steam turbine uprates - concordia universityusers.encs.concordia.ca/~kadem/ge_steam_turbine.pdf ·...

GE Power Systems

Steam TurbineUprates

J. F. LesiukGE Power SystemsAtlanta, GA

GER-4199

g

Contents

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1Dense Pack Section Replacement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

High Efficiency Steam Path . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3Sustained Heat Rate and Solid Particle Erosion (SPE) Resistance. . . . . . . . . . . . . . . . . . . . . . . 7

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Steam Turbine Uprates

GE Power Systems � GER-4199 � (10/00) i


GE Power Systems � GER-4199 � (10/00) ii

SummaryThe thermodynamic performance of a steamturbine is primarily determined by the steampath components. Because the efficiency of theentire power plant cycle is largely dependent onthe efficiency of the energy conversion in theturbine, it is important to minimize aerody-namic and steam leakage losses in the steampath.

Nozzle and bucket aerodynamic-profile losses,secondary-flow losses, and leakage lossesaccount for roughly 80-to-90 percent of the totalstage losses. In order to ensure high-efficiencyturbine designs without sacrificing turbine reli-ability, it is necessary to use both highly-efficientnozzle and bucket designs to minimize profileand secondary losses, and advanced clearancecontrols to minimize leakage flows.

GE’s development effort has been underway formore than a decade in order to reduce theselosses. The objective of this long-term programis to develop specific design features for bothnew turbines and retrofits of existing units thatmaximize overall turbine efficiency and main-tain a high degree of reliability and cost effec-tiveness. The development program has includ-ed activities conducted for all of the steam tur-bine product lines in cooperation with othercompany components such as Aircraft Engines,Gas Turbine, and Corporate Research &Development. The result of this developmenteffort was the introduction of the AdvancedDesign Steam Path, ADSP, in 1995.

Since 1995, GE has installed over 40 first- andsecond-generation ADSP retrofits on a variety ofunits. Field test data from these units has indi-cated steam path efficiency improvements rang-ing from 1.5 percent to 3.0 percent. Along withthis, GE has been developing new NPI features,such as enhanced advanced-vortex designs,integral cover buckets, and brush seals. These

features achieve additional improvement in effi-ciency.

As a natural progression from the ADSP experi-ence, GE introduced the “Dense Pack” redesignapproach to achieve additional section efficien-cy improvement.

BackgroundDuring the mid-1980s and early 1990s the prin-ciple reason for making steam turbine upratedecisions was the aging of power plants, (seeFigure 1), and the attendant poorer reliability ofolder equipment, as shown in Figure 2.

Utilities evaluated their aging fleet and con-cluded it was more economical to extend thelife of the equipment than to retire the unit.This resulted in turbine life extension evalua-tions to identify equipment that needed repairor replacement in order to operate beyond thegenerally-accepted design life of 40 years.

Talk of deregulation of the power industry dur-ing the mid-1990s created considerable uncer-tainty among the utilities. Decisions regardingthe sale of generating and/or transmission anddistribution assets were analyzed. One clearconclusion from the deregulation discussions


GE Power Systems � GER-4199 � (10/00) 1

1970

5

0

10

15

20

25

30

35

1995 2000 1990 1985 1980 1975

Year

Per

cen

t o

f C

apac

ity

Figure 1. Utility dependence upon units greaterthan 30 years old

was that the utilities were going to be in anextremely competitive environment in thefuture.

The initial reaction to deregulation within theutility industry was to take a “wait and see” posi-tion. As decisions regarding the retention orselling of assets were made, new drivers in thecost model developed. Evaluations for replace-ment power during peak periods soared and itbecame clear that additional capacity andimproved efficiency were the principle objec-tives for the future. The low cost producer of

electricity with additional capacity would be thevictor in the new, unregulated market.

Today, in addition to addressing the aging fleetissue, utilities are looking for the competitiveedge that additional capacity and better per-formance will provide. Other factors that enterinto the economic model are the desire for sus-tained performance with minimal degradationover period of at least ten years and the desireto extend time between major overhauls to atleast ten years. The balance of the paper dis-cusses the technologies and product solutionsdeveloped to address these utility needs.

Dense Pack Section ReplacementThe Dense Pack turbine section performance,as shown in Figure 3, is the latest evolution of GEsteam design that began in 1903 with a 5000 kWturbine. The design limits, operational experi-ence, and rugged dependability synonymouswith GE turbine design is not changed with aDense Pack.

The design goal of a Dense Pack retrofit is toput the most efficient steam path into an exist-



0 5 10 15 20 25 30 35 40 45 50

Age in Years

0

1

2

5

9

7

6

4

3

8

Fo

rced

Ou

tag

e

Rat

e

(%)

Figure 2. Typical aging reliability trend

Figure 3. Efficiency evolution of a single HP stage

98

97

96

95

94

93

92

91

90

89

881960’s

Free-vortexdesign

1970’sImproved

Vane profiles

1980’sStep RTSS,

SPE, CV stgs

1990’sAV1, RTSS

adv

Late 1990’sAV2, ICBs

opt. clearance

2000’sDense Pack

2000’sDense Pack

w/adv Sealing

HP

Ste

am P

ath

Effi

cien

cy -

%

ing high-pressure outer shell. The high efficien-cy steam path will produce a lower heat rate andincreased output for the same steam flow. Indesigning the most efficient steam path, twointeresting by-products also resulted. Thedesign parameters utilized to increase efficien-cy, such as bucket and nozzle solidity andreduced rotor diameters, also had the benefit ofreducing solid particle erosion. This led to asteam path that not only is more efficient butalso has a greater, sustained efficiency. Coupledwith the supply of a solid rotor (elimination ofthe rotor bore), the sustained efficiency resultsin a turbine section that should not requireinternal repair or inspection for ten or moreyears. These complementary, interacting bene-fits produce additional megawatts at a lower life-cycle cost.

High Efficiency Steam Path In the early 1990s, GE produced an AdvancedVortex bucket and diaphragm retrofit known asAdvanced Design Steam Path (ADSP), Figure 4.Although ADSP remains a cost-effective effi-ciency retrofit, the potential efficiency benefitof ADSP is limited because the retrofit is

installed onto the existing steam turbine rotorand inner shell. This limited the designers’flexibility because the number of turbine stagesand rotor diameters was fixed.

With more than 40 ADSP packages currently inoperation, the ADSP program provided a goodexperience base for the Dense Pack. ADSPincorporates some of the features to be includ-ed in Dense Pack such as steam flow manage-ment, optimized packing clearances, andadvanced shaft sealing. Pending deregulationand decreasing reserve margin have driven theutility marketplace to seek ever more aggressivemethods of lowering bus bar cost and increas-ing capacity. Armed with the experience gainedfrom the ADSP program, improved turbinedesign and modeling tools, shared technologyfrom GE’s aircraft engine and gas turbine prod-uct lines, and customer desire for steam pathenhancements, the Dense Pack team wasformed.

The Dense Pack team began by reviewing theturbine to determine where to focus designefforts to improve efficiency. Figure 5 is a Paretochart of unrecoverable turbine losses by turbine



Advanced aero buckets All stages, with advanced bucket tip sealing

Advanced aero partitions All diaphragms

Additional Diffusion coated 1st stage nozzle with SPEProfile Partition

Diamond Tuff™ HVOF coating (stage 1 buckets, 1st reheat diaphragm,1 st

reheat buckets)

Figure 4. Advanced design steam path

section. Unrecoverable losses are essentiallyinefficient use of the steam energy. A review ofthe Pareto chart shows that except for the heatlosses of the condenser, the HP and IP turbinesections have the greatest amount of irre-versibility.

The efficiency upgrade focused on HP and IPsections and attacked the tip leakage and steampath secondary and profile losses.

The Dense Pack team drew on cross-functionalGE engineering resources from steam and gasturbine design, Aircraft Engines, and CorporateResearch and Development (CRD). Theirobjective was to design the most efficient steampath that could be installed into an existingouter shell that included incorporating addi-tional staging into the turbine section. Hence,the name “Dense Pack,” for additional stagingin the same axial spacing.

Given the latitude of supplying a new steam tur-bine rotor and inner shell, the team capitalizedon adding staging and reducing the rotor stagediameter. Both of these concepts result in aninherently more efficient steam path. By addingstages, the energy may be extracted in smaller,

more controlled, and more efficient incre-ments. Reducing the stage root diameter resultsin increased bucket and nozzle partitionheights. Secondary losses decrease as the parti-tion height increases. In addition, reducing therotor diameter results in a reduction in shaftsealing leakage area, further improving effi-ciency. The smaller steam path diameter alsoresults in lower steam velocity and reduces theeffect of any solid particle carryover.

The steam path design incorporates reducedsolidity concepts or less buckets and nozzles perrow. The reduction in steam velocity coupledwith fewer partitions results in lower profile loss-es per stage. Typical efficiency gains vs. sourcesof irreversible losses for a conventional andDense Pack steam path are shown on Figure 6.

The first step in the steam path redesign is toestablish the mechanical constraints. The min-imum rotor diameters are determined throughan analysis of the torsional stress requirementsand a rotor dynamics investigation including arigorous review of rotor stability. Once the min-imum rotor packing diameters are determined,the minimum steam path diameter can beestablished from the known radial height



0.0

0.5

1.0

1.5

2.0

2.5

3.0

Hea

t r

ejec

tio

n

HP

tu

rbin

e

con

den

sor

loss

es

IP t

urb

ine

LP

tu

rbin

e

fee

dw

ater

hea

ters

LP

exh

aust

reh

eat

cycl

e

valv

es

cro

ss o

ver

LP

inle

t

HP

exh

aust

IP e

xhau

st

feed

pu

mp

s

HP

inle

t

IP in

let

pip

ing

loss

es

irre

vers

ibili

ty, %

inle

t av

aila

bili

ty

(48.9%)Typical 700 MW unit

HP sectionirreversibility

Figure 5. Efficiency loss by system

requirements for the various bucket attach-ments available to the turbine designer. Thisallows the steam path optimization to proceed.

GE has always utilized some reaction levels inthe design of high-pressure steam turbine buck-ets. Building on ADSP experience, the DensePack increases the bucket reaction to approxi-mately 20-25 percent. Modern design and mod-eling tools allow the computational iterationsnecessary to individually optimize both the

stage number and bucket reaction levels foreach steam path. Figure 7 shows a typical opti-mization contour plot of stage count vs. stageroot reaction. Additional parameters are com-pared in order to arrive at the optimum valuefor the key parameters that define the DensePack steam path.

Figure 8 shows the results of the Dense Pack opti-mization approach as applied to the redesign ofa single-flow, high-pressure section. The base-



valv

es

inle

t /b

rid

ge

C.S

. N

ozz

le P

rofi

le

C.S

. N

ozz

le S

eco

nd

ary

C.S

. N

ozz

le T

rail

ing

Ed

ge

C.S

. B

uck

et P

rofi

le

C.S

. B

uck

et S

eco

nd

ary

C.S

. B

uck

et T

rail

ing

Ed

ge

C.S

. T

ip L

eaka

ge

C.S

. R

oo

t L

eaka

ge

C.S

. P

acki

ng

Lea

kag

e

C.S

. W

ind

age

CS

. N

otc

h B

lock

C.S

. E

nd

turn

aro

un

d d

uct

No

zzle

Pro

file

No

zzle

Sec

on

dar

y

No

zzle

Tra

ilin

g E

dg

e

Bu

cket

Pro

file

Bu

cket

Sec

on

dar

y

Bu

cket

Tra

ilin

g E

dg

e

Tip

Lea

kag

e

Ro

ot

Lea

kag

e

Pac

kin

g L

eaka

ge

Win

dag

e L

oss

No

tch

Blo

ck L

oss

exh

aust

leak

age

and

mix

ing

Sp

ecif

ic Ir

reve

rsib

ility

, BT

U/lb

m baselinedense pack

Figure 6. Typical loss mechanism comparison

5 6 7 8 9 10 11

5

10

15

20

25

30

Number of Stages

Ro

ot

Rea

ctio

n

Increasing Eff’y

Figure 7. Optimization contour plot

line design employed a 6-stage section with adouble-flow first stage. By lowering the steampath diameters and increasing the stage reac-tion, a more efficient 14-stage steam pathresults. The major components affected by thisredesign include replacement of the bucketedrotor, diaphragms, inner shell, and end packingheads. In addition, the original double-flow,first-stage nozzle box is replaced with either asingle-flow nozzle box, a nozzle plate, or a noz-

zle diaphragm, depending on unit size andnumber of control-stage admissions.

Figure 9 is a photograph of an 11-stage, single-flow high-pressure rotor that was redesignedfrom a baseline configuration of 7 stages includ-ing a double-flow first stage.

Once the high efficiency steam path is funda-mentally designed for the number of stages, hassolidity and has sufficient rotor root diameter,



Figure 9. Dense Pack redesign of a single-flow high-pressure rotor

BaselineDouble flow

inlet

6 stages

Dense PackSingle flow

inlet

More stagesLonger bucketsIntegral covers

Smaller rotor and wheel diameters

Adv seals

Single flow control stage

Figure 8. Dense Pack redesign of a single-flow high-pressure section

BaselineDouble flow

inlet

6 stages

then design engineering adds sealing featuressuch as optimized packing clearances, integralcovered buckets, and advanced tip sealing (SeeFigure 10). Bimetallic seal rings complete theoffering and maximize steam efficiency.

The Dense Pack design approach has been ver-ified through a rigorous Six Sigma Design ofExperiment process in GE’s Steam Turbine TestVehicle (STTV) located in Lynn, Massachusetts.This state-of-the-art test facility was completedin 1999 and the first test was completed in Aprilof that year. The STTV is a fully-instrumented,multi-stage test turbine using high-pressuresteam. The initial test in April of 1999 was toestablish a baseline representing the installedfleet designs. This was followed by a series oftests using the Dense Pack design approachapplied to the same steam conditions. Figure11shows the test facility and Figure 12 shows thetest rotors for the baseline and one of the DensePack tests.

Sustained Heat Rate and Solid ParticleErosion (SPE) Resistance Another GE design objective with the DensePack redesign was to supply a steam path thatwould be resistant to SPE degradation. In the

1980s GE performed extensive computer mod-eling and analysis of solid particle trajectories inthe steam path. The comprehensive develop-ment effort resulted in the following:

� A fundamental understanding of theerosion mechanism in the steamturbine steam path.

� Development of plasma spray anddiffusion coatings that significantlyenhanced the SPE resistance of thesteam path.

� A combination of design changesidentified by the trajectory analysisand SPE-resistant coatings that led tominimizing particle carryover damage.



Integral Covered Buckets Integral Covered Buckets w/Adv Sealin g

Dense Pack Diaphragms

Figure 10. Dense Pack features

Dense Pack Diaphragms

Integral Covered Buckets Integral Covered Buckets w/Adv Sealing



Figure 11. Steam Turbine Test Vehicle

Figure 12. Baseline and Dense Pack test rotors

These design changes included increased axialclearance between the stationary nozzles androtating buckets for the reheat section,increased nozzle scale factor for all stages, andredesigned nozzle profiles that allow smootherand less damaging passage of solid particlesthrough the steam path.

The study resulted in the redesign of the high-pressure nozzle partitions to distribute the solidparticle impact over a broader area of the parti-tion. Customer experience indicated a 75%reduction in the degradation attributed to solid

particle erosion. Clearly, the model and the GEanalysis were on target. The high-efficiencysteam path of the Dense Pack will virtually elim-inate solid particle erosion. Figure 13 depicts thesolid particle erosion of a conventional steampath nozzle partition. The severity of the ero-sion varies from zero, shown in blue, to thehighest shown in red.

Figure 14 shows the benefits of the reduced noz-zle count, lower solidity, and the redesignednozzle partition. These features (lower nozzleand bucket solidity, redesigned partitions,



reduced steam velocities, and solid coatings)yield a steam path that is more resistant to solidparticle carryover. This will allow units to beoperating for longer periods between majoroverhauls because the rate of performancedegradation will be significantly reduced, if notvirtually eliminated.

ConclusionsThe Dense Pack section replacement is the lat-est option in GE’s long history of steam pathefficiency improvements. Incorporating tech-nologies from gas turbine, Aircraft Engines,and Corporate Research and Development, the

Figure 13. Solid particle erosion on conventional steam path

Figure 14. Solid particle erosion on Dense Pack steam path



Dense Pack provides an alternative to theAdvanced Design Steam Path design. The DensePack alternative supplies the user with aredesigned steam path including a new bucketedrotor, diaphragms, and inner shell. The increasein output and the reduction in heat rate addressthe two major competitive issues facing the utili-ty industry today.

The inherent features of Dense Pack, including alower solidity design steam path, and fewer noz-zles and buckets per row, combined with GE’s

proven solid particle erosion protection features,address a utility concern of sustained perform-ance. The resultant damage to the steam pathfrom particle carryover is virtually eliminated,enabling utilities to extend the time betweenmajor overhauls and reduce life-cycle costs.

AcknowledgementThe author wishes to acknowledge the contribu-tions of the Dense Pack team led by JamesMaughan.

References1. J.I. Cofer, IV, J.K. Reinker, W.J. Sumner, “Advances in Steam Path Technology,” GE Power Systems

Paper, GER-3713.

2. R.S. Couchman, K.E. Robbins, P. Schofield, “GE Steam Turbine Design Philosophy andTechnology Programs,” GE Power Systems Paper, GER-3705.

3. M.F. O’Conner, “Upgrade Opportunities for Steam Turbines,” GE Power Systems Paper, GER-3696.

List of FiguresFigure 1. Utility dependence upon units greater than 30 years old

Figure 2. Typical aging reliability trent

Figure 3. Efficiency evolution of a single HP stage

Figure 4. Advanced design steam path

Figure 5. Efficiency loss by system

Figure 6. Typical loss mechanism comparison

Figure 7. Optimization contour plot

Figure 8. Dense Pack redesign of a single-flow high-pressure section

Figure 9. Dense Pack redesign of a single-flow high-pressure rotor

Figure 10. Dense Pack features

Figure 11. Steam Turbine Test Vehicle

Figure 12. Baseline and Dense Pack test rotors

Figure 13. Solid particle erosion on conventional steam path

Figure 14. Solid particle erosion on Dense Pack steam path



steam turbine uprates - concordia universityusers.encs.concordia.ca/~kadem/ge_steam_turbine.pdf ·...

Documents