., ACCELERATED DISTRIBUTION DEMONST$A.TION SYSTEM REGULA . Y INFORMATION DISTRIBUTIO. SYSTEM (RIDS) , ~ '7 ACCESSION NBR:9203230281 DOC.DATE: 92/03/13 NOTARIZED: NO DOCKET FACIL:50-387 Susquehanna Steam Electric Station, Unit 1, Pennsylva 05000387 50-388 Susquehanna Steam Electric Station, Unit 2, Pennsylva 05000388 AUTH. NAME AUTHOR AFFILIATION KEISERgH.W. Pennsylvania Power & Light Co. RECIP.NAME REC1PIENT AFFILIATION MILLER.C.L. Project Directorate I-2 R SUBJECT: Forwards util revised response to Station Blackout Rule per NRC 920114 Safety Evaluation w/answers attached to all but I one NRC recommendation. Query on CR instrument cabinet temp to be answered no later than 920501. D, DISTRIBUTION CODE: A050D COPIES RECEIVED:LTR ENCL j SIZE: g5 f ~ f $ TITLE: OR Submittal: Station Blackout (USI A-44) 10CFR50.63, MPA A-22 / 05000387 05000388 A D RECIPIENT ID CODE/NAME PD1-2 PD INTERNAL: ACRS NRR PD2-4PM TAM NRR/DST/SELB NRR/DST/SRXB8E EXTERNAL: NRC PDR NOTES: COPIES LTTR ENCL 1 1 1 1 1 1 3 3 1 1 1 1 2 2 RECIPIENT ID CODE/NAME RALEIGHiJ. AEOD/DSP/TPAB NRR/DET/ESGB 8D NRR DST/ PLB8D1 G FILE 01 NSIC COPIES LTTR ENCL 1 1 1 1 2 2 3 3 1 1 1 1 D D NOTE TO ALL "RIDS" RECIPIENTS: PLEASE HELP US TO REDUCE WAS'ONTACTTHE DOCUMENT CONTROL DESK. ROOM Pl-37 (EXT. 20079) TO ELIMINATE YOUR NAME FROM DISIRIBUTION LINIS FOR DOCUMENTS YOU DON'T NEED! A D D TOTAL NUMBER OF COPIES REQUIRED: LTTR 19 ENCL 19
Transcript
., ACCELERATED DISTRIBUTION DEMONST$A.TION SYSTEM
REGULA . Y INFORMATION DISTRIBUTIO. SYSTEM (RIDS), ~
'7
ACCESSION NBR:9203230281 DOC.DATE: 92/03/13 NOTARIZED: NO DOCKETFACIL:50-387 Susquehanna Steam Electric Station, Unit 1, Pennsylva 05000387
50-388 Susquehanna Steam Electric Station, Unit 2, Pennsylva 05000388AUTH.NAME AUTHOR AFFILIATION
KEISERgH.W. Pennsylvania Power & Light Co.RECIP.NAME REC1PIENT AFFILIATION
MILLER.C.L. Project Directorate I-2R
SUBJECT: Forwards util revised response to Station Blackout Rule perNRC 920114 Safety Evaluation w/answers attached to all but Ione NRC recommendation. Query on CR instrument cabinet tempto be answered no later than 920501. D,
DISTRIBUTION CODE: A050D COPIES RECEIVED:LTR ENCL j SIZE: g5 f ~ f $TITLE: OR Submittal: Station Blackout (USI A-44) 10CFR50.63, MPA A-22
PLEASE HELP US TO REDUCE WAS'ONTACTTHE DOCUMENT CONTROL DESK.ROOM Pl-37 (EXT. 20079) TO ELIMINATEYOUR NAME FROM DISIRIBUTIONLINIS FOR DOCUMENTS YOU DON'T NEED!
A
D
D
TOTAL NUMBER OF COPIES REQUIRED: LTTR 19 ENCL 19
Pennsylvania Power 8 Light CompanyTwo North Ninth Street ~ Allentown, PA 18101-1179 ~ 215/774-5151
Harold W. KeiserSenior Vice President-Nuclear215/774<194
NR l 3 1992
Director of Nuclear Reactor RegulationAttention: Mr. C.L. Miller, Project DirectorProject Directorate I-2Division of Reactor ProjectsU.S. Nuclear Regulatory CommissionWashington, D.C. 20555
SUSQUEHANNA STEAM ELECTRIC STATIONRESPONSE TO STATION BLACKOUTSAFf"TYEVALUATIONPLA-3745 FILE R41-2
Reference: RESPONSE TO THE STATIONBLACKOUTRULE FOR SUSQUEHANNA STEAMELECTRIC STATION, UNIT1 AND2 PAC NOS. M68613 ANDM68614) DatedJanuary 14, 1992.
Dear Mr. Miller:
This letter provides the Pennsylvania Power &Light Company (PP&L) revised response to theStation Blackout (SBO) Rule as required by the referenced NRC Safety Evaluation.
This response (attached) revises diesel generator target reliability to 0.975 based on yourposition, and provides the requested justification to support PP&L's original position that SSESis only required to cope with a SBO event for 4 hours. However, it should be noted that athorough evaluation was undertaken to review the staff s concerns regarding the need and abilityfor SSES to cope for 8 hours. Results of this evaluation concluded SSES has the capability tocope for 8 hours and longer ifrequired.
With the exception of a final technical resolution to your question regarding Control Roominstrument cabinet temperatures, the attachment responds in full to each of yourrecommendations. Our resolution to the cabinet temperature concern willbe forwarded to youno later than May 1, 1992.
9203230281 920313PDR ADOCK 05000387P PDR
-2- FILE R41-2 PLA-3745Mr. C. L. Miller
Questions regarding this revised response should be directed to Mr. A.K. Maron at(215) 774-7852.
Very truly yours,
H. W. Keiser
Attachment
cc: NRC3)ocnment:Control DeaR (original)NRC Region IMr. G. S. Barber, NRC Sr. Resident Inspector - SSESMr. J. J. Raleigh, NRC Project Manager - Rockville
~ ~ q ~
,9203230281 ATTACHMENTTO PLA-3745
I~NTR DUCTI N
The Station Blackout Rule (10 CFR 50.63) was instituted in 1988 and required licensees toassess their ability to cope with a station blackout (SBO) of a specified duration. In 1989, PP&Lsubmitted the results of our coping study to the NRC, concluding that Susquehanna SES (SSES)must be able to cope with a station blackout for 4 hours and maintain an Emergency DieselGenerator (EDG) reliability of 0.975 (97.5%). In February of 1991, PP&L revised its EDGtarget reliability value from 0.975 to 0.95 based on a spray pond bypass valve modification.On January 14, 1992, NRC issued its Safety Evaluation of the SSES SBO submittal concludingthat SSES was an 8 hour coping plant requiring EDG reliability be maintained at 0.975. Thefollowing is an item by item response to the recommendations identified in the NRC SafetyEvaluation.
c-".STATION,::::>SL'ATCKOUT:;,::::DUR'ATION>"''.lt NR RK MMENDATION: The licensee needs to change the EDG reliability targetfrom 0.95 to 0.975 and the coping duration from 4 hoursto 8 hours.
P~PRL R
A) Coping Duration
One input to the determination of required SBO coping duration is the "return time" ofextremely high winds(>125 mph). As part of our original coping assessment, PP&Lcontracted with Dames & Moore Consulting Engineers for the calculation of this "returntime" for SSES. Dames &Moore determined this value to be -6.7E-4/yr. (about once in1500 years) using data specific to SSES. Any return time value less than 1.OE-3/yr,coupled with our severe weather and off-site power design classification, places SSES in a4 hour coping category.
The NRC evaluation did not credit use of site specific data due to this data being applicablefor winds at 10 meters off the ground, rather than the required assessment height of 30meters from the ground (average transmission tower height). It was therefore concluded,based on NUMARC Table 3.2, that the return time for SSES was more frequent than onceper 1000 years and that SSES must cope with a SBO for 8 hours,
Page 1
~ ) ~ ~
(
ATTACHMENTTO PLA-3745
To address this coping duration concern, PP&L investigated the basis of Table 3.2 inNUMARC 87-00 and contracted again with Dames &Moore to determine the return timeof wind speeds at 30 meters. Conversations with both NUMARCpersonnel and NRC staffindicated that the use of site specific data is acceptable. The NRC cautioned that the useof such data should account for wind speeds of 125 mph at 30 meters and consider NationalBureau of Standards (NBS) publications 118 and 124, as well as several National Oceanicand Atmospheric Administration (NOAA)documents. Note that the use of site specific datais encouraged in NUMARC 87-00.
NBS 118 provides a method of scaling wind speeds to various heights and providesmeasured weather data from 129 meteorological stations across the US mainland. It is thisdata which PP&L and Dames & Moore believe provides the best estimates of wind speedreturn times at SSES. Using the method of NBS 118, the 125 mph "fastest mile" windspeed at 30 meters is scaled to a "fastest mile" wind speed of 107 mph at 10 meters (thenormalized height of all reported weather data). Using the data for meteorological stationsclosest to SSES, NBS 118 provides the following "return times" for various fastest milespeeds:
Fastest Mile Wind Speed (mph)Return Time
ears
1,000
5,000
10,000
50,000
100,000
500,000
1,000,000
Scranton
60.86
67.34
70.12
76.58
79.36
85.82
88.60
95.06
97.84
Harrisburg
70.57
80.49
84.75
94.64
98.90
108.79
113.05
122.95
127.21
Page 2
ATTACHMENTTO PLA-3745
In addition, Dames &Moore have calculated the probability of exceeding various wind speedswithin 1000 years, also based on the data and methods in a paper by H.C.S.Thorn:
Probability ofExceedance in
1000 rs Scranton Harrisburg
Fastest MileWind Speed (mph)
0.500
0.250
0.100
0.050
0.005
72
75
79
82
92
87
92
99
103
117
From the first table above, one can see that the return time of a wind speed of 107 mph at 10meters is expected to be greater than 1 million years at Scranton and almost 50,000 years atHarrisburg. Table 2 shows'that the probability of exceeding the 107 mph wind speed within1000 years is less than 1% at Scranton and about 3% at Harrisburg. Using the data fromHarrisburg in Table 1, the expected return time of a 125 mph wind at 30 meters is -37,500years. PP&L also-reviewed NBS, 124 for applicability. NBS 124 relies on the extrapolation ofcoastal weather data to infer wind speeds inland. Further, this method of extrapolation assumesintervening terrain to be open and grass covered. Since SSES is located within a valleyseparated from the coast by approximately 100 miles of hills and forest, the extrapolation ishighly inaccurate. Thus, PP&L views NBS 124 as valid only for scoping calculations andshould only be used in the absence of better techniques/data.
PP&L considers the preceding arguments and data sufficient justification for not using Table 3.2ofNUMARC 87-00 for determining our ESW category. Further, this data shows that the returntime of winds in excess of 125 mph at SSES is highly likely to be greater than 1000 years.Thus, it is concluded that the ESW category of "2" originally reported in our coping study isfullyjustified (the data actually justifies an ESW classification of "1"), and that SSES remainsa "Pl" plant (per NUMARC 87-00) requiring a SBO coping time of 4 hours.
B) EDG Target Reliability
In 1991, PP&L informed the NRC that for purposes of complying with the SBO rule ourtarget EDG reliability was to be 0.95 (95%). In making this determination, PP&L reliedon the use of "staggered operation" of RHR pumps to cool both suppression pools.Staggered operation is required because, although in principle any two EDG s can cool bothunits, in actuality there are two combinations ofEDG's (A and C, or B and D) which resultin only one RHR pump in each unit available to alternately cool the suppression pools.
Page 3
i
eATTACHMENTTO PLA-3745
The NRC noted that the use of staggered operation did not meet the "connectabilitycriterion" and was determined to be an unacceptable increase in operator burden. Thiscriterion was explained in documentation provided by the NRC to NUMARCafter submittalof the SSES SBO analysis. The NRC concluded that to avoid use of staggered operation,3 of the 4 EDG's would be required.
Further, the NRC noted that ifonly diesels A and B start, no control structure HVACwould be available. PP&L has performed a calculation of steady state control roomtemperature using the method in NUMARC 87-00 and assuming that the measured, normalcontrol room heat load exists. The result of this calculation is that the control roomtemperature will not rise above 111'F in the absence of normal HVAC. Becausetemperature remains less than 120'F, the control structure environment remains acceptable.
Based on the inability to take credit for staggered operation, PP&L concurs with the staff'sposition in"requiring 3 of 4 EDGs and the reliability target value of 0.975.
NRC made the following four recommendations based on their previous determination thatPP&L had to address the need for SSES to cope with an 8 hour Station Blackout.
1) The licensee needs to conform to an 8 hour coping duration and increase the EDGreliability target from 0.95 to 0.975.
2) The licensee should provide a procedure to refill the CST from the RWST duringan SBO event.
3) The licensee should add the portable AC generator to the list of SBO equipment,provide procedures for its utilization, and apply to it an appropriate QA program.The portable ac generator should meet the criteria in Appendix B of NUMARC87-00. Also the licensee should replace battery 1D650 with a higher capacitybattery or provide charging capability to the existing battery to extend its supportfor the 8 hour SBO duration, and recovery thereafter. The licensee shouldinclude all the analyses and related information in supporting documentation thatis to be maintained by the licensee for possible staff review.
4) The licensee should provide for staff review a full description, including thenature and objectives of any modification required. The analyses and relatedinformation should also be included in the supporting documentation that is to bemaintained by the licensee in support of the SBO submittals.
Page 4
Sli
pl
(i
i0
~ R
P~PLR N R
'2 R
ATTACHMENTTO PLA-3745
As addressed in the initial section of this response, PP&L concludes that SSES must cope witha SBO event for 4 hours. This conclusion is supported by. the use of site specific weather data(at the required assessment height). As for the EDG reliability target value, PP&L has reviewedthe NRC concerns and has concurred with the staff's finding that the configuration of SSESmandates an EDG reliability target value of 0.975. This reliability value has been included inthe EDG Reliability Program developed in accordance with NUMARC 87-00 Appendix D.
PP&L has thoroughly evaluated the ability of SSES to cope 8 hours with an SBO event,including all areas of concern identified in the NRC Safety Evaluation. PP&L is confident thatSSES has the ability to cope for 8 hours and longer ifrequired. Since PP&L has demonstratedthat SSES is a 4 hour coping plant this information will not be provided in support of ourrevised submittal, but is available for review.
NR REC MMENDATI The licensee should: I) provide additional informationand/or technical justification for the initial conditions andassumptions used in the heat-up analysis for each area ofconcern, 2) with regard to COTTAP computer code,provide detailed information to address the staff's concernsas identified above, and 3) re-perform the heat-up analysisfor each area of concern and for an 8 hour coping durationtaking into account the non-conservatism as identified in theSAIC TER.
P~PRL R N R — CCPPAP2 C 1
The use of the Compartment Temperature Transient Analysis Program (COITAP) computercode has been presented to the staff as part of our submittals to resolve steam leak detectionTechnical Specification changes. Attachment A contains a user's manual for the COTTAPcomputer code and a copy of a recent paper published in Nuclear Technology which describesthe methodology used in the COTI'AP program and presents some of the verification calculationswhich have been performed. The user's manual presents some of the calculations which wereperformed against problems that have exact analytical solutions. The referred paper presentsthe methodology along with calculations which have been benchmarked against calculationsperformed with the CONTAIN computer program. In addition, the program and computationpackage have been independently reviewed by Gilbert Associates. PP&L also maintains aQuality Assurance file/package for the COTTAP computer code.
A
Page 5
~t of
ATTACHMPITTO PLA-3745
In the original coping assessment, two basic COTTAP2 calculations were performed: anassessment ofDominant Areas of Concern (DACs); and an evaluation of control room cabinets.For BWRs, the DACs are the HPCI and RCIC rooms, and the main steam tunnel (NUMARC87-00). The main steam tunnel is considered because, apparently at some plants, HPCI andRCIC are isolated on high temperature in the tunnel. At SSES, the HPCVRCIC isolations donot come from main steam tunnel temperature but from sensors located on the 683 foot elevationof the reactor building common to both HPCI and RCIC piping. During SBO, only the RCICisolation logic is powered. Thus, for SSES, the main steam tunnel is not a true DAC. Thecommon piping area, called the RHR piping area in the calculation, is a DAC.
PP&L recalculated the DAC temperatures using CO1TAP2 and "conservative" inputs. Inputsincluded use of "maximum normal" room temperatures per the FSAR. Outside air temperaturewas assumed to be a constant 95'F. The influence. of hot piping (including flued heads) wasadded to the HPCI, RCIC, RHR piping area, and the main steam tunnel. (The absence of thishot pipe loading caused the cooldown of the main steam tunnel noted in the SAIC TechnicalEvaluation Report). No engineering reference for a con'crete thermal 'conductivity of 0.7 couldbe found. However, this value was changed from 1.0 to 0.7, per the TER. The actual inputdeck, and the justification for all input values used, appears in the detailed calculation.
The results of the COTTAP2 calculations are presented in the tables below.
Original Submittal:
Temperature ('F)
New Calculation:ROOM
8 hours 72 hours 8 hours 72 hours
HPCI
RCIC
RHR Piping
MS Tunnel
113
118
123
114
117
117
114
107
125
150
119
130
171
From Table 3, the temperatures of the DACs remain less than the 180'F operability limit, evenat 72 hours. The inclusion of the hot pipe. loads does cause, significant increases in tunneltemperatures.
Page 6
0ATTACHMENTTO PLA-3745
Temperature ('F)
ROOM
RHR Piping
MS Tunnel
COTTAP2 at 72 hours
130
171
NUMARC 87-00
176
Table 4 presents a comparison of the two hottest DAC temperatures as calculated by bothCOTTAP2 and the method of NUMARC 87-00. While it appears that the NUMARC methodproduces "conservative" results, it must be noted that the NUMARC calculation produces asteady state, infinite time result. The COTTAP2 results are not steady state but time dependent,and at 72 hours the temperatures in these rooms are still increasing. At longer and longer times,one would expect better agreement between the two methods. The results of the above tableshow that the agreement between the two methods is quite good.
The TER made reference to "oscillatory" temperature profiles. Review of the originalCOTI'AP2 work revealed no such profiles. The reviewers may be referring to temperatureprofiles which peak and drop in the short term, then continue a long term temperature rise(Figure 1). The large early peak is caused by AC motor heat loads which decay away. At latertimes, the room is heated by surrounding walls. This result is consistent with expected behavior.
The reviewers questioned PP&L's use of COTTAP2 for calculation of instrument cabinettemperatures and several assumptions used in these calculations. The original impetus for usingCORI'AP2 to calculate cabinet temperatures was the desire to avoid opening control structurecabinet doors and not impose unnecessary operator burden.
PP&L concurs with the NRC that modifications are needed to two assumptions used in thecabinet temperature calculations. The NRC questioned our use of 120'F as the control roomtemperature, implying such a temperature was overly conservative. In response, the infinite timecontrol room temperature, assuming measured normal operating heat loads, has been calculatedusing the method of NUMARC 87-00. The resulting control room temperature is 111'F. TheTER questioned use of 180'F as the operability limitof control room instruments. Based oninformation received from equipment manufacturers, we currently believe the correct limit is140'F, and are performing a reevaluation on this basis. This evaluation willbe completed andsubmitted to the NRC no later then May 1, 1992,
Page 7
<r
ATTACHMENTTO PLA-3745
';:;CONTAPC41i22lT,!ISOLATION','R
RE MMEND ATION'he licensee should list the valves identified in an
appropriate procedure and identify the actions necessary toensure that these valves can be fullyclosed, ifcontainmentisolation is required during an SBO event. The valveclosure should be confirmed by position indication (local,mechanical, remote, process information, etc.)
P&L R
The penetrations which have been identified by the NRC as requiring to be proceduralized arethe Residual Heat Removal (RHR) and Core Spray (CS) suction lines along with theContainment Spray line. Containment isolation of these lines has been addressed and approvedby the NRC prior to this submittal. The following identifies that approved approach.
Susquehanna SES FSAR section 6.2.4.3.6 states in part that "Containment isolation provisionsfor certain lines in engineered safety feature or engineered safety feature-related systems mayconsist of a single isolation valve outside containment. A single isolation valve is consideredacceptable ifit can be shown that the system reliability is greater with only one isolation valvein the line, the system is closed outside containment, and a single active failure can beaccommodated with only one isolation valve in the line." Additionally, section 6.2.4.3.6.3states, "Although strictly speaking the HPCI, RCIC, CS, and RHR pump suction lines do notconnect directly to the primary containment, they are nevertheless evaluated to 10 CFR 50Appendix A, General Design Criteria 56. These lines are each provided with one remotemanually motor operated gate valve external to the containment and use the respective pipingsystems as the second isolation,barrier.. For the RHR and CS valves the hand switches are keylocked".
Further investigation into this issue reveals that section 6.2.4 of the NRC Safety EvaluationReport (NUREG 0776) for Susquehanna SSES documents the NRC approval of meeting thealternative acceptance criteria specified in section 6.2.4 of the Standard Review Plan. Thissection summarizes these alternative acceptance criteria along with specifically identifying thelines found acceptable via this method.
Based on the above explanation we believe that containment isolation is established andcontainment integrity willbe maintained.
Page 8
t g ~
0
4
'L
ATTACHMENTTO PLA-3745
:1'R'O.CEDURFS.:;::lANDl,TRAXNING,',
RE MME ATI The staff expects the licensee to implement the appropriatetraining to assure an effective response to an SBO event.
PP&LR N E
Appropriate plant personnel willbe trained on any new or revised procedures in accordance withthe requirements of Initiative 2, NUMARC 87-00 and Reg.Guide 1.155, section 3.4.
RE MME ATI The staff expects that the plant procedures willreflect theappropriate testing and surveillance requirements to ensurethe operability of the necessary SBO equipment,
'P&L'f
4
It is PP&L s intent to satisfy the Quality Assurance (QA) requirements of Reg. Guide 1.155 byupgrading an existing procedure to incorporate Station Blackout. This procedure addresses allthe Reg. Guide QA requirements and will require the necessary Inspections and Tests to beperformed in accorda'nce with the Operational Quality Assurance Program.
::-;ED6'!RELIA'SILIIYiPROGRAM::::..":
NR RK MMENDATI N'he licensee should complete the implementation of anEDG reliability program which meets the guidance of RG1.155, Section 1.2 and provide a schedule for itscompletion. Confirmation that such a program is in placeor will be implemented should be included in thedocumentation supporting the SBO submittals that is to bemaintained by the licensee.
Page 9
0
0
'I
0
ATTACHMENTTO PLA-3745
PP&L R
Reg. Guide 1.155 specifies that each utilityestablish an EDG performance monitoring program.NUMARC87-00 Appendix D contains guidance for the development and implementation of sucha program. PP&L has committed to implement a program of reliability monitoring and, asindicated above, PP&L must maintain an EDG reliability at or above 97.5% as part of our SBOcoping strategy.
The Reg. Guide and NUMARCprovide "trigger values" for determining compliance with targetreliability. NRC reviewers indicated that lack of this data in our submittal hindered assessmentof SSES EDG reliability. At the 97.5% reliability level, compliance is assumed ifthe failuresto start/load are less than or equal to 3, 4, and 5 out of the last 20, 50 and 100 start attempts,respectively. As of 2/10/92 the failures to start/load in each category were 0,0, and 3,respectively. Thus, today, PP&L can accept the increased reliability target of 97.5%.
PP&L's Emergency Diesel Generator reliability monitoring program has been developed anddocumented in Nuclear Department Administrative Procedure-QA-0401 entitled "EmergencyDiesel Generator Monitoring Program." This procedure complies with the reliabilityrequirements delineated in Appendix D of NUMARC 87-00, Rev. 1. Reliability will bemonitored against a set of "trigger values" with actions specified for various levels of triggervalue exceedance.
Page 10
l~ r v
O.
d
4
ROOM TEMPERATURE RESPONSE TO A STATION BLACKOUT
200
180
160
140~ I—
120
I—
100
8010 20 30 40
TIME (HRS)50 60 70 80
Legendg HVAC EQUIP RM
0 EXH fAN RM
~ HVAC EQUIP RM
0 HVAC EQUIP RM
6 RECIRC PLENUM
COTTAP: A COMPUTER CODEFOR SIMULATIONOF THERMALTRANSIENTS IN SECONDARYCONTAINMENTS OF BOILINGWATER REACTORS
~ ~ ~
"7MARKA. CHAIKO and MICHAELJ. MURPHYPennsylvania Power & Light Company, Allentown, Pennsylvania 18101
Received December I, 1989Accepted for Publication September 12, 1990
The Compartment Transient Temperature AnalysisProgram (COTTAP) was developed by the Pennsylva-nia Power &Light Company forpostaccident boilingwater'reactor (BWR) secondary containment thermalanalysis. The code makes use ofpreviously developedimplicit temporal integration methods and sparse ma-trixinversion techniques to allow modeling ofan en-tire BWR secondary containment. Investigations weremade with a model consisting of 121 compartmentsand 767 heat-conducting slabs. The simulation pre-sented involves the numerical integration of20 101 or-dinary differential equations over a 30-h simulationperiod. Two hours ofCPU time were required to carry
out the calculation on an IBM3090 computer. TheCOTTAP code considers natural convection and radi-ation heat transfer between compartment air and walls
'hrough a detailedflnite difference solution ofthe slabconduction equations. Heat addition from hot pipingand operating equipment, and cooling effects associatedwith ventilation flows and compartment heat removalunits are also included. Additional capabilities ofCOTTAP include modeling ofcompartment heatup re-sulting from steamline breaks and simulation ofnat-ural circulation cooling in compartments with flowpaths at differing elevations.
I. INTRODUCTION
Under postaccident conditions, boiling water reac-tor (BWR) secondary containment ventilation systemstypically isolate to prevent fission product release tothe environment. Since cooled air is no longer circu-lated through the secondary containment, increasedcompartment temperatures result. Predictions of post-accident compartment temperatures are necessary todetermine whether safety-related equipment is sub-jected to temperatures that'exceed its maximum designvalues. Safety-related equipment must be operable un-der postaccident conditions in order to effect the safeshutdown of the reactor.
After an accident, the secondary containment
ventilation system operates in a recirculation modeto promote air mixing between compartments andto dilute locally concentrated radioactive isotopes.Original design calculations for Pennsylvania Power& Light Company's (PP&L) Susquehanna SteamElectric Station (SSES) assumed that air recircula-tion provided enough mixing to produce a fairlyuniform temperature distribution throughout all sec-ondary containment compartments. For this reason,a single-compartment transient. model was used in thesimulation of postaccident conditions. Recent investi-gations based on steady-state calculations have shown,however, that significant temperature variations canexist between compartments. These temperaturevariations were large enough to prompt a detailed
NUCLEAR TECHNOLOGY VOL. 94 APR. 199l
Chaiko and Murphy POSTACCIDENT BWR SECONDARY CONTAINMENTTHERMALANALYSIS
multicompartment transient analysis of the secondarycontainment.
To reanalyze the postaccident transient behavior ofthe SSES secondary containment, PP&L developed theCompartment Transient Temperature Analysis Pro-gram (COTTAP). Development of this program beganafter an evaluation of available codes revealed thatnone were capable of performing a sufficiently detailedsimulation owing to the large number of heat-conduct-ing structures found in the SSES secondary contain-ment. For example, the CONTEMPT code,'hich isprobably the most widely used containment analysisprogram, can model as many as 999 compartments butis limited to 99 heat-conducting slabs. In contrast,COTTAP can model up to 1200 heat-conducting slabsand 300 compartments. It also contains models thatdescribe heat dissipation from operating electricalequipment and process piping. A COTTAP model ofthe SSES-1 and -2 secondary containment structuresconsists of -120 compartments and 800 heat-conduct-ing slabs.
The CONTAINcode2 is a more recently developedcontainment simulation program with complex mod-eling capabilities. It is, however, designed specificallyfor primary containment simulation and is not wellsuited for secondary containment modeling because ithas no provisions for energy input to compartmentsfrom heat loads such as electrical panels, lighting, mo-tors, and hot piping.
A description of the COTTAP code, including as-sumptions, governing equations, numerical solutionmethods, and code limitations is given in Sec. II. Rep-resentative results of the SSES-1 and -2 secondary con-tainment analysis are presented in Sec. III, and codeverification is discussed in Sec. IV.
II. DESCRIPTION OF THE COTTAP CODE
II,A. Compartment Mass and Energy Balances
The COTTAP code allows for air and water vapormass transfer between compartments by means offorced ventilation, leakage, and natural, circulationflows. A forced ventilation flow model describes heat-ing/ventilating/air conditioning systems, and a leakagemodel simulates intercompartment flows that hre gen-erated by pressure differentials. In addition, a naturalcirculation model simulates gravity4riven flows betweencompartments connected by flow paths at differingelevations. Steam can also be added to a compart-ment as a result of pipe breaks or removed throughcondensation and rain-out. Airand water vapor massconservation equations for a compartment with N„ventilation paths, NI leakage paths, and N, natural cir-culation paths are given by
H„ NcV g WojYoj + g WgYIJ + g Woj(Y<j Y)
J —1 j~i jmi
and
dp Ivu Jvl
V—"= g W„j(1 —Y,j) + g Wg(1 —YIJ)dl jai jaiH~
+ g Wy(Y—Ycj) + W~ —Woold —Wlo >j=l(2)
where
V = compartment volume (m3)
t = time (s)
p„p„= compartment air and water vapordensities, respectively (kg/m3)
partments for ventilation path jand leakage path j, respectively
Y~ ——mass fraction of air in adjoiningcompartment associated with cir-.
culation path jWq, = rate of steam addition due to pipe
breaks (kg/s)
W„„d = steam condensation rate (kg/s)
W„= rain-out rate (kg/s).
The values Wj and Wlj are positive for flow into thecompartment and negative for flow out of the com-partment, whereas the circulation rate Wj is always apositive quantity. Ventilation paths are described by
'heirassociated mass flow rates and identificationnumbers of source and receiving compartments. Ven-tilation flows can be tripped offor on at any time dur-ing a transient by supplying appropriate trip-logic data.Leakage, circulation, and pipe break models are dis-cussed in Sec. II.C along with other special purposemodels.
In formulating the compartment energy balance, itis assumed that air behaves as an ideal gas.
Moreover,-'or
the transients of interest, partial pressures of wa-ter vapor are typically (I atm. Therefore, it is assumedthat the steam specific enthalpy depends only on tem-perature, i.e., the vapor enthalpy is equal to the en-thalpy of saturated steam at the temperature of the.gasmixture. The partial pressure of water vapor within acompartment is computed from the ideal gas equationof state, and the total compartment pressure is calcu-lated as the sum of the air and water vapor partialpressures. With these assumptions, the compartmentenergy balance becomes
NUCLEAR TECHNOLOGY VOL. 94 APR. 1991
Pb„,k = total compartment pressure if pipecontains saturated liquid (Pa)
Pb«ak = pipe fluid pressure if pipe containssaturated steam (Pa)
hg(Pb„,k) = specific enthalpy of saturated watervapor at pressure Pb„,k (J/kg)
hi(T) = specific enthalpy of saturated liquidwater at temperature T(J/kg)
TJ —- temperature in adjoining compart-ment associated with circulation pathj (K).
Compartment heat loads from lighting, electrical pan-els, motors, and miscellaneous equipment are main-tained constant unless they are tripped on, off, orexponentially decayed during the transient. Hot pipingand room cooler loads vary with compartment temper-ature and can also be tripped on or off. In addition,hot piping heat loads can be exponentially decayedusing the heat load decay model discussed in Sec.
(3) II.C.7.+ (1 —Yy)hg(T))—(1 —Y)hg (T)],
where
Chaiko and Murphy POSTACCIDENT BWR SECONDARY CONTAINMENTTHERMALANALYSIS
Cp,(T) = specific heat of air at temperature T(J/kg K)
hg(T) = specific enthalpy of saturated watervapor at temperature T (J/kg)
R, = ideal gas constant for air (288.7 J/kg K)
R = ideal gas constant for water (461.4J/kg K)
Qligbt, Qpanel» Qmotor> Qcooler» Qpiping > Qrnisc
= compartment heat loads due to light-ing, electrical panels, motors, aircoolers, hot piping, and miscellane-ous equipment (J/s)
Q,i,b = rate of heat transfer to compartmentair/water vapor mixture from sur-rounding slabs (J/s)
II.B. Slab Model
In the secondary containment of a BWR, compart-ment walls, ceilings, and floors are generally concreteslabs that range in thickness from -0.3 to -2 m. Todetermine the heat transfer rate between a compart-ment atmosphere and the bounding concrete slabs, theone-dimensional heat conduction equation
(4)
is solved for each slab. Here, T, (K) is the slab temper-ature, and x (m) is the spatial coordinate. Since thethermal diffusivityns (m /s) is supplied as input foreach slab, materials other than concrete can be mod-eled provided that slabs are of uniform material com-position. This one-dimensional description assumesthat slab edge effects do.not significantly affect theoverall rate of heat transfer.
Boundary conditions on slab temperature are givenby
Qb„,k = heat transfer rate to air/water vapormixture from liquid exiting break asit cools to compartment temperature(J/s) and
[Tl(r) Ts(0 r)]aT, h,Bx „o ks
Wb, ——mass flow rate of steam exiting break(kg/s)
46 NUCLEAR TECHNOLOGY VOL. 94 APR. 1991
Chaiko and Murphy
where
T> (t), T2(t) = temperatures of compartments ad-jacent to the slab
k, = slab conductivity (J/m s K)
L, = slab thickness (m)
h~, h2 = heat transfer coefficients (J/mz s K).
The solution of Eq. (4) subject to Eqs. (5) and (6) givesthe rates of energy transfer from the slab surfaces tothe adjacent gas mixtures.
The coefficients hi and hz account for naturalconvection, radiation, and condensation heat transfer.In the absence of condensation, the coefficient hl canbe expressed as
hi ——ht + h/p, (7)
where h>„and h~, are the natural convection and ra-diation components, respectively.
Natural convection coefficients are expressed interms of the Nusselt number, which in turn is a func-tion of the Rayleigh and Prandtl numbers. For the co-efficient hl„, the appropriate relation is
Nu = —= f(Ra,Pr),h)„Ct.k (g) (~, +1)h/I: '4 + a + b —c)elm,auaTau ~ (10)
wherewhere
Ct. ——slab characteristic length
k = gas thermal conductivity
and the Rayleigh and Prandtl numbersture are, respectively, defined by
o = Stefan-Boltzmann constant (5.669 x 10 s J/mz s K4)
free convection from a vertical plate. For horizontalslabs, free-convection coefficients depend on whetherthe surface is being heated or cooled by the surround-ing gas mixture. As recommended by Holman,4 thecorrelation of Fujii and Imuras is used with the mod-ified characteristic length proposed by Goldstein et al.~to compute the coefficient for an arbitrarily shapedslab with heated surface facing upward or cooled sur-face facing downward. In cases where the upper sur-face is cooled or the lower surface is heated, thecorrelations of Lloyd and Moran7 are used.
Diatomic gases such as nitrogen and oxygen are es-sentially transparent to thermal radiation; however, theemissivity of water vapor with respect to thermal radi-ation is significant. In COTTAP, radiant energy ex-change between a slab surface and water vapor containedwithin the surrounding gas mixture is modeled throughthe use of an effective radiation heat transfer coeffi-cient [see Eq. (7)]. For the applications of interest, tem-perature differences between a slab surface and thesurrounding gas mixture are relatively small (typically(5 K). Therefore, the following approximate relationproposed by Hottel and Sarofim for small tempera-ture differences is used to compute the radiation coef-ficient:
where
g = acceleration due to gravity (9.8 m/sz)
p = coefficient of thermal expansion (K ')v = kinematic viscosity (mz/s)
n = thermal diffusivity (m /s)
p = dynamic viscosity (kg/m s)
Cv = specific heat of the airhvater vapor mixture- ~ (J/kg K).
T = gas temperature (K)
T„„~ = slab surface temperature (K)
e„,„= emissivity of water vapor evaluated at T,u.
The Cess-Lian'quations, which give an analyticalapproximation to the emissivity charts of Hottel andEgbert," are used to compute the water vapor emis-sivity. In Eq. (10), c has the value 0.45, and a and b areobtained through differentiation of the Cess-Lian emis-sivity equations
Gas mixture properties used in the calculation of freeconvection coefficients are evaluated at the thermalboundary layer temperature, which is taken as the av-erage of the slab surface temperature and the bulk gastemperature.
For vertical slabs, coefficients are calculated fromthe correlation proposed by Churchill and Chu3 for
and
81n[e„(T,P„P„,P„L )]a
Bin(P„L )
8 ln [e„(T,P„P„, P„L„,)]8 ln(T)
(12)
NUCLEAR TECHNOLOGY VOL. 94 APR. 1991 47
losure', for in-nd condensa-
g walls. For a-nsation alonebecoming sat-ain-out) form
g s mpartment rel-ative humidity less than or equal to unity, the rainoutrate W (kg/s) is calculated from the following empir-ical model:
surface temperature drops below the dew point (thesaturation temperature of water evaluated at the par-tial pressure of water vapor in the compartment) of theair/water vapor mixture. Heat transfer coefficients forcondensation conditions are calculated using the exper-imentally determined Uchida" correlation, which in-cludes the diffusional resistance effect of noncondensible
W, = 200 (RH —0.99)max(W„C,i)
ifRH) 0.99
andgases on steam condensation rates.
In COTTAP, initial compartment temperatures,pressures, and relative humidities are specified as in-put data. An initial slab temperature profile is deter-mined by computing the steady solution to Eqs. (4),(5), and (6) corresponding to the initial compartmentconditions. This implies that compartments have beenmaintained at their initial conditions long enough forslabs to attain steady-state temperature profiles.
W, = 0.0 ifRH s 0.99, (16)
where
RH = relative humidity
Ws = total steam flow rate into the compartment„(kg/s)
C,i = constant that is supplied as part of the inputdata (kg/s).
Chaiko and Murphy FOSTACCIDENT BWR SECONDARY CONTAINMENTTHERMALANALYSIS
where isolation of a pipe break (due to valve c
P, = a'r pa t'al prcssu e (Pa)stance) a compartment begins to cool ation continues to occur on surroundin
P„= water vapor partial pressure (Pa) sufficiently fast cooldown rate, condedoes not prevent compartment air fromurated, and thus moisture droplets (r
Condensation on a slab surface occurs when the within the a mixture. To maintain co
II.C. Special Purpose Models
The COTTAP code includes specialized models tosimulate the effects of pipe breaks, hot piping, andcompartment air coolers. Leakage and natural circu-lation models are also included to describe intercom-partment mass transfer. In addition, the code includesa simplified slab model, a heat load decay model, anda compartment model in which temperature, pressure,and relative humidity are specified as a function oftime.
II.C.l. Pipe Break Model
Within the scope of the present model, pipes maycontain steam or saturated liquid water. Input data de-fine the total mass flow through the break Wb, (kg/s)along with the time at which the break develops andthe length of time over which fluid loss occurs. Forpipes containing saturated liquid, the steam flow rateWb, exiting the pipe (kg/s) is calculated from the en-ergy balance
Wbihy(P>) = Wbslig(P) + (Wbi —Wbs)h/(P), (14)
which describes the isenthalpic expansion of fluid frompipe pressure P~ to compartment pressure P. The liq-uid fraction, which does not flash as it leaves the pipe,is assumed to cool to compartment temperature, andthe dissipated sensible heat is transferred directly to thecompartment air/water vapor mixture. For the casewhere a pipe contains steam, all of the mass and energyexiting the break is deposited directly into the compart-ment gas mixture.
Rain-out phenomena can be important in compart-ments containing pipe breaks. For example, following
48
II.C.2. Hot Piping Model
In many secondary containment compartments,the major heat source consists of piping that containsreactor steam or coolant. The heat addition rate to acompartment airhvater vapor mixture from a hot pipeis calculated from
alp(<g Up 7rLpDp[Tj'(t )] > (I7)
where. ~
Up = overall heat transfer coefficient (J/m2 s K)
L~ = pipe length (m)
D~ = outside diameter of the pipe (or insulation ifthe pipe is insulated) (m)
Tj ——pipe fluid temperature (K)
T = compartment temperature.
The overall heat transfer coefficient is calculated by thecode based on initial compartment conditions; the co-efficient is then maintained constant throughout thetransient.
II.C.3. Air Cooler Model
Cooling units are used in a number of secondarycontainment compartments to remove heat generatedby equipment such as emergency core cooling systems(ECCS) injection pumps and high-voltage buses andtransformers. Heat removal rates of cooling units arecalculated from
Qcool(t ) Ccool (T(t) Tcool(t )j ~ (Ig)
NUCLEAR TECHNOLOGY VOL. 94 APR. 199t
Chaiko and Murphy POSTACCIDENT BWR SECONDARY CONTAINMENTTHERMALANALYSIS
where
T„,/(t) = average of the inlet and outlet coolingwater temperatures
C„,/ = constant that is computed from spec-ified initial values of the cooling loadQ,/, the inlet cooling water tempera-ture, the cooling water flow rate, andthe compartment temperature T.
An energy balance on the cooling water yields the out-let cooling water temperature.
II.C.4. Leakage Models
The COTTAP leakage model simulates pressure-induced intercompartmental mass transfer throughopenings such as doorways and ventilation ducts. In-tercompartment leakage is calculated by balancing thepressure differential between the compartments with anirreversible pressure loss. Thus, the leakage rate sat-isfies
P (t) —P, (t) Kinesia(t)[ Win(t)](19)
2p//(l)Awwhere
P1, Pz = pressures of the compartments associatedwith the leakage path (Pa)
WN = leakage rate (kg/s)
K~ = irreversible pressure loss coefficient
A//, ——leakage area (m )
p//, = gas density within the compartment sup-plying the leakage flow (kg/m ).
It is assumed that inertial effects do not significantlyaffect leakage rates.
II.C.5. Natural Circulation Model
A natural circulation model simulates gravity-driven mixing in compartments connected by flowpaths at differing elevations. The circulation rate W,(kg/s) is obtained from
This model also describes intercompartment, gravity-driven circulation flows that can develop at open door-ways (see the analysis of Brown and Solvason'.
II.C.6. Thin Slab Model
The detailed slab model discussed in Sec. II.B isnot required to describe heat transfer through thinslabs that have little thermal capacitance. Slabs of thistype, e.g., refueling floor walls, have nearly linear tem-perature profiles, and thus the heat flow through a thinslab can be calculated by the use of an overall heattransfer coefficient U„. The rate of heat transferthrough a thin slab is obtained from
q/s(r) = UisA [T1 (>) —T2(/)], (21)
where
A„= thin slab heat transfer area (m )
Tj Tz = temperatures of the compartments sepa-rated by the slab (K).
Values of U„(J/m s K) are supplied as part of thecode input data (one value for each vertical slab andtwo values for each horizontal slab). For horizontalslabs, two values of U„are required because free-convection film coefficients depend on the direction,upward or downward, of heat flow through the slab.
II.C.7. Heal-Load Decay Model
Cooling of a component such as a pipe filled withhot stagnant fluid or a pump that has ceased operat-ing is simulated through the use of a lumped-param-eter heat transfer model. Most compartments in thesecondary containment have a large thermal capacitybecause of the bounding concrete slabs. It is thereforeassumed that the component temperature changes ona faster time scale than the compartment air temper-ature; i.e., the air temperature is assumed to remainfairly constant during the cooldown of the component.With this assumption, the component heat dissipationrate Qc(t) is governed by
7'Q'" =-Q(/) (22)
d/
W — g['() '()](" )
K//[Alp2(t)] + KN/[ANpi (/)] J
where
Qc(/o) = Qco (23)
where
p1, pz ——densities of the air/water vapor mixtureswithin the two adjacent compartments(kg/m ) (here it is assumed that p2 is thegas density for the cooler compartment)
E„,E/—- elevations of the upper and lower flowpaths (m)
A„,A/——upper and lower flow path areas (m ).
NUCLEAR TECHNOLOGY VOL. 94 APR. 1991
McCwVc
UcAc(24)
where
M, = mass of the component (kg)
Ci~ = specific heat of the component (J/kg K)
49
and 7, (s '), the thermal time constant of the compo-nent, is given by
Chaiko and Murphy POSTACCIDENT BWR SECONDARY CONTAINMENTTHERMALANALYSIS
U, = overall heat transfer coefficient (J/m2.s K)
A, = component heat transfer area (m2).
In Eq. (23), to (s) is the time at which the cooldownprocess begins, and Q«, which is supplied as inputdata, is the heat dissipation rate prior to cooldown. So-lution of Eqs. (22) and (23) gives the exponential-decayapproximation used in COTTAP to model heat dissi-pation of cooling components. The component timeconstant y, is specified as input data except in the caseof hot piping, where it is calculated by the code fromthe piping description data.
II.C.8. Time-Dependent Compartment Model
With the time-dependent compartment (TDC)model, environmental conditions within a compart-ment are specified as a function of time; i.e., temper-ature, pressure, and relative humidity versus time aresupplied as tabular input data. This model is particu-larly useful for representing outside air conditions, in-cluding solar and thermal radiation effects. Theinfluence of solar and long-wave atmospheric radiationon exterior buildup surfaces can be described by spec-ifying the effective Sol-Air temperature'n the TDCinstead of the actual outside air temperature. In sec-ondary containment analysis, the TDC model is alsouseful for describing transient conditions within theprimary reactor containment, which are generallyknown from the results of detailed licensing basis cal-culations.
dTsl =GT ps sxx (25)
where
i = 1,2,3,...,N, the number of equally spacedgrid points
Tp ——slab temperature at grid point iT p
= finite difference approximation to thesecond-order spatial derivative at grid pointi.
Following the approach used by Pirkle andSchiesser'3 in the MOL solution of parabolic equa-
50
II.D. Numerical Solution Nlethods
An energy balance and two mass balances are solvedfor each compartment to determine gas temperature,air mass, and water vapor mass. In addition, the one-dimensional heat conduction equation is solved foreach slab. Before computing the numerical solution ofthe governing equations, partial differential equationsdescribing heat flow through slabs are approximatedby sets of ordinary differential equations (ODEs). Thisis accomplished through application of the method oflines (MOL). In the MOL, a finite.difference approx-imation is applied only to the spatial derivative inEq. (4), giving
tions, fourth-order central difference formulas are usedto compute T t at interior grid points:
A six-point sloping difference formula is used to ap-proximate T p at i = 2 and i = N —1:
ITsxx2 = —
2 (10Tsi 15Ts2 4Ts3 + 14Ts4
—'6T,s+ T,6) + O(~ ) (27)
and
1
TsxxN l 2 (10TSN 15TsN-l 4TsN-2
+ 14TsN-3 —6TsN-4 + TsN-5)
+O(~4) . (28)
For the end points, where the normal derivativesare specified through convective boundary conditions,the following finite difference approximations, recom-mended by Pirkle and Schiesser,'3 are used to com-pute T
I 415 32T = ———T i + 96T2 —36T3+ —T4sxxt 1262 6s s S
3s
Tss 50t3,Tsxl + O(h ) (29)3 4
and
I 415Ts N= ———TN+ 96TN i —36TN 2sxx
32 3+ —TsN-3 ——TsN-4 + 506TsxN2
+O(a) (30)
In Eqs. (29) and (30), the normal derivatives Tsxi andT~ are evaluated in accordance with Eqs. (5) and (6),the convective boundary conditions; i.e.,
Chaiko and Murphy POSTACCIDENT BWR SECONDARY CONTAINMENTTHERMALANALYSIS
Allgovernirig equations are now expressed in terms ofODEs of the form
dy—= F(y,t) with y(0) =yedL'32)Solutions of Eq. (32) exhibit rapid initial adjust-
ments in compartment air temperature caused by therelatively small thermal capacitance of the air containedwithin the compartment. Moreover, slab temperaturesundergo rapid initial changes in narrow regions nearthe boundaries, resulting in the formation of spatialthermal boundary layers. In the numerical integrationof Eq. (32), small time steps are required to simulatethese initial transients. As the initial transient responsedecays, however, it is desirable to increase step sizes inorder to reduce the computation time required to fol-low the slowly varying part of the solution. Equations,such as Eq. (32), which exhibit initial temporal bound-ary layer structures are termed stiffdifferential systems(see the discussion in Ref. 16), and because of stabil-ity limitations, they cannot be solved efficiently withexplicit integration schemes. For this reason, an im-plicit scheme was selected for COTTAP.
Numerical integration of the governing Eq. (32) iscarried out with the LSODES code,'hich uses theimplicitbackward differentiation methods proposed byGear for the solution of stiff systems. The LSODEScode also employs sparse'matrix inversion techniquesin solving the implicit finite difference equations. Withthese numerical integration features, it is feasible tocarry 'out the integration of the large differential sys-tems that arise in the simulation of secondary contain-ment transients. As an illustration of the problemdimension, simulation of the SSES-1 and -2 secondarycontainments under postaccident conditions requiredthe solution of 20101 coupled ODEs.
For these large-scale problems, reevaluation ofcode-calculated slab heat transfer coefficients at everytime step leads to unacceptably long computationtimes. To alleviate this difficulty-, the frequency of re-evaluation (number of steps between reevaluation ofcoefficients) is a parameter supplied as input to thecode. Sensitivity calculations on small-scale problemsrepresentative of postaccident secondary containmenttransients indicate that coefficients can be reevaluatedas infrequently as once per ten steps without introducingsignificant errors in the results. The CPU time require-ments were reduced by a factor of 4 when coefficientswere reevaluated at every tenth time step.
1. Fission product transport among compartmentsis not modeled.
II.E. Code Limitations. in Modeling Accident Scenarios
The following modeling limitations have been iden-tified in the current version of the COTTAP code:
2. Cooler modeling does not describe moisture re-moval under conditions where the cooling coil temper- .
ature is below the dew point of the inlet gas mixture.
3. Pipe break modeling is valid only for lines con-taining steam or saturated liquid; breaks involving therelease of subcooled liquid cannot be described.
4. Compartment flooding events cannot be simu-lated because all liquid is assumed to exit through com-partment floor drains.
III. RESULTS OF SSES SECONDARY CONTAINMENTANALYSIS FOR POSTACCIOENT CONDITIONS
This section gives representative results for a COT-TAP simulation of the combined SSES-1 and -2 sec-ondary containments under postaccident conditions.The thermal responses of the Units 1 and 2 secondarycontainments are coupled by heat transfer throughcommon walls that separate the two structures. TheSSES model consists. of 105 compartments, 16 time-dependent compartments, 767 slabs, 38 thin slabs, and505 heat loads. The simulation was carried out for 30 hand required 124 min of CPU time on an IBM 3090computer. Note that most of the CPU time is requiredto simulate the rapidly varying part of the transientthat occurs within the first few hours of the event.Thus, substantially longer simulation times do not sig-niflcantly increase CPU time requirements.
For this analysis, it is assumed that a loss-of-coolant accident (LOCA) occurs in SSES-1 and a falseLOCA signal (a spurious signal that indicates loss ofreactor coolant and leads to ventilation system isola-tion and operation of ECCS injection pumps) is gen-erated on SSES-2. Under postaccident conditions,ECCS injection pumps comprise the key equipmentwithin the secondary containment structure. The ECCSconsists of the residual heat removal (RHR), corespray, and high-pressure coolant injection (HPCI) sys-tems. These systems receive electrical power from high-voltage buses contained within emergency switch gearand load center rooms. Figure I shows the calculatedtemperature response within a SSES-1 RHR pumproom (each unit contains two RHR pump rooms andtwo core spray pump rooms). Initially, the air temper-ature increases rapidly because of the small thermal ca-pacitance of the air within the compartment. As airtemperature increases, a balance between compartmentheat sources and losses to compartment air coolers andslabs begins to develop. At this time, air.temperaturestarts to increase on the slow time scale governed bythe slab thermal capacity and transport properties. Aninitial rapid temperature rise followed by a muchslower temperature increase is characteristic of all com-partment heatup transients. After 1 h of operation, thisparticular RHR pump switches from the injectionmode of operation to the suppression pool cooling
NUCLEAR TECHNOLOGY VOL. 94 APR. 1991 51-
Chaiko and Murphy POSTACCIDENT BWR SECONDARY CONTAINMENTTHERMALANALYSIS
hC
320
E 316I-E3 316
o 314
3120 5 10 15 20 25 30
Time (h)
Fig. 1. Simulation of postaccident temperature responsewithin SSES-I RHR pump room for LOCA onSSES-I and false LOCA on SSES-2.
317
~ 316.
~~ 315
I- 314Eo 313
E 312
Q 311
X310
0 5 10 15 20 25 30
Time (h)
Fig. 3. Simulation of postaccident temperature responsewithin SSES-I HPCI pump room for LOCA inSSES-I and false LOCA in SSES-2.
mode. As a result of increased compartment heat loadsassociated with the change in operating mode, the tem-perature again increases rapidly until a new balancebetween the heat-generation and heat-loss rates is at-tained.
The temperature response within a SSES-I corespray pump room is shown in Fig. 2. Core spray op-eration begins at the start of the event and ceases I hlater. Temperature decreases rapidly at this point be-cause, once pump operation is terminated, no signif-icant heat loads remain in the compartment. Figure 3illustrates the temperature response of the SSES-I
HPCI system, which also begins operation at the startof the accident. In this case, however, compartmenttemperature continues to increase when the systemceases operation at I h into the transient. This occursbecause piping heat loads within this compartment aresubstantial. When HPCI pump operation stops, an as-sociated room cooling unit also ceases operation. Uponshutdown of the cooling unit, slowly decaying pipingheat loads rapidly increase compartment temperatureuntil a balance between heat generation and heat lossesto compartment slabs is approached. Figure 4 gives thetemperature within a SSES-I load center room that
~ 317P
E~ 316
I-E3 315
CC
~ 314
V)
313O 0 5 10 15 20 25 30
Time (h)
Fig. 2. Simulation of postaccident temperature responsewithin SSES-I core spray pump room for LOCA inSSES-I and false LOCA in SSES-2.
—309
ejE 308I-E3
cc
~ 3078CO
3060 5 10 15 20 25 30
Time (h)
Fig. 4. Simulation of postaccident temperature responsewithin SSES-I load center room for LOCAin SSES-Iand false LOCA in SSES-2.
52 NUCLEAR TECHNOLOGY VOL. 94 APR. 1991
supplies electrthis compartmstant throughout the transient.
From the results of this analysis, it is determinedthat under postaccident conditions, some of the equip-ment within the secondary containment would be ex-posed to temperatures that exceed their qualificationvalues. Consequently, components were reassessed foroperation at higher temperatures, and in some in-stances equipment was relocated to compartments withless severe environmental conditions. Furthermore, aprocedure was developed to instruct plant operators toshed nonessential electrical loads within 24 h after anaccident in order to moderate the temperature re-sponses within secondary containment compartments.
K 310P
~~ 305
COTTAP———CONTAIN
300
IChaiko and Murphy POSTACCIDENT BWR SECONDARY CONTAINMENTTHERMALANALYSIS
ical power to emergency equipment. In 315ent, heat loads remain essentially con-
IV. EVALUATION OF CODE ACCURACY
As part of the verification process for the COT-TAP code, calculational results were compared withthose obtained with the CONTAIN (Ref. 2) program,which has been verified through comparison with ex-perimental data.' Although the CONTAIN codedoes not accommodate a direct heat input (such asfrom operating mechanical or electrical equipment) toa compartment, useful problems can nevertheless beformulated in order to investigate the modeling andcomputational accuracy of COTTAP. Two such prob-lems were formulated for code verification. The firstproblem tests the CO%I'AP compartment mass and en-ergy balance calculations and the slab heat transfersimulation. This problem consists of a single compart-ment that has a 1000-m3 volume and contains air at300 K and 101 325-Pa initial temperature and pressure.Concrete slabs, which range in thickness from 0.1 to1 m, form the walls of the compartment. Allslabs havea uniform, initial temperature of 300 K. To add heatto the compartment, the air in contact with the outersurface of one slab (the slab that is 0.1 m thick) is sud-denly increased to 400 K at t = 0. In addition, at 50 sinto the transient, air with a temperature of 500 K is in-jected into the compartment at a 0.26 kg/s flow rate.Outer surface temperature rise and air injection con-ditions were selected to effect significant, but not ex-cessive, temperature and pressure response.
Figures 5 and 6 present a comparison of the COT-TAP and CONTAIN calculation results for the firsttest problem. The temperature and pressure simula-tions both show excellent agreement; note that thepressure response curves given in Fig. 6 completelyoverlap. In -Fig. 5, the initial temperature increase,which is due to injection of hot air into the compart-ment, begins to level offat -0,5 h. Heat addition bymeans of conduction through the externally heated slabthen begins to occur, causing a further but less rapidincrease in temperature.
The second test problem considered for code ver-
0.20
0.18
0.16
0.14Q-
0.12—COTTAP———CONTAIN
0.100 2 4 6 8 10
Time (h)
Fig. 6. Comparison of COTTAP and CONTAIN.compart-ment pressure simulations for test problem l.
ification involves modeling of compartment tempera-ture and pressure behavior under conditions wherehigh-energy steam is injected into the compartment. Inthis problem, condensation effects strongly influencethe rate of temperature and pressure increase. Com-partment physical description data are the same as thatfor test problem 1. In this case, however, the only heatsource is the steam entering the compartment at a0.20 kg/s flow rate and a 2.7756 x 106 J/kg enthalpy.This flow rate and enthalpy are characteristic of asmall steam leak within a secondary containment com-partment. Figures 7 and 8 show a comparison of the
0 2 4 6 8 10
Time (h)
Fig. 5. Comparison of COTTAP and CONTAINcompart-ment temperature simulations for test problem I,
NUCLEAR TECHNOLOGY VOL. 94 APR. I99t 53
Chaiko and Murphy POSTACCIDENT BWR SECONDARY CONTAINMENTTHERMALANALYSIS
450
~
4oo'50
300
—COTTAP———CONTAIN
ACKNOWLEOGMENTS
The authors thank Jack G. Refling, James E. Agnew,Mark R. Mjaatvedt, and Leonard J. West for their manyhelpful suggestions during the course of this work. We alsothank Lisa Walsh for typing the manuscript.
REFERENCES
1. C. C. LIN, C. ECONOMOS, J. R. LEHNER, G.MAISE, and K. K. NG, "CONTEMPT4/MOD4: A Multi-compartment Containment System Analysis Program,"BNL-NUREG-51754, Brookhaven National Laboratory(1984).
0 5 10 15 20
Time lh)
Fig. 7. Comparison of COTTAP and CONTAINcompart-ment temperature simulations for test problem 2.
0.6
os
g 0.4
IPn 0.3-
Q.
2. K. K. MURATAet al., "User's. Manual for CONTAIN1.1: A Computer Code for Severe Nuclear Reactor AccidentContainment Analysis," NUREG/CR-5026, Sandia Na-tional Laboratories (1989).
3. S. W. CHURCHILLand H. H. S. CHU, "CorrelatingEquations for Laminar and Turbulent Free Convectionfrom a Vertical Plate," Int. J. Heat Mass Transfer, 18, 1323(1975).
4. J. P. HOLMAN, Heat Transfer, 4th cd., p. 250,McGraw-Hill Book Company, New York (1976).
5. T. FUJII and H. IMURA,"Natural Convection HeatTransfer from a Plate with Arbitrary Inclination," Int. J.Heat Mass Transfer, 15, 755 (1972).
6. R. J. GOLDSTEIN, E. M. SPARROW, and D. C.JONES, "Natural Convection Mass Transfer Adjacent toHorizontal Plates," Int. J. Heat Mass Transfer, 16, 1025(1973).
0.2
0.1
COTTAP———CONTAIN7. J. R. LLOYDand W. R. MORAN, "Natural Convec-
tion Adjacent to Horizontal Surface of Various Planforms,"ASME 74-WA/HT-66, American Society of MechanicalEngineers (1974).
0 5 10 15 20
Time (h)
Fig. 8. Comparison of COTTAP and CONTAINcompart-ment pressure simulations for test problem 2.
COTTAP and CONTAIN simulation results. The re-sults show good agreement even though the codes em-ploy considerably different approaches in thecalculation of condensation rates on slab surfaces. TheCOTTAP code uses the experimentally determinedUchida'ondensation coefficient, while CONTAINcarries out a detailed computation of the thermal re-sistances associated with the gas boundary layer andthe condensate film.
8. D. Q. KERN, Process Heat Transfer, p. 690, McGraw-HillBook Corupany, New York (1950).
9. H. C. HOTTEL and A. F. ballot'tM, RadiativeTransfer, McGraw-HillBook Company, New Yoit (1967)
'1~
10. R. D. CESS and M. S. LIAN,"ASimple Parameteriza-tion for the Water Vapor Emissivity," Int. J. Heat Transfer,98, 676 (1976).
11. H. C. HOTTEL and R. B. EGBERT, "Radiant HeatTransmission from Water Vapor," Am. Inst. Chem. Eng.,38, 531 (1942).
12. H. UCHIDA, A. OYAMA,and Y. TOGO, "Evalua-tion of Post-Incident Cooling Systems of Light-WaterPower Reactors," Proc. 3rd Int. Conf. Peaceful Uses ofAtomic Energy, Geneva, Switzerland, 1964, Vol. 13, p. 93,United Nations (1965).
54 NUCLEAR TECHNOLOGY VOL. 94 APR. 1991
Chaiko and Murphy
13. W. G. BROWN and K. R. SOLVASON, "Natural Con-vection Through Rectangular Openings in Partitions-I Ver-tical Partitions," Int. J. Heat Mass Transfer, 5, 859 (1962).
14. ASHRAE Handbook 1985 Fundamentals, AmericanSociety of Heating, Refrigerating and Air-Conditioning En-gineers, Atlanta, Georgia.
15. J. C. PIRKLE, Jr. and W. E. SCHIESSER, "DSS/2: ATransportable FORTRAN 77 Code for Systems of Ordinaryand One, Two and Three-Dimensional Partial DifferentialEquations," presented at 1987 Summer Computer Simula-tion Conference, Montreal, Canada, 1987.
18. K. K. MURATAand K. D. BERGERON, "Experimen-tal Validation of the CONTAIN Code," Proc. 11th LWRSafely Information Mtg., Gaithersburg, Maryland, October24-28, 1983, SAND-83-1911C, Sandia National Laborato-ries (1983).
19. K. K. MURATAet al., "CONTAIN:Recent Highlightsin Code Testing and Validation," Proc. Int. Mtg. Light WaterReactor Severe Accident Evaluation, Cambridge, Massa-chusetts, August 28-September I, 1983, American NuclearSociety (1983).
16. C. W. GEAR, Numerical Initial Value Problemsin Or-dinary Differential Equations, Chap. 11, Prentice-Hall, En-glewood Cliffs, New Jersey (1971).
17. A. C. HINDMARSH, "ODEPACK, A SystematizedCollection of ODE Solvers," Scientific Computing, Vol. I,p. 55, R. S. STEPLEMAN et al., Eds., IMACS Transac-tions on Scientific Computation, North-Holland PublishingCompany, Amsterdam (1983).
Mark A. Chaiko [BS, 1980, and MS, 1983, chemical engineering, Penn-sylvania State University (PSU); PhD, applied mathematics, Lehigh Univer-sity, 1989] is a project engineer-nuclear systems at the Pennsylvania Power &Light Company. His current technical interests include boiling water reactorstability analysis and thermal-hydraulic modeling of reactor systems.
Michael J. Murphy (BS, mechanical engineering, 1982, and MS, nuclearengineering, 1986, PSU) is a project engineer-nuclear systems with the Penn-sylvania Power & Light Company. He is currently involved in simulation ofanticipated transient without scram and severe accident analysis.
