Otakar J onas is a Consultant with J onas, Inc., in Wilmington, Delaware. Heworks inthefieldof industrial andutilitysteamcycle corrosion, water andsteamchemistry, reliability, andfailureanalysis. After periods of R&D at Lehigh University and engineering practiceatWestinghouseSteamTurbineDivision, Dr. J onasstarted hiscompanyin1983. Thecompanyisinvolvedintroubleshooting, R&D (EPRI, GE, Alstom), failureanalysis, andintheproduction of special instrumentsandsamplingsystems. Dr. J onas has a Ph.D. degree (Power Engineering) fromthe Czech Technical University. He is a registered Professional Engineer intheStatesof DelawareandCalifornia. LeeMachemer isaSenior Engineer atJ onas, Inc., inWilmington, Delaware. Hehas13yearsof experiencewithindustrial andutility steam cycle corrosion, steam cycle and water chemistry, and failureanalysis. Mr. Machemer received a B.S. degree(Chemical Engineering) fromtheUniversity of Delawareand is a registered Professional Engineer intheStateof Delaware. ABSTRACT This tutorial paper discusses the basics of corrosion, steam and deposit chemistry, and turbineand steamcycledesign and operation—as they relate to steam turbine problems and problemsolutions. Major steam turbine problems, such as stress corrosion crackingof rotorsanddiscs, corrosionfatigueof blades, pitting, and flow accelerated corrosion are analyzed, and their root causesandsolutionsdiscussed. Alsocoveredare: lifeprediction, inspection, andturbinemonitoring. Casehistories aredescribed for utilityandindustrial turbines, withdescriptionsof rootcauses andengineeringsolutions. INTRODUCTION This tutorial paper discusses steamturbinecorrosionanddepo- sition problems, their root causes, and solutions. It also reviews designandoperation, materials, andsteamanddeposit chemistry. Referencesareprovidedat theendof thepaper. Withanincreaseof generatingcapacityandpressureof individ- ual utilityunitsinthe1960sand70s, theimportanceof largesteam turbinereliability andefficiency increased. Theassociatedturbine sizeincreaseanddesignchanges (i.e., larger rotors anddiscs and longer blades) resulted in increased stresses and vibration problems and in the use of higher strength materials (Scegljajev, 1983; McCloskey, 2002; Sanders, 2001). Unacceptable failure rates of mostly blades anddiscs resultedininitiationof numerous projectstoinvestigatetheroot causesof theproblems(McCloskey, 2002; Sanders, 2001; Cotton, 1993; J onas, 1977, 1985a, 1985c, 1987; EPRI, 1981, 1983, 1995, 1997d, 1998a, 2000a, 2000b, 2001, 2002b, 2002c; J onasandDooley, 1996, 1997; ASME, 1982, 1989; Speidel and Atrens, 1984; Atrens, et al., 1984). Some of these problems persist today. Cost of corrosion studies (EPRI, 2001a, Syrett, et al., 2002; Syrett andGorman, 2003) andstatistics(EPRI, 1985b, 1997d; NERC, 2002) determined that amelioration of turbine corrosion is urgently needed. Same problems exist in smaller industrial turbines and the same solutions apply (Scegljajev, 1983; McCloskey, 2002; Sanders, 2001; Cotton, 1993; J onas, 1985a, 1987; EPRI, 1987a, 1998a; J onasandDooley, 1997). The corrosion mechanisms active in turbines (stress corrosion cracking, corrosion fatigue, pitting, flow-accelerated corrosion) areshowninFigure1. Figure1. CorrosionMechanismsActiveinSteamTurbines. Purpose, Design, andOperationof SteamTurbines Thesteamturbineis thesimplest andmost efficient enginefor convertinglargeamountsof heat energyintomechanical work. As thesteamexpands, it acquireshighvelocityandexertsforceonthe turbineblades. Turbinesrangeinsizefromafewkilowattsfor one stage units to 1300 MW for multiple-stage multiple-component units comprising high-pressure, intermediate-pressure, and up to three low-pressure turbines. For mechanical drives, single- and double-stage turbines are generally used. Most larger modern turbines are multiple-stage axial flow units. Figure 2 shows a typical tandem-compound turbinewith acombined high pressure (HP), intermediatepressure(IP) turbine, andatwo-flowlow-pressure (LP) turbine. Table1(EPRI, 1998a) providesalternateterminology for several turbinecomponents. 211 STEAMTURBINE CORROSION ANDDEPOSITSPROBLEMSANDSOLUTIONS by Otakar J onas Consultant and LeeMachemer Senior Engineer J onas, Inc. Wilmington, Delaware Figure2. Typical TandemCompound, SingleReheat, Condensing Turbine. (Courtesyof EPRI, 1998a) Table1. AlternateTerminologyfor TurbineComponents. Steamentersfromthemainsteamlinesthroughstopandcontrol valvesintotheHPsection. Thefirst(control) stageisspacedslightly apart fromsubsequent stages to allowfor stabilization of theflow. After passingthroughtheHP turbine, coldreheat pipingcarriesthe steamtothereheater (if present) andreturnsinthehot reheat piping to theintegratedHP andIP cylinder to pass throughtheIP turbine section. Theflowexits theIP turbinethrough theIP exhaust hood andthenpassesthroughcrossover pipingtotheLP turbineandexits tothecondenser throughtheLP exhaust. Thetypical modernsteam turbine has a number of extraction points throughout all sections wherethesteamisusedtosupply heat tothefeedwater heaters. Duringits expansionthroughtheLP turbine, thesteamcrosses thesaturationline. Theregionwherecondensationbegins, termed the phase transition zone (PTZ) or Wilson line (Cotton, 1993; EPRI, 1997c, 1998a, 2001b), is the location where corrosion damagehas been observed. In singlereheat turbines at full load, thiszoneisusually at theL-1stage, whichisalsointhetransonic flowregion where, at thesonic velocity (Mach =1), sonic shock waves can be a source of blade excitation and cyclic stresses causing fatigue or corrosion fatigue (EPRI, 1997c; J onas, 1994, 1997; Stastny, et al., 1997; Petr, et al., 1997). Design Because of their long design life, steamturbines go through limited prototype testing where the long-termeffects of material degradation, such as corrosion, creep, and low-cycle fatigue, cannotbefullysimulated. Inthepast, whendevelopmentwasslow, relativelylong-termexperiencewastransferredintonewproducts. Withnewturbinetypes, larger sizes, newpower cycles, andwater treatmentpracticescomingfastinthelast25years, experiencewas short and limited, and problems developed, which need to be correctedandconsideredinnewdesignsandredesigns. Whilethe turbineseemstobeasimplemachine, itsdesign, includingdesign against corrosion, iscomplex. Therearefiveareasof designthat affect turbinecorrosion: • Mechanical design (stresses, vibration, stress concentrations, stress intensity factor, frictional damping, benefits of overspeed andheater box testing) • Physical shape(stressconcentration, crevices, obstaclestoflow, surfacefinish, crevices) • Material selection(maximumyieldstrength, corrosionproperties, material damping, galvanic effects, etc.) • Flowandthermodynamics(flowexcitationof blades, incidence angle, boundary layer, condensation and moisture, velocity, location of the salt zone, stagnation temperature, interaction of shock wavewithcondensation) • Heat transfer (surface temperature, evaporation of moisture, expansionversusstress, heatedcrevices) Recognitionof theseeffectsledtoaformulationof rudimentary designrules. Whilethemechanical designiswell advancedandthe material behavior is understood, theflowexcitationof blades and the effects of flow and heat transfer on chemical impurities at surfacesarenot fully includedindesignpractices. Selectionof somecombinations of thesedesignparameterscan lead to undesirable stresses and impurity concentrations that stimulatecorrosion. In addition, somecombinations of dissimilar materialsincontact canproducegalvanic corrosion. Design and material improvements and considerations that reduceturbinecorrosioninclude: • Welded rotors, large integral rotors, and discs without keyways—eliminateshighstressesindisc keyways. • Replacement of higher strength NiCrMoV discs with lower (yieldstrength<130ksi (896MPa) strengthdiscs. • Repair welding of discs and rotors; also with 12%Cr stainless steel weldmetal. • Mixedtunedbladerowstoreducerandomexcitation. • Freestanding and integrally shrouded LP blades without tenon crevicesandwithlower stresses. • TitaniumLP blades—corrosionresistant inturbineenvironments except for NaOH. • Lower stressandstressconcentrations—increasingresistanceto SCC andCF. • Flowpathdesignusingcomputerizedflowdynamicsandviscous flow—lower flow induced vibration, which reduces susceptibility toCF. • Curved (banana) stationary blades that reduce nozzle passing excitation. • Newmaterialsfor bladepinsandbolting—resistantagainstSCC. • Flow guides and double-ply expansion bellows—reduces impurity concentration, better SCC resistance. • Moisture extraction to improve efficiency and reduce flow- acceleratedcorrosion(FAC) andwater droplet erosionanduseof alloy steelstoreduceFAC. LP Rotor andDiscs Therearethreetypesof constructioninusefor LP rotors: • Built-up(shrunk-ondesign) withforgedshaft onto whichdiscs areshrunk andkeyed, PROCEEDINGSOFTHETHIRTY-SEVENTHTURBOMACHINERY SYMPOSIUM•2008 212 Machinedfromonesolidpiece(most common), or • Thediscsweldedtogether toformtherotor (Figure3). The rotor and disc construction is governed by the practices of individual manufacturers, capabilities of steel mills, cost, and, duringthelast fewdecades, by their resistanceto SCC. Thesolid andweldedrotorsdonot haveaproblemwithdisc boreSCC. The threetypes, showninFigure3, havelittleeffectontheSCCandCF susceptibilitiesof thebladeattachments. Figure3.ThreeTypesof Rotor Construction. (Courtesyof EPRI, 1998a) LP Blades Blade and blade path design and material selection influence bladeCF, SCC, pitting, andother forms of damageinmany ways (Sanders, 2001; EPRI, 1981, 1997c, 1998a, 2001b; BLADE-ST™, 2000). Themaineffectsof thebladedesignoncorrosion, corrosion fatiguestrength, stresscorrosioncrackingsusceptibility, andpitting resistanceinclude: • Vibratory stressesandtheir frequencies. • Maximumservicesteady stressesandstressconcentrations. • Flowinducedvibrationanddeposition. • Mechanical, frictional, andaerodynamic damping. Rotating LP turbine blades may be “free standing” (not connectedtoeachother), connectedingroups, or all bladesinthe row may be “continuously” connected by a shroud. Connections made at the blade tip are termed shrouds or shrouding. Shrouds may be inserted over tenons protruding above the blade tips and thesetenons thenriveteddowntosecuretheshrouds, or they may consist of integrally forged or machined stubs, which, during operation, provide frictional damping of vibration because they touch (Figure 4). This design also eliminates the tenon-shroud crevices where corrosive impurities could concentrate. In some cases, long 180 degree shrouds are used or smaller shroud segmentsareweldedtogether. Figure4. Typical TurbineShrouds. (Courtesyof EPRI, 1998a) Blades are connected at the root to the rotor discs by several configurations (Figure 5). There are several types of serrated attachments: the fir tree configuration, which is inserted into individual axial slotsinthedisc; andtheT-shape, whichisinserted intoacontinuouscircumferential slotinthedisc. The“finger”type attachment is fitted into circumferential slots in the disc and securedbyaxiallyinsertedpins. All of thebladeroot designshave geometries that result in higher local stresses at radii and stress concentrations that promoteSCC and CF. Thegoal of thedesign shouldbetominimizetheselocal tensilestresses. Figure5.Typesof BladeRootAttachments. (Courtesyof EPRI, 1998a) Theairfoil of blades may beof constant cross-sectionfor short blades, andtwistedfor longer ones. Thelongest bladesfor thelast few rows of the LP are twisted to match the aerodynamics at different radii and improve aerodynamic efficiency. The longer blades are usually connected at the point of highest vibration amplitude to each other by a tie or lashing wire, which reduces vibrationof theairfoil. Toreducerandomexcitation, mixedtuning of longrotatingbladeshasbeenusedinwhichadjacentbladeshave different resonant frequencies(EPRI, 1998a). Stationary blades in LP stages are typically arranged in diaphragms of cast or welded construction. In wet stages, diaphragmsmay bemadewithhollowbladevanesor other design features as ameans of drawingoff moisturethat wouldotherwise lead to liquid droplet erosion (EPRI, 2001b). Recently, some stationary blade designs have also been leaned or bowed, improving flow and efficiency and lowering theexcitation forces onthedownstreamrotatingblades. Casings Turbine casings must contain the steampressure and maintain support and alignment for the internal stationary components. They aredesigned to withstand temperatures and pressures up to themaximumsteamconditions. Their designhasevolvedover the years andcasings arenowmultiplepressurevessels (for example, an inner and outer casing in the HP and IP cylinder, or a triple casing) allowingsmaller pressuredropsandthinner wall thickness. Thesethinner cross-sectionsallowfor alower temperaturegradient across thecasingsectionandthus lower thermal stresses. TheLP casing may also be a multiple part design with the inner casing containing the supports for the diaphragms and the outer casing directingtheexhaust tothecondensers. DesignRecommendationsfor CorrosionControl New designs, redesigns, and failed components should be checked to determine if they meet allowable corrosion-related design specifications and other corrosion related requirements (J onas, 1985c). 213 • STEAMTURBINE CORROSION ANDDEPOSITSPROBLEMSANDSOLUTIONS Innewdesignsandredesignsof turbinecomponents, useshould be made of the new design tools, including 3D finite element stress and vibration analysis and 3D viscous flow analysis, and consideration of condensation and impurity behavior. Blade resonancefrequenciesshouldbeverifiedbytelemetry. Toensurea corrosion-freedesign, acorrosion engineer and achemist should beconsultedduringthedesignactivity. Thefollowingshouldbeconsideredindesignof steamturbines: • Stresses (J onas, 1985c)—The mechanical design concepts for avoidanceof SCCandCF shouldincludeanevaluationof thematerial corrosionpropertiesanddefectsthat influencesusceptibility toSCC and CF, i.e., threshold stress (σ SCC ), threshold stress intensity (K ISCC and⌬K thCF ), crackgrowthrate((da/dt) SCC and(da/dN) CF ), corrosion fatigue limit, pitting rate, and pit depth limit. True residual stresses (micro and macro) should also be considered. Because SCC and CF initiate at surfaces, the maximumsurface stresses must becontrolled, usually by control of theelastic stress concentration factor, k t . The stresses should be the lowest in the “salt zone” regionwherecorrosionismost likely. • Vibratory stresses are rarely accurately known, except when telemetryonoperatingturbinesisperformed. Thedesignapproach should be to minimize flow excitation, tune the blades, provide maximumdamping, and performlaboratory and shop stationary frequency testing. Heater box overspeed and overspeed testing during operation are generally beneficial in reducing operating stressesby local plastic deformation. • Heat transfer and flow (EPRI, 1997c)—Surface temperature resultingfromheattransfer andflowstagnationshouldbeconsidered alongwithitseffect onthermodynamicconditionsof theimpurities and water filmat surfaces (i.e., evaporation of moisture). Flow effects onbladevibrationanddeposit formationarecomplex, and thereareover 15flowbladeexcitationmechanismstobeconsidered. • Flowof moisture—To avoid flow-accelerated corrosion (EPRI, 1996; Kleitz, 1994; J onas, 1985b; Svobodaand Faber, 1984) and water droplet erosion (Ryzenkov, 2000; Pryakhin, et al., 1984; Rezinskikh, et al., 1993; Sakamoto, et al., 1992; Povarov, et al., 1985; Heyman, 1970, 1979, 1992), theflowvelocity of wet steam shouldnot exceedtheallowablevelocity specific to thematerials and moisture chemistry. Regions of high turbulence should be avoided or higher chromium steels should be used. Blade path moisturecanbeextracted. • Crevices—Crevicescanact asimpuritytrapsandconcentrators, facilitate formation of oxygen concentration cells, and may generatehighstresses by theoxidegrowthmechanism. Theworst crevicesarethosewithcorrosiveimpuritiesandmetal temperature within the “salt zone.” Some disc bore and keyway and blade tenon-shroudcrevicesfall intothiscategory. • Galvaniceffects—Whendissimilar materialsarecoupledtogether, corrosionof bothmaterialscanbeaffectedbytheassociatedshift in corrosion potentials into the stress corrosion cracking (SCC) or pitting regions. The more active of the two materials may suffer galvanic corrosion. In addition, in some environments, the potential shift could beinto theregion whereoneof thecoupled materialsissusceptibletostresscorrosioncrackingor pitting. • Inspectability—In designing turbine components, the question of inspectability should be addressed. In particular, crevice and high stress regions should be reachable using available inspectiontechniques. • Chemical compounds used during machining, cleaning, nondestructivetesting (NDT), and other activities—Many different chemical compoundsareusedduringmanufacture, storage, erection, and inspection of turbine components. Some of them contain chlorineand sulfur as impurities or as apart of theorganic matrix. Duringthermal decomposition of theresidues of thesecompounds, hydrochloric, sulfuric, carbonic, andorganic acids canform(J onas, 1982). It isrecommendedthat their compositionandusebecarefully controlled to minimize the risk of residual contamination and subsequent corrosion. For thecompoundsthat canremainonturbine surfaces during operation, chlorine and sulfur levels should be restricted to lowppmlevels in that compound. Of specific concern are: MoS 2 (Molylube™) (Turner, 1974; Newman, 1974), Loctite™, threadcompounds(Cu, Ni, graphite), andchlorinatedsolvents. MaterialsandCorrosionData Thereislittlevariationinthematerialsusedfor blades, discs, rotors, and turbine cylinders, and only a few major changes have been introducedinthelast decade. Titaniumalloy blades arebeingslowly introduced for the last LP stages, and better melting practices to providecontrol of inclusions and traceelements arebeing evaluated for discsandrotors.Table2(Jonas, 1985a) listscommonmaterialsand thetypical corrosionmechanismsfor thevariousturbinecomponents. Table 2. LP Turbine Components, Materials, and Related CorrosionMechanisms. LP rotors are typically constructed of forgings conforming to ASTM A293 (Class 2 to 5) or ASTM A470 (Class 2 to 7), particularly 3.5NiCrMoV. Shrunk-on discs, when used, are made fromforgingsof similar NiCrMoV materialsconformingtoASTM A294(GradeB or C), or ASTM 471(Classes1to3). Thestrength and hardness of turbinecomponents must belimited becausethe stronger andharder materialsbecomeverysusceptibletoSCC and CF (EPRI, 1998a); particularly turbine rotors, discs, and blades cannot bemadefromhighstrengthmaterials. The crack propagation rate increases exponentially with yield strengthat highyieldstrengthvaluesandSCCstartsbeinginfluenced byhydrogenembrittlement. Becauseof thissensitivitytohighyield strength, practically all turbinediscs, fossil andnuclear, withyield strength higher than ~140 ksi (965 MPa) have been replaced with lower strengthmaterials. Figure6isacorrelationof crackpropagation ratesversusyieldstrengthfor several operatingtemperatures(Clark, etal., 1981). Thistypeof datahasbeenusedtopredicttheremaining lifeandsafeinspectioninterval. Thereisalsoanupper temperature limit for LP rotor anddiscsteels, ~650ЊF (345ЊC) aimedat avoiding temper embrittlement (EPRI, 1998a). Figure 6. Average Crack Growth Rates Versus Yield Strength for Several Operating Temperatures for NiCrMoV Disc Steel. (Courtesyof Clark, et al., 1981) PROCEEDINGSOFTHETHIRTY-SEVENTHTURBOMACHINERY SYMPOSIUM•2008 214 Sincethe1930s, mostLP turbinebladeshavebeenmanufactured froma12%Cr stainlesssteel—typicallytypesAISI 403, 403-Cb, 410, 410-Cb, and422, dependingonthestrengthrequired. Types403and 410havebetter SCC andCF resistancethanType422, animportant characteristic for usein thewet stages of theLP turbine. Thereare numerousspecificallycustomizedversionsof thesegenericmaterials (Carpenter H-46, J etheteM152(modified403), etc.). Theprecipitationhardenedstainlesssteelsdesignated17-4PH, 15-5PH, and PH13-8Mo have been used for some fossil and nuclear LP turbine blades. The composition of 17-4PH is 17 percent Cr, 4 percent Ni, and 4 percent Cu. These steels may be difficult toweldandrequirepost-weldheat treatment. Thecopper richzonesinthecopper bearingstainlesssteelsareoftensubject to selectivedissolution, forming pits filled with corrosion products. Thesepitscanbecrack initiationsites. Titaniumalloys, primarilyTi-6Al-4V, havebeenusedfor turbine blades sincetheearly 1960s (EPRI, 1984d, 1984e, 1985c). There arenumerousbenefitstousingtitaniumalloyblades, includingthe ability to use longer lasting stage blades, favorable mechanical propertiesinapplicationsinvolvinghighstressesat lowtemperatures, excellent corrosionresistance, andresistanceto impact andwater droplet erosion damage. Drawbacks to titanium include higher cost, difficult machining, andlowmaterial damping. LP turbinecasings aretypically constructedof weldedandcast components. Materials acceptablefor lower temperatures, suchas carbonsteel plate, areused. Considering the typical steamturbine design life of 25 to 40 years and the relatively high stresses, these materials have been performingremarkablywell. TurbinesteelsaresusceptibletoSCC and CF in environments such as caustic, chlorides, acids, hydrogen, carbonate-bicarbonate, carbonate-CO 2 , and, at higher stressesandstrengthlevels, inpurewater andsteam. CorrosionData Corrosion data should provide allowable steady and vibratory stresses andstress intensities for defineddesignlifeor inspection intervals. It is suggested that SCC data include threshold stress (σ SCC ), threshold stress intensity (K ISCC ), crack growth rate (da/dt), and crack incubation and initiation times. Corrosion fatiguedatashouldincludefatiguelimits for smoothandnotched surfacesandproper stressratios, crack growthdata, andcorrosion fatiguethresholdstressintensities. Examplesof thetypeof dataneededareshowninFigures7to9 for theNiCrMoV disc material andinFigure10for 12%Cr blade steel. Properly heat treated12%Cr bladesteel (yieldstrength 85 ksi, 600 MPa) is not susceptibleto stress corrosion cracking and stress corrosiondataarenot needed. ThedatashowninFigures 7 to 10 can be used in turbine disc and blade design in which the allowable stresses and stress intensities should be below the thresholdvaluesfor SCC andCF. Theuseof thesedataisoutlined in(J onas, 1985c). Figure7. StressCorrosionBehavior of NiCrMoVDiscSteel Versus Yield Strength for “Good” Water and Steam (Compiled from Published Data); K ISCC —Threshold Stress Intensity, σ SCC — Threshold Stress, and da/dtq—Stage 2 Crack Growth Rate. (Courtesyof J onas, 1985a) Figure 8. Corrosion Fatigue (Goodman) Diagramfor NiCrMoV Disc Steel; Testedto10 8 Cycles. (Courtesyof Haas, 1977) Figure9. Air FatigueStrength Reduction of NiCrMoV Disc Steel Caused by Pitting (Courtesy of McIntyre, 1979)—Effects of Pit DensityWereNot Investigated. Figure 10. Corrosion Fatigue (Goodman) Diagram for Three StainlessSteel BladingAlloys. (Courtesyof Atrens, et al., 1984) STEAMTURBINE CORROSION ANDDEPOSITSPROBLEMSANDSOLUTIONS 215 Theneededdataaredifficult to obtainbecauseof thelongtime neededfor testingandbecauseof thegreatrangeof possibleservice environments and temperatures. There does not seem to be any universallyacceptedacceleratedtestor environment. TheK ISCC test in hydrogen sulfide gives a reasonable approximation of K ISCC for lowalloy steels, andultrasonic fatiguetests giveusabledatato ahighnumber of cyclesthat may beusablefor turbinedesign. As aruleof thumb, theelasticity limit at temperature(usually between 0.4 and 0.6 of the 0.2 percent yield strength) can be used as a good estimateof theSCC threshold stress for low and mediumstrengthmaterialsinmildly corrosiveenvironments. This isconsistent withtheoxidefilmruptureor straininducedcracking theory of stresscorrosioncracking. Shot peening (EPRI, 2001b)—has been used as a means of reducing high surfacetensilestresses. At asufficiently high shot peeningintensity, asurfacelayer of residual compressivestresses is produced. One turbine vendor uses shot peening and other surface treatments extensively and they have almost no SCC of discs and corrosion fatigue (CF) of turbine blades. There is a concernthat incorrosiveenvironments, pits cangrowthroughthe compressive stress layer into the subsurface region with much higher tensilestresses. Other surfacetreatments for protection against corrosion, such as coatings andelectroplating, havebeenevaluated(EPRI, 1987a, 1993a, 2001b; J onas, 1989) and sometimes used. There are now several suppliers of steamturbinecoatings (EPRI, 1987a, 1993a; J onas, 1989). Environment—Stress—Material Turbinestresscorrosioncrackingandhigh- andlow-cyclecorrosion fatigue mechanisms are typically governed by a combination of environmental effects (steamchemistry, temperature, etc.), steady and vibratory stresses, and material strength, composition, and defects(Figure11). Itshouldbenotedthatevenpurewater andwet steamcan causecracking of turbinematerials, particularly in the lowalloy rotor anddisc steels, andthat mediumandhighstrength materialsareverysusceptibletoenvironmentallyinducedcracking inany environment, includingpurewater andsteam. Figure 11. Three Components of Turbine Stress Corrosion and CorrosionFatigueCrackinginTurbines. Turbine environment plays a major role in corrosion during operationandlayup. Theuniqueness of this environment is caused bythephasechangesof theworkingfluidandtheimpuritiescarried by thesteam(steam, moisture, liquid films, and deposits). Within the steamflow path and on the turbine component surfaces, the parameterscontrollingcorrosion, suchaspH, concentrationof salts andhydroxides, andtemperature, canchangewithinabroadrange. Even though steamimpurity concentrations are controlled in the lowparts per billion(ppb) range, theseimpurities canconcentrate by precipitation, deposition, and by evaporation of moisture to percent concentrations, becoming very corrosive (Figures 12 and 13) (EPRI, 1994a, 1997b, 1997c, 1999; J onas, et al., 1993). Figure12. Physical-Chemical ProcessesinLP Turbines. Figure13. CrossSectionof anLPTurbinewiththeLocationsof the ProcessesListedinFigure15-11. SteamChemistry Steamchemistry or purity, together with the thermodynamics and flow design, determines corrosiveness of the deposits and liquid films on turbine component surfaces (J onas, 1982, 1985a, 1985d; J onasandDooley, 1996, 1997; EPRI, 1984b, 1994a, 1997b, 1997c, 1999; J onas, et al., 1993; J onas and Syrett, 1987; Schleithoff, 1984). Infossil units, theLPturbinerequiresthelowest concentrationof impuritiesinthecycle, thatis, lowpartsper billion concentrations (1 ppb is 1 µg/liter). Steampurity is controlled by the purity of makeup, condensate, and feedwater and in drum boilersbyboiler water chemistry, boiler pressure, andcarryover. As a minimum, steam purity should be monitored by isokinetic samplingandbyanalysisof sodiumandcationconductivity(EPRI, 1986, 1994c, 1998b, 1998c, 2002a; J onas, 2000). Thecorrosivenessof thesteamturbineenvironmentiscausedby oneor moreof thefollowing: • Concentration of impurities fromlow ppb levels in steamto percent levels insteamcondensates (andother deposits) resulting intheformationof concentratedaqueoussolutions • Insufficient pH control and buffering of impurities by water treatment additivessuchasammonia • High velocity and high turbulence flow of low-pH moisture droplets(FAC) ThesituationisgeneralizedinFigure14, whichisaMollierdiagram showingtheLP turbinesteamexpansionlineandthermodynamic regions of impurity concentrations (NaOH, salts, etc.) and resultingcorrosion. Lowvolatilityimpuritiesinthe“salt zone” are presentasconcentratedaqueoussolutions. TheNaCl concentration can beas high as 28 percent. Notethat theconditions at thehot turbine surfaces (in relation to the steamsaturation temperature) PROCEEDINGSOFTHETHIRTY-SEVENTHTURBOMACHINERY SYMPOSIUM•2008 216 can shift fromthewet steamregion into thesalt zoneandabove. This can be the reason why disc stress corrosion cracking often occurs inthewet steamregions. Thesurfaces may behot because of heat transfer through the metal or because of the stagnation temperatureeffect (zeroflowvelocityat thesurfaceandchangeof kinetic energy of steamintoheat). Figure 14. Mollier Diagramwith LP SteamExpansion Line and Thermodynamic Regions for Impurity Concentration and CorrosionMechanisms. Theimpurity concentrationmechanismsinclude: • Precipitationfromsuperheatedsteamanddeposition. • Evaporationanddryingof moistureonhot surfaces. • Concentrationonoxidesby sorption. • Nonhomogeneousnucleationof concentrateddropletsandcrystals onsurfaces. Dissolvedimpuritiesdeposit fromsuperheatedsteamwhentheir concentrationexceedstheir solubilities, whichsharplydecreasesas thesteamexpands. Dependingontheir vapor pressure, theycanbe present as a dry salt or an aqueous solution. In the wet steam region, they areeither dilutedby moistureor couldconcentrateby evaporationonhot surfaces. Theregionof passivity for ironandlowalloy andcarbonsteels isnarrow, fallingwithinthepH rangeof 6to10. SincepH control inapower plant ismostly for theprotectionof thepreboiler cycle and the boiler, it often does not match the needs of the turbine surfaces. ThepHintheturbinedependsontemperature, mechanical and vaporous carryover of impurities, and water treatment chemicalsfromtheboiler andtheir volatility(distributionbetween thevapor andthesurfacefilm). Whenhydrochloricacidformsincycleswithammoniaall-volatile treatment (by decompositionof chlorinatedorganics, Cl Ϫ leakage frompolishers, or seawater or other cooling water leakagein the condenser), ammoniumchlorideforms in thewater, and theacid may beneutralized. However, becauseof itsvolatility, ammonium chloride is transported with steam into the turbine where it hydrolyzes, formingNH 3 gasandHCl. Deposits on turbine surfaces in units with sodium phosphate boiler water treatment (most drumboilers) areless corrosive(J onas andSyrett, 1987; EPRI, 1984b; J onas, 1985d). Sodiumphosphateis a better neutralizing agent through the cycle; fewer acids are transported into the turbine and phosphate frequently codeposits with harmful impurities, providing in-situ neutralization and passivation. This is most likely the reason for lower frequency of turbine corrosion in systems with phosphate water treatments. Measurementsof pittingpotential of discandbladingalloysconfirm thebeneficial effectsof sodiumphosphateinthepresenceof NaCl. Besides the corrosion during operation, turbines can corrode during manufacture (corrosive products from machining fluids, exposure to tool tip temperatures), storage (airborne corrosive impurities, preservatives containing Cl and S), erection (airborne impurities, preservatives, cleaning fluids), chemical cleaning (storage of acid in hotwells), nondestructive testing (chlorinated cleaning and NDT fluids), and layup (deposits plus humid air). Manyof thesecorrosivesubstancesmaycontainhighconcentrations of sulfur andchloridethat couldformacids upondecomposition. Decomposition of typical organics, such as carbon tetrachloride occursat about 300ЊF (150ЊC). Thecompositionof all of theabove compounds should becontrolled (maximumof 50 to 100 ppmS andCl eachhasbeenrecommended), andmost of themshouldbe removedbeforeoperation. Molybdenum disulfide, MoS 2 , has been implicated as a corrodent in power systemapplications (Turner, 1974; Newman, 1974). It can cause stress corrosion cracking of superalloys and steels by producingan acidic environment. Its oxidation products form low pH solutions of molybdic acid and even ammonium molybdate, whichcanformduringoperation, causingrapidattack of turbinesteels. MoS 2 has beenusedas athreadlubricant andin theprocess of disc-rotor assemblingwhenthediscs arepreheated and shrunk on the rotors. Analysis of disc bore and keyway surfaces oftenreveals thepresenceof molybdenumandsulfur. In steam, MoS 2 reduces thenotchstrengthof disc steel, to about 30 percent of itsstrengthinair. It hasalsobeenimplicatedinbolt and rotor shaft failures. Layupcorrosionof unprotectedturbinesincreasesrapidlywhenthe relativehumidity of air reachesabout 60percent. Whensalt deposits arepresent, pitting during unprotected layup is rapid. Pit growth in turbinebladeandrotor alloysinchloride-metal oxidemixturesinwet air isabout asfast asinaboilingdeaerated28percent NaCl solution. Turbinelayupprotectionbycleandryair isrecommended. Progress incontrollingturbinecorrosionthroughbetter control of thesteamchemistry includes(J onas, 1982, 1985d, 1994; EPRI, 1984b, 1986, 1994a, 1994c, 1997b, 1997c, 1998b, 1998c, 1999, 2002a; J onas, et al., 1993, 2000; J onas and Syrett, 1987; Schleithoff, 1984, “Progressin...,” 1981): • Decreasingconcentrationof corrosiveimpuritiesinmakeupand feedwater, lower air inleakageandcondenser leakage, etc. • Oxygenated water treatment for once-through fossil units for excellent feedwater chemistry andcleanboilers. • Layupprotection. • Turbine washing after chemical upsets to remove deposited impurities. • Reduction or elimination of copper and its oxides and their synergistic corrosion effects by reducing oxygen concentration, operating with a reducing (negativeoxidation-reduction potential [ORP]) environment and a low ammonia concentration, or by replacingcopper alloyswithsteel or titanium. PROBLEMS, THEIR ROOT CAUSES, ANDSOLUTIONS Steam turbine corrosion damage, particularly of blades and discs, has long been recognized as a leading cause of reduced availability (Scegljajev, 1983; McCloskey, 2002; Sanders, 2001; Cotton, 1993; J onas, 1985a, 1987; EPRI, 1981, 1998a, 2001b; J onas and Dooley, 1996, 1997; NERC, 2002). It has been estimated that turbinecorrosionproblemscost theU.S. utility industry asmuchas one billion dollars per year (EPRI, 1985b, 2001a; Syrett, et al., 2002; Syrett andGorman, 2003; J onas, 1986) andthat thecost for industrial turbines, whichsuffer similar problems, isevenhigher. In this section, themain corrosion problems found in LP turbines and their root causes are summarized, and solutions to reduce or eliminateeachproblemarediscussed. Thefieldmonitoringequipment showninFigure15canbeusedtodiagnoseandpreventmanycommon LP turbinecorrosion and deposition problems (EPRI, 1997c, 2001b; Jonas, 1994; Jonas, et al., 2007). Inaddition, therearealso monitors availabletodetect vibration, bladeandrotor cracking, steamleaks, air inleakage, rotor position, andwear of bearings(Jonas, et al., 2007). STEAMTURBINE CORROSION ANDDEPOSITSPROBLEMSANDSOLUTIONS 217 Figure 15. LP Turbine Troubleshooting Instrumentation Can IdentifySpecificCorrosiveConditions. (Courtesyof EPRI, 1997c) An overview of the low-pressure turbine corrosion problems together with erosion and other problems is given inTable 3. The problems arelistedaccordingto their priority andimpact withdisc rimblade attachment stress corrosion cracking being the highest impact problemtoday because of the long time required for weld repair or procurement of anewdiscor newrotor (uptosixmonths). Crackingof discs, corrosionfatigueof therotor shaft, andfatigueor corrosionfatigueof longblades canbecomeasafety issuebecause they can lead to perforation of the casing and other destructive events(EPRI, 1981, 1982a, 1998a; J onas, 1977; Turner, 1974). It is estimated that inadequate mechanical design (high steady and vibratorystresses, stressconcentration, andvibration) isresponsible for about 50 percent of the problems, inadequate steamchemistry for about 20 percent, and nonoptimumflow and thermodynamic design for about 20 percent. Poor manufacturing and maintenance practicesaccount for theremaining10percent of theproblems. Table3. LP TurbineCorrosion, Erosion, andDepositionProblems. Whenacorrosionproblemisdiscoveredduringinspectionor by equipment malfunction, thefailuremechanismandtheroot causes arenot alwaysknown. Evenwhenthedamagefitsadescriptionof awell-knownproblem(disc or bladecracking), replacement parts maynot bereadilyavailableandthedecisionfor what todohasto be made quickly. The main objectives in handling identified and potential problems are maintaining safety and avoiding forced outages. Thequestionsshouldbeasked: canweoperatesafelyuntil the next planned inspection or overhaul? If not, what is the safe inspectioninterval? If repeatedfailuresarelikelyandrepair would takealongtimeandleadtoalargelossof production, sparerotors may beagoodeconomical solution. Data collected by the North American Electric Reliability Council (NERC) for 1476 fossil units between 1996 and 2000 shows that LP turbines were responsible for 818 forced and scheduled outages and deratings, causing the utilities a 39,574 GWh production loss. The outages are often characterized as “low frequency high impact events.” (NERC, 2002) shows the components that were responsible for the failures as well as the MWh losses associated with the failures. Many of these outages arecausedby corrosion(except for bearings). Table4. Forced and ScheduleOutages and Deratings Caused by LP TurbineComponents for theYears 1996 to 2000 (1476 Units, 168Utilities). LifePredictionandInspectionInterval Experience shows that pits and ground-out stress corrosion crackscanremainin-servicefor several years, dependingonstress and environment. However, components containing high-cycle corrosion fatiguecracks should not beleft in-service. Procedures for predictionof residual lifeanddeterminationof asafeinspection interval have been developed for all major failure mechanisms including SCC, CF, fatigue, FAC, and creep. The procedures for SCC of turbinediscs (Clark, et al., 1981; EPRI, 1989; Rosario, et al., 2002), lowcyclecorrosionfatigue, andFAC(EPRI, 1996) have been successfully applied because all variables influencing these mechanisms can be reasonably predicted or measured. However, life prediction for high cycle corrosion fatigue and fatigue has not been so successful because the vibratory stresses and the corrosivenessof theenvironmentareusuallynotaccuratelyknown. Life prediction is based on results of inspection, fracture mechanicsanalysisof componentswithdefects, andapplicationof SCC andCF crack growthdata. Timeor number of loadcyclesto reach ductileor brittlefractureis predicted and asafety factor is applied to determine the time for the next inspection. In the procedure used by OEMs and Nuclear Regulatory Commission (NRC) for nuclear turbinesfor determiningtheinspectioninterval for turbine discs under SCC conditions, the safety factor of two wasappliedtothepredictedtime-to-failure. StressCorrosionCrackingof Discs Stress corrosion cracking of LP turbine disc keyways and blade attachments havebeen thetwo most expensivegeneric problems in largesteamturbines(CheruvuandSeth, 1993; EPRI, 1982a, 1982b, 1984a, 1984c, 1985a, 1985d, 1987b, 1989, 1991a, 1997a, 1998d; J onas, 1978; Speidel and Bertilsson, 1984; Clark, et al., 1981; Rosario, et al., 2002; Nowak, 1997; Kilroy, et al., 1997; Amos, et al., 1997; Turner, 1974; Newman, 1974; Parkins, 1972; Holdsworth, 2002). The keyway cracking problem has been resolved by redesigns of theshrunk-onor bolted-ondiscs, material replacement withlower strengthmaterial, andbyeliminationof thecorrosivedisc bore lubricants, based on molybdenumdisulfide (MoS 2 ), used in assemblingtherotor. SCC of bladeattachmentsof variousdesignsis still aproblem(EPRI, 1997d, 1998a, 2002c; Nowak, 1997). PROCEEDINGSOFTHETHIRTY-SEVENTHTURBOMACHINERY SYMPOSIUM•2008 218 There are several corrosion damage mechanisms and many factorsaffectingdiscs(Figure16). Typical locationsandorientations of SCC cracks inLP discs areshowninFigures 17and18. There has also been SCC in pressure balance holes. Most incidents of disc rimattachment cracking have been found in nuclear units, however, there have also been problems in fossil units. In an independent survey, 13 of 38 (35 percent) boiling water reactors (BWRs) and 28 of 72 (39 percent) pressurized water reactors (PWRs) reporteddiscrimattachment crackingwhile29of 110(26 percent) supercritical fossil units and only 20 of 647 (3 percent) subcritical units reportedcracking(EPRI, 1997d). Disc rimblade attachment cracking occurred in multiple-hook (steeple), fir tree attachment designs (typically occurs inthecorners of thehooks), in straddlemount dovetail and pinned-finger attachments, and in T-roots. It isalwaysassociatedwithstressconcentrations. Figure16. CrackLocations inTurbineDiscs withProbableImpurity ConcentrationandCorrosionMechanismsandCorrodents. (Courtesy of J onas, 1985a) Figure17. Typical LocationsandOrientationsof SCCFoundinLP TurbineDiscs. (Courtesyof EPRI, 1982a) Figure18. Typical Locations of Disc RimCracking. (Courtesy of EPRI, 1982a) Materials—All low alloy steels used for LP turbine rotors and discs are susceptible to SCC and CF in numerous turbine environments including pure water and wet steam. The strongest material factor influencing SCC is yield strength. At higher yield strength, SCCcrackgrowthratecanbeseveral ordersof magnitude higher than for lower yield strength materials. The purity of the material and the steel melting practices mostly influence the fracturetoughness, whichdeterminesthemaximumtolerablecrack sizebeforeadisc brittlefractureandburst. Fractography—SCC cracks of low alloy steels often initiate frompitsandpropagateintergranularlywithbranching. Theinitial part of thecrack canbecorrodedandfilledwithmagnetite. There couldbebeachmarksor stretchmarks, causedby overloadingthe crackduringoverspeedtesting. Another typeof beachmarkcanbe causedbychangesof theenvironment or byfatigue. Dependingon theratioof thesteady andcyclic stresses, thedisc crackscanbea mixture of intergranular and transgranular cracking. Figure 19 shows anSCC crack initiatingat theradius of theupper serration of L-1bladesteeple. Theintergranular crack initiatedfromapit. Figure 19. SCC Crack in the Upper Steeple Serration-L-1 Blade Attachment. The factors that determine the SCC crack initiation time and propagation rate include material yield strength, surface stress, temperature, and the local chemical environment. Some of these relationships are shown in Figures 6 and 7. At yield strengths above~135ksi (930MPa), theselow-alloy steels showhighSCC growthrates. PittingofteninitiatesSCC. Whencorrosivedepositsarepresent, pitting during unprotected layup can befaster than pitting during operation. This is because during the layup, there can be 100 percent relative humidity and there is oxygen present. At high stresses, above the elasticity limit of the material, pitting is enhanced through the mechanism identified as stress induced pitting(Parkins, 1972). In somecases, bladeattachment cracking isacombinationof stresscorrosioncrackingandcorrosionfatigue becauseof theeffectsof bladevibration. Root Causes SCC of discs (at keyways, bores, and blade attachments) is caused by a combination of high surface stresses, a susceptible material, andoperational andshutdownenvironments. Design-related root causes are the most important and prevalent. They include high surfacetensilestresses and stress concentrations, and useof highstrengthmaterials. Sourcesof stressesthat contributetoSCC of discsinclude: • Basic centrifugal load caused by rotor rotation. Locally high concentrationof centrifugal loads causedby variationinthegaps (gauging) betweenbladeanddisc rimattachment. • Residual machiningstresses. • Vibratorystresses—interactionof SCCandcorrosionfatigue. Also, vibratory stresses reducethelifeof thecracked disc when theflaws reachasufficient sizethat fatiguebecomesadominant mechanism. Steamchemistry root causesof SCC andCF crackinginclude: • Operating outside of recommended steam purity limits for long periods of time; sometimes caused by organic acids from decompositionof organic water treatment chemicals. STEAMTURBINE CORROSION ANDDEPOSITSPROBLEMSANDSOLUTIONS 219 Condenser leaks—minor butoccurringover alongperiodof time. • Condenser leaks—major ingress, generally one serious event, andthesystemandturbinenot subsequently cleaned. • Water treatment plant or condensate polisher regeneration chemicals(NaOH or H 2 SO 4 ) leak downstream. • Improperly operated condensate polisher (operating beyond ammoniabreakthrough, poor rinse, etc.). • Shutdownenvironment: poor layuppracticespluscorrosivedeposits. Sodium hydroxide is the most severe SCC environment encountered in steam turbines. The sources of NaOH include malfunctioning condensate polishers and makeup systems and improper control of phosphate boiler water chemistry combined withhighcarryover. Many other chemicalscanalsocauseSCC of low alloy steels. The chemicals used in turbine assembly and testing, such as molybdenumdisulfide (lubricant) and Loctite™ (sealant containing high sulfur), can accelerate SCC initiation (Turner, 1974; Newman, 1974). Solutions In most cases where material yield strength is <130 ksi (895 MPa), the solution to disc SCC is a design change to reduce stressesat critical locations. Thishasbeenachievedbyeliminating keyways or even disc bores (weldedrotors) andby larger radii in thebladeattachments. Higher yieldstrength(>130ksi, 895MPa) low alloy steel discs should be replaced with lower strength materials. Thegoal istokeeptheratioof thelocal operatingstress toyieldstressaslowaspossible, ideallyaimingfor theratiostobe less than 0.6. Minimizing applied stresses in this manner is most beneficial inpreventinginitiationof stresscorrosioncracks. Once cracks begin to propagate, a reduction in stress may be only marginally effectiveunless thestress intensity can bekept below ~10 to 20 ksi-in 1/2 (11 to 22 MPa-m 1/2 ). This is because of the relativeindependenceof thecrack growthrateover abroadrange of stressintensities. For many rimattachment designs, suchlevels of appliedstressintensity areimpossibletoachieveonceaninitial pit or stress concentration has formed. An emerging solution to disc rimstress corrosion cracking is a weld repair with 12%Cr stainless steel. Another solution has been to shot peen the blade attachmentstoplacethehook fit regionintocompression. Goodcontrol of thesteampurityof theenvironment canhelpto prevent or delay theSCC. Maintainingtherecommendedlevelsof impurities during operation and providing adequate protection duringshutdowncanhelpminimizetheformationof depositsand corrosive liquid films, and lengthen the period before stress corrosion cracks initiate. The operating period(s), events, or transients that are causing excursions in water and steam chemistry shouldbeidentifiedusingthemonitoringlocationsand instrumentationrecommendedintheindependent water chemistry guidelines(EPRI, 1986, 1994c, 1998b, 1998c, 2002a; J onas, et al., 2000) andspecial monitoringasshowninFigure15andelsewhere (EPRI, 1997c, 2001b; J onas, et al., 2007). CorrosionFatigueand StressCorrosionCrackingof Blades LP turbine blades are subject to CF, SCC, and pitting of the airfoils, roots, tenonsandshrouds, andtiewires(Holdsworth, 2002; EPRI, 1984c, 1984d, 1985c, 1985d, 1987c, 1991b, 1993b, 1994b, 1998d; J affe, 1983; Evans, 1993; Singh, et al.; BLADE-ST™, 2000). Figure 20 depicts the typical locations on an LP turbine rotatingbladethatareaffectedbylocalizedcorrosionandcracking. Inaddition, thebladesurfacesarealsosubjecttofatigue, deposition, water dropleterosion, andforeignobjectdamage. CF istheleading mechanismof damage. It is aresult of thecombination of cyclic stressesandenvironmental effects. Therearealwaysenvironmental effectsinfatiguecrackinginLPturbines(EPRI, 1984d, 1984e) and all fatigue cracking should be considered corrosion fatigue. The fatigue limit of all turbine materials in turbine environments includingpuresteamis lower thantheair fatiguelimit. Corrosion fatiguecracksoftenoriginatefrompits. Figure20. Typical Locationsof CrackingandLocalizedCorrosion onLP TurbineRotatingBlades. ThereHasAlsoBeenSCC andCF CrackingintheTiewireHoles. (Courtesyof EPRI, 1998a) Figures 21 and 22 show corrosion fatigue cracks of L-1 blade root and airfoil, respectively. Figure 23 illustrates pitting in the blade tenon-shroud area, which sometimes initiates corrosion fatiguecracking. Fractographyshowsthat CF bladecrackingoften initiates fromapit, continuingfor 50to150mils (1.3to3.8mm) by intergranular cracking and then proceeding as a flat fatigue fracturewithbeachmarksandstriations. Figure21. CorrosionFatigueof L-1BladeAttachment. (Courtesy of EPRI, 1998a) Figure 22. Corrosion Fatigue of L-1 BladeAirfoil. (Courtesy of EPRI, 1981) PROCEEDINGSOFTHETHIRTY-SEVENTHTURBOMACHINERY SYMPOSIUM•2008 220 • Figure23. PittingintheL-1BladeTenon-ShroudArea. (Courtesy of EPRI, 1981) Damageby corrosionfatigueoccurs inthelast fewrows of LP turbines, mostly in the phase transition zone (PTZ) (salt zone shown in Figure14). ThePTZ moves according to load changes, but istypicallynear theL-1rowinmost LP fossil turbines(Figure 24). Units that increasecycling duty may besubject to worsened corrosion fatigue problems. As the unit is ramped up and shut down, theblades pass throughresonancemorefrequently andthe phasetransitionzoneshifts, potentiallyaffectingmorestagesduring the transients. In addition, the steam purity can be significantly worseduring transients than during steady-stateoperation, and if theunit is shut down as part of thecycling operation, significant degradation of the local environment can occur (deposits and humidair). Figure24. Distributionof BladeFailuresinU.S. Fossil Turbinesby Row. (Courtesyof Power, 1981) Causesof bladeandbladeattachment failuresarelistedinTable5 (J onas, 1985a). Tofindthetruecausesof corrosion, it isessential to analyzethelocal temperature, pressure, chemistry, moisturedroplet flow, andstressconditions. Theseanalysesareoftenneglected. Table5. Causesof BladeFailuresinLP, IP, andHP SteamTurbines. Root Causes Highstressesandmarginal steamchemistryactingtogether arethe most frequent root causes. Corrosion fatigue cracks are driven by cyclic stresses withhighmeanstress playingalargerole. Figure10 shows how fatiguestrength is reduced at high mean stresses. It has been estimated that the corrosion fatigue limit for long LP turbine bladesintheareaof highmeanstressisaslowas1ksi (7MPa). Rough surfacefinishandpittingoftenshortenthetimetocrack initiation. Cyclic stresses are caused by turbine startups and shutdowns (low number of cycles, high cyclic stresses), by the turbine and bladesrampingthroughcritical speedsat whichsomecomponents areinresonance(highamplitude, highfrequency), andby over 15 causes of flow induced blade excitation during normal operation that include: • Synchronous resonanceof theblades at aharmonic of theunit runningspeed. • Nonuniformflows. • Bladevibrationinducedfromavibratingrotor or disc. • Self-excitationsuchasflutter. • Randomexcitation-resonancewithadjacent blades. • Shock waves in the transonic flow region and shock wave- condensationinteraction. • Bad blade design with wrong incidence flow angle and flow separation. Elevated concentrations of steam impurities (particularly of chloride, sodium, and sulfate) and the resulting deposition and concentrationby evaporationof moistureareunderlyingcausesof corrosion fatigue. When feedwater, boiler water, and steam impurity levels exceed recommended limits, cycle chemistry can beacontributor or eventheroot cause. Poor shutdownandlayup procedures are primary contributors to aggressive environments that canleadto pittingandcorrosionfatigue. Highsteamsodium or cation conductivity may indicate conditions that can lead to rapid accumulation of deposits and concentrated liquid films on bladesurfaces. Solutions Thesolutionstobladecorrosionfatigueproblemsinclude: • Design of blades that arenot in resonancewith running speed anditsharmonic frequenciesor withany of theexcitationsources listedabove. • Designwithfrictiondamping. • Eliminationof thesourcesof excitation. • Reductionof meanandalternatingstressesbydesign(lower stress concentrations, etc.). • Better materials, suchasby avoidinghighstrengthalloys, using materialswithhighmaterial damping, or usingtitaniumalloys. • Improvement of steamchemistry. Long-termactionsfor dealingwithcorrosionfatiguebeginwith economic and remaining life assessments. Depending upon the severityof theproblemandthecoststoeliminateit, somesolutions maynotbepractical for all circumstances. Theavailableprevention strategies fall into four main categories: redesigning thebladeto reduce resonance, redesigning the blade or attachment to reduce stresslevels, improvingsteampurity, and/or changingthematerial or surface(better surfacefinish, shot peening) of theblade. Stressreductionoptionsinclude: • Changing the vibration resonance response of the blade by design modification (adding or reducing weight of the blade, STEAMTURBINE CORROSION ANDDEPOSITSPROBLEMSANDSOLUTIONS 221 changing to free standing, grouped or fully connected shrouded blades, moving tiewires or tenons, shroud segment integrally machinedwiththeblade—notenons, etc.). • Changingtheresponseof thebladesby mixedtuning. • Increasingthedampingof theblades. • Changingoperatingprocedures; for example, avoidingoff-design operation such as very low load, high backpressure operation (which can cause stall flutter), and changing rotational speed duringstartups. Optionstoimprovetheturbineenvironment include: • Controllingimpurityingress(reduceair inleakage, plugleaking condenser tubes, etc.). • Changingunit operatingprocedures, particularly for shutdowns andstartups. • Control of boiler carryover bydrumdesignandwater chemistry. • Optimizingor changingfeedwater andboiler water treatment to reduceconcentrationof impuritiesinsteam. Another option to improve the environment at surfaces is to improvethesurfacefinishof blading. Depositionandsubsequent concentration of impurities areafunction of bladesurfacefinish (EPRI, 2001b), andthisimprovementmayhelpslowtheaccumulation of impurities. Improving the continuous monitoring of steam chemistry (sodium, cation conductivity) will help to verify improvementsintheenvironment. If such changes are not sufficient, then changing to a more corrosion resistant material, such as a material with a higher chromiumcontent or atitaniumalloy, isgenerally recommended. It shouldbenotedthat higher strengthmaterials areoftenmore susceptible to stress corrosion cracking. It has been shown that 403SS, with yield strength above ~90 ksi (620 MPa), becomes susceptibletoSCC. Pitting Pittingcanbeaprecursor tomoreextensivedamagefromCF and SCC, althoughextensivepittingof bladescanalsocausesignificant loss of stageefficiency by deteriorating thesurfacefinish (EPRI, 2001b). Pitting is found in awiderangeof components. It occurs most prevalently during shutdown when moisture condenses on equipment surfacesandasaresult, it canbefoundinstagesof the turbine that are dry during operation. However, it can also occur duringoperation, particularly increvices(crevicecorrosion). In LP turbines, pitting is primarily found on theturbineblades andblade-disc attachments, particularly inthe“salt zone” (Figure 14). Thereislittlepittingonwetstagesbecausecorrosionimpurities arewashedawaybythesteammoisture. Pittingisfrequentlyfound inthebladetenon-shroudcrevicesbecause, oncecorrosiveimpurities enter, they cannot be removed. Titanium alloys are the most resistant topittingcorrosion, followedbyduplexferritic-austenitic stainless steels (Fe-26Cr-2Mo), precipitation hardened stainless steels(Fe-14Cr-1.6Mo) and12%Cr steels. Root Causes Steamand deposit chemistries are the main causes of pitting, withchloridesandsulfatesbeingthemaincorrodents. Copper and iron oxides acceleratepitting by providing thematrix that retains salts. Copper oxides also transport oxygen to thecorrosion sites. Dissolvedoxygendoesnot concentrateintheliquidfilmsforming onbladesurfaces duringoperationbut does, however, accumulate in the liquid films and wet deposits that can form during unit shutdown if proper layuppractices arenot used. Therehavebeen casesof severebladepittingrequiringbladereplacement onbrand new rotors fromwhich preservatives were stripped leaving them exposedtoseasalt duringprolongederectionperiods. Solutions The first line of defense against pitting is controlling steam purity. This will improve the local environment produced during operation and decrease the amount of deposition. The chemistry guidelinesestablishedbyanindependent researchanddevelopment firm(EPRI, 1986, 1994c, 1998b, 1998c, 2002a; J onas, etal., 2000) particularlyfor cationconductivity, chloride, andsulfate, shouldbe followed. Thereshouldbealayupprotectionof theLP turbinesby dehumidifiedair, vapor phaseinhibitors, or nitrogen. After asteamchemistry upset, such as alargecondenser leak, theturbineshouldbewashedonlineor after adisassembly. Doing nothing may result in multimillion dollar corrosion damage requiringrotor replacement. Flow-AcceleratedCorrosion Flow-acceleratedcorrosionof carbonandlow-alloysteelsinthe steam path two-phase flow has been less widespread in fossil plants than in nuclear plants; however, it has occurred at some locations (EPRI, 1996; 1998a; Kleitz, 1994; J onas, 1985b; SvobodaandFaber, 1984) suchas: • Wet steamextractionpipesandextractionslots. • Exhaust hoodandcondenser neck structure. • Casings. • Rotor glandandother seal areas. • Disc pressurebalanceholes. • Rimandsteeplesof last rowdisc. • Rotor shaft—last disc transition. • Leakinghorizontal joint. • Transitionbetweenthestationary bladesandthebladering. Whilemostcasesof flow-acceleratedcorrosiondamageareslow to develop and are found during scheduled inspections, FAC of piping and turbine casing horizontal joints can lead to leaks and FAC of rotorsanddiscscaninitiatecracking. Root Causes Root causes of FAC in theturbine(EPRI, 1996, 1998a; J onas, 1985b) include: • Susceptiblematerial: carbonsteel or low-chromiumsteel. • Locally highflowvelocitiesandturbulence. • Highmoisturecontent of thesteam. • Lowlevelsof dissolvedoxygen, excessoxygenscavenger. • LowpH of moisturedroplets. • Water/steamimpurities. Solutions It is recommended that acomprehensiveFAC control program be implemented, including an evaluation of the most susceptible pipingandother components(EPRI, 1996). Areasof local thinning need to be periodically inspected and repaired. Approaches to piping repair include replacement with low alloy steels, weld overlay, and plasmaarc and flamespraying to protect susceptible surfaces. Thematerial appliedshouldhavehighchromiumcontent. If a component is replaced, the material of the new component should contain some chromium. For example, carbon steel pipe should be replaced with 1.25 percent or higher chromiumsteel. Little can be done about changing the moisture concentration in fossil turbines. Steam chemistry improvements through better control of feedwater andboiler water chemistry, suchasreduction of organic acids, could result in increase of pH of the early condensateandlessFAC (EPRI, 1997b, 1997c, 1999). PROCEEDINGSOFTHETHIRTY-SEVENTHTURBOMACHINERY SYMPOSIUM•2008 222 Treatment of feedwater, such as maintaining high pH levels (above 9.6) and elevated oxygen concentrations, can also reduce FAC intheturbine. Other Phenomena(Noncorrosion) Although not specifically corrosion related, there are other problems that occur in steam turbines including: deposition on bladesurfaces; water dropleterosionof wetstageblades; lowcycle thermal fatigueof heavyhightemperaturesectionsof rotor, casing, and pipes; solid particleerosion of turbineinlets and valves; and water inductionandwater hammer. DepositiononBladeSurfaces Deposits are the result of impurities in the feedwater, boiler water, andattemperatingwater beingcarriedover into theturbine (J onasandDooley, 1997; EPRI, 1997b, 2001b; J onas, et al., 1993; J onas, 1985d). All impurities are soluble in superheated and wet steamand their solubility depends on pressure and temperature. The steam leaving the steam generator is at the highest steam pressure and temperature in the cycle. As it passes through the turbine, the pressure and temperature decrease, the steamloses its ability to hold the impurities in solution, and the impurities precipitateanddeposit ontheturbinebladesandelsewhere. The main impurities found in turbine deposits are magnetite, sodiumchloride, andsilica. It takesonlyafewhoursof achemical upset, suchas amajor condenser leak or aboiler carryover event, to build up deposits, but it takes thousands of hours of operation withpuresteamtoremovethem. The impact of deposits on turbine performance is the most pronounced in the HP section. Performance loss depends on deposit thickness, their location(steampressure), andtheresulting surface roughness (EPRI 2001b). Deposits will change the basic profile of the nozzle partitions resulting in losses caused by changes inflow, energy distribution, andaerodynamic profiles, as well as by surfaceroughness effects. Thesechanges can result in large megawatt and efficiency losses. With replacement power typicallyover $100/MWhandcostingasmuchas$7000per MWh inthesummer of 1998, thesavings fromreducingthis deposition canbevery high. IntheLP turbine, deposits areoftencorrosive, they canchange theresonant frequency of blades, increasethecentrifugal loadon blade shrouds and tenons and, in the transonic stages, they can influencethegenerationof shock waves. Solutions—Optimization of cycle chemistry (J onas, 1982, “Progress in...,” 1981; EPRI, 1986, 1994c, 1998b, 1998c; 1985d, 2002a; J onas, etal., 2000, 2007; ASME, 2002) istheeasiestmethod for reducing impurity transport and deposition. Theoptimal cycle chemistry will result inreducedcorrosionandminimizedimpurity transport. Thisisespeciallyimportantif copper alloysarepresentin thesystembecausetheoptimal feedwater pH for copper alloysand ferrousmaterialsarenotthesameandtheincorrectpHcanresultin highlevelsof ironor copper transport. Other methodsfor managing depositiononbladesurfacesinclude: • Specify goodsurfacefinishes (polishedblades) onall newand replacement blades. • Determine the effect of erosion and deposition on maximum MWandefficiencywithavalveswideopen(VWO) test. Thedata fromseveral VWOtests canbecomparedtodeterminetherateof MW loss and MW versus chemistry and operation. This can be usedtooptimizethesystem. • Turbine washing to reduce deposition and MW losses and improveefficiency. Both, aturbinewash of an assembledturbine andawashof adisassembledrotor canbeusedtoremovesoluble deposits. To remove corrosive salt deposits, several days of washing may be needed. A wash is usually completed when the concentrationof corrosiveimpurities, suchassodiumandchloride, inthewashwater islessthan50ppb. Water Droplet Erosion In thelast stages of theLP turbine, thesteamexpands to well belowsaturation conditions andaportionof thevapor condenses intoliquid(EPRI, 2001b; Ryzenkov, 2000; Oryakhin, et al., 1984; Rezinskikh, et al., 1993; Sakamoto, et al., 1992; Povarov, et al., 1985; Heyman, 1970, 1979, 1992). Although the condensed dropletsarevery small (0.05to1µm, 2to40µindiameter), some of themaredepositedontosurfacesof thestationary bladeswhere theycoalesceintofilmsandmigratetothetrailingedge. Herethey aretornoff bythesteamflowintheformof largedroplets(5to20 microns, 0.2 to 0.8 mils). These droplets accelerate under the forcesof thesteamactingonthemand, whenthey arecarriedinto theplaneof rotationof therotatingblades, theyhavereachedonly afractionof thesteamvelocity.Asaresult, thebladeshitthemwith avelocitythat isalmost equal tothecircumferential velocityof the blades, whichcanbeas highas 640m/s (2100ft/s) inafossil LP turbine. Water droplet erosion typically occurs in the last two to threerows of theLP turbines in fossil fired units. Thedamageis most commonontheleadingedgeandtipof thebladesandalong theshroud. In turbines operating at low load for long periods, such as cycling and peaking units, reversed flow of steam caused by windageandactivationof hoodspray canerodethetrailingedges of blades. Thin trailing edges with erosion grooves can become fatigue or corrosion fatigue crack initiation sites. Less frequent erosion damage locations include LP turbine glands and seals, stationary blades, blade attachment sections of discs, disc flow holes(impulsedesign), andtheLP rotor at thegland. The effects of steamand early condensate chemistry on water droplet erosionarenot known, but studieshavefoundthat NaCl in thedropletssignificantlyreducestheincubationperiodfor erosion. In addition, pH was found to have a strong impact on both the incubation period and the erosion rate. At higher pH, both the maximumandsteadystateerosionratesdecreasewhiletheincubation periodincreases(Ryzenkov, 2000; Povarov, et al., 1985). Solutions—There are several options available for reducing the amountof damagefromliquiddropletimpact. Therearetwoprincipal options: protection of the leading edge by a hard material and collectionanddrainageof moisture. Table6outlinestheseoptions. Table6. Long-TermActionsfor ReducingMoistureErosiononLP TurbineBlades. TurbineInspectionandMonitoring Adequate turbine inspection methods are available to detect corrosion and deposition. These NDT methods include visual, magnetic particle, ultrasonic, dye penetrant, eddy current, and radiographic techniques. Modern turbine designs consider the accessibility of individual componentsby inspectionprobes. Monitoring techniques include stress and vibration monitoring (i.e., vibrationsignature), temperatureandflowmeasurement, and water and steamchemistry sampling and analysis. (J onas, 1982, 1985d, 1986, 1994; EPRI, 1984b, 1994a, 1994c, 1997b, 1997c, 1998c, 1999, 2002a; J onas, et al., 1993, 2000; J onas and Syrett, 1987; Schleithoff, 1984; “Progress in...,” 1981). An advanced expert systemhas been developed (EPRI, 1994c) for the use by stationoperatorsandchemists, whichautomaticallydeterminesthe STEAMTURBINE CORROSION ANDDEPOSITSPROBLEMSANDSOLUTIONS 223 problems and recommends corrective actions. There are also monitoring methods for online diagnosis of the environments on turbinesurfaces andcorrosion(J onas, 1994; EPRI, 1994a, 1997b, 1997c, 1999; J onas, et al., 1993). MISSINGKNOWLEDGE It is estimated by the author that 70 percent of knowledge to solve and prevent corrosion problems in steam turbines is available. Thepercentageof availableknowledgefor understanding theeffects of stress and environment is much lower than that for solving the problems, about 40 percent. The knowledge that is missingor needsimprovement includes: • ThresholdstressrequiredtoinitiateSCC inbladeattachments. • Effects of steeplegeometry (stress concentrations and size) on SCC andCF. • Effects of overloads during heater box and overspeed tests onstress redistributionandSCC insteeples, bladeroots, anddisc keyways. • Effectivenessof grindingout SCC andCF cracksasacorrective measure. • Effectsof organicwater treatmentchemicals, andorganicimpurities onSCC, CF, andpittingandcompositionof water droplets. • Effectsof electrical chargescarriedbywater dropletsoncorrosion. • Effectsof galvanic couplingof dissimilar materials, suchasthe blade-steeple, oncorrosion. • Effects of residues of preservatives andLoctite™ on SCC, CF, andpittingof bladeattachments. • Effectsof bladetrailingedgeerosiononcracking. • Acceleratedstresscorrosiontesting. • Effects of variable amplitude loading on CF crack initiation andpropagation. • Effectsof water dropletpHandcompositiononerosionandhow topredict erosion. • Effectsof shotpeeningtoreducestressesandSCCof bladeroots anddisc steeples. • Understanding of the basic mechanisms of stress corrosion, corrosionfatigue, fatigue, andstressinducedpitting. CASE HISTORIES StressCorrosionCrackinginFinger-StyleDovetails Unit: Thestationconsistsof three805MW, once-throughboiler, supercritical, coal-fired units that went into operation between 1974and1976. Eachof thethreeunits has two LP turbines (LPA andLPB). Problem: In 1995, in Unit 1, abladefailedin theL-1rowat a tiewire hole of the leading blade of a four-blade group (Nowak, 1997; Kilroy, et al., 1997). Threedamagedgroups of blades were removed for replacement. A wet fluorescent magnetic particle examinationof thefinger-stylediscattachmentsfoundhundredsof crack-likeindications, whichwerelater identifiedasSCC. Cracks ranbothaxially andradially, andweredeep(Figure25). Figure 25. Locations in Finger and PinAttachments Where SCC HasBeenFound. (Courtesyof EPRI, 1997) Inspection: Inspection procedures and the acceptance criteria that would be applied were developed before the outage. Eventually NDE inspectionfoundsomedamageineachof the12 ends of all six rotors. Severecrackingwas foundinall four-rotor endsinUnit 2andinUnit 3onbothendsof LPB. Next inseverity werebothendsof Unit1LPB andUnit3LPA; withafewindications in Unit 1 LPA. In the heavily cracked rotors, damage was most severeinFingers3, 4, and5withlittleor nocrackinginFingers1 or 6. Cracking was evenly distributed between admission and dischargesides in Fingers 3 and 4 with somewhat morecracking ontheadmissionsideinFinger 5. Results of metallurgical analysis: A metallurgical evaluation confirmedthepresenceof extensivepitting. Crackswerefoundtobe intergranular, highly branched, and oxide-filled. The metallurgical examinationwasunabletodetectthepresenceof specificcontaminants on the fracture surface. Sampling of deposits, which had occurred elsewhere in the cycle (crossover piping, bucket pins, and bucket fingers) duringprior outages, hadfoundindicationsof sodium, sulfur, andchlorideandsodiumhydroxidewasfoundby x-raydiffractionin crossover piping. Analysis of samples fromtwo rotors showed that thechemical compositionwaswithinthespecificationfor ASTMA470Class7 material. Tensilestrengthaveraged126.5ksi (870MPa) for thetwo specimens; yieldaveraged112.5ksi (775MPa). Review of cycle chemistry: Several cycle chemistry changes had been madeover timein responseto events such as improved technology andinformation, andupsets causedby condenser tube leaks, demineralizer breaks, andvariationsinboiler water makeup. Themainwater chemistryproblemwasoperationwithmorpholine, whichresultedinpoor condensatepolisher performanceandhigh concentrations of sodium hydroxide in feedwater and steam. A heavy deposit of sodiumhydroxide was found in the crossover piping. Theunit hadbeenchangedto oxygenatedtreatment about thetimeof thediscovery of thestresscorrosioncracks, whichhas sinceresultedinbetter water andsteampurity. Resultsof stressanalysis: A finiteelement stressanalysisshowed that Fingers 3, 4, and5of thegroupwerethemost highly stressed. Themaximumequivalent elastic stresses aroundthepinholes were ~222ksi (1530MPa), ~114ksi (786MPa) at theinner land, and~89 ksi (614MPa) at theouter land(closer todiscouter diameter [OD]). Thedifferencesbetweeninner andouter landstressesagreedwiththe observation that field cracking was more severe at the inner land. However, no SCC was discovered in the flat portion of the disc fingersaroundtheholes, wherethestresswasnearly twiceashigh. Closer examinationfoundthat theflat surfacenear thepinholes did not show cracks nucleating from the bottom of the pits, whereas intergranular cracks appeared in thepits along theledge wherethepitswerelinkedtoformcontinual flaws. Root causes: There were two root causes of this massive problem: high design stresses and improper feedwater chemistry using morpholine, which resulted in the presence of NaOH and other impurities in steam. One can also speculate about the contributionsof thelocal stressconcentration, poor surfacefinish, andresidual machiningstresses. Economic analysis: An economic analysis was performed and the costs considered included: replacement of rotors in-kind, replacementwithanimprovedrotor steampaththatwouldimprove unit heat rate by 1.2 percent, rotor weld repair, outage duration costs, performance changes, reduced generation capacity from pressureplates, fuel pricing, andreplacement energy costs. Actions: As a temporary fix, the affected blade rows were removed and nine pressure plates were installed out of the 12 possible locations. A pressure plate is a temporary device that provides a pressure drop when installed as a replacement for a removedbladerow. A decisionwasmadetopurchasetwonewfully bladed rotors and to refurbish theexisting rotors. For Unit 1, new rotors were purchased fromthe original equipment manufacturer (OEM). ThetwoLP rotorsremovedfromUnit 1wereweldrepaired PROCEEDINGSOFTHETHIRTY-SEVENTHTURBOMACHINERY SYMPOSIUM•2008 224 using12percent chromiummaterial appliedwithasubmergedarc weldprocess, andinstalledinUnit 3. Material testingandanalysis were used to determine the expected lifetime of the refurbished partsagainstdamagebySCC, andbothhigh- andlow-cyclefatigue. Turbineandcrossover pipeswerecleanedtoremovedeposits. MassiveSCC of Disc Rim—Supercritical Once-ThroughUnit after OnlyFiveYearsof Service Problem: Stress corrosion cracking of the L-1 stage disc was discoveredduringroutineinspection(Figure26). Figure 26. Massive SCC of L-1 Disc Caused by High Concentrationof NaOH inSteam. Root cause: The SCC was caused by poor performanceof the condensate polishers that operated in H-OH form through ammonia breakthrough when they released Na + . The sodium reactedwithwater formingNaOH anddepositedintheturbine. Actions: Rotor replaced and operation of condensate polishers andmonitoringwasimproved. StressCorrosionCrackingof Bolted-onDiscsinOneTypeof LP Turbine Problem: In theeffort to accommodatelonger L-0 blades, one typeof LP turbinewas originally designedwiththebolted-onlast disc madeof NiCrMoV low-alloy steel andheat treatedto ayield strengthof upto 175ksi (1200MPa). This steel was foundto be susceptibletoSCC. Root cause: Theroot causeof this problemwas design with a highstrengthmaterial that isvery susceptibletoSCC. Solution: All of these discs in many power plants had to be replaced with a lower strength material because of the danger of SCC failuresinall typesof steamenvironments. CorrosionFatigueof aBladeAirfoil Unit: A 400 MW reheat unit with a once-through boiler, seawater cooling, andmixedbedcondensatepolishers. Problem: During a three-month period, condenser cooling water leakage occurred periodically. Cation conductivity in the condensateandfeedwater increasedupto2µS/cmfor about 30to 60minutesper day. At theendof thethree-monthperiod, vibration wasdetectedintheLP turbine(EPRI, 1998a) Damage: Theturbinewas openedandfivebrokenfreestanding L-2 rotating blades were found. According to the turbine design data, the broken blades were in the phase transition zone. The bladeswerebrokenatthetransitionbetweenreddishdepositsatthe bladefoot and clean metal in theupper part of theblades (phase transitionontheblade). A laboratory investigationfoundchloride andsodiumatthecracksiteandconfirmedcorrosionfatigueasthe underlyingmechanism. A calculationof vibrationfrequencies did not showabnormal conditions. Root cause: Improper water chemistry conditionscausedby the condenser tubeleak. Actions: The broken blades were replaced, the condenser leak was repaired, and adampening (lacing) wirewas introduced into thebladedesign. By doing so, both theenvironmental and stress contributorswerereduced. CorrosionFatigueof NumerousModificationsof L-1Blades Problem: Onetypeof LP turbineexperiencedcorrosionfatigue failures of the L-1 blade airfoil, which was modified and redesigned over 10 times. Fatigue failures occurred in periods rangingfromsixweekstoover 10yearsof operation. Thepresence of chloride in the blade deposits at concentrations above 0.25 percent caused pitting, which accelerated crack initiation. In one case, new16inch(40cm) bladeswereforgedintwodiesandmost of thebladesforgedinoneof thedies(butnonefromtheother die) failedwithinabout six weeks. Rootcause: Therootcauseof thisfastCF failurewasanoff-design blade geometry caused by wrong die dimensions, which brought thesebladesintoresonance. Thefailureaccelerationwascausedby corrosiveimpuritiesinsteam, mainly chloride. Actions: Redesigned, better tuned blades were installed and improvedcontrol of water andsteamchemistry wasinitiated. MassivePittingof aTurbine(HP, IP, andLP) after BrackishWater Ingress Problem: A separation of the welded end of the condensate sparger caused breakage of condenser tubes and ingress of brackish water into a once-through boiler cycle. Because of the poor reliability of the water chemistry instrumentation, the instrument readingswereignoredandthetroublewasnoticedafter therewasalmost noflowthroughthesuperheater becauseof heavy deposits. Theturbinewith seasalt deposits was left assembled in thehighhumidity environment andwasonly opened11daysafter thecondenser tubefailure. It wasfoundthat thewholeturbinewas severely rustedandpittedandit waseventually replaced. Mixed bed condensate polishers were not able to protect the cycle against the massive ingress of brackish water. They were exhaustedwithinafewminutes. Root cause: Theroot causeof theimpurityingresswasafailureof thecondensatesparger. Poor water chemistrymonitoringandcontrol and the long delay in beginning turbine damage assessment and cleaningsignificantly contributedtotheamount of damagecaused. Actions: Therotorswerereplacedandnewchemistrymonitoring instrumentationwasinstalled. L-0BladeCorrosionFatigue CrackingCausedbyTrailingEdgeErosion Problem: After 14to18yearsof serviceof onetypeof 400MW turbine, there were five cases of L-0 33.5 inch (85 cm) long shroudedbladefailuresabout 5inches(12.7cm) belowthetip. The CF cracksoriginatedat theerodedandthinnedtrailingedge. When thebladetipsseparated, thebladefragmentshadsuchkineticenergy that they penetratedover 2inches (5cm) of carbonsteel condenser struts. Thebladematerial wasmartensitic12%Cr stainlesssteel with a Brinell hardness of ~345. All affected units were similar drum boiler units, some on all volatile treatment (AVT) and some on phosphatetreatment. Steamchemistryintheaffectedunitswasgood anddidnot play aroleintherateof erosionor cracking. Root cause: Erosion caused by frequent operation at low load with the hood sprays on and reversed steamflow, lack of proper early inspection, andbladedesignwiththintrailingedge. Actions: Heavily eroded and cracked blades were replaced, shallow erosion damage was polished, and similar turbines were inspectedfor damage. StressCorrosionCrackingof Dovetail Pins Problem: A dovetail pin, which penetrates both the rotating bladeandthewheel dovetails, holdsabucket inplaceonthewheel of arotor. Inmanyplantsduringthe1970sandearlier, dovetail pins hadsufferedstresscorrosioncracking, althoughnolost bladeshad resulted. The material originally used for the dovetail pins was similar toASTMA681GradeH-11tool steel, withachemistry of Fe-5.0Cr-1.Mo-0.5V-0.4C at a strength level of 250 to 280 ksi STEAMTURBINE CORROSION ANDDEPOSITSPROBLEMSANDSOLUTIONS 225 (1715to1920MPa). Thismaterial isstill usedinrarecaseswhere thehighest strengthisrequired. Root cause: Useof highstrengthmaterial, whichis susceptible toSCC, combinedwithhighbendingstresses. Actions: The approach to solving the cracking problem was twofold: changing the material chemistry and strength and using steel ball shot peeningtoimpart acompressivelayer tothesurface of the finished pins. The new high strength material is 5CrMoV lowalloy steel at astrengthlevel of 240to 270ksi (1645to 1850 MPa). For lower stressapplications, anew1CrMoV lowalloysteel withastrengthof (170to200ksi) isused. Results: Nocrackinghasbeenobservedindovetail pinssincethe changeinmaterialsandtheintroductionof theshot peeningpractice. TurbineDestruction—StickingValves Problem: After only16hoursof operationof anew6MWsteam turbineinstalledinafertilizer plant, anaccidental disconnectof the electrical loadonthegenerator ledtoadestructiveoverspeed. The overspeed occurred becausehigh boiler carryover of boiler water treatment chemicals, including polymeric dispersant, introduced these chemicals into the bushings of all turbine control valves, gluingthevalvesstuck intheopenposition. Root cause: Poor control of boiler operation (drum level) together with the use of the polymeric dispersant that, after evaporation of water, becomes astrong adhesive. Controls of the electric generator allowedaccidental disconnect. Actions: New turbine generator installed, boiler and generator controls fixed, turbine valves reused after dissolution of the bushingdepositsinhot water. CONCLUSIONS • Steamturbines can be a very reliable equipment with life over 30 years and overhaul approximately every 10 years. However, about 5 percent of the industrial and utility turbines experience corrosionanddepositionproblems. MostlyduetoLPbladeandblade attachment (disc rim) corrosionfatigueor stresscorrosionfailures. • Theroot causes of thebladeanddisc failures includedesignwith highstresses, badsteamchemistry, anduseof highstrengthmaterials. • Other steam turbine problems include: low cycle thermal fatigue, pitting during unprotected layup and operation, loss of MW/HPandefficiencyduetodeposits, water dropleterosion, flow acceleratedcorrosion, solidparticleerosionbymagnetiteparticles exfoliatedfromsuperheater, turbinedestructiveover speedcaused by the control valves stuck open because of deposits in the bushings, andwater induction-water hammer. • All theproblemsarewell understood, detectable, andpreventable. Monitoring, inspection, anddefectsevaluationmethodsareavailable. Thesemethods includedesign reviews and audits of operation and maintenance, NDT, life prediction, vibration monitoring, vibration signature analysis, water, steam, and deposit chemistry monitoring andanalysis, valveexercise, andcontrol of superheater temperatures. • Steamcycle design and operation influences turbine problems by causinghighsteady andvibratory stresses, by thermal stresses related to load and temperature control, and by water and steam purity andboiler carryover. REFERENCES Amos, D., Lay, E., and Bachman, S., 1997, “Qualification of Welding Rotors with 12Cr Stainless Steel to Improve SCC Resistance,” Proceedings: Steam Turbine Stress Corrosion Workshop, EPRI, PaloAlto, California, TR-108982. ASM International, 1989, ASMMetalsHandbook, 9thEdition, 17, NondestructiveEvaluationandQualityControl. ASME, 1982, ASME Committee on Water in Thermal Power Systems, Industrial Subcommittee, SteamPurity Task Group, Results of 1982 Survey, American Society of Mechanical Engineers, NewYork, NewYork. ASME, 1989, The ASME Handbook on Water Technology for Thermal Power Systems, American Society of Mechanical Engineers, NewYork, NewYork. ASTM 2001, 2001Annual Book of ASTM Standards, Section 3: Metals Test Methods and Analytical Procedures, 03, NondestructiveTesting, West Conshohocken, Pennsylvania. ASME, 2002, Consensus for the Lay-up of Boilers, Turbines, Turbine Condensers, and Auxiliary Equipment, ASME Research Report, CRTD-Vol. 66, 2002. American Society of Mechanical Engineers, NewYork, NewYork. Atrens, A., et al., 1984, “SteamTurbine Blades,” Corrosion in Power GeneratingEquipment, PlenumPress. Baboian, R., Editor, 1995, “Corrosion Tests and Standards— Application and Interpretation,” ASTM, West Conshohocken, Pennsylvania. BLADE-ST™ Version 3, 2000, STI Technologies, Rochester, NewYork. Cheruvu, N. and Seth, B., 1993, “Key Variables Affecting the Susceptibility of Shrunk-On Discs to Stress Corrosion Cracking,” The SteamTurbine Generator Today: Materials, FlowPath Design, Repair, and Refurbishment, ASME, New York, NewYork. Clark, W. G., etal., 1981, “Proceduresfor EstimatingtheProbability of Steam Turbine Disc Rupture from Stress Corrosion Cracking,” ASME/IEEE Power Generation Conference, Paper No, 81-J PGC-PWR-31, NewYork, NewYork. Cotton, K., 1993, Evaluating and Improving Steam Turbine Performance, NewYork, NewYork: CottonFact, Inc. EPRI, 1981, “SteamTurbineBlades: ConsiderationsinDesignand aSurvey of BladeFailures,” PaloAlto, California, CS-1967. EPRI, 1982a, “SteamTurbineDisc CrackingExperience: Volume 1—Literature and Field Survey, Volume 2—Data Summaries and Discussion, Volume 3—Stress Corrosion Cracking of Low-Alloy Steels, Volume4—Factors Determining Chemical Composition of LowPressureTurbineEnvironments, Volume 5—Characteristics and Operating Histories of U.S. Power Plants, Volume 6—Description of Turbine Rotor Models, Volume7—Metallurgical Analysis of Cracked Discs from10 U.S. Power Plants,” PaloAlto, California, NP-2429-LD. EPRI, 1982b, “Metallurgical Evaluation of a Failed LP Turbine Disc,” PaloAlto, California, NP-2738. EPRI, 1983, “EPRI Research Related to the Steam Turbine Generator,” PaloAlto, California, NP-3288-SR. EPRI, 1984a, “Properties of Turbine Disc Materials,” Palo Alto, California, NP-3634. EPRI, 1984b, “TheEffectsof PhosphateEnvironmentsonTurbine Materials,” PaloAlto, California, CS-3541. EPRI, 1984c, “StressCorrosionCrackinginSteamTurbineDiscs: Survey of Data Collection, Reduction, and Modeling Activities,” PaloAlto, California, NP-3691. EPRI, 1984d, “CorrosionFatigueof SteamTurbine-BladingAlloys inOperational Environments,” PaloAlto, California, CS-2932. EPRI, 1984e, “TheEffect of Variables ontheFatigueBehavior of Ti-6Al-4V,” PaloAlto, California, CS-2934. EPRI, 1985a, “StressCorrosionCrackinginSteamTurbineDiscs: Analysisof FieldandLaboratory Data,” PaloAlto, California, NP-4056. EPRI, 1985b, “Surveyof SteamTurbineBladeFailures,”PaloAlto, California, CS-3891. PROCEEDINGSOFTHETHIRTY-SEVENTHTURBOMACHINERY SYMPOSIUM•2008 226 EPRI, 1985c, “Proceedings: Steam Turbine Blade Reliability Seminar andWorkshop,” PaloAlto, California, CS-4001. EPRI, 1985d, “StressCorrosionCrackingandCorrosionFatigueof SteamTurbineMaterials,” PaloAlto, California, NP-4074M. EPRI, 1986, “Interim Consensus Guidelines on Fossil Plant Chemistry,” PaloAlto, California, CS-4629. EPRI, 1987a, “Guidefor Useof Corrosion-Resistant Coatings on SteamTurbineBlades,” PaloAlto, California, CS-5481. EPRI, 1987b, “Stress Corrosion Cracking of A471 Turbine Disc Steels,” PaloAlto, California, NP-5182. EPRI, 1987c, “Steam Turbine Blade Reliability Seminar and Workshop,” PaloAlto, California, CS-5085. EPRI, 1989, “Guidelines for PredictingtheLifeof SteamTurbine Discs Exhibiting Stress Corrosion Cracking,” Palo Alto, California, NP-6444. EPRI, 1991a, “Proceedings: Fossil SteamTurbine Disc Cracking Workshop,” PaloAlto, California, GS-7250. EPRI, 1991b, “FloridaPower andLight UsesBLADE Programto Analyze Low Pressure Turbine Blade Failures,” Palo Alto, California, 100229. EPRI, 1993a, “Solid Particle Erosion Technology Assessment,” Final Report, PaloAlto, California, TR-103552. EPRI, 1993b, “Proceedingsof theSteamandCombustionTurbine Blading Conference and Workshop—1992,” Palo Alto, California, TR-102061. EPRI, 1994a, “Steam Chemistry and Corrosion,” Palo Alto, California, TR-103738. EPRI, 1994b, “BLADE CodeHelpsTVA DiagnoseBladeFailure Quickly and Avoid Blade Replacements,” Palo Alto, California, IN-10420. EPRI, 1994c, “Cycle Chemistry Guidelines for Fossil Plants: OxygenatedTreatment,” PaloAlto, California, TR-102285. EPRI, 1995, “Low-Pressure Turbine Blade Design Evaluation,” PaloAlto, California, TR-105839. EPRI, 1996, “Flow-AcceleratedCorrosion in Power Plants,” Palo Alto, California, TR-106611. EPRI, 1997a, “Proceedings: Steam Turbine Stress Corrosion Workshop,” PaloAlto, California, TR-108982. EPRI, 1997b, “Turbine Steam, Chemistry and Corrosion: Experimental TurbineTests,” PaloAlto, California, TR-108185. EPRI, 1997c, “Steam, Chemistry, and Corrosion in the Phase Transition Zone of Steam Turbines: Volume 1: Key Results, Summary, andInterpretation; Volume2: Individual Contributions of Participants,” PaloAlto, California, AP-108184. EPRI, 1997d, “Low-Pressure Rotor Rim Attachment Cracking Surveyof UtilityExperience,”PaloAlto, California, TR-107088. EPRI, 1997e, “MoistureNucleationinSteamTurbines,” PaloAlto, California, TR-108942. EPRI, 1998a, “TurbineSteamPathDamage: TheoryandPractice,” PaloAlto, California, AP-108943. EPRI, 1998b, “Cycling, Startup, Shutdown, and Layup—Fossil Plant Cycle Chemistry Guidelines for Operators and Chemists,” PaloAlto, California, TR-107754. EPRI, 1998c, “InterimCycleChemistry Guidelinesfor Combined CycleHeat Recovery SteamGenerators (HRSGs),” PaloAlto, California, TR-110051. EPRI, 1998d, “Proceedings: Workshop on Corrosion of Steam TurbineBladingandDiscsinthePhaseTransitionZone,” Palo Alto, California, TR-111340. EPRI, 1999, “TurbineSteam, ChemistryandCorrosion-Generation of Early Liquid Films in Turbines,” Palo Alto, California, TR-113090. EPRI, 2000a, “Productivity Improvement Handbook for Fossil Steam Power Plants: SecondEdition,” PaloAlto, California, TR-114910. EPRI, 2000b, “Corrosion of Low Pressure Steam Turbine Components,” PaloAlto, California, 1000557. EPRI, 2001a, “Cost of Corrosion in theElectric Power Industry,” PaloAlto, California, 1004662. EPRI, 2001b, “SteamTurbine Efficiency and Corrosion: Effects of Surface Finish, Deposits, and Moisture,” Palo Alto, California, 1003997. EPRI, 2002a, “Cycle Chemistry Guidelines for Fossil Plants: All-VolatileTreatment,” PaloAlto, California, 1004187. EPRI, 2002b, “Advances in Life Assessment and Optimization of Fossil Power Plants,”Proceedings, PaloAlto, California, 1006965. EPRI, 2002c, “Priorities for Research and Development for the Electric Power Industry,” PaloAlto, California, 1007274. Evans, D., 1993, “Comparison of Unstalled Flutter Predictions and Field Measurements for Steam Turbine Blades,” Proceedings of the Steamand Combustion Turbine-Blading Conference and Workshop—1992, EPRI, Palo Alto, California, TR-102061. Haas, H., J une 1977, “Major Damage Caused by Turbine or Generator Rotor FailuresintheRangeof theTrippingSpeed,” Der Maschinenschaden. Heymann, F., 1970, “Toward Quantitative Prediction of Liquid Impact Erosion,” ASTM STP 474, American Society for TestingandMaterials. Heymann, F., 1979, “ConclusionsfromtheASTM Interlaboratory Test Program with Liquid Impact Erosion Facilities,” Proceedings of theFifth International Conferenceon Erosion byLiquidandSolidImpact, CavendishLaboratory, University of Cambridge, UnitedKingdom. Heymann. F., 1992, “Liquid Impingement Erosion,” ASM Handbook, Volume 18: Friction, Lubrication, and Wear Technology, ASM International. Holdsworth, S., 2002, “Prediction and Prevention of Stress Corrosion and Corrosion Fatigue Cracking LP Steam Turbines,” Advances in LifeAssessment and Optimization of Fossil Power Plants—Proceedings, EPRI, Palo Alto, California, 1006965. J affee, R., Editor, 1983, CorrosionFatigueof SteamTurbineBlade Materials, NewYork, NewYork: PergamonPress. J onas, O., 1977, “Major Damage to Steam-Turbosets,” Panel Discussion at the 9th Allianz Forum, “Technology and Insurance,” 1976, andDer Maschinenschaden, 1977. J onas, O., 1978, “TaperedTensileSpecimen for Stress Corrosion Threshold Stress Testing,” ASTM J ournal of Testing and Evaluation, 6, (1), pp. 40-47. J onas, O., September 1982, “Beware of Organic Impurities in SteamPower Systems,” Power, 126, (9). J onas, O., February 1985a, “SteamTurbineCorrosion,” Materials Performance. J onas, O., March 1985b, “Control Erosion/Corrosion of Steels in Wet Steam,” Power. J onas, O., 1985c, “DesignAgainst Localized Corrosion,” Second International Symposium on Environmental Degradation of Materials in Nuclear Power Systems—Water Reactors, Monterey, California. STEAMTURBINE CORROSION ANDDEPOSITSPROBLEMSANDSOLUTIONS 227 J onas, O., 1985d, “Transportof ChemicalsinSteamCycles,”Paper No.245, Corrosion/85, NACE. J onas, O., 1986, “Costof CorrosionandScaleinElectricUtilities,” 1985Fossil Plant Water Chemistry Symposium: Proceedings, EPRI, PaloAlto, California, CS-4950. Jonas, O., 1987, “Corrosionof SteamTurbines,”MetalsHandbook, 9th Edition, 13, Corrosion, ASM, International, MetalsPark, Ohio. J onas, O., May 1989, “Developing Steam-Purity Limits for Industrial Turbines,” Power. J onas, O., 1994, “On-LineDiagnosisof TurbineDepositsandFirst Condensate,” 55th Annual International Water Conference, Pittsburgh, Pennsylvania. J onas, O., 1995, “Steam Generation,” Corrosion Tests and Standards—Application and Interpretation, ASTM, West Conshohocken, Pennsylvania. J onas, O., 1997, “Effects of Steam Chemistry on Moisture Nucleation,” Moisture Nucleation in SteamTurbines, EPRI, PaloAlto, California, TR-108942. J onas, O. andDooley, B., 1996, “SteamChemistry andits Effects onTurbineDeposits andCorrosion,” 57thInternational Water Conference, Pittsburgh, Pennsylvania. J onas, O. andDooley, B., 1997, “Major TurbineProblemsRelatedto Steam Chemistry: R&D, Root Causes, and Solutions,” Proceedings: FifthInternational ConferenceonCycleChemistry inFossil Plants, EPRI, PaloAlto, California, TR-108459. J onas, O. andSyrett, B. C., 1987, “Chemical TransportandTurbine Corrosion in Phosphate Treated Drum Boiler Units,” International Water Conference, Pittsburgh, Pennsylvania. J onas, O., Dooley, B., andRieger, N., 1993, “SteamChemistryand Turbine Corrosion—State-of-Knowledge,” 54th International Water Conference, Pittsburgh, Pennsylvania. J onas, O., etal., 2000, “EPRI ChemExpert: CycleChemistryAdvisor for Fossil Power Plants,” EPRI 6th International Conferenceon CycleChemistry inFossil Plants, Columbus, Ohio. J onas, O., et al., March2007, “Water Hammer andother Hydraulic Phenomena,” Power. Kilroy, R., et al., 1997, “A 12% Chrome Weld Repair Increases StressCorrosionCrackingResistanceof LP FingerTypeRotor Dovetails,” Proceedings of the International J oint Power GenerationConference, 2, Power. Kleitz, A., 1994, “ErosionandCorrosionPhenomenainWetSteam Turbines,” ESKOM Power Plant Chemistry Symposium. Kramer L. andMichael, S., September 1975, “ServiceExperience andStress Corrosionof Inconel 600Bellows ExpansionJ oins inTurbineSteamEnvironments,” MaterialsPerformance. McCloskey,T., 2002, “TroubleshootingTurbineSteamPathDamage,” Advances in LifeAssessment and Optimization of Fossil Power Plants—Proceedings, EPRI, PaloAlto, California, 1006965. McIntyre, P., 1979, “U.K. Experimental Work onTurbineDiscand Rotor Cracking: An Overview,” Working on SCC in Turbine RotorsandDiscs, EPRI-CEGB, Leatherhead, England. NERC, 2002, pc-GAR (GeneratingAvailability Report, Statistical Data1982-2000), NorthAmericanElectricReliabilityCouncil. Newman, J., 1974, “TheStressCorrosionof TurbineDiscSteelsin DiluteMolybdateSolutions and Stagnant Water,” CERL Lab NoteNo. RD/L/N 215/74. Nowak, 1997, “Low PressureTurbine Stress Corrosion Cracking Investigation at theNavajo Generating Station,” Proceedings: SteamTurbine Stress Corrosion Workshop, EPRI, Palo Alto, California, TR-108982. Parkins, R., J anuary 1972, “Stress Corrosion Spectrum,” British CorrosionJ ournal, 7. Petr, V., Kolovratnik, M., J iricek, I., and J onas, O., 1997, “Experimental Investigationof theEffectsof SteamChemistry on Droplet Nucleation,” Moisture Nucleation in Steam Turbines, EPRI, PaloAlto, California, TR-108942. Povarov, O., et al., J une 1985, “Basic Regularities of Erosion of Steam-Turbine Blade Materials under Impingement of Water Droplets,” Strojarstvo. “Progress in Understanding Improves Outlook for Control of Turbine-SteamChemistry,” March1981, Power. Pryakhin, V. A., et al., October 1984, “Problems of Erosionof the RotatingBladesof SteamTurbines,” Teploenergetika. Rezinskikh, V., Vaiman, A., and Melamed, M., November 1993, “TheMechanismof DamagingtheMetal of Blades of Steam Turbines Working in the Zone of the Phase Transition,” Teploenergetika. Rosario, et al., 2002, “Evaluation of LP Rotor Rim-Attachment Cracking Using LPRimLife,” Advances in Life Assessment andOptimizationof Fossil Power Plants—Proceedings, EPRI, PaloAlto, California, 1006965. Ryzenkov, V., 2000, “The State of the Problem and Ways for Increasing the Wear Resistance of the Power Equipment at Thermal Power Stations,” Thermal Engineering, 47, (6). Sakamoto, T., Nagao, S., andTanuma, T., 1992, “Investigation of Wet Steam Flow for Steam Turbine Repowering,” Steam Turbine-Generator Developments for the Power Generation Industry, ASME Power, 18. Sanders, W., 2001, SteamTurbinePathDamageandMaintenance Volumes1(February2001) and2(J uly2002), Pennwell Press. Scegljajev, V., 1983, Parni Turbiny (Steam Turbines), Prague, Czechoslovakia: SNTL. Schleithoff, K., 1984, “Influence of Water SteamChemistry on Stress Corrosion Cracking of Steam Turbine LP Discs,” CorrosioninPower GeneratingEquipment, PlenumPress. Singh, M., Matthews, T., and Ramsey, C., “FatigueDamageof SteamTurbine Blade Caused by Frequency Shift Due to Solid Buildup—A Case Study,” http://www.dresser- rand.com/e-tech/tp110/tp110prt.htm Speidel, M. and Atrens, A., Editors, 1984, Corrosion in Power GeneratingEquipment, PlenumPress. Speidel, M. andBertilsson, J., 1984, “StressCorrosionCrackingof Steam Turbine Rotors,” Corrosion in Power Generating Equipment, PlenumPress. Stastny, M., J onas, O., andSejna, M., 1997, “Numerical Analysis of theFlowwithCondensationinaTurbineCascade,”Moisture Nucleation in SteamTurbines, EPRI, Palo Alto, California, TR-108942. Svoboda, R. and Faber, G., 1984, “Erosion-Corrosion of Steam Turbine Components,” Corrosion in Power Generating Equipment, PlenumPublishingCorporation. Syrett, C. and Gorman, J. A., February 2003, “Cost of Corrosion in the Electric Power Industry—An Update,” Materials Performance. Syrett, C., Gorman, J. A., Arey, M. L., Koch, G. H., andJ acobson, G. A., March 2002, “Cost of Corrosion in theElectric Power Industry,” MaterialsPerformance. Turner, D., 1974, “SCC of LP Turbines: The Generation of Potentially Hazardous Environments from Molybdenum Compounds,” CERL LabNoteNo. RD/L/N 204/74. PROCEEDINGSOFTHETHIRTY-SEVENTHTURBOMACHINERY SYMPOSIUM•2008 228
Report "Steam turbine and deposits problems and solutions"