ABSTRACT Modern GE steam turbine designs for electrical

power generation are the result of more than 90years of engineering development. The productline of fossil-fueled, reheat steam turbines for both50Hz and 60Hz applications extends from 125-1100MW and is based on a design philosophy and com-mon characteristic features that ensure high relia-bility, sustained high operating efficency and caseof maintainance. This paper identifies GE’s currentproduct line for 50 and 60 Hz applications includ-ing High Power Density Designs™ incorporatingadvanced steam path design, installation and main-tainence features which continue to make GE theprefered choice for power generation equipment.

INTRODUCTION Modern GE steam turbines for electrical power

generation are the result of more than 90 years ofengineering development. The first GE productionturbine was rated 500 kW and went into operationin 1901. Just two years later, a unit rated 10 timeslarger was placed in service at Commonwealth Edi-son’s Fisk Street Station. Advances in the technolo-gy have continued since that time, and today a fullproduct line is offered for both 50 Hz and 60 Hzapplications, with ratings from 100 to over 1300MW for fossil-fueled, reheat cycles, and from 600 toover 1500 MW for nuclear applications. Through-out the range of sizes and applications, GE steamturbines reflect a consistent philosophy of designand include common characteristic features thatensure high reliability, sustained high efficiency andease of maintenance. This paper will describe theproduct line of GE steam turbines for electricpower production utilizing steam from fossil-fuelfired boilers and nuclear reactors.

RATINGS ANDCONFIGURATIONS

Fossil UnitsFossil utility steam turbines are that class of

large, reheat units used almost exclusively for elec-tric power generation. Because of their large size,usually greater than 200 MW, these units utilize

higher steam pressures (2400 psi/165 bar) andhigher), have several stages of regenerative heatingand incorporate other design features to maximizeperformance, reliability and availability. GE’s cur-rent product offering of utility steam turbines isbased on the availability of new longer 50 and 60Hz last-stage buckets and other recent advancesmade in steam turbine technology.

Last-Stage BucketsHistorically, increases in steam turbine ratings

have been accompanied by longer last-stage bucketsin order to maintain an economical unit size.Longer last-stage buckets can accommodate largersteam flows and loadings at relatively the same per-formance level by maintaining or reducing exhaustlosses, without increasing the number of low pres-sure turbine flows. In the late 1960s, GE introducedits first continuously-coupled, last-stage bucket(LSB), the 60 Hz, 33.5-inch/851 mm LSB. Thisunique design utilized a cover, a supersonic tip sec-tion and other features resulting in the highest effi-ciency level of any LSB designed.

Its continuous coupling and loose constructionresulted in exceptional damping and unsurpassedreliability. Because of the unequaled performanceof the continuously-coupled design, GE underwenta redesign program to incorporate the features ofcontinuous coupling in its existing families of 50and 60 Hz LSBs.

With the successful implementation of theredesign program in the 1980s, development priori-ty was given to the evolution of longer LSB designs.Although unit ratings have stablilized, longer buck-ets would result in more compact, cost-effectiveunits. The longer LSBs include a 40-inch/1016mm, 60 Hz titanium LSB and a 42-inch/1067 mm,50 Hz LSB (Figure 1), which is a direct scale up ofthe modern 33.5-inch/851 mm, 60 Hz LSB. A 48-inch/1219 mm 50 Hz LSB scaled from the 40-inch/1016 mm design has also been introduced.

All fossil utility offerings, therefore, utililize theunique GE continuously-coupled, last-stage bucketdesigns.

High Power Density Designs™

The benefit of the increased annulus area associ-ated with the longer last-stage buckets is demon-

1

STEAM TURBINESFOR LARGE POWER APPLICATIONS

J.K. Reinker and P.B. MasonGE Power Systems

Schenectady, NY

GER-3646D

strated in Figure 2. Each High Power Densitydesign has the equivalent performance of the previ-ous design, but with the benefits of a more compactsteam turbine configuration. These benefitsinclude a compact, cost-effective station design,faster, easier maintenance because of the fewer cas-ings and components, and fewer spare parts tomaintain. Reliability and availability of these sim-

pler designs is, therefore, expected to exceed thatof previous designs.

Additionally, recent technologial advances such asimproved steam paths and rotor dynamics haveevolved resulting in futher improvments in unit config-urations. These advances are discussed in more detailin “Advances in Steam Path Technology” and “HighPower Density Steam Turbine Design Evolution.”

2

GER-3646D

GT24510

Figure 2. High power density–last-stage bucket impact

RDC27265-02-11

Figure 1. Double-flow, low-pressure rotor with 42-inch (1067 mm) last-stage buckets

GER-3646D

3

The complete family of High Power Density unitsfor single reheat application with conventionalsteam conditions is shown in Figure 3 and represen-tative cross sections are shown in Figures 4 -10. Eachturbine type is available for a range of ratings withthe selection of steam conditions and exhaust annu-lus area appropriate to the individual technical andeconomic conditions. The limit in rating shown foreach configuration is approximate and will dependon such variables as steam conditions, exhaust pres-sure, number of admissions, number and locationof extractions, and flow margin. All configurationsare available with the capability to operate continu-ously at 5% overpressure if specified.

For the smallest utility ratings, the two-casingunit is available with a single-flow exhaust as illus-trated in Figure 4. These designs utilize double-shell high pressure inlet construction, with therugged nozzle plate design and direct actuatedindividual control valves with Admission ModeSelection (AMS) for both full and partial arc admis-sion.

The next larger units are two-casing designscombining high and intermediate pressure sec-tions in a single casing and double-flow low pres-sure turbine sections. This design uses double-shell,nozzle plate construction with shell-mounted con-trol valves at the lower turbine ratings (Figure 5)and off-shell mounted control valves at the higher.For high temperature and pessure applications,the triple-shell nozzle box construction with a sepa-rately mounted stop/control valve chest is available(Figure 6).

For higher ratings, or for applications requiringadditional annulus area, a unit with two double-flow low pressure sections is available. The three-casing design is shown in Figure 7, and the higherpressure design is shown in Figure 8. The higheroutput four-flow designs utilize separate high andintermediate sections as shown in Figure 9.Depending on the needed admission require-ments, these units may have single or double-flowfirst-stage designs.

The highest rated single reheat units utilizingthree double-flow low pressure turbines, Figure 10,use separate high and intermediate pressure sec-tions and are available for units over 1200 MW. Forthese large ratings, cross compund units are alsoavailable if desired.

In addition to the designs shown, special designsto meet unusual conditions are also available. Forexample, sites with unusual heat rejection require-ments may require a design suitable for exhaustpressures up to 15 inches HgA/381 mm HgA. Spe-cially designed 50 and 60 Hz last-stage buckets areavailable for such applications.

Advanced Steam ConditionsGE continues to be the leader in the develop-

ment of high performance large steam turbines.In the late 1960s, GE introduced into service thefirst of several highly efficient, double reheat unitswith supercritical steam pressures and advancedsteam temperatures. Double reheat units utilize anopposed-flow, high pressure, first reheat section

GT22243B

Figure 3. Fossil turbine arrangements

Single Reheat

Double Reheat

4

GER-3646D

RDC27035

Figure 4. Two-casing, single-flow steam turbine

RDC27015-03

Figure 5. Two-casing, double-flow steam turbine with shell-mounted valves

Figure 6. Two-casing, double-flow steam turbine with off-shell valvesRDC270153

5

Figure 7. Three-casing, four-flow steam turbine

Figure 8. Three-casing, four-flow steam turbine

RDC27045

RDC27091

GER-3646D

Figure 9. Four-casing, four-flow steam turbineRDC27199

6

GER-3646D

and are available at steam pressures up to 4500psi/310 bar and ratings up to 900 MW in both four-(Figure 11) and six-flow low pressure configura-tions.

Highly efficient units with advanced steam con-ditions continue to be quoted on a selective basis.Reference 3 describes a 1000 MW class ultra super-critical unit for a Japanese customer with steamconditions in excess of 1100 F/593 C. Figure 12 is a750 MW unit four-casing design recently offeredwith 48-inch/1219 mm titanium LSBs. This unithas ultracritical steam conditions of 3860 psi and1071F/1112F (266 bar and 577C/600C) andincludes special materials and cooling arrange-ments to accommodate the higher steam condi-tions. Performance is further enhanced by the useof state-of-the-art steam path technology includingimproved leakage controls, advanced air foils,advanced hood design, and other features to maz-imize efficiency.

Nuclear While nuclear turbines are available for almost

any capacity rating, the licensing requirements fornuclear reactors and the pressures of economy ofscale have dictated applications almost exclusivelyat the larger ratings, utilizing four- and six-flowexhausts. However, by some projections futurereactor designs may be smaller, on the order of 600

MW, and double-flow designs using 52-inch/1321mm last-stage buckets will be suitable.

52-inch/1321 mm last-stage bucket designs areavailable for both 50 and 60 Hz applications. Thefirst application of the 50 Hz, 52-inch/1321 mmlast-stage bucket is a six-flow, 1356 MW unit pow-ered by GE’s Advanced Boiling Water Reactor(ABWR), which is currently in operation. The 60Hz design is being used in the Electric PowerResearch Institute/Department of Energy fundedFOAKE (First Of A Kind Engineering) ABWRplant.

Figure 13 shows the nuclear turbine configura-tions with moisture separator reheaters (MSRs)between the high-pressure and low-pressure sec-tions. Figure 14 is a cross section for a typical six-flow design.

As with fossil turbines, each nuclear turbine isdesigned to meet the individual utility requirementsin terms of rating, reactor steam conditions, andfeedwater temperature and cycle parameters, suchas steam reheating, reactor feedpump turbines, andfeedwater heater and drain arrangements. Nuclearsteam turbines are rated and designed for flow-pass-ing capability in the same manner as fossil turbines.However, they are designed to be suitable for thepart-load pressure characteristics of the particularreactor steam supply and are not usually designedfor a throttle pressure 5% above rated pressure atvalves-wide-open flow.

Figure 10. Five-casing, six-flow steam turbineRDC24265-7

Figure 11. Four-casing, four-flow, double-reheat steam turbineRDC24265-04

Figure 14. Six-flow nuclear steam turbineRDC24265-1

GER-3646D

7

Figure 12. Ultrasupercritical steam turbineGT24018A

Figure 13. Nuclear turbine arrangementsGT23242

• Advanced Steam Conditions- 3860PSI/1071F (266 bar/577C) Throttle- 1112F (600C) Reheat

• Advanced Airfoils• Improved Clearance

Control• Advanced Hoods • 48” Titanium Buckets

(1219 mm)

MAJOR DESIGN FEATURES GE steam turbines across the range of ratings

and applications have a number of consistent char-acteristic features. As designs are developed, thereare almost always conflicting considerations, andmajor design features result from many years ofexperience with many units in operation. The rea-son a particular feature is adopted over alternativedesign approaches is not always obvious.

Impulse Staging with Wheel-and-Diaphragm Construction

The single most important factor relating todesign features is the use of impulse stage design,which in turn leads to a construction known aswheel-and-diaphragm. This is in contrast to themajor alternative technology of reaction stagedesign with a drum-type rotor and related construc-tion features. GE developed the impulse designtechnology after joining forces in 1896, withCharles G. Curtis, who held basic patents.

In a pure impulse stage, the entire stage pressuredrop is converted into velocity in the fixed nozzles.There is no pressure drop across the moving buck-ets, which only impose a change in direction of thesteam and absorb energy by momentum exchange.In a reaction turbine, some portion of the stagepressure drop, typically 50%, takes place across themoving blades, increasing the velocity of the steamand imparting energy to the blades by reaction, aswell as momentum exchange. Peak efficiency isobtained in an impulse stage with more work perstage than in a reaction design (Figure 15), assum-ing the same diameter. It can be deduced from Fig-ure 15, that a reaction turbine design will requireeither twice as many stages or 40% greater stagediameters, or some combination thereof, for peakefficiency. GE turbines employ significantly lessreaction and have approximately 40% fewer stagesin the HP and IP sections than is typical of reactiondesigns. The contrast is less in the low-pressuresection where the long bucket length results in asignificant increase in velocity of the bucket fromthe root to the tip. An efficient design requires anincrease in the degree of reaction from the root tothe tip, and the low-pressure stage designs ofimpulse and reaction turbines tend to be similar.

In the GE stage design, the buckets are mountedon the periphery of wheels and the nozzle parti-tions are supported in diaphragms, as shown in Fig-ure 16. Because of the relatively large pressure dropthat exists across the moving blades in the reactiondesign, a very high thrust force would exist on therotor if the blades were mounted on wheels with

faces exposed to the pressure differential. A drum-type rotor, as shown in Figure 17, is used in reac-tion-type turbines to avoid excessive thrust.

The significant differences that are associatedwith these two basic constructions can be separatedinto those affecting efficiency and those affectingmechanical integrity.

Figure 16. Typical impulse stages, wheel-and-diaphragm construction

Figure 15. Ideal stage efficiency as a function ofvelocity ratio for impulse and reactionstage designs

8

RDC27218

Figure 17. Typical reaction stages, drum rotorconstruction

GER-3646D

RDC27217

T20472A

Efficiency Minimizing stage leakage flow is important to

stage efficiency. With less pressure drop across thebuckets, the loss due to leakage at the bucket tip isobviously much less for an impulse design than fora reaction design, as shown schematically in Figure18.

Greater pressure drop exists across the stationarynozzles in an impulse design than in a reactiondesign. However, the leakage diameter is typically25% less and, therefore, the cross-sectional area forleakage is less. Also, with fewer stages there is suffi-cient space between wheels to mount spring-backedpackings with generous provision for radial move-

9

GER-3646D

GT20475B

Figure 18. Tip leakage for impulse and reaction stages

GT20476A

Figure 19. Root leakage for impulse and reaction stages

ment and a large number of labyrinth packingteeth. In total, the leakage at the shaft packing ofan impulse stage is less than that of a reactionstage. The efficiency advantage is even greater thanthat suggested by the difference in leakage, howev-er, because, as shown in Figure 19, the leakage flowin the impulse stage passes through a balance holein the wheel and does not reenter the steam path.Because the construction of a reaction stage pre-cludes the use of balance holes, the leakage flowmust reenter the steam path between the fixed andmoving blades causing a disturbance of the mainsteam flow leading to a significant, additional loss.

In high-pressure turbine stages typical of mod-ern designs, tip leakages are two to four timesgreater and shaft packing flows are 1.2 to 2.4 timesgreater for a reaction design than for an impulsedesign for turbines of equal rating. The total effi-ciency loss is even greater due to the reentry effectof the shaft packing flow inherent with the drumrotor. The effect of leakage losses on stage perfor-mance, of course, becomes smaller as the volumeflow of the stages increases for both reaction andimpulse designs. On a relative basis, however, theleakage losses on a reaction stage will always begreater than those on an impulse stage designedfor comparable application. This is also significantfrom the standpoint of sustained efficiency becausethe impulse design is less sensitive to the effects ofincreased packing clearances that might occur inoperation.

With more energy per stage, steam velocities inan impulse stage are higher than in a reactionstage. These higher velocities have the potential ofresulting in profile losses that could offset theeffects of reduced leakage loss if poor nozzle andbucket profiles were used. This was a legitimateconcern in the early days of steam turbine develop-ment with only very simple bucket profiles used.Profile losses, however, are very amenable to reduc-tion with increased sophistication of nozzle andbucket profiles. With current computer analysismethods and aerodynamic testing techniques (Fig-ure 20), significant gains continue to be made inreducing profile and other secondary losses.

With an impulse design, the pressure dropacross the diaphragm of the first stage of the reheatand low-pressure sections is high relative to thevelocity head of the steam in the inlet pipe, ensur-ing a uniform flow distribution through the stage.With the lower pressure drop of a reaction stage,poor flow distribution in the first stage of a sectioncan cause performance losses, and complex meanssuch as inlet scrolls are sometimes used to improveflow distribution. Such designs have very little ben-

efit with the GE impulse design.

MechanicalImpulse stage design with wheel-and-diaphragm

construction lends itself to a rugged, reliabledesign because the pressure drop occurs across sta-tionary, rather than moving, parts and because theneed for fewer stages permits space for sturdydiaphragm design.

Because of the low stage thrust, a balance pistonis not required as it is with reaction turbines.Thrust bearings are used with conservative loadingwithout resorting to large sizes.

Thermal stresses in high-temperature rotorslimit the rate at which a turbine-generator canchange load. These stresses, which are greatest atthe rotor surface, depend heavily upon the diame-ter of the rotor body and the corresponding stressconcentration factors. The wheel-and-diaphragmdesign results in significantly smaller rotor bodydiameter and permits ample axial spacing betweenstages for generous fillet radii at the intersection ofthe packing diameter and the side of the wheels,resulting in low stress concentration factors at thepoint of maximum thermal stress. In contrast, thestress concentration factors on drum rotors are rel-atively high because of the intricate geometryrequired for blade attachment. The wheel-and-diaphragm construction, therefore, leads to signifi-cantly lower rotor thermal stresses and greatercapability for load cycling operation.

Furthermore, the wheel-and-diaphragm designseparates the region of maximum rotor thermal

10

GER-3646D

Figure 20. Advanced three-dimensional aerostages

GT25603

stress from the bucket dovetail region. The dovetailregion of the rotor is most likely to be affected bycreep resulting from the combination of high tem-perature, intricate geometry, and tensile stress dueto the centrifugal load of the buckets. In the drumconstruction these areas are at the same location,as can be seen from Figure 19, and any creep dam-age will be additive to low cycle fatigue damagecaused by temperature cycling.

One additional advantage of wheel-and-diaphragm construction arises because high pres-sure and reheat inner shells with heavy joint flangestend to undergo distortion due to uneven thermalexpansion. Interstage and tip seals are generallysupported directly from the inner shells in reactionturbines, and distortion of the shell results in move-ment of the seals, exacerbating the problem of lim-iting leakage flow. It is GE’s practice to supportboth of these seals from the diaphragm which isunaffected by any distortion of the shell. Thisarrangement can be seen in Figure 16. Various con-structions have been developed with reaction tur-bines to eliminate or minimize the distorting effectof the joint flange on seal clearances. In oneapproach, the two halves of the inner shell are heldtogether by a series of rings installed with a shrinkfit, creating inward radial forces. This can eliminatethe horizontal joint flange, but makes assembly anddisassembly difficult. The problem that this designaddresses does not exist with GE diaphragm con-struction.

It is significant to note that although reactionstage design virtually dictates the use of a drumrotor, with an impulse stage design either wheel-and-diaphragm or drum construction could beused. Wheel-and-diaphragm construction is the

choice because of its many mechanical and efficien-cy-related advantages.

Opposed-Flow, High-Pressure/Intermediate-Pressure Design

The single-span, opposed-flow HP/IP design,shown in Figure 21, was introduced by GE in 1950,in a turbine rated 100 MW. Today there are over500 turbines with this feature in operation. It is ahighly-developed design with a maximum ratingthat has increased over the years. The present limitis approximately 650 MW for partial arc units, and750 MW for full arc machines. This arrangementresults in a significantly more compact turbine andstation arrangement than that of a unit with thehigh-pressure and reheat sections in separate bear-ing spans. There is also one less turbine section tobe maintained.

High-pressure steam enters the center of the sec-tion and flows in one direction (to the left in Figure21), while steam reheated to similar temperaturealso enters near the center and flows in the oppo-site direction. This arrangement confines the high-est temperature steam to a single central locationand results in an even temperature gradient fromthe center toward the ends, with the coolest steamadjacent to the end packings and bearings.

The opposed-flow design is more compact than adesign with separate high-pressure and reheat sec-tions. Tests have shown that this leads to a lowerrate of temperature decay after overnight andweekend shutdowns permitting more rapid restart-ing.

Although a number of factors affecting perfor-mance, including stage packing diameters, shaft-

11

GER-3646D

Fig. 21. Arrangement of opposed-flow, high-pressure and intermediate-pressure sectionsGT23809

end packing leakage, and volume flow effects, aredifferent in the opposed-flow design and a designwith separate HP and IP sections, the net differ-ence in performance is essentially zero at all rat-ings.

Reliability statistics on the entire fleet of GE tur-bines operating in the United States indicate asmall but consistent advantage for the opposed-flowdesign over a design with separate sections at thesame rating.

The bearing span for the opposed-flow rotor isgreater than the bearing span for either rotor of aunit with separate high-pressure and reheat sec-tions. Also, the shaft diameter tends to be some-what larger as a result of designing for similardynamic characteristics. This could be a disadvan-tage at the very largest ratings if the boiler andother plant equipment have a greater capability forrapid starting and loading, and if the unit will cyclefrequently. When carefully studied, however, this isseldom found to be the case. In most cases the GEopposed-flow design with wheel-and-diaphragmconstruction will have starting and loading capabili-ty comparable to a drum-type design with separatehigh-pressure and reheat sections. Nevertheless, anarrangement with separate HP and IP sections canbe provided in the larger ratings when it is believedthat the disadvantages are justified by a need forbetter starting and loading characteristics.

Inlet ConfigurationsWith the exception of the very largest units, and

other special cases which do not warrant it, GE util-ity units have individually actuated control valveswith Admission Mode Selection (AMS), whichallows the unit to operate with the benefits of eitherfull or partial arc operation.

With partial-arc admission, the first-stage nozzlesare divided into separate nozzle arcs with each arcindependently supplied with steam by its own con-trol valve. For units operating with constant initialpressure, load is reduced by closing these valves insequence. For smaller units, all four valves wouldoperate in sequence providing four consecutiveadmissions. For the largest units in a given configu-ration, three valves would initially operate togetherand one separately to provide two admissions.Intermediate-size units would have two valves clos-ing together with the remaining two closing insequence to give a three-admission unit. Theimpact on part load performance for these admis-sion modes is illustrated in Figure 22.

With a single-admission (or full arc) machine,load is controlled by throttling on all of the admis-sion valves equally, and all control valves connect

into a common chamber ahead of the first-stagenozzles. As load is decreased on the single-admis-sion unit, an increasing amount of throttling takesplace in the control valves. In a partial-admissionunit on the other hand, less throttling loss occurs atreduced load because the valves are closed sequen-tially, and only a portion of the steam admitted atany given load undergoes throttling, while theremaining flow passes through fully-open valves.

With AMS, the unit can be used for starting andloading in full arc admission, reducing themalstresses, and converting to partial arc admission forimproved steady-state performance.

Variable-pressure operation, using boiler pres-sure to vary load at a fixed valve position, is nowcommon, and the question is sometimes raised asto whether the partial-arc admission feature is eco-nomical.

If load is reduced by varying pressure with valves-wide-open, load increase can only be achieved byincreasing boiler pressure, which is a relatively slowprocess, and the unit cannot participate in systemfrequency control. These shortcomings can beovercome with a hybrid method (Figure 22) ofoperation in which load is reduced approximately15% at constant pressure, providing some “throttlereserve” before beginning to reduce pressure. Withpartial-arc admission, it is attractive to fully closeone valve and then vary pressure. If a greater capa-bility for rapid load increase is desired, two valvescan be closed. In either case, partial-arc admissionyields a better heat rate than full throttling, evenwith variable pressure operation.

Solid Particle Erosion ResistanceCarryover of iron oxide particles from boiler

superheater and reheater tubes can cause severeerosion to turbine nozzles and buckets. Solid parti-cle erosion (SPE) has a major economic effect in

12

GER-3646D

GT24516

Figure 22. Effect of admission modes and throttle pressure programs on heatrate

loss of sustained efficiency, and in causing need forlonger and more frequent maintenance outages.Extensive efforts to understand the erosive mecha-nisms in the turbine steam path and develop resis-tant coatings have led to substantial improvementsin the erosion resistance of GE turbines.

Analysis of particle trajectories in steam as afunction of density and velocity has led to changesin geometry of nozzle partitions and relative spac-ing between nozzles and buckets in the first high-pressure and reheat stages, that result in dramaticdecreases in the rate of erosion. These features,along with either plasma spray or diffusion-appliedhard coatings in the same regions, are available onreheat turbines operating with fossil-fuel fired boil-ers with steam temperature of 1000F/538C orgreater.

Centerline SupportTurbine components undergo considerable

thermal expansion as they undergo changes intemperature. The various stationary componentssurrounding the rotor in GE turbines are support-ed at, or very close to, the centerline, and are freeto expand radially to maintain concentricity. Asshown in Figure 23, all diaphragms are positioned

by means of radial keys inside inner shells and, in asimilar manner, inner shells are positioned insideouter shells, or hoods, by means of radial surfacesat the horizontal joint and at the vertical centerline.Finally, the outer shells are supported by the rotorbearing standards at their true horizontal center-lines.

Number of BearingsGE has considerable experience both with tur-

bine designs employing two bearings per rotorspan and with designs that employ fewer bearings.There are advantages and disadvantages to bothapproaches, but overall the use of two bearings perrotor on large turbines is considered to have suffi-cient advantage to justify the additional cost and,sometimes, added length. The benefits are lessclear, however, on smaller units and three turbinebearings in two-casing machines with single-flowexhausts or small double-flow exhausts arecurrently used.

The use of two bearings per rotor gives thedesigner flexibility to accurately establish rotor criti-cal speeds by selection of bearing span. It results inshorter bearing span and, therefore, smaller rotorbody diameter, which is beneficial to efficiency and

13

GER-3646D

Figure 23. Method of location of stationary componentsGT20468

starting and loading capability. The shorter, stifferrotors between bearings and the added damping ofthe additional bearings reduce susceptibility torotor instability.

With two bearings per rotor, each rotor can beprecision, high-speed balanced on its own journalsin the factory. The result is a finely-balanced tur-bine rotor that can be assembled to the otherrotors in the field, and in nearly all cases startedwithout additional balancing. If additional balanc-ing is necessary at start-up or following a turbineoutage, it can easily be accomplished with a mini-mum of balance shots and downtime because ofthe relatively small dynamic interaction betweenadjacent rotors. The imbalance can be located andcorrected with a small impact on availability of theunit.

The shared-bearing rotor design results in arotor system that is more sensitive to imbalance andmore difficult to field balance. Turbines with onebearing per span have rotors that are factory bal-anced using a stub shaft or temporary journal. Thisprocedure creates a difference between the rotoroperating conditions during factory balance andthe actual conditions, when the rotors are fullyassembled in the turbine-generator. The result isthat rotors may require some rebalancing afterassembly in the field, and since there is moredynamic interaction between adjacent rotors, bal-ancing is more difficult to accomplish.

General maintenance and bearing inspectionare easier with two bearings per rotor. An auxiliarybearing is not required for support when a cou-pling is broken. As a result, coupling alignment canbe more accurately established, further contribut-ing to smooth operating characteristics.

The major benefit of using fewer bearings, otherthan cost, is that some reduction in overall lengthof the unit can be achieved. Designing for two bear-ings per rotor requires some additional length toachieve adequate rotor flexibility between adjacentbearings and thereby, tolerance for misalignment.

CrossoverGE’s fossil turbines use a single crossover to

transport steam from the intermediate-pressure tur-bine exhaust to the low-pressure turbine inlets.Pressure-balanced expansion joints, as shown inFigure 24, are provided to permit differential ther-mal expansion between the crossover and the sta-tionary parts it connects without imposing largeaxial forces due to steam pressure. The stainlesssteel bellows have high reliability in this applica-tion, since there is no load imposed in torsion orbending.

Compared to use of a cross-around pipe on eachside of the turbine, this design makes for a less clut-tered turbine arrangement with unobstructedaccess at the floor level. There is also a reliabilityadvantage in that only half as many bellows arerequired.

In most cases there is also an advantage inreduced bearing span of the IP section and overalllength of the machine by making the IP exhaustconnection in the upper half and the reheat inletand extraction connections in the lower half (Fig-ure 21).

Because of the very large volume flow betweenthe high-pressure exhaust and the low-pressureinlets with nuclear turbines, and the presence ofmoisture separator reheaters between the samepoints, from four to eight cross-around pipes areused. The routing is generally three-dimensional,therefore inherently flexible, and the temperaturesinvolved are low, so that expansion bellows are notrequired.

AccessoriesGE continually refines the accessories of the tur-

bine to ensure the highest levels of reliability andlowest initial investment for the owner.

In the area of controls, the modern triply-redun-dant Mark V control system provides unit controland interface to the plant DCS. The reliability ofmodern control systems allows for the use of dualpath electronic overspeed protection versus the tra-ditional mechanical overspeed protection withbackup electrical system. The speed and the over-speed are independently monitored by three sepa-rate sensors for each signal.

The lube oil system has been continually refinedto provide higher levels of reliability. Recentimprovements include the optional full-flow filtersand conversion from a shaft-driven pump to a mod-ern all-motor pumping system.

The remaining accessories such as hydraulicpower unit and steam seal equipment are highlypackaged to allow for ease of installation and main-

14

GER-3646D

GT20479

Figure 24. Crossover expansion joint

tenance. Each of these packaged systems allows fortailored selection of optional features.

Maintainability FeaturesThe extent to which a turbine can be maintained

without disassembly and the ease with which it canbe disassembled and reassembled for inspectionand maintenance directly impact availability andare important design considerations.

Disassembly of GE turbines is facilitated by a gen-erally uncluttered arrangement, the use of twobearings per span, the use of opposed-flow HP/IParrangements, to minimize the number of casings,and a minimum number of piping connections tothe upper half shell. Any special tools or liftingdevices required are provided. Optional featuresthat can be provided include special hydraulically-extended coupling bolts and horizontal joint studsfor outer shells and low pressure inner casings. Inaddition, a small jib crane installation for liftingbearing parts without removing the crossover isavailable, as shown in Figure 25.

Features that reduce the frequency with whichmajor disassembly is required are, if anything, evenmore important. These include the SPE-resistantfeatures that have been described, provisions forfield balancing in all rotors, full-flow lube oil filters,positive-pressure, variable-clearance packings thatprovide increased clearance during start-up andnormal close clearances at load, and access portsfor steam path inspection by borescope coupledwith long term maintenance packages, 10 year sec-tionalized inspection intervals are available.

Assembled ShipmentThe practice for larger steam turbine configura-

tions with two or more casings has been to fielderect the unit. With installation time varying with

the number of casings, number of shifts, experi-ence of the installer, etc., the total costs associatedwith the installation can be significant, and meansof reducing installation costs have, therefore, beenimplemented on a continuing basis. Main steamvalves, lube oil tanks, and other skid items havebeen packaged to minimize their installation costsand cycles. Innovations such as full flow lube oil fil-ters and hydro-flushing, laser alignment tech-niques, and electronic measurement devices havealso reduced the critical path installation cycletime.

Significant savings in both installation time andcosts can now be realized by shipping the high pres-sure and intermediate sections of these larger unitsfactory-assembled. High Power DensityTM turbinesare designed to be shipped assembled and can be,site transportation facilities permitting. Installingassembled high and intermediate pressure casingscan typically eliminate 25-35 critical path installa-tion days in the erection cycle, while reducing con-struction costs by 15% to 25%. In addition, storageand inventory control requirements are greatlyreduced, and the turbine hall crane is available forother uses.

Assembled sections (Figure 26) are shipped withthe rotor installed, diaphragms and other partsinstalled and aligned, and the shell “hot” bolted. Insome cases, the front standard can be included aspart of the shipment.

Procedures to facilitate the installation of thesesections have been prepared and are derived fromGE’s vast experience with the installation ofshipped assembled sections for single and multi-casing designs for industrial and combined-cycleapplications. To date, the largest unit shippedassembled has been the high and intermediatepressure sections of two 600 MW four-casingdesigns.

15

GER-3646D

Figure 25. Jib crane for bearing maintenanceGT20480

Figure 26. Shipped assembled HP/IPRDC26298-4-12

CONCLUSIONGE offers a full range of steam turbine-genera-

tors for both fossil-fueled and nuclear 50 and 60Hz applications. Many basic design features arethe result of an overall consistent design philoso-phy that emphasizes efficiency, reliability, andmaintainability. Many years of development efforthave gone into the present product line and it isexpected that this evolutionary process will con-tinue in the future.

REFERENCES1. Cofer, IV, J.I., Koenders, S., and Sumner, W.J.,

Advances in Steam Path Technology, GER-3713C, 38th General Electric Turbine State-Of-The-Art Technology Seminar, August 1994.

2. Moore, J.H., High Power DensityTM Steam Tur-bine Design Evolution, GER-3804, 38th GeneralElectric Turbine State-Of-The-Art TechnologySeminar, August 1994.

3. Retzlaff, K.M., and Ruegger, W.A., Steam Tur-bines for Unltrasupercritical Power Plants, GER-3945, 39th General Electric Turbine State-of-the-Art Technology Seminar, August 1996.

16

GER-3646D

© 1996 GE Company

LIST OF FIGURES

Figure 1. Double-flow, low-pressure rotor with 42-inch/1067 mm last-stage bucketsFigure 2. High power density–last-stage bucket impactFigure 3. Fossil turbine arrangementsFigure 4. Two-casing, single-flow steam turbineFigure 5. Two-casing, double-flow steam turbine with shell-mounted valvesFigure 6. Two-casing, double-flow steam turbine with off-shell valvesFigure 7. Three-casing, four-flow steam turbineFigure 8. Three-casing, four-flow steam turbineFigure 9. Four-casing, four-flow steam turbineFigure 10. Five-casing, six-flow steam turbineFigure 11. Four-casing, four-flow, double-reheat steam turbineFigure 12. Ultrasupercritical steam turbineFigure 13. Nuclear turbine arrangementsFigure 14. Six-flow nuclear steam turbineFigure 15. Ideal stage efficiency as a function of velocity ratio for impulse and reaction stage designsFigure 16. Typical impulse stages, wheel-and-diaphragm constructionFigure 17. Typical reaction stages, drum rotor constructionFigure 18. Tip leakage for impulse and reaction stagesFigure 19. Root leakage for impulse and reaction stagesFigure 20. Advanced three-dimensional aero stages Figure 21. Arrangement of opposed-flow, high-pressure and intermediate-pressure sectionsFigure 22. Effect of admission modes and throttle pressure programs on heat rateFigure 23. Method of location of stationary componentsFigure 24. Crossover expansion jointFigure 25. Jib crane for bearing maintenanceFigure 26. Shipped assembled HP/IP

GER-3646D

P. B. MasonPaul B. Mason joined GE’s Field Engineering Program in 1980

progressing to become a specialist in startup and controls. In1989 he joined the Steam Turbine Application Engineeringresponsible for all technical aspects of proposal engineering fornew industrial, combined cycle and central power plant steam tur-bine generators, becoming the Technical Leader of the group in1996. Paul graduated from Southampton University in Englandwith a B.S. in Mechanical Engineering in 1977.

John K. ReinkerJohn K. Reinker joined GE’s Power Generation Business in

1984 as a Steam Turbine Engineer. He worked in several designoffices culminating in his assignment of Manager New UnitDesign for Small and Large Steam Turbines. He was named to hiscurrent position of Product Line Leader — Large Steam Turbinein June 1996.

A list of figures appears at the end of this paper.