THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS 88-GT-51345 E. 47 St., New York, N.Y. 10017
00) The Society shall not be responsible for statements or opinions advanced in papers or in dis-cussion at meetings of the Society or of its Divisions or Sections. or printed in its publications.Discussion is printed only if the paper is published in an ASME Journal. Papers are available
from ASME for fifteen months after the meeting.
Printed in USA.
Steam Injection of Frame 5 Gas Turbines for PowerAugmentation in Cogeneration Service
DAVID A. LITTLELiburdi Engineering International Corporation
Hamilton, Ontario, CanadaJ. P. RIVES
Rhone-Poulenc ChimieUsine Du Pont-De-Claix, France
ABSTRACT
A large chemical plant in Francegenerates both electricity and steam using afleet of Frame 5 gas turbines. During thewinter months when maximum generation capacityis utilized, excess steam may be produced.Instead of venting to atmosphere, the excesssteam will now be injected into two gasturbines through a steam injection system.Power increases of up to 29.6% were achievedduring testing with a corresponding increasein cogeneration fuel conversion efficiency ofup to 4.8%.
This paper describes the preliminaryengineering, design, fabrication, install-ation, test and operation of the steaminjection system for the two Frame 5 gasturbines.
INTRODUCTION
Rhone-Poulenc Chimie, Pont-de-Claixgenerates all of its steam and most of itspower requirements using a fleet of Frame 5Model 'LA's with unfired waste heat boilers,Model 'N's with optionally fired waste heatboilers, and an electric boiler.
Because the price of electricity toRhone-Poulenc from EDF (Electricitie deFrance) is higher on designated 'winter' daysthan on 'summer' days, there is a tremendousvariation in power plant operation. Duringsummer, when it is cheaper to purchaseelectricity than to produce it, turbines arerun to satisfy chemical plant steamrequirements (with the production of minimumpower). However, purchased electricity isvery expensive in the winter so all turbinesare base loaded to generate the total chemicalplant power demand. If the chemicalproduction cannot utilize all the steamproduced, excess steam is vented to atmosphereto hold the main steam header pressure.
Rhone-Poulenc contacted LiburdiEngineering in mid-1983 during aninvestigation into gas turbine steam injectionfor power augmentation. Discussions on thetopic led to the awarding of a preliminaryengineering contract to Liburdi Engineering inOctober of 1984.
PRELIMINARY ENGINEERING
To supply information necessary for Rhone-Poulenc to evaluate the feasibility of steaminjection, a computer model of the Frame 5 gasturbine and its optionally fired boiler wasconstructed. The model was calibrated againstG.E. Frame 5 Model 'P' quoted information, andthen component performances were modified toduplicate measured performance. Using thecomputer model, the effects of steam injectionon: power, exhaust flow, fuel consumption, andcogeneration fuel conversion efficiency werecalculated and plotted, as illustrated forpower in Figure 1.
The gains in power in the winter monthswhen the ambient varies between 15°C and -5°Cwould be severely restricted by the existinggenerator limit; thus the generator ratingswere reviewed with the manufacturer. It wasfound that the generator could absorb the fullgear box limit of 30 MW when the ambient isbelow 20°C. Therefore, power augmentationwould be possible in the winter.
The piping drawings for the turbinecompartment were used to find a preliminaryrouting for the steam injection system. Asimple direct routing appeared to be possiblewith the use of flexible hoses.
The economics of steam injection werestudied by Rhone-Poulenc using the preliminarypower, fuel, and efficiency data and adecision was made to install steam injectionsystems on two Frame 5 'N' units. Projectresponsibility was shared as shown in Figure 2to ease logistics and utilize the expertise ofeach firm.
Presented at the Gas Turbine and Aeroengine CongressAmsterdam, The Netherlands—June 6-9, 1988
Downloaded From: https://proceedings.asmedigitalcollection.asme.org/ on 01/29/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
30
28
w3a. 24U
22
N
w20
GEARBOX LIMIT
a.^-•-.-•_..•-
FIGURE I PRELIMINARY GAS TURBINE POWER WITH STEAMINJECTION.
RESPONSIBILITY
Liburdi RhoneTASKEngineering - Poulenc
Detailed design of steaminjection system within Xturbine enclosure
Preliminary layout of plant Xsteam injection system
Detailed design of plant steam Xinjection system
Supply steam injection nozzles Xand manifolds
Supply plant steam injection Xsystem
Turbine modifications, Supervision Xnozzle and manifoldinstallation Implementation X
Installation of plant steam Xinjection system
Operating guidelines X
Supervision XAcceptance testing
Implementation X
FIGURE 2 STEAM INJECTION PROJECT ORGANIZATION.
DETAILED ENGINEERING
Detailed engineering commenced with avisit to Grenoble and a thorough examinationinside the turbine enclosure of each candidateturbine. The locations of all bleed, coolingair, and vent piping, and coolers in thevicinity of the combustor shell were measuredon both engines.
The overall plant layout was studied andin conjunction with Rhone-Poulenc personnel, apossible routing from the main plant steamheader to each turbine was defined. Thisrouting and the existing piping inside theturbine enclosure both favoured a system with
entry through the top of the enclosure.Figure 3 shows a schematic of the steaminjection system which was agreed upon andimplemented.
The piping and components outside theturbine enclosure were designed and fabricatedby Rhone-Poulenc to the material and componentstandards used for steam piping throughout theplant. Thermal expansion loops wereincorporated as necessary, with the top of theturbine enclosure being used as the rigidlocating point for the turbine end of thesystem. All components inside the enclosurewere detailed and the steam injectionmanifolds and nozzles fabricated by LiburdiEngineering. Differential thermal expansionbetween the combustor shell and the steaminjection piping was provided for through theuse of flexibles between the steam injectionnozzles and manifold. Type 316L was specifieddownstream of the strainer to prevent thepossibility of rusting in the piping whensteam injection is not in use and to minimizeexposure to stress corrosion cracking.
The injection system in the turbinecompartment bypassed all existing piping, waseasy to install, and ensured that any steamcondensate leaking past the strainer duringengine shut down would drain out the bottom ofthe manifold through a thermostatic trap. The10 steam injection nozzles were designed sothat any possible water entering during upsetconditions would spray along the inside of thecombustor shell and away from the transitionswhere thermal shock damage could occur. Thenozzles were located between transitions toensure good mixing of steam and compressordelivery air flowing to the liners.
The diameter of the holes in the nozzleswas sized to pass a maximum of 30 t/h (perunit) with the Masoneilan control valve fullyopen, the gas turbine IGV's fully open, thegas turbine on exhaust temperature control,and the compressor inlet temperature at -6°C.
Existing Main Steam Header (30 Oar, 300°C Steam)
Steam Injection System
300 MM Heads, Eatenaian IMet L
rinleelionI
^NOxz lea
P Tso M.
Duplicate MM MM I 75 MMSystemFo, 2nd
unit
I L InjectionT
To V n' rNouleaStack IMM L
SYMBOLS: IrPlant4 Piping F^
Conirel Room Op.rated SINat-ON Valvefff^^^
13 MM I
Automatic central valveI I".,e°
T."....T
Hand Vetoes_D IC_IEneloau ra
Varbt Meter
LII Steiner
❑T TMrmoetetic Trop
FIGURE 3 STEAM INJECTION SYSTEM SCHEMATIC.
Downloaded From: https://proceedings.asmedigitalcollection.asme.org/ on 01/29/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
ASSEMBLED AS FABRICATED
t
I
WITH BAROMETER = 984 MBARIGV's=OPEN
_ TEGEARBOXEXHAUST TEMP. CONTROL LIMIT
°C,30B.EFF. /^^STEAM=300,'^^X
\\x x\ \5t5^ j
/
\\ / \ xi o
s Q'o
MAXIMUM —INJECTIONINJECTION RATE 0 y(5% OF COMPRESSORMASS FLOW) y^ y °'S^^tQ
FIGURE 5 GAS TURBINE POWER WITH STEAM INJECTION.
GEARBOXWITH: BAROMETER - 984 MBAR / \ / LIMIT
IGV's = OPEN / xEXHAUST TEMP. CONTROL'-..../STEAM =300"C, 30B.EFF
/ x
T ^°
SrFgM ^^5MAXIMUMINJECTION
RATE /T'yi o as
FIGURE 6 GAS TURBINE FUEL CONSUMPTION WITH STEAMINJECTION.
'0
28
9 26
W
24
Wz
22I-FU,
c5 20
I= 85I-
00
0 80
as
0 75UJW
LL 70Wzmtr- 65
cal
6C
FIGURE 4 STEAM INJECTION NOZZLE.
Thirty t/h provided 40% extra capacity overthe nominal 5% of compressor inlet flow limit.The nozzles thread into the combustor shell asshown in Figure 4, and are restrained againstrotation by the locking wire and pins.
The integrity of the combustor shell withten 2-inch NPT holes was verified as follows:
A395 ductile cast iron used by G.E. inthe combustor shell can be used between -30°Cand 350°C and can contain a pressure of up to6.9 MPa. The highest predicted compressordelivery temperature of 349°C will occur at35°C compressor inlet temperature with 30 t/hof injection steam. But, because the shell isnot insulated, the metal will operateconsiderably below 349°C due to enginecompartment ventilation, thus overtemperatureof the material is not a problem. Internalpressure is obviously well below 6.9 MPa withor without steam injection.
The maximum allowable stress for A395 is82.8 MPa according to ASME Section VIII, TableUCD-23 (ASME BOILER AND PRESSURE VESSEL CODE;PART UCD - REQUIREMENTS FOR PRESSURE VESSELSCONSTRUCTED OF CAST DUCTILE IRON). Using themaximum internal pressure seen by thecombustor shell at -10°C compressor inlettemperature and 30 MW gas turbine power, theshell thickness required for 82.8 MPaallowable stress is 12.6 mm (in the plane ofthe steam injection holes).
According to ASME Section VIII, UG-37(REINFORCEMENT REQUIRED FOR OPENINGS IN SHELLSAND FORMED HEADS), the remaining shellthickness (19-12.6 =6.4 mm) acts as areinforcement of each hole. In addition,according to PART UG-43 (METHODS OF ATTACHMENTOF PIPE AND NOZZLE NECK TO VESSEL WALL), thesteam injection nozzle attached to the shellwith ANSI B2.1 full penetration threadprovides additional hole reinforcement.
Using the code guidelines presented inUG-37, Figure UG-37.1, the total holereinforcement area contributed by the excessshell thickness and the nozzle walls is 2.5times the minimum code requirement. Thus, thecombustor shell modifications more thansatisfy the requirements of ASME SECTION VIII,ensuring the integrity of the combustor shellwith steam injection.
The predicted performance of the turbineswith steam injection was refined using animproved boiler model and modifying gasturbine component performances to duplicatedata taken on each engine in March, 1986.
The additional power resulting from steaminjection with the gas turbine operating onexhaust temperature control is shown in Figure5. The existing MK I exhaust control curvewas not changed as calculations showed thatthe turbine inlet temperature would stay at,or drop slightly below current levels as steamwas introduced. The suggested maximuminjection rate, as well as the system capacityof 30 t/h are both included. At 0°C,increases of 4.1 MW (17%) at maximum injectionrate and 5.5 MW (23%) at system capacity wereexpected. The increase in fuel consumptionnecessary to reach exhaust temperature controlwith steam injection is shown in Figure 6. Atsystem capacity, up to 9% extra fuel would beburned. Because extra exhaust flow passesthrough the boiler, up to 7% extra steam wouldbe produced with maximum steam injection, asshown in Figure 7, thus partially offsettinginjection steam consumption.
Downloaded From: https://proceedings.asmedigitalcollection.asme.org/ on 01/29/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
-?SrgjEN\ j - \
MAXIMUMINJECTION
RATE t.A
^Z
o
NO BOILER FIRINGINJECTION STEAM=300°CBOILER STEAM 307'C
30 B.EFF4.0 T/H OF 7B FLASH
STEAM PRODUCTION
_ 53I
I—
Z 52
D0 51ma
. 50I-InC 49
w 48J
°m47
WITH: BAROMETER =984MBARIGV's • OPENEXHAUST TEMP CONTROL
FIGURE 7 BOILER 30BAR STEAM PRODUCTION WITH STEAMINJECTION.
WITH: BAROMETER 948 MBARIGV's =OPENEXHAUST TEMP. CONTROL
- NO BOILER FIRINGINJECTION STEAM=300°C,30EEFFBOILER STEAM=307°C,30B.EFF.,,\4.0 T/H OF 7B FLASH /STEAM PRODUCTION
The net effect of increased fuelconsumption, greater gas turbine power, andadditional boiler steam production is shown bythe cogeneration fuel conversion efficiency inFigure 8, where cogeneration fuel efficiencyis defined as:
17=! MW generated + 30 bar steam flow x available energy
+ 7 bar steam flow x available energy
(gas turbine + boiler)fuel flow x LHVwhere:available energy=enthalpy at steam conditions
-enthalpy of feed water
It can be seen that as steam is injected,cogeneration efficiency increases by up to 2.5points (at 0°C) at system capacity. The reasonfor the increase is that the excess steambeing injected into the turbine is consideredas "free", seeing it would have been vented toatmosphere if not injected.
INSTALLATION AND TEST
For the first engine, a Liburdi engineerwas present to oversee the installation, by anexperienced Aisthom crew, of the Liburdisupplied components in the engine compartment.The installation proceeded between June 24 andJuly 4, 1986 as follows:1. remove turbine cover.2. disassemble all piping in the vicinity of
the combustor shell.3. using air gun and custom fixturing, drill
pilot hole followed by tap hole throughcombustor shell wall.
4. use hand operated tapered reamer to formtapered shape for NPT thread.
5. hand tap hole using progressively sizedtaps.
6. check penetration with gauge plug andadjust depth of tap according.
7. repeat steps 3 to 6 for each of theremaining 9 holes.
8. remove combustors and transitions frombottom of unit and clean out the metaldebris left from the drilling and tappingoperations.
9. install steam injection nozzles.10. re-assemble all standard piping.11. assemble steam injection manifolds and
support from exhaust plenum.12. bolt steam injection flexibles between
nozzles and manifolds.13. replace engine cover.
The installation demonstrated that thesystem could be retrofitted without liftingthe combustor shell.
The manifolds' inlet flanges were blankedoff and the turbine was put back into oper-ation until completion of the plant piping.
Rhone-Poulenc fabricated and installedthe plant piping during September and October,1986 and the same Alsthom crew repeated thenozzle installation on the second unit ontheir own in October.
The entire system was tested during thelast 2 weeks of October, 1986. The tests wereconducted by obtaining baseline performancedata from plant instrumentation and thengradually (in 5-tonne steps) increasing thequantity of steam being injected and recordingall performance data at each increment. Figure9 shows the measured increase in power withsteam injection for both units with IGV's open(80 0 ) and closed (60 0 ). The dashed line showsthe predicted power increase with IGV's at80 0 . The measured increase fell within 0.4 MWof the predicted increase at maximum steamflow, indicating a slight discrepancy in theperformance model.
As no compressor surge line informationwas available, the units were ramped slowlypast the suggested maximum steam flow of 5% ofcompressor inlet flow while instrumentationwas being monitored in the engine controlcompartment. There was never any indicationof instability or incipient surging as thesteam throttle valve was fully opened.
80
78ILw
74
wu- 72
s
70wzw
v ^
Downloaded From: https://proceedings.asmedigitalcollection.asme.org/ on 01/29/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
LEGENDOGN502, VIGV=80°,&GN502, VIGV=60°❑ GN50I, VIGV=80° }VGN501, VIGV=60° F-----PREDICTED MAXIMUM
AT 10°C COMP. RECOMMEN INLET TEMP. STEAM FLO / Q
^ VIGV=80° 60
WITH: BAROME ER =984 MBAREXHAUS TEMP. CONTROL
QW
7/
X
/7 _
1
FIGURE 9 MEASURED GAS TURBINE POWER INCREASE WITHSTEAM INJECTION.
W
0
z
w
W
UZ
00 5 10 15 20 25 28.5STEAM INJECTION (T/H)
(vWITH: IGV •80•
E CHAUST T. CONTRO
60OGN502❑ GN50I
Sc?
40
3
O20 -
0
10
O0 10 20 31
STEAM INJECTION (T/H)(a)
FIGURE 10 EXHAUST TEMPERATURE SPREAD WITH STEAMINJECTION.
U
IiicaU)WIrI-F-9Wa2WI—
DSxW
60
Q 50WWaU)W 40
WW 30n.2W
20U)
I 10W
m
O0 10 20 30STEAM INJECTION (T/H)
(b)
V GN 501, VIGV • 60• +EXH.T. CONTROLGN502,VIGV•60°EXH. T. CONTROGN 502,VIGV-100•<EXH.T. CO TROL
+GN502.VIGV•MW • CONSTAN
t t
It was found, however, that the maximumflow rate through the steam injection systemwas 28.5 t/h rather than the design 30.0 t/h.The 5% shortfall was likely due to differencesin the final vs design plant piping, and aslightly optimistic assumption of the nozzle'sflow coefficient.
It was interesting to note that for thesame quantity of injection steam, there was a16% greater increase in power with the IGV'sclosed than with the IGV's open. The reasonis that the same quantity of steam expandingthrough the turbine in either case generatesabout the same power. However, the extrapower needed to boost the compressor flow tothe higher pressure ratio with steam flowingthrough the turbine varies with compressorflow. The compressor flow with IGV's closedis 84% of that with IGV's open, thus lesspower is absorbed leaving more of the powerfrom steam expansion to drive the generator.
Examination of the 18 exhaust thermo-couple readings did uncover an importantrelationship. In Figure 10(a), the exhausttemperature spread was between 20°C and 30°Cfor all steam flows when both units had IGV'sopen and were on exhaust temperature control.However, Figure 10(b) shows that past 25 t/hthe exhaust temperature spread jumped to the70°C level. The tests included all runs withIGV's closed, plus part load with the IGV'sopen.
Plotting the 18 readings revealed thatthe increased spread was due to a lowtemperature 'hole' developing in the lowerportion of the exhaust manifold (up exhaust).It was felt at first that this could have beendue to exhaust diffuser inlet swirl variation,but turbine aerodynamic analysis showed thediffuser inlet angle to vary only slightlywith increasing steam flow.
Combustor mass flow was plotted againstair/fuel ratio for the tests of Figure 10 andit was found in some runs that high spreadoccurred as the combustor moved into a morestable region of the stability plot, thereforeincipient blowout was discarded as a possibleexplanation.
Finally, all significant gas turbinemeasurements were plotted against steam flowfor Figure 10's runs, and all parametersexcept fuel flow changed smoothly as steam wasincreased. In those cases where the exhausttemperature spread increased past 25 t/h ofsteam, the fuel flow also took a sharp jumpcorresponding to a 2.3% to 2.7% reduction incombustion efficiency. It is felt that theincreased spread and fuel consumption bothoccurred when the combustor flame was cooledbelow the level required for efficient com-bustion (Note: no changes in the combustionsystem had been made to cater for steaminjection). The drop in efficiency occurredin all cases when the steam/fuel weight ratioexceeded 4.6:1.
A steam flow limit of 25 t/h was instit-uted for all IGV closed, and all IGV open,part-load operation to ensure efficientcombustion under all operating conditions.
Downloaded From: https://proceedings.asmedigitalcollection.asme.org/ on 01/29/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
COMPRESSOR INLET T.=15°C F IRINGRINGEXHAUST_T MP. CONTROL
NO BOILER FIRING _ —
VIGV =60° ❑
0O
0 5 10 15 20 25 28.5STEAM INJECTION (T/H)
FIGURE 12 MEASURED COGENERATION FUEL CONVERSIONEFFICIENCY WITH STEAM INJECTION.
>-I.--j00
W
W
X
8E
UZ
84WILWO 82
N
> 8CZ0U
w 7EDU_ZO 7EQitW
w 74
0U
The measured increase in fuel consumptionwith steam injection was within 4% of thepredicted for IGV's open, as shown in Figure11. The increase was much larger with theIGV's closed but it should be remembered thatthere was also more of an increase in powerwith IGV's closed. In fact, at the maximumrecommended steam flows for IGV's closed andopen, the percentage increase in fuelconsumption for each 1.0% increase in powerwas identical at 0.42%.
Combined with the measured boiler steamflow, the power and fuel data gave thecogeneration fuel conversion efficienciesshown in Figure 12. The efficiency increased
OGN502, VIGV =80°= 1200AGN502,VIGV=60°
I-M ❑ GN50I, VIGV=80°rVGN50I, VIGV=60°Z ----PREDICTED AT
IOCOMP. INLETma°C-
Z 1000 TEMP.,VIGV =80 0 1Q MAXIM
/RECOMMENDED 3STEAM FLOW
800D WITH: BAROMETER =984 MBARZ EXHAUST TEMP. CONTROL ' g
^i WJ 600 1-
W
? 400 XX
cr 200
fWi Q
UZ
5 10 15 20 25 28.5STEAM INJECTION (T/H)
FIGURE II MEASURED GAS TURBINE FUEL CONSUMPTIONINCREASE WITH STEAM INJECTION.
an average 3.5 points (4.6%) with maximumsteam injection. This increase is about 1.0point more than was predicted in Figure 8 andwas primarily due to more boiler steamproduction than had been expected. Withmaximum boiler firing, maximum steam injectionand IGV's closed, an efficiency of 84.9% wasmeasured.
The steam injection ramp rate was variedduring testing and for the quantities of steaminvolved (up to 7.5% of compressor inlet massflow), it was found that steam could be addedas fast as the update interval of the steamvortex meter allowed.
OPERATION
A number of general guidelines for systemoperation were suggested to Rhone-Poulenc;namely,1, inject steam only when gas turbine output
exceeds 5 MW.2. the most efficient plant operation is to
inject all available excess steam equallyinto both units and allow the existinggas turbine control systems to vary fuelflow and IGV position on both units inparallel, in response to generationdemand.
3. during periods when no excess steam isavailable, keep the carbon steel supplylines hot by allowing a very small flowof steam through the system. Tiny pilotholes in the control valve and theautomatic shut-off valves will allow asmall leakage of condensate which willdrain from the strainer periodicallythrough the thermostatic trap.
During the winter of 1986 - 1987, the steaminjection system operated to the satisfactionof Rhone-Poulenc and ensured that excess steamwas never vented to atmosphere. Economicanalysis of the data confirmed the expectedfast payback.
The beneficial effects of steam injectionon NO,, emissions were checked by Rhone-Poulencand the NO,, corrected to 15% 02 in the exhaustwas found to drop from 83 ppm without steam to12 ppm with 20 t/h of steam. When plotted asNO,, ratio vs steam/fuel weight ratio (Figure13), the unexpectedly strong NO,, reductioncharacteristic of this steam injection systemis realized. The dotted line shows thepredicted NO,, reduction when EPA correctionmethods for combustor inlet humidity,temperature and pressure, and %O 2 in theexhaust are considered. The dashed line showsLM5000 test data published in ASME paper 86-GT-231 (Ref.1) for steam injected in thenozzle for the purpose of NO. reduction.Because the prediction indicates the effect ofincreased humidity air reaching the liners andthe LM5000 data indicates the effect ofprimary zone flame temperature reduction, itcan be concluded that a large portion of thesteam injected in the Rhone-Poulenc testreached the primary zone without beingingested into the secondary and dilutionregions of the liners (probably due to theshape and location of the steam injectionnozzles). Thus it appears as if steam
Downloaded From: https://proceedings.asmedigitalcollection.asme.org/ on 01/29/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
W
0.9
Z 20.8OH
0-)Ld ? 0.7z
Wv~i0.6HH
0
I- HO.t3 3x x00
ZZ0.4
0
Q 0.?C:
0Z 0.,
— RHONE-POULENC TEST
1 ^^ I — —RHONE-POULENC PREDICTIONBASED ON EPA METHODEXTRAPOLATION
1 ----LM5000 TEST DATA (REF. I)
^^1
\\ \
CONCLUSIONS
Power augmentation using steam injectionwas successfully implemented on two Frame 5gas turbines in cogenerative service. Themeasured power, fuel, and efficiency changesresulting from steam injection closely matchedthe predicted, and up to a 5.5 MW (29.6%)power increase was possible when injectingsteam equivalent to 7.5% of compressor inletflow. With steam injection, NO, emissionswere reduced to 14% of the pre-injectionlevels. Steam injection was found to have noadverse impact on the hot gas path blading.
REFERENCE
1. Burnham, J.B., Giuliani, M.H., Moeller,D.J., "Development, Installation andOperating Results of a Steam InjectionSystem (STIGTM) in a General ElectricLM5000 Gas Generator", ASME Paper No.86-GT-231, June, 1986.
0.5 1.0 1.5 2.0 2.5 3.0
STEAM/FUEL WEIGHT RATIO
FIGURE 13 EFFECT OF STEAM INJECTION ON NOX EMISSIONS.
injection of the Frame '5' for NO, control upto the previously recommended limit of 4.6:1steam/fuel weight ratio is possible, as longas the additional power can be utilized.
As part of a maintenance program, one ofthe units was opened in mid-1987 with thefollowing findings:1. traces of impact on the turbine buckets
which were thought to have come fromdrilling chips not completely cleanedout of the combustor shell during nozzleinstallation. This finding suggests thateither more care be taken to clean outthe combustor shell after drilling, orthe drilling should take place on a dis-assembled casing.
2. no vane or bucket erosion or cracking dueto the introduction of steam. The steamwas taken from the process steam with nospecial treatment. Natural gas (91%methane, 7% ethane by volume) was thefuel, but the concentration of Na andother contaminants in either the steam orthe fuel gas were unknown.
3. the penetration of each nozzle into thecombustor shell casing was not identical.On some, the jets impinged on the edge ofthe hole drilled through the shell. Thenozzles were adjusted to achieve uniformpenetration.
Downloaded From: https://proceedings.asmedigitalcollection.asme.org/ on 01/29/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
