run
23 (a) Calibration Of A Rosemount Differentiampl Pressure Gaugebull and (b) Venturi lube
24 Demonstrated Linearit7 Of Venturi Plow Meter At 05 Meters of Water lull Scample Deflection
25 Telog Data Output For 5-second Averaging At 32 gpa Wominampl Plow Rate
26 Wave Record For Sensor 1 During The Practice Run
27 Wave Record Por Sensor 2 During The Practice Run
28 Magnitude Spectrum
29 Wave Record Por Unit 2 Sensor 4 During Offshore Teats
30 Magnitude Sprectrua
31 One-Third Significant Wave Heights Obtained During The Offshore Teat And During The Practice Run
32 Fluid Recoverr Rages For The Jramo Skimmer From Flow Rate Measurements
33 Fluid Recoverr Rates For The Jramo Skimmer From Tank Depth Measurements
34 Fluid Recoverr Rates For The GT185 Skimmer From Flov Rate Measurements
35 Fluid Recover) Rates For The GT185 Skimmer From Tank Depth Measurements
36 Results Of Laboratorr Anal7sia Of In-Line Samples For Percent Water
31 Results Of Laboratorr Analysis Of Stratified Tank Samples For Percent Water
38 Operational Data Logs From Practice Run On 92187
39 Operational Data Logs From Offshore Tests On 92487
vii
LIST OF ABBRBVIATIOIS AID SYMBOLS
ABBREVIATIOfS
OIDCSElr -- Oil and Hazardous Materials Simulated Enrtromientampl Teat Tank
CCG -- Canadian Coast Guard
SIXBOLS
ACKIOWLEDGMDT
Thill vorlr vabull supported b7 Tamprious aember aaenciebull of the US EPAbull OU amplld Hazardous Materialbull Simulated BnTiromental Teat faalr (OHMSEll) Intershyaa-cr Technical COllllittee (OITC)
The OITC currentl7 includes the US EPA US Coast Guard the Mineralbull Management Senice the Canadian Consenation and Protection SerTice and the US lfaT)
The support and encourage-nt ot n11111eroua personnel troa OITC aaencies are aclrnovledged In particular ve Yish to express appreciation to the tolshylaving sponsors and clients tor their continuing support
o US EnTironmental Protection Agenc7 shyRobert Billger Project Otticer and Richard Griffiths
o U S Minerals Manage-nt SerTice shyEdvard TeDD7110D
o EnTironmental Canada - Barrr Whittalrer
o Canadian Coast Guard - William Ryan and Warne Balle7
We are grateful tor the supreme ettorts that the OHllSElf statt proTided during the vorlr at the OBMSETT tacilit7 and in Canada Tralr the project YU
a team ettort the manr and Taried ideas cOIDllleDts and extra ettorts ot the entire statt contributed in no small measure to the succes1 ot the program
ix
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
COhElllS
Forevordbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 111 A~act bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1T bhibita bullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbull bullbullbullbullbullbullbullbullbullbullbullbull - bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bullbullbullbull bull bullbullbullbullbullbullbullbullbullbullbullbullrl AbbreTiationa and S)mbolbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullixActnovledgment bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullx
1 Introduction and Background bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull1 2 Conclusionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 3 Recommendationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 4 Materials and Methodbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Boom Selectionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boca Performance Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Flov and RecoTbulllJ Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull PhotoVideo Instrumentationbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Laboratory Equipaent and Materiamplbullmiddotmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other lquipnent and Materials bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
5 Experimental Proceduresbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Booa Illllersion Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Boom Tension Measurementbull bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 1lov and RecoTbullrJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull LaboratOlJ Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Other Measurementbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
6 Results and Discussionbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Rererencebullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull Appendices
A PlampIIIled OBMSETT Personnel Duties and Stations During the Ottabore Trialbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
B Computer Programs used to Analyze Teat Datbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
Glos aampl) bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull bull
T
EXHIBITS
1 Transducer Mounting Locations AloDg the Boom
2 Detail Of Transducer MountiDg
3 Cable Harness Arrangement
4 Strain Load Link Arrangement
5 Layout Of Venturi Tube On RecoverT Ve1111el
6 Ex8111ple Of Rav Data output From The Calibration
1 Calibration Curve Of Transducer lo 1
8 Calibration Curve Of Transducer lo 2
9 Calibration Curve Of Transducer lo 3
10 Calibration ot Transducer lo 4
11 Calibration Curve Of Transducer lo 5
12 Calibration Curve For Transducer lo 6
13 Calibration Curve For Transducer lo 1
14 Calibration Curve For Transducer lo 8
15 Example of Rav Data Plot tor the Second Calibration Series
16 Wave Record (Plot Of Immersion Data) From Transducer 1 During The Fixed-Sensor Tests
17 Magnitude-Frequency Spectrum For Transducer 4 Fixed Sensor With Waves
18 Magnitude-Frequency Spectrum tor Transducer 4 Moving Sensor With Waves
19 Tank Testing Of Both Boom Sections
20 Wave Records For Transducers During Initial BoOlll Tests With Waves and Applied Tension
21 Magnitude-Frequency Spectra For Transducers During Tank Tests
vi
22 Measured Tendon On The Boom Por (a) Pirat Ralf Of Teat And (b) Second Bamplf Of Teat During The Practice Run
23 (a) Calibration Of A Rosemount Differentiampl Pressure Gaugebull and (b) Venturi lube
24 Demonstrated Linearit7 Of Venturi Plow Meter At 05 Meters of Water lull Scample Deflection
25 Telog Data Output For 5-second Averaging At 32 gpa Wominampl Plow Rate
26 Wave Record For Sensor 1 During The Practice Run
27 Wave Record Por Sensor 2 During The Practice Run
28 Magnitude Spectrum
29 Wave Record Por Unit 2 Sensor 4 During Offshore Teats
30 Magnitude Sprectrua
31 One-Third Significant Wave Heights Obtained During The Offshore Teat And During The Practice Run
32 Fluid Recoverr Rages For The Jramo Skimmer From Flow Rate Measurements
33 Fluid Recoverr Rates For The Jramo Skimmer From Tank Depth Measurements
34 Fluid Recoverr Rates For The GT185 Skimmer From Flov Rate Measurements
35 Fluid Recover) Rates For The GT185 Skimmer From Tank Depth Measurements
36 Results Of Laboratorr Anal7sia Of In-Line Samples For Percent Water
31 Results Of Laboratorr Analysis Of Stratified Tank Samples For Percent Water
38 Operational Data Logs From Practice Run On 92187
39 Operational Data Logs From Offshore Tests On 92487
vii
LIST OF ABBRBVIATIOIS AID SYMBOLS
ABBREVIATIOfS
OIDCSElr -- Oil and Hazardous Materials Simulated Enrtromientampl Teat Tank
CCG -- Canadian Coast Guard
SIXBOLS
ACKIOWLEDGMDT
Thill vorlr vabull supported b7 Tamprious aember aaenciebull of the US EPAbull OU amplld Hazardous Materialbull Simulated BnTiromental Teat faalr (OHMSEll) Intershyaa-cr Technical COllllittee (OITC)
The OITC currentl7 includes the US EPA US Coast Guard the Mineralbull Management Senice the Canadian Consenation and Protection SerTice and the US lfaT)
The support and encourage-nt ot n11111eroua personnel troa OITC aaencies are aclrnovledged In particular ve Yish to express appreciation to the tolshylaving sponsors and clients tor their continuing support
o US EnTironmental Protection Agenc7 shyRobert Billger Project Otticer and Richard Griffiths
o U S Minerals Manage-nt SerTice shyEdvard TeDD7110D
o EnTironmental Canada - Barrr Whittalrer
o Canadian Coast Guard - William Ryan and Warne Balle7
We are grateful tor the supreme ettorts that the OHllSElf statt proTided during the vorlr at the OBMSETT tacilit7 and in Canada Tralr the project YU
a team ettort the manr and Taried ideas cOIDllleDts and extra ettorts ot the entire statt contributed in no small measure to the succes1 ot the program
ix
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
EXHIBITS
1 Transducer Mounting Locations AloDg the Boom
2 Detail Of Transducer MountiDg
3 Cable Harness Arrangement
4 Strain Load Link Arrangement
5 Layout Of Venturi Tube On RecoverT Ve1111el
6 Ex8111ple Of Rav Data output From The Calibration
1 Calibration Curve Of Transducer lo 1
8 Calibration Curve Of Transducer lo 2
9 Calibration Curve Of Transducer lo 3
10 Calibration ot Transducer lo 4
11 Calibration Curve Of Transducer lo 5
12 Calibration Curve For Transducer lo 6
13 Calibration Curve For Transducer lo 1
14 Calibration Curve For Transducer lo 8
15 Example of Rav Data Plot tor the Second Calibration Series
16 Wave Record (Plot Of Immersion Data) From Transducer 1 During The Fixed-Sensor Tests
17 Magnitude-Frequency Spectrum For Transducer 4 Fixed Sensor With Waves
18 Magnitude-Frequency Spectrum tor Transducer 4 Moving Sensor With Waves
19 Tank Testing Of Both Boom Sections
20 Wave Records For Transducers During Initial BoOlll Tests With Waves and Applied Tension
21 Magnitude-Frequency Spectra For Transducers During Tank Tests
vi
22 Measured Tendon On The Boom Por (a) Pirat Ralf Of Teat And (b) Second Bamplf Of Teat During The Practice Run
23 (a) Calibration Of A Rosemount Differentiampl Pressure Gaugebull and (b) Venturi lube
24 Demonstrated Linearit7 Of Venturi Plow Meter At 05 Meters of Water lull Scample Deflection
25 Telog Data Output For 5-second Averaging At 32 gpa Wominampl Plow Rate
26 Wave Record For Sensor 1 During The Practice Run
27 Wave Record Por Sensor 2 During The Practice Run
28 Magnitude Spectrum
29 Wave Record Por Unit 2 Sensor 4 During Offshore Teats
30 Magnitude Sprectrua
31 One-Third Significant Wave Heights Obtained During The Offshore Teat And During The Practice Run
32 Fluid Recoverr Rages For The Jramo Skimmer From Flow Rate Measurements
33 Fluid Recoverr Rates For The Jramo Skimmer From Tank Depth Measurements
34 Fluid Recoverr Rates For The GT185 Skimmer From Flov Rate Measurements
35 Fluid Recover) Rates For The GT185 Skimmer From Tank Depth Measurements
36 Results Of Laboratorr Anal7sia Of In-Line Samples For Percent Water
31 Results Of Laboratorr Analysis Of Stratified Tank Samples For Percent Water
38 Operational Data Logs From Practice Run On 92187
39 Operational Data Logs From Offshore Tests On 92487
vii
LIST OF ABBRBVIATIOIS AID SYMBOLS
ABBREVIATIOfS
OIDCSElr -- Oil and Hazardous Materials Simulated Enrtromientampl Teat Tank
CCG -- Canadian Coast Guard
SIXBOLS
ACKIOWLEDGMDT
Thill vorlr vabull supported b7 Tamprious aember aaenciebull of the US EPAbull OU amplld Hazardous Materialbull Simulated BnTiromental Teat faalr (OHMSEll) Intershyaa-cr Technical COllllittee (OITC)
The OITC currentl7 includes the US EPA US Coast Guard the Mineralbull Management Senice the Canadian Consenation and Protection SerTice and the US lfaT)
The support and encourage-nt ot n11111eroua personnel troa OITC aaencies are aclrnovledged In particular ve Yish to express appreciation to the tolshylaving sponsors and clients tor their continuing support
o US EnTironmental Protection Agenc7 shyRobert Billger Project Otticer and Richard Griffiths
o U S Minerals Manage-nt SerTice shyEdvard TeDD7110D
o EnTironmental Canada - Barrr Whittalrer
o Canadian Coast Guard - William Ryan and Warne Balle7
We are grateful tor the supreme ettorts that the OHllSElf statt proTided during the vorlr at the OBMSETT tacilit7 and in Canada Tralr the project YU
a team ettort the manr and Taried ideas cOIDllleDts and extra ettorts ot the entire statt contributed in no small measure to the succes1 ot the program
ix
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
22 Measured Tendon On The Boom Por (a) Pirat Ralf Of Teat And (b) Second Bamplf Of Teat During The Practice Run
23 (a) Calibration Of A Rosemount Differentiampl Pressure Gaugebull and (b) Venturi lube
24 Demonstrated Linearit7 Of Venturi Plow Meter At 05 Meters of Water lull Scample Deflection
25 Telog Data Output For 5-second Averaging At 32 gpa Wominampl Plow Rate
26 Wave Record For Sensor 1 During The Practice Run
27 Wave Record Por Sensor 2 During The Practice Run
28 Magnitude Spectrum
29 Wave Record Por Unit 2 Sensor 4 During Offshore Teats
30 Magnitude Sprectrua
31 One-Third Significant Wave Heights Obtained During The Offshore Teat And During The Practice Run
32 Fluid Recoverr Rages For The Jramo Skimmer From Flow Rate Measurements
33 Fluid Recoverr Rates For The Jramo Skimmer From Tank Depth Measurements
34 Fluid Recoverr Rates For The GT185 Skimmer From Flov Rate Measurements
35 Fluid Recover) Rates For The GT185 Skimmer From Tank Depth Measurements
36 Results Of Laboratorr Anal7sia Of In-Line Samples For Percent Water
31 Results Of Laboratorr Analysis Of Stratified Tank Samples For Percent Water
38 Operational Data Logs From Practice Run On 92187
39 Operational Data Logs From Offshore Tests On 92487
vii
LIST OF ABBRBVIATIOIS AID SYMBOLS
ABBREVIATIOfS
OIDCSElr -- Oil and Hazardous Materials Simulated Enrtromientampl Teat Tank
CCG -- Canadian Coast Guard
SIXBOLS
ACKIOWLEDGMDT
Thill vorlr vabull supported b7 Tamprious aember aaenciebull of the US EPAbull OU amplld Hazardous Materialbull Simulated BnTiromental Teat faalr (OHMSEll) Intershyaa-cr Technical COllllittee (OITC)
The OITC currentl7 includes the US EPA US Coast Guard the Mineralbull Management Senice the Canadian Consenation and Protection SerTice and the US lfaT)
The support and encourage-nt ot n11111eroua personnel troa OITC aaencies are aclrnovledged In particular ve Yish to express appreciation to the tolshylaving sponsors and clients tor their continuing support
o US EnTironmental Protection Agenc7 shyRobert Billger Project Otticer and Richard Griffiths
o U S Minerals Manage-nt SerTice shyEdvard TeDD7110D
o EnTironmental Canada - Barrr Whittalrer
o Canadian Coast Guard - William Ryan and Warne Balle7
We are grateful tor the supreme ettorts that the OHllSElf statt proTided during the vorlr at the OBMSETT tacilit7 and in Canada Tralr the project YU
a team ettort the manr and Taried ideas cOIDllleDts and extra ettorts ot the entire statt contributed in no small measure to the succes1 ot the program
ix
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
LIST OF ABBRBVIATIOIS AID SYMBOLS
ABBREVIATIOfS
OIDCSElr -- Oil and Hazardous Materials Simulated Enrtromientampl Teat Tank
CCG -- Canadian Coast Guard
SIXBOLS
ACKIOWLEDGMDT
Thill vorlr vabull supported b7 Tamprious aember aaenciebull of the US EPAbull OU amplld Hazardous Materialbull Simulated BnTiromental Teat faalr (OHMSEll) Intershyaa-cr Technical COllllittee (OITC)
The OITC currentl7 includes the US EPA US Coast Guard the Mineralbull Management Senice the Canadian Consenation and Protection SerTice and the US lfaT)
The support and encourage-nt ot n11111eroua personnel troa OITC aaencies are aclrnovledged In particular ve Yish to express appreciation to the tolshylaving sponsors and clients tor their continuing support
o US EnTironmental Protection Agenc7 shyRobert Billger Project Otticer and Richard Griffiths
o U S Minerals Manage-nt SerTice shyEdvard TeDD7110D
o EnTironmental Canada - Barrr Whittalrer
o Canadian Coast Guard - William Ryan and Warne Balle7
We are grateful tor the supreme ettorts that the OHllSElf statt proTided during the vorlr at the OBMSETT tacilit7 and in Canada Tralr the project YU
a team ettort the manr and Taried ideas cOIDllleDts and extra ettorts ot the entire statt contributed in no small measure to the succes1 ot the program
ix
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
ACKIOWLEDGMDT
Thill vorlr vabull supported b7 Tamprious aember aaenciebull of the US EPAbull OU amplld Hazardous Materialbull Simulated BnTiromental Teat faalr (OHMSEll) Intershyaa-cr Technical COllllittee (OITC)
The OITC currentl7 includes the US EPA US Coast Guard the Mineralbull Management Senice the Canadian Consenation and Protection SerTice and the US lfaT)
The support and encourage-nt ot n11111eroua personnel troa OITC aaencies are aclrnovledged In particular ve Yish to express appreciation to the tolshylaving sponsors and clients tor their continuing support
o US EnTironmental Protection Agenc7 shyRobert Billger Project Otticer and Richard Griffiths
o U S Minerals Manage-nt SerTice shyEdvard TeDD7110D
o EnTironmental Canada - Barrr Whittalrer
o Canadian Coast Guard - William Ryan and Warne Balle7
We are grateful tor the supreme ettorts that the OHllSElf statt proTided during the vorlr at the OBMSETT tacilit7 and in Canada Tralr the project YU
a team ettort the manr and Taried ideas cOIDllleDts and extra ettorts ot the entire statt contributed in no small measure to the succes1 ot the program
ix
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
SECTIOI 1
IllTRODUCTIOll
BOC11 PERFORMABCE EVALUATIOll
The OHMSETT Interagency Technicampl Committee (OITC) has been supporting
oil spill research and spill containment procedures development tor mB111
Much ot this work has been performed at the US EPAbull Oil and Hazardoua
Materiampls Simulated Test Tank (OllM3ETT) in Leonardo llJ and by the OHMSEll
contractors at other locations A summary ot work performed through 1979 b7
the OllM3ETT contractor has been published (l) Since 1980 a variety ot reshy
search programs have been conducted to evaluate the performance ot oil conshy
tainment booms to determine the environmentamp] and operationamp] factors that
attect boom performance to determine whether boom performance can be quanshy
titatively measured and reliably predicted and to develop a protocol tor
eTampluating boom performance without the need tor spilling oil in open-water
tests A brief sllllllllampry ot the considerable previous work leading to the efshy
forts described in this report is presented below
The large amount ot data acquired at the OHMSETT facility on a broad
variety ot booms have demonstrated that performance (defined as the ability ot
a moving boom to contain oil) depends on several operational characteristics
Booa Test Report August 10 1988 Page 1
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
includill8 the following
o Speed of the moving boom relative to the surface current
o Kinematic properties of the water or oilwater med11111 including
wave height wave period wave uniformitr deneitr and rtecoeitr
o Mechanical and kinetic properties of the boom
The prertoue work has shown that whereas a complete analreb of bocm
design from first principles would be exceedi~ difficult or impossible it
lllQ be possible to evaluate boom performance br selected testing under conshy
trolled conditions SpecificaUr it has been postulated that bOOlll perforshy
llampDce can be demonstrated reliablr br measuring the relative motions of boom
canponente and the water medi1111 using the sea state as a forcing function
without the need for spillill8 oil in open waters
At the time the current project was initiated a methodology for conshy
ducting boom performance testing and for analrzing the data had been proposed
in the form of a draft Boom Teet Protocol ( 2 ) In its purest form the
protocol would allow boom performance to be predicted solely from the seakeepshy
ing ability of the boom in open waters as measured by the motion of the boom
relative to the water If this analrsis proved to be intractable the
protocol includes provisions for conducting both in-tank tests (where spilled
oil would be controlled) and open ocean tests without the need for releasing
oil
Boom Test Report August 10 1988 Page 2
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
As a result or the previoua vork it haa been poatulated that the
aeakeeping ability of the boom can be quantified tram the relative motiona of
boom and medium by measuring the frequency and energy auociated Yith each
110ving mass From these motion spectra the ratios or power den1itie1 for
the boom and the water at each diacrete frequency of oscillation can be
derived The Boom Teat Protocol specifies that these ratios called Response
Amplitude Operators (RAO) as measures or the wave-riding or wave-following
ability of the boom are related directly to the ability of the boom to conshy
tain oU
The boom motion spectra can be obtained from direct accelerometer
measurements or trom amplitude (depth) records over time for critical segshy
ments of the boom The motion spectra of the medium can be obtained trom
siailar instrumentation attached to a wave-rider buoy
The work described in this report stemmed trom a recognition that the
methodqlogy for evaluating boom performance had not been tested adequately unshy
der open-sea conditions Additional data were needed during boom deployment
in sea states typical of those encountered in response actions and during acshy
tual oil containment in open water
OIL RECOVERY EVALUATIONS
The gross performance of oil recovery equipment is 110st simply evaluated
as the volume of oil recovered per unit time The two pamprlllleters which must
BoOlll Test Report August 10 1988 Page 3
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
be -aeured for such evaluations are the recoveey rate of product (generaJ4
an oll-vater mixture) and the recoveey efficiencr or the relative amounts of
oil and water in the recovered product These parameters can be determined
dznamically from measurements of flow rate and collection of in-line samples
for analraia or ataticalJy from volume measurements and stratified sampling
of bulk recovered product Considering the nuaerous operational parameters
that directly affect recoveey methods significant variation in recoveey rate
and efficiencr is expected over even short periods of time Therefore
recoveey performance is beat evaluated by integration of dynamic measurements
A large bodr of information on oil recoveey evaluations has been genshy
erated at the OHMSETT facilitr and elsewhere Extensive testing has been pershy
forwed vithin test tanks and contained test pools BoweTer fewer measureshy
ments have been made for open-sea spills The instances in which recoveey of
ampDT accidental release can be instrumented are rare and there have been only
a limited nU1Dber of cases where performance of recoveey equipnent has been
documented adequately
OFFSHORE OPERATIONAL TESTS AND REPORT
For over two years plans have been under development br OITC member
~encies for conducting a aeries of open-water tests in which various oil conshy
tainment and recovery equipment could be evaluated These teats have been
designed to allow evaluation of oil contaimnent boom performance and oil
recovery equipment and procedures within an operational spill scenario The
overall planned test objectives procedures and schedule are described in Teat
Bolta Test Report August 10 1988 Page 4
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
Protocol for Offshore BoOlll and Skimmer Trials 3rd Draft SL Ross Environshy
11eataJ Research Limited May 1987 The offshore trials vere Tieved as a
realistic opportunity to test the boOlll performance enluation methodology
deTeloped by OITC-spon11ored research oTer the past several years as well as
to evaluate quantitatively the recovery efficiency of different types of sltimshy
aer equipnent Accordingly the OlMSElr contractor vu tasked with providing
technical support for the offshore tests in the areas of boom performance and
oi1 recovery instrumentation
bull
The purpose of this report is to provide a summary of the systems that
were developed and deployed the measurements that were made and the data
and results obtained during the OHMSJllT portions of the Offshore Boom and
Skmlller Trials Jo attempt bas been made to describe or document the overall
offshore test program of which the OHMSJllT activities were ~ a small part
OBJECTIVES
The objectives of the OHMSJllT project were to
1) Design instrumentation for collection of boom performance data
and to prepare and deliver to the Offshore Trial assembly point an
operational length of oil boom
2) Incorporate testing of the instrumented boom for wave-following
properties within the context of the overall Offshore Boan Trial
plans
Boom Test Report August 10 1988 Page 5
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
3) AnamplJze the bOOlll performance data according to the draft boca test
protocol 1 1111d
4) Provide instrumentation methods and personnel tor collecting data
on the rate of oil recover during comparative skimmer operations
AUTHORITIES
Thill project was authorized br Work Assignment lo oa7204 Task 2
(dated July 17 1987) under Contract llo 68-03-3450 trca the USEPA to Roy F
Weston Inc the operating contractor tor the USEPA OBMSETl tacUitr in
Leonardo NJ and br Contract No KE144-7-60480l-SS from the Environment
Cbullnitda to WESTON The USEPA project primarily involved preparation of the inshy
strumental systems and the Canadian portion of the project dealt excluaivelr
vith direct technical support to the offshore operations Thia report has
been prepared to aatisy the requirements of both contracts
Boca Teat Report August 10 1988 Page 6
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
SECTIOf 2
COllCLUSIOllS
OVBRALL COllCLUSIORS
The overall conclusions of this project from the standpoint of OBMSETT
involvement in the Offshore Boom and Skimmer Trials are as folloVB
1 The feasibilit7 of instrumenting an oil containllent boOll collectshy
ing data in an open-water environment and analrzing the results
in accordance with procedures specified in the draft Boom Test
Protocol was demonstrated
2 Off-the-shelf instrumental srstems are available for measuring the
boom performance parameters required br the Protocol
3 An alternate methodologr is available for an~is of boom perforshy
mance in the event that data on the sea state are unavailable
4 Relative recoverr rates and recoverr efficiencies of oil ski11111ing
equipnent can be measured in an operational scenario by a comshy
bination of flow meters and oilwater ratio product analysis
Boca Test Report August 10 1988 Page 7
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
-
SPBCIFIC CONCUJSIOBS
1 First response strain-gauge bridge preSBure transducers are apshy
propriate sensors for detennination of boom motions under condishy
tions of open-sea deployment
2 Digital data loggers of appropriate resolution and memory capacit7
are available for recording of data required b7 the Boom Test
Protocol
3 The conditions existing during the tests vere not ideal for
evaluation of oil retention booma according to the proposed
protocol In particular tov speeds vere generall7 too high and
vere not varied through the stages of oil retention first loss
and groBB loss as planned The selected boom did not perfonn veil
under the conditions of high tov speed and high sea state and
difficulties vere encountered in maintaining proper boom tension
and configuration
4 Despite the operational difficulties the general methodology for
instrumenting a boom to measure and record vave-folloving abilit7
vas tested successfull7 Instrumenting a boom to provide very
precise data on immersion and motion characteristics while
withstanding the rigors of extensive testing vas shown to be
feasible
Boom Test Report August 10 1988 Page 8
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
5 The methodoloa tor computiDg boom performance parameters appears
to be valid tor actual oil-spill data
6 The methods chosen to measure oil recover rates and etticieney
appear to be acceptable
Boom Test Report August 10 1988 Page 9
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
SEXTIOR 3
RECOMMEllDATIORS
Although the overall feasibilit7 of instrumenting oil containment and
recovery equipnent to provide quantitative evaluation of performance measures
was shown during this program it was also indicated that improvements are
possible Additional work on tension and recovery flow measurements should be
accomplished before procedures are fixed or standards are developed
Sealed pressure transducers with no capillary tube vent are recommended
for use in performing boom immersion and motion measurements The small drift
that might be experienced for the closed system rill insignificantly affect
subsequent data interpretation and the problems encountered in maintaining a
drr transducer element would be avoided
Boom Test Report August 10 1988 Page 10
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
SECTION 4
METHODS AllD MATERIALS
S~TION OF THE BOOM
Beginning in July 1987 OBMSElr began preparing and calibrating inshy
strumentation tor measuring boom performance during the offshore trials AJshy
thoaamph the general requirements both tor instrumentamp design and tor specific
components was known tre111 previous OBMSlrT work the exact specifications and
methods tor ruggedizing and mounting of components depended on the exact bocn
structure that would serve as the instrumental platform during the tests
Thus the initial step was to obtain a typical oil containment boom tor which
instrumentation would be procured installed and tested
The available and suitable options tor booms were reviewed at the onset
ot ~be program This research quickly led to selection of the Globe Oil Fence
48 boom tor severampl reasons
(1) A significant length 152111 of the Oil Fence 48 was on hand at the
OHMSETT facility and an additional length of 90 meters could be
supplied immediately by the manufacturer Thus the required
length of 200-250 meters to realisticallr simulate a typical
deployment was made available at no cost to the program
Boca Test Report August 10 1988 Page 11
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
(2) The selected boom was expected to be capable of performing the
operations denoted within the Offshore Trial Test Protocol
Specificamplly it Yampll felt that the bean could be towed at 05 to l
kt (relative to the surface current) and was expected to retain a
major portion of the released oil volWlle (approrlmately 80 M3) at
the lover end of this speed range
(3) Booms of this design had undergone extensive testing in the
OHMSETT test tank amp11 described in Section 1 Thus comparisons
between at-sea and tank testing would be po1111ible following the
Offshore Trials
( 4) The Globe Oil Fence 48 contains rigid structural ribs upon which
the instrumentation and cables could be mounted
Above all based on previous OHMSETT testing it vas believed that the
boom would be suitable for oil containment under the planned operational conshy
ditions (relative speed of 05-1 kt during tests sea state of 2-3 and maxishy
1111111 wind velocity of 15 kt)
Following selection of the Globe Oil Fence 48 work began immediately on
preparing the boom for installation of the instrWllentation and cables Work
began by replacing broken ribs and noats and patching torn sections of the
skirt on the 152-meter length of boom available at the OBMSElr facility
Spare parts and hardware were provided by Globe International along with
BOOllll Test Report August 10 1988 Page 12
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
three additionampl 30-meter sections ot boom All sections of the boom (8 secshy
tions totamplling 244 meters) were present at the OHMSETT tacilit7 b7 Auguat 15
and all outfitting was completed b7 August 28 1987
BOOM INSTRUMENTATION AlID WIRING
Meaaurement of wave-following ability
To determine the wave-following ability of the boom ie to measure the
motion ot the boom relative to the waves pressure transducers with reasonable
accuracy and a relativel7 fast response time (compared to the height and
period ot the waves) were required Drawing upon experiences gained in preshy
Tious OHMSElI work capacitance-bridge transducers were selected tor this purshy
pose These bridge-type transducers proTide acceptable response time and acshy
curacy but require one Bide of the bridge to be maintained at atmospheric
pressure For transducers to be located on the boom skirt under water this
link would be provided by a capillary tube vented above the water level It a
short tube were used deployment of the boom and the action ot rough seas can
resuJt in water entering the capillary which (at best) provides an erroneous
reference level and (at worst) an inoperable transducer if water reaches the
sensing element Thus transducers were ordered with cables long enough to
contain an unbroken and intact capillary tube ertending from each transducer
to the tow boat Eight Druck Model PlX160D transducers two each with cable
lengths of 76 107 137 and 168 meters were ordered with direct HERMIT 2000
connectors and sealed end caps
Boom Test Report August 10 1988 Page 13
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
Eight transducers were mounted near the bottom ot the boOlll skirtbull at apshy
prozimately 30-meter intervals trcm the center Tvo transducers vere mounted
near the center ot the boom (iebull near the apex ot a catenarr formed by the
booll) to provide useful information from the location vbere oil is contained
andor lost A schematic diagram ot transducer locations as ther would apshy
pear operationally is shown in Exhibit 1 When the booa is deployed each
transducer is located at a nominal depth ot 057 meters belov a calm vater
surface Each halt ot the boom vas wired separately with cables trcm the tour
transducers on each side running back to separate recording systems on each
tov boat The split cabling resulted in reduced lengths ot cable runs
reduced bulk ot the wiring harness along the boom and maintenance ot mechanishy
cal syn111etry along the boan As noted previously each transducer ordered vas
specified with the required length ot cable hermeticallr sealed br the
manufacturer completely tram the transducer to the required connector tor the
recording system on each tov boat
To record the output of the pressure sensors in the desired digital forshy
mat Hermit 2000 (TM) data recorders were used Two ot these units were cusshy
tom modified by In-Situ Incbull and provided with the Druck pressure
transducers described above as a complete unit with all matching cables and
connector assemblies Each ot the two HERMIT recorders vas modified by Inshy
Situ to allow tour-channel recording at a constant interval ot 025 seconds
The internal memory in each HERMIT allowed storage ot tour channels of 16384
12-bit words Thus two units provided a fU1l hour ot recording capabilitr tor
eight channels with the desired resolution
Boom Test Report August 10 1988 Page 14
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
Mounting Of The Transducers on the Skirt
The body of each pressure transducer was attached to a boom structural
rib at the desired location near the bottom of the skirt using two clamps and
the bolts which originally attached the rib to the boom skirt Additional
pb7sical protection of the transducer and cable attachment point was provided
bT surrounding the transducer with a short length of heaTT-valled aluminUlll
conduit which was mounted to the structural rib using U-bolts Holes were
driiled in the conduit to allov free passage of vater to the sensing element
Tests were performed in the OHMSElT test tank to verity that the protective
casing had no observable effect on pressure-time measurementbull for vave motion
A close-up view of the pressure transducer mount with surrounding conduit and
attached cables is shovn in Exhibit 2
The cables from transducers on each half of the boom vere harnessed
together and the bundle of one to four cables (depending on position along the
boom) was covered vith split automotive heater hose wrapped vith plastic
electrical tape and secured with plastic wire ties A photograph of this
operation is given in Exhibit 3 The cabling vas attached to each structural
rib 122 meters along the boom allowing sufficient slack between attachment
points for stretching and folding of the boom The cable assembly vas atshy
talthed loosely along the tov line and the excess cable (approximately 30
meters) vas bundled into a tight coil The completed assembly afforded conshy
siderable protection for the cable runs both between transducers and from the
boom to the tow boats
Boom Test Report August 10 1988 Page 15
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
Although each cable connector vas supplied b) In-Situ Incbull with an oshy
rinamp-sealed protective cap additional protection YU provided by inserting
each connector in a plastic bottle using a tight-titting split rubber stopshy
per and completing the seal by covering with moldable 3M self-vulcanizing
rubber tape This seal vas shown to be leskproot at T-toot depth in the
OIDCSETT Test TampIllt tor a period of 12 hours
Boom Tension Instrumentation
A tov line ot 318-mm diameter twisted-braid Nylon 152li aeters long YU atshy
tached to each end of the boom through a Ii 5-m steel cable choker The tow
line ends terminated in shackles attached through strain links to the towing
bit of each boat The strain links were Metrox Model No TLlOl-lOK load cells
proTiding a full-scale tension range ot 10000 lb The load cells were orshy
dered with 915 meters of custom-sealed cable and connectors The 4-20 ma
output of each load cell was amplified by a Metrox Model 2060-00 Signal
ConditionerAmplifier which also provided power to the strain-gauge bridge
within each load cell
The output of one load cell was recorded directly on a strip chart reshy
corder with a 5-volt range The recorder response for a calibrated tension
input of 9500 lb was found to be 38 volts The other load cell was conshy
nected to a TELOG Model 2107 0-5 volt data logger operated at 5-second
t~ internals with 10-bit resolution and 1628-vord memoey Programming
was input to and data was dumped from the TELOG data using a Zenith Z181
Boom Test Report August 10 1988 Page 16
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
portable computer with noppy diskette storage
A photograph ot the load cell within the tow line is given in Exhibit 4
DYBAMIC OIL RECOVERY INSTRUMENTATION
To provide a quantitative measure ot the pertormance ot various skilllshy
aers in-line tlow measurements were made during recovery operations The
systems and instrumentation deployed tor these measurements were as tallows
o A custom-made venturi tube with 102-cm diameter inlet and outlet
tubes and a throat diameter ot 80-cm was used to measure now
trom the skimmers The pressure ditterential across the throat
was measured using a Rosemount Model 1151 Ditterential Pressure
transmitter providing a li-20 ma output over a preset pressure
range ot 0-254 mm ot water (tlow rate ot 0-45 m3hr) The output
ot the Rosemount unit was monitored by a Telog Model 2101 current
data logger operated at 5-second sampling intervals and conshy
trolled by an IBM PCrr microcomputer The reading on the
taceplate ot the Rosemount gauge also was monitored during the
tests A photograph ot the venturi arrangement is shown in Exshy
hibit 5
o A Rosemount Model 1151 transmitter with one side open to the at shy
mosphere was connected at the inlet end ot the venturi tube to
measure the discharge pressure ot the skimmer pump The output ot
Boom Test Report August 10 1988 Page 17
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
this Rosemount unit adjusted tor 50 psig tull-scale output was
monitored bJ a Telog Model 2101 0-5 volt data lauer The Telog
was programmed by an IBM PCrr microcomputer which served tor
storage ot data tiles on tloppy disk The reading on the
faceplate ot this Rosemount gauge also was monitored during the
tests
o Samples ot recovered product vere obtained periodically through a
stopcock located downstream tram the outlet end ot the venturi
tube Approximately 100-mL samples (in duplicate whenever tlov
rate permitted) vere collected in 125-mL polyethylene bottles tor
subsequent determination ot percent water in the product
STATIC RECOVERED-PRODUCT MEASIJR]MENTS
The depth ot product within the recovery tanks was measured using a 3shy
meter marked pole designed tor measuring the contents ot gasoline service stashy
tion tanks Samples ot the stratified product in the recovery tanks vere colshy
lected by custom-made Johnson samplers a 3-meter long chambered sampling
tube Chambers vere termed by o-rings at 152-cm intervals along the plunger
rod The volume ot each chamber approximately 30-mL was be separately colshy
lected as the plunger was removed The samples allowed a determination ot
percent water in each 15-cm layer in the recovery tanks
Boca Test Report August 10 1988 Page 18
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
PBDTOVIDEO INSTRUMEBTATION
Pretest test and post-test operations conducted bT OHMSETT personnel
were documented using still photographr and video tape The equipment used
during these exercises included Nikon F3 and FE cameras for color print and
slide photographr and two Sony Model CCD-VllO 8-111111 color cameras for video
coTerage
LABORATORY EQUIPMENT AND MATERIALS
The percent water in recovered product both for in-line samples and
saples of the stratified recovery tank contents was determined bT ASlM
Method No 1796 using a DamonIEC Model 1111-S centrifUge with 125-llL graduated
tubes Make-up solvent used for samples of insufficient volume was ACS
reagent-grade toluene Viscosity and specific gravity of initial and
recovered materials were determined using ASTM Methods 2983 (Brookfield Model
LVT Viscometer) and br ASTM Method Dl298 (glass hrdrometers) Viscosity
standards used during the laboratorr work obtained frC1111 Brookfield were 100
cps (lot No 111585) and 975 cps (lot No 100385) fluids
OTHER EQUIPMENT AND MATERIALS
The distances between tow boats and between each boat and various tarshy
gets on the boom were monitored during the tests using Ranging Inc Model
1200 and Model 620 optical rangefinders The rangefinders were calibrated and
expected uncertainties were determined using targets at fixed distances prior
Boa111 Test Report August 10 1988 Page 19
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
to the sea trials
The relative speed between each tow boat and the surface water vu
measured using floating wood chips Distances of 61 and 915 meters were
measured and marked along the rail of the tow boats bull Wooden blocks measuring
l25cm x 50cm x T6cm and painted nuorescent orange were dropped at one
mark and the time required to traverse the distance between marks was measured
b7 a stopwatch The distance traversed was divided by the time of traverse to
determine tow boat speed
Boom Test Report August 10 1988 Page 20
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
SECTION 5
EXPERIMENTAL PROCEDURES
BOOM IMMERSION MEASUREMENTS
Calibrations
Although each pressure transducer was receiTed with a recent
JDampDufacturers calibration certification (in the form of a scale factor to
be used in programming the recording system) bull static llnearitr calibration
checks at known depths were performed and dynamic records of wave spectra in
the OHMSETT test tank were obtained The calibration and response were
measured over a depth range of zero to 12 meters and for a representative
wave period of about 3 seconds Because the HERMIT 2000 data loggers are
continuous-recording devices with slow data dump capabilitr the calibrations
were performed at various depths of immersion while operating the recorders
continuously
The pressure transducer calibrations were checked in two groups or four
on separate days The transducers were clamped to a crosspiece attached to
a 3 meter long pipe with markings at 254 mm intervals The pipe was lowered
to immerse the sensors to a given depth in the test tank tor 15 seconds
providing 60 data points at the rate of 4 per second During this period the
Boom Test Report August 10 1988 Page 21
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
Hermit 2000 digitized and recorded the output of the transducers The aashy
seably vaa then lovered to the next designated leTel The procedure vas
repeated until the transmitters had traversed the operating range anticipated
ill the offshore trials In this calibration series meaaurements vere made at
15 depths betveen zero and 122 meters of immersion
The data stored in the Hermit vere dUIDped to an IlM PCAT computer using
the RS232C interface and the output vas vritten to a fioppy disk The data
in the tile vere imported to a Lotus 123 spreadsheet and edited to remove
transient data ie data points recorded at intermediate levels betveen the
designated depths An example of data records prior to the editing is shovn ill
Exhibit 6 The average datUID at each immersion level was calcuJated Because
the change in immersion is the parameter of interest the data vere tared to
the first average vhich forced the first value to be zero A least-squares
regression analysis vas made of additional measured ilDlllersion and the knovn
added immersion The regression results are shovn vith plots of the data for
the eight transducers in Exhibits 7 through 14
The calibrations vere performed in the open Test Tank The vind-induced
surface chop vas estimated at about 61D111 (14 inch) during the calibration
procedure The vavelet action undoubtedly affected the accuracy of the
knovn added immersion the effect being more pronounced at small depths
The vavelets also affected the precision of the shallov measurements defined
as the coefficient of variation However the regression analysis showed that
the transducer outputs are linear vith depth and that added immersion is
measured accurately by the transmitters The lowest value calculated for the
Boom Test Report August 10 1988 Page 22
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
coefficient of correlation is greater than 0 999 indicating that at vorst
approximately 005 of the scatter in the edited data is not explained b7 the
changing level of immersion
No measurable difference in the data was found during experiments in
which the sensors vere being raised or being lowered The data obtained vith
a given sensor while raising the sensor was equivalent to that obtained while
lovering the sensors
Tank Tests
Several tests vere performed using the wavemalting capabilities of the
OIDEETT test tank to evaluate the performance of the pressure transducers In
the first test series the dynamic response of the transducers as vaveshy
bulleasuring devices was documented A group of four transducers vere mounted on
the calibration rod which vas affixed to the moveable bridge spanning the
tank vith the depths set at a nominal value of 06 meters The test tank
vave generator vas operated at 20 RPM vith an ll4-cm stroke to generate waves
vith an amplitude of about 03 meters and a period of about 3 seconds After
about 6 minutes the bridge began moving forward into the waves at a speed of
15 bullmin (055 knots) A plot of approximately 4000 data points from one of
the transducers is given in Exhibit 16 over a time period including the time
prior to initiation of waves during waves vith the bridge motionless and
folloving the initiation of bridge motion The transducer data vere fastshy
Fourier transformed using the Basic program PROT05 ( see Appendix B) to obtain
the magnitudefrequency spectra for a stationary transducer (Exhibit 17) and
Booa Test Report August 10 1988 Page 23
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
tor a moving transducer (Exhibit 18) AlJ eight transducers produced verr
siailar result11 in tvo sets or such measurements
The single major peak in the magnitude spectrum or Exhibit 17 vith verr
little fine structure (harmonics) corresponds to the output expected for tank
vaves (ie monochromatic waves vith a fundamental frequenc7 or 033 Hz (20
rpm)) Because the transducers vere at rest the frequencies or principal and
renected waves were identical With the transducer in motion (Exhibit 18)
there is a positive Doppler shirt in the fundamental frequenc7 or the prinshy
cipal wave and a corresponding negative shirt for the reflected wave Beshy
cawse the end or the tampDk contains a simulated beach to dampen reflections
the reflected wave is much lover in magnitude than the principal vave For a
sensor (bridge) velocit7 or 15mmin (025 msec) and the observed vave
velocit7 or 523 msec the expected Doppler frequencies are
rdsfo(l +- 025523) bull 033(1 +- 0047)
td (forward) = 0346 Hz
rd (reflection) bull 0315 Hz
The Doppler frequencies correspond very closei to the frequency or the
peaks shown in Exhibit 18 ie 035 and 031 Hz
Arter completion or the tests described above one or the transducers
was encased in the aluminum conduit sheath that had been designed to provide
protection when mounted on the boom The test with waves and bridge motion
vas repeated There was no observable difference in the resulting wave or
Boom Test Report August 10 1988 Page 24
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
frequency spectrum occasioned by the presence or the protective sheath
Following installation or tour transducers on each 122-meter length or
boom the units vere placed in the test tank and arrayed linearly along the
long (North-South) dimension The booms vere tied to a moveable bridge at the
south end or the tank and the other ends vere fastened vith lengths or tov
line through strain gauge load links at the north end or the tank A
photograph or the arrangement is shown in Exhibit 19 The moveable bridge
was used to apply several hundred pounds or tension to the boom waves vere
generated and the vave spectra tor the tour transducers vere recorded A
representative vave record and corresponding magnitude spectrum are shown in
Exhibits 20 and 21
Several characteristics or the spectra produced by the boom motions in
the test tank are of interest First there is a decided tine structure
present as noted in the magnitude-frequency spectrum shown in Exhibit 21
vhen compared to the spectrum obtained tor waves using fixed transducers
(Exhibit 17) At least four harmonics vere noted in each or the magnitude
spectra indicating that boom motions vere somewhat more complex than simple
wave motions However the major component or energy (which is proportional
to the area under the magnitude-frequency curves) remains at the fundamental
frequency or the waves (the measured wave frequency tor this test was 0303
Hz)
An additional point to be made regarding the spectra is that there are
components at very lov frequencies on the order or 001-004 Hz This feature
Boom Test Report August 10 1988 Page 25
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
m81 relate to the resonant frequency of the tank The tank is 203 meters in
length and 2 4 meters deep and a seiche wave would have a fundamental
frequency of about 0012 Hz The low-frequency component mq also be due to a
vibrational mode of the stretched boom This explanation is supported by the
observation that the low-frequency modes are more prominent in the boom
spectra than in the fixed-sensor spectra In either case the amount of
energy represented by frequencies other than the fundamental is very small
A final series of tank tests was performed prior to removal of the boom
sections from the tank for packing and transport to St Johns Newfoundland
In this sequence a test was performed over a period of time approximating the
duration of testing planned during the offshore tests (one hour)
Approximately 14000 data points were recorded over 555 minutes for
each of four sensors on the East boom section which was placed under 3100
Newtons (700 pounds) of tension These data were analyzed separately for each
of 13 segments of 1024 data points
Practice Run
The instrumented boom was deployed during the practice run conducted off
the coast of Newfoundland near Torb~ Point on September 21 1987 Data were
recorded for 55 minutes on all eight transducers during the practice run The
bocm was maintained in a catenary configuration that was roughly symmetrical
although the gap of the catenary varied considerably The opening of the boom
varied between 50 and 180 meters because of the difficulty the tow boats en-
Boom Test Report August 10 1988 Page 26
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
countered in maintaining course and spacing For most of the test duration
the gap was roughly twice the desired distance of 811 +- 15 meters and the
booa gap ratio was much greater than desired In addition the speed of the
boca through the water generally appeared to be too great for retention of
oil
Offshore Test
Following repair of the cables on one half of the boOJll which had been
seTered by tow boat propellers during the practice run the boom vas deployed
as part of the actual offshore oil-release exercise on September 24 1987
Forty-six minutes of data were recorded during the offshore test Two of the
transducers on which repairs had been attempted became inoperable throughout
and provided no useable results Of the six remaining sensors four provided
a full 46 minutes of data transducers l and 3 on the Port side and
tranaducers 3 and 4 on the starboard side The remaining two transducers on
the port side recorded useful data only for about 5 minutes after which a
twist in the boOJll placed both sensors out of the water for the remainder of
the test
BOOM TENSION MEASUREMENTS
Calibration
The Metrox Model TL-lOlOK strain links were received from the manufacshy
turer with calibration data which was used directly in establishing output
Boca Test Report August 10 1988 Page 27
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
rampD8eS of the instrumentation Calibration data are as folloVll
o SN 2668 (Tow boat CG212)
sensitivity bull 1744 mvv shunt calibration
bull 42072 Newtons (9458 lb)
o SN 2667 (Tow boat CG2o6214)
sensitivity bull 1744 mvv shunt calibration
bull 42393 Newtons (9530 lb)
The gain I of each amplifier vu adjusted to provide an output corshy
responding to 11128 Newtonsvolt (2500 lbvolt) by use of the shunt
calibration function and a digital voltmeter Thus a full-scale tension of
li4li84 Newtons produced an output signal which was BOS of the 5-volt range
used for the strip chart recorder and the TELOG data logger
Practice Run
No records of tension on the boom were obtained from the load cell conshy
nected to the strip chart recorder during the practice run Apparently the
strain gauge amplifier was overloaded by radio-frequency interference from the
boats generator resulting in full scale deflection of the recorder at all
times
Records obtained from the Telog recorder show that the measured tension
oscillated considerably even over the 5-second sampling interval that was
Boca Test Report August 10 1988 Page 28
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
used A plot of the data averaged over 25-second intervals is shown in Exshy
hibit 22 Throughout the practice run the tension vas never within the range
expected for a towed boom of this size and length A static tension level on
tbe order of 9000-11000 Newtons was expected whereas the data shoYll a varishy
ab1e tension between 900 and 4500 Newtons
Offshore Tests
Essentially no data were obtained from the load links on either side of
the boom during the offshore tests It is possible that the twists in the tow
and shackle arrangement that occurred during the tests may have loosened the
cab1e connection at the strain link During the test it was observed that
there vas o~en slack in the tow lines as the tow boats Jockeyed to maintain
proper direction and speed
FLOW AND RECOVERY MEASUREMENTS
Calibration
The 4-inch Venturi tube and associated Rosemount pressure gauges were
calibrated at OHMSETT prior to shipment to St Johns The Rosemount transshy
mitters were calibrated by adjusting the full-scale output to correspond to an
applied pressure head by means of a water-filled manometer tube The faceshy
p1ate readings at several values of applied pressure were obtained to
degtOnstrate that the expected square-root relationship between readings and
Boan Test Report August 10 1988 Page 29
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
applied pressure was valid see Exhibit 23 It was determined that the
Rosemount signal was proportional to the square root or the differential presshy
sure and therefore should be directly proportional to flow through the Venshy
turi tube
The linear relation between flow and Rosemount gauge readings was tested
by pumping salt water through the Venturi at various pumping rates and measurshy
ing the output or the pressure transmitter The results using the race-plate
scale or the Rosemount are summarized in Exhibit 24 for two sets of tests at a
tu1J-scale output of 050 meters of water
The Telog data recorder was used during the calibration series to
evaluate this selection for recording the Venturi data The Telog unit colshy
lected data over a programmable time interval providing the maximum and minishy
lllUlll values or data collected each second and the average over the interval
and was programmed to provide tabular flow rate data directly in gallons per
minute An example or the output is provided in Exhibit 25 exactly as the
data were output from the microcomputer controller The right side of the
data display in the Exhibit 25 provides a rough plot of the data
The Rosemount transmitter used to measure the pressure of the skimmer
pump discharge line was calibrated directly in pressure units so that a full
scale reading corresponded to 345 x 105 Nevtonsm2 bull
The depth of fluid in the recovery tanks was measured by a 3 meter (10shy
ft) pole with 64 lDID markings No attempt was made to calibrate either the
Boom Test Report August 10 1988 Page 30
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
measuring device or the volume-depth relationship of the recovery tanks
Rather the geometry of the cylindrical tanks (6li meters in length and 213
meters in diameter) was used to calculate the volume using the following
relation
Volumebull LR2 ((Q573) - sin Q cos Q
Cos Q bull 1 - (HR)
where H is the height of liquid in the tank of length L and radius R and the
angle Q is computed in degrees
Practice Run
No measurements of now rate or recovered product were made during the
practice run on September 21 1987
Offshore Tests
Flow and total recovery measurements of oil collected by various skimshy
mers were made during the offshore tests No nov discharge pressure or
tank accumulation was observed during attempts to operate the Heavy Oil Skimshy
mer
During the 23-minute period of operation of the Framo skimmer the venshy
turi flow rates were recorded by the Telog data recorders using 5-second
Boom Test Report August 10 1988 Page 31
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
averaging the discharge pressure was recorded manually at approximately 10shy
minute intervals and the depth of recovered product in the storage tank was
measured at approximately 5-minute intervals
Similar measurements were made during the 29 minutes that a Terling
GT185 skimmer was operated
Boom Test Report August 10 1988 Page 32
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
LABORATORY MEASUIUMEBTS
Duplicate in-line samples were collected eveey 5-10 minutes during
operation of the Framo skimmer and single samples were collected over apshy
proximately 10-minute intervals during operation of the Terling GT185 skimmer
These samples were subsequently analyzed by ASTM methods to determine
the percent of water in the recovered materials
OTHER MEASUREMENTS
Tow Speed was measured using wood chips and distance between tow boats
and to various points on the boom were measured by optical rangefinders during
the offshore tests The tow speed was measured by timing a given distance
traveled by floating wood chips dropped over the side of each tow boat Fixed
distances were marked along the rail of each tow boat To determine the
precision and accuracy of this method several series of tests were performed
using the technique on the moving bridge of the OBMSETl test tank These
results are shoved that a set of measurements would be expected to obtain
values within +- 20J of the mean and the true speed
The Rangefinders used to determine distance were calibrated by adjustshy
ment to provide the best image at a known distance of 100 meters Several
sets of measurements were made using the rangefinders to determine marked disshy
tances from 10 meters to 200 meters It was determined that an accuracy or a
precision of +- 20J could be expected
Boam Test Report August 10 bull 1988 Page 33
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
SECTION 6
RESULTS AND DISCUSSION
BOOM MOTION MEASUREMENTS
Practice Run
Typical boom wave records over short and longer periods are given in Elcshy
hibits 26 and 27 The increased complexity ot the boom motions on the ocean
surtace compared to that obtained tor monochromatic waves in the test tank
is noted
The 55 minutes of data allowed 13 segments each containing 1048 data
points to be analyzed tor each ot the eight sensors A typical magnitudeshy
frequency spectrum obtained by proce1111ing the boom wave data is given in Exshy
hibit 28 It is noted that most of the energy from the boom motion is conshy
tained within a band of frequencies from Ol to 07 Hz Although there is
some activity over higher frequencies (ie above l Hz) the magnitude is at
or near the noise level and individual harmonics are less than 001 meters
Some of the spectra exhibited significant very-low-frequency activity in the
range of zero to Ol Hz vhich may be related to the erratic variations in tov
speed and catenary conformation that occurred during the test
Boom Test Report August 10 1988 Page 34
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
The spectra from the various transducers along the boom were remarkabiy
siailar for ampDY given time segment exhibiting similar variations to those obshy
serTed between different time segments at the same transducer The one-third
significant wave height calculated as 283 times the square root of the area
under the magnitude spectrum was found to be within the range of 006 to 009
meters at all boom locations These results indicate that the motion of the
boom is sufficiently similar throughout its length and over time that
statistical an~sis is possible
Offshore Test
A typical boom wave record obtained during the test is presented in lJshy
hibit 29 The magnitude-frequency spectrum for the same 1048 data point segshy
ment is given in Exhibit 30
The spectra obtained during the offshore tests are similar basical1Y
to those obtained during the practice run Most of the energy is contained in
the range of zero to 10 Hz with considerable noise but no discrete sigshy
nificant peaks at higher frequencies There is considerable energy in the
verr low frequency band
The magnitude of the boom spectrum at any given frequency is conshy
siderablY greater than that obtained during the practice run as would be exshy
pected for the greater sea state that was present There was a decrease in
magnitude at the significant frequencies (0-1 Hz) as the test progressed inshy
dicating a somewhat diminishing swell during the time when data vere being re-
Boom Test Report August 10 1988 Page 35
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
corded The one-third significant wave height throughout the spectra are conshy
sistent~ between 014 and 017 feet A table of B (13) obtained during both
the practice run and the offshore tests is given in Exhibit 31
OIL RECOVERY MEASUREMERTS
Offshore Test
Recovery information for the Framo skimmer is summarized in Exhibit 32
As evident from the plot the product recove-q rate of the Framo skimmer was
approxllllate~ 075 mmin (200 gpm) over a significant portion of the time it
vas deployed The results obtained by measurement of the tank volumes (see
Exhibit 33) are in close agreement to those obtained from the venturi now
rates and indicate an average overall now rate of o72 mmin (192 gallons
per minute) The total height of product in the tank was 1079 meters (425
inches) corresponding to 1158 cubic meters (3070 gallons) The height of
remaining product following draining of free-standing water was 0705
meters corresponding to 660 cubic meters (1750 gallons) Of the product
remaining in the tank the subsequent an~sis of stratified samples (see Secshy
tion D) shoved that about 40 was emulsified water Therefore the Framo
skimmer was found to recover 377 +- 075 m3 (1000 +- 200 gallons) of oil at
an overall rate of 017 m3 (45 gallons) of oil per minute and a recovery efshy
ficiency of 32
A summary of the now data obtained for the GT185 skimmer is given in
Boom Test Report August 10 1 1988 Page 36
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
Exhibit 34 The Venturi data indicated an average nov rate of near 038
bull3ain (100 gpn) after an initial startup period and the tank sounding data
(Exhibit 35) shoved an overall average recovery rate of about 032 m3min (85
gpa) The final depth of product in the tank was 0914 meters corresponding
to a volume of 935 cubic meters (2480 gallons) Bo tree-standing water was
observed in the recovered material but subsequent anamplTsis shoved that the
product contained an average of 54 of emulsified water Thus the GT185
sklmmer was found to recover 415 +- 038 cubic meters (1100 +- 100 gallons)
of oil at an overall rate of 0 14 m3 ( 38 gallons) of oil per minute and a
recovery efficien07 of about 44
LABORATORY MEASUREMEllTS
The results obtained by analysis of in-line samples for percent water
are given in Exhibit 36 There was considerable difficultr in obtaining these
samples and because of the large volume of discharge lines an uncertainty
regarding which skinaner produced the sampled product at any given time The
data do not provide conclusive evidence of skinaner performance The data from
samples of the Framo skimmer (which are less uncertain) generally agree with
the conclusion that approximately 34 of the product was oil
Stratified samples of the recovered product in the storage tanks were
also collected following stripping of free-standing water
The samples collected within depth intervals in the recovery tanks also
were analyzed The data obtained from analysis of these samples from the
Boom Test Report August 10 1988 Page 37
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
Premo skimmer and the GT185 skimmer are provided in Exhibit 31 It is noted
that the samples tra11 the GT185 contain lligniticantlr 110re emulsified water
(averaging 55) than the samples trom the Framo sk1-er (averaging 31)
Also all ot the water found in the product recovered by the GT185 vaa emulshy
sified water
The data tor the GT185 skimmer shovs the correct nUJ11ber (6) ot six-inch
saaples to correspond to the observed 36-inch depth ot fiuid in the tank
However the number ot six-inch samples (5) trom the Framo skimmer does not
correlate with the measured depth ot fluid in the tank 275 inches Also
the volumes found in the individual compartments varied considerablr trom the
expected 30 mL It is possible that (1) the sampler leaked between compartshy
ments prior to sample collection1 (2) the samples were collected at a slant
rather than vertically a likely occurrence since ther were taken atop the
tank on a rolling pitching deck (3) exact separations between compartments
were not obtained during the sample collection andor (4) the sampler was
closed at the deep end ot the tank during a pitch 110tion or the ship Each ot
these factors will attect the results derived from the laboratorr data
generallr by lovering the calculated recoverr etticiencr ot the skimmer
OTHER MEASUREMENTS
Practice Run
The sparse data obtained tor compass reading (heading) distances beshy
tween boats and to various points on the boom and through-the-water speed
Boom Test Report August 10 1988 Page 38
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
during the practice run are provided in Exhibit 38 Because of the difficulty
encountered in maintaining orientation and tov speed most aeasure11111nts were
variable and erratic Activities neceBSampry to maintain a proper boom conshy
figuration took precedence over the distance and speed data collection efshy
forts
The small amount ot speed and orientation data obtained during the offshy
shore test is provided in Exhibit 39 Because major activity focused on keepshy
ing the tov lines clear of the boat props and maintaining observation to
detect boom twist and roll only a limited number of measurements were made
Attempts to measure boat heading were forsaken entirel) becamp1111e the boom
direction changed often and rapidly as needed to maintain the boom location
vitb respect to the oil slick These changes in boat conformation made voodshy
chip measurements erratic and most attempts resulted in obtaining no data
So- of the reportable results indicated that the boom was being towed within
the range of 05 to l knot However most observations denoted more rapid
movement of the boom
Boom Test Report August 10 1988 Page 39
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
REFERE1CES
l Smith GF and Lichte HW Summary ot US Environmental Protection
Agencys OlM3Ell Testing 1974 - 1979 EPA-6o09-8l-007 us Environshy
mental Protection Agency Cincinnati Ohio 1981 341 pp
2 Borst MJ and Nash JH Oil Containment Boom Test Protocol (In
preparation) US Environmental Protection Agency Cincinnati Ohio
1988
Boca Test Report August 10 1988 Page 40
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41
GLOSSARY
BOCIC PERJORMANCB - The ability of an oil containment device to contain an
oil slide during deployment
RECOVERY BFFICIENCI - The volume of oil recovered by a skimmer device
divided by the total volume of recovered product
within a given time interval
Boom Test Report August 10 1988 Page 41