RenewableandSustainableEnergyReviews12(2008)19741987AreviewonsilterosioninhydroturbinesMamataKumariPadhy,R.P.SainiAlternateHydroEnergyCentre,IndianInstituteofTechnology,Roorkee247667,IndiaReceived27November2006;receivedinrevisedform27November2006;accepted5January2007AbstractErosivewearofhydroturbinerunnersisacomplexphenomenon,whichdependsupondifferentparameters such as silt size, hardness and concentration, velocity of water, and base materialproperties. Theefciencyof theturbinedecreaseswiththeincreaseintheerosivewearandnalbreakdown of hydro turbines. Various researchers have conducted experiments to study the effect ofthese parameters on erosive wear, but most of these experiments are on small-size samples in differenttypesoftestrigstosimulatetheowconditionsintheturbine, butactualowconditionsandthephenomenon of erosive wear are too complex to simulate. In the present paper, studies undertaken inthis eld by several investigators have been discussed extensively. Based on literaturesurvey variousaspectsrelatedtosilterosioninhydroturbines, differentcausesforthedeclinedperformanceandefciencyofthehydroturbinesandsuitableremedial measuressuggestedbyvariousinvestigatorshavebeendiscussed.r 2007ElsevierLtd.Allrightsreserved.Keywords:Hydroturbine;Silterosion;Erosivewear;Performance;EfciencyContents1. Introduction..................................................... 19752. Theoreticalinvestigations ............................................ 19773. Experimentalstudies ............................................... 19794. Casestudies ..................................................... 19835. Conclusions ..................................................... 1986References ...................................................... 1986ARTICLEINPRESSwww.elsevier.com/locate/rser1364-0321/$ - seefrontmatter r 2007ElsevierLtd.Allrightsreserved.doi:10.1016/j.rser.2007.01.025Correspondingauthor.E-mailaddress:mamatapadhy@rediffmail.com(M.K.Padhy).1. IntroductionEnergyisoneofthekeyfactorsthatinuencethedevelopmentofanationprovidingeconomic and social benets to its people. This is more important in developing countrieslike India, for economic development at micro level necessitating the need for theARTICLEINPRESSNomenclaturea averagegrainsizecoefcientofsuspendedsedimentwithabaseof0.05 mmC solidconcentrationc meansandconcentrationCwfractionofsolidbyweightd particlesized50medianparticlesize(m)f(dp50) functiondeningparticlesizek normalcomponentofparticleimpactvelocityneededtoinitiatetheerosionk1,k2,k3coefcienttoaccountforshape,hardnessandabrasionresistanceofbasemetal,respectivelyM thetotalmassofimpactingparticlesMrcoefcientofwearresistanceofbasematerialPE modiedsuspendedsedimentcontentP theaverageannualsuspendedsedimentcontenting/lp aconstantq quartzcontent(hardparticlecontent)S1coefcientofsiltconcentrationS2coefcientofsilthardnessS3coefcientofsiltparticlesizeS4coefcientofsiltparticleshapeTA turbineabrasionV velocityofparticlev relativejetvelocityvfowvelocity(m/s)W wearrate(kg/h)Wddeformationwearatnormalimpactconditionz exponentforrelativevelocitya exponentforannualsuspendedsedimentscontentb exponentforaveragegrainsizecoefcienty impactangled abrasionrate(mm/h)e deformationfactorK, b, g and j constants whose values depend on the properties of the erodent as wellasthetargetmaterialZmturbinepeakefciencywithsedimentladenowZwturbinepeakefciencywithcleanwaterM.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1975availability of secure and sustainable energy. The different sources of energy arehydro, thermal, nuclear and non-conventional energy resources like wind, solar andbiomass. Presently in India the total installed capacity including all the resources is126,089 MW, of which the share of the hydro energy is 26%[1], whereas an idealhydrothermal mixshouldbe40:60. Amongthedifferent sourcesofenergy, hydropoweris recognized as a renewable source of energy, which is economical, non-pollutingand environment friendly. India is blessed with immense amount of hydroelectricpotential andonly17%of thehydropotential hasbeenharnessedtill now[1]. Hence,a lot of importance is being giventodevelopthe hydropower potential. Ministry ofPower has been entrusted to develop large hydropower resources and Ministry ofNon-conventional Energy Resources has been promoting small and mini hydroprojects (p25 MW) soas toprovide energy toremote andhilly areas. The manage-ment of thesmall hydropower plants for achievinghigher efciencyof hydroturbineswithtimeisanimportant factor, but theturbinesshowdeclinedperformanceafterfewyears of operationas they get severely damageddue tovarious reasons. One of theimportant reasons is erosive wear of the turbines due to high content of abrasivematerial during monsoon. This problemis a major concern in the case of smallhydropowerplants, asmostoftheseplantsarerun-of-riverschemesandaresituatedinsteephillyterrains. Duringthe rainyseason, alarge amount of sediment (as highas20,000 ppm) is present inwater andit becomes difcult toremove all these sedimentsbeforepassingthroughtheturbine.Thesiltmainlyconsistsofquartz(7098%),whichisextremely hard (hardness 7 in Mohs scale) causing severe damage to the turbinecomponents [24]likewater passagecomponents,guidevane,topand bottomringliners,labyrinths, runner blades, inlet valve seals, etc. Erosionwear effect inhydroturbinesresults in alteration of blade prole, increased vibration, fatigue damage, inefcientoperationandsystemfailure.Erosive wear of the turbine blades is a complex phenomenon that depends on (i) erodingparticles, their size, shape, hardness andconcentration(ii) substrates, chemistry, elasticproperties, surface hardness and surface morphology and (iii) operating conditions,velocity and impingement angle [5,6]. It should be emphasized that sand erosion even withgrainsizelessthan60 mmhasledtoseveredamageof needlesandnozzlesof aPeltonturbine, whichis causedbythestrongturbulenceinthehighjet velocitybringingtheparticles to oscillate and rotate in circles causing collisions with the steel surface. If coarsesand is present, the Pelton buckets are also severely eroded while the damage of the nozzlesmay be less serious. The reason may be explained by the extreme acceleration of a particlepassing the Pelton bucket. A ow analysis shows that an acceleration of 100,000 m/s2mayoccurforsmallbucketsinahighheadturbine[5].Theerodent shapeisanimportant property, but itseffect isdifcult toquantifyfornatural particles. For impact of spherical particles onductilematerials, Hutchings andWinter [7] assumed ploughing of material forming lips around the crater, which breaks upinsubsequentimpacts. Inanotherstudy, WinterandHutchings[8] haveshownthattheangularparticlesremovethematerial byploughingandmicrocuttingforleadandmildsteel. Desaleetal. [9]haveshownthatthesurfacemorphologyofsubstratematerialhasdeepcratersandhighervalueofaveragesurfaceroughnessforangularerodentparticlescomparedwiththeblockyshapeerodents.Thepresentpaperdiscussesthestudiescarriedoutbyvariousinvestigatorsinordertodeterminetheeffectoferosivewearandidentifythegapsforfuturestudies.ARTICLEINPRESSM.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 19762. TheoreticalinvestigationsGenerally,erosiondamageisconsideredasthegradualremovalofmaterial causedbyrepeateddeformationandcuttingactions. Sanderosionisdesignatedasabrasivewear.Thistype ofwearwillbreak downtheoxide layeron theow guidingsurfaces and partlymake the surfaces uneven which may also be the origin for cavitation erosion. Sand erosiontherefore may be both a releasing and contributing cause for damages that are observed inpower plants with a large transport of wearing contaminants in the water ow. The actualmechanismof erosivewearwas not fullyunderstood. Therefore, asimple, reliableandgeneralized quantitative model for erosion could not be developed. Most commonexpression for the erosive wear was based on experimental experiences. The hydro-abrasivewear was commonly quantied by means of wear rateW, dened by loss of mass per unittimeandgenerallyexpressedasW f properties oferoding particles; properties ofbase material and operating conditions.Kjolle[5] studiedthecauses of damages inhydroturbines andfoundthat themaincauses of damage of water turbines were due to cavitation problems, sand erosion, materialdefects and fatigue. The turbine parts exposed to cavitation are the runners and draft tubecones for the Francis, Kaplan and the needles, nozzles and the runner buckets of the Peltonturbines. The effect of cavitationerosionwas foundtobe reducedby improvingthehydraulicdesignandproduction ofcomponents,adoptingerosion resistantmaterialsandarrangement of the turbines for operations within the good range of acceptable cavitationconditions.NeilsonandGilchrist[10] studiedthetotalwearatnormalimpactangleassumingthetotalweartobecontributedbyonlydeformationwear,whichgivesadeformationfactor(e)asbelow: Wd12MVsin y k2 , (1)where Wdis deformationwear at normal impact condition, M is the total mass ofimpacting particles, V is velocityof particle, y is impact angle andkis the normalcomponent of particleimpact velocityneededtoinitiatetheerosionwhichis generallyneglected,consideringverysmallcomparedtotheimpactvelocity.Bain et al. [12,13]have attemptedto develop a correlation for the estimationof erosionratebasedonextensivedatacollectedinabenchscaletestrig. Thegeneral formofthecorrelationcanberepresentedas,W KVbdgCj, (2)where W is erosion rate, V is velocity of particle, d is particle size, C is solid concentration,andK, b, gand jareconstantswhosevaluesdependonthepropertiesoftheerodentaswell as the target material. For different erodents, the effect of particle size has beennormallyconsideredas aparameter affectingthewear andtheexponent valueg wasfoundtoliebetween0.3and1.6[1117].ThehydraulicperformancetestsonaFrancisturbinemodelwithsedimentladenow,conducted in Japan and reported by Okamura and Sato [18], showed that the turbines bestefciency decreased in direct proportion to the increase in solids concentration. TheARTICLEINPRESSM.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1977efciencywascorrelatedbythefollowingexpression:Zm 1 0:085CwZw; (3)where Zmis turbinepeakefciencywith sediment-ladenow, Zwis turbinepeakefciencywithcleanwater,andCwisfractionofsolidbyweight.Sundararajan [19] presented a comprehensive theoretical model for erosion, valid for allimpactanglescombiningtheconceptoflocalizationofplasticdeformationleadingtolipformationandthegeneralizedenergyabsorptionrelationsvalidforallimpactanglesandallshapesoferodingparticles.Krause andGrein[20] reportedthat the abrasionrate onconventional steel PeltonrunnermadeofX5CrNi13/4wasasgivenbelow,d pqcv3:4f dp50, (4)wheredis abrasionrate (mm/h), pis aconstant, qis quartz content, c is meansandconcentration,andvisrelativejetvelocity;f(dp50)isfunctiondeningparticlesize.Since the above equation had been proposed for X5 CrNi 13/4, it is applicable to turbinecomponentsmadeofthismaterialonly.Naidu[21]suggestedthefollowingexpressionforpredictingthesilterosionrate:W S1S2S3S4Mrvx, (5)where S1 is coefcient of silt concentration, S2 is coefcient of silt hardness, S3 is coefcientof silt particle size, S4 is coefcient of silt particle shape, Mr is coefcient of wear resistanceofbasematerial,andvisrelativevelocityofwater.ThevaluessuggestedfortheexponentXare:3forFrancisrunner,2.5forguidevanesandpivotringliner,2.5forPeltonnozzleand1.5forPeltonrunnerbuckets.Wear occurs by the plastic displacement of surface and near surface material and by thedetachmentofparticlesthatformweardebris.Thisiscausedbythefrictionalcontactofowingwaterorbysolidsilt particlesentrainedinowingwater. Severeimpact of siltparticles on the blade surfaces contributes signicantly to the erosion and wear. When thesilt is composedof hardabrasive particles, wear is causedbycuttingactionof theseparticles[22].TheturbineabrasionwasexpressedbyAsthana[23]asTA f PE;vz, (6)wherePEismodiedsuspendedsedimentcontent, visrelativevelocitybetweenowingwaterandturbinepartswhereabrasionissevere,andzisexponentforrelativevelocity.Themodiedsedimentcontent(PE)isexpressedbythefollowingequation:PE Paabk1k2k3, (7)where P is the average annual suspended sediment content in gm/l. It is based on the long-termmeasurements intheriver; ais exponent of P representingcorrectionfactor forsuspendedsediment concentration. It is takenas 1for concentrationupto5 g/l; aisaverage grain size coefcient of suspended sediment with a base of 0.05 mm; b is exponentof a representingcorrectionfactorforaverageparticlesize, whichwastakenas1forparticleupto0.6 mmandcurvedow;k1,k2,k3,representthecoefcienttoaccountforshape, hardness and abrasion resistance of base metal, respectively. k1 is taken as 0.75, 1.0and1.25dependingonirregularitiesrangingfromfewtosevere, k2wastakenas1forARTICLEINPRESSM.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1978hardnessgreaterthan3(onMohsscale)and0.5forlessthan3, k3wastakenas1for13Cr4Nisteel.Schneider and Ka chele [24] proposed that wear rate, W was a function of a multitude ofparametersasshowninthefollowingalgebraicrelation:Wcqf d50 vnf, (8)where c (kg/m3) is sand concentration, q (kg/kg) is hard particle contents, d50 (m) is medianparticlesize,andvf(m/s)isowvelocity.Theauthorssuggestedthatthevalueofnvariedconsiderablyfromabout2.1tomorethan 3. This range of values was reported to reect the limitations of the algebraic relation,which considered neither the material parameters of the eroded body nor the owparametersorthevariousmaterialparametersofthesiltparticles.Macket al.[25]suggesteda numericalmodeltopredicttheerosion on guide vanesandon labyrinth seals in hydraulic turbines. The prediction of erosion, based on theLagrangian calculation of particle paths in a viscous ow, was described for twocomponents of aFrancis turbine, for whichresults of eldtests wereavailable. It wasshownthattheerosionlevelwasstronglydependentontheparticlesize.Afully3Dowand erosion calculation around the guide vanes of the same Francis turbine was presented.It wasshownthat thegeometryof theguidevaneincludingtipclearance, support andllets leadtoacomplexoweld, which, as aconsequence, resultedinacomplicateduidparticle interaction strongly affecting the erosion pattern. There was a goodagreement between the numerically obtained erosion pattern and the eld testmeasurements.Kecket al. [26] presentedastudyonthe utilizationof CFDmethodtopredict theerosionpatterninahydraulicturbineandcomparedit witheldmeasurementsof theerosion. Sand erosion was modeledby applying the Lagrange method i.e. tracking a largenumberofindividualparticlesintheoweld.Themotion oftheparticleswasdescribedby the BassetBoussinesqOseen equation whereby experimentally based correlations wereusedforthedragandtheinuenceofturbulentmotion.DuringtheLagrangiantracking,the number of particles impinging ona surface was recorded. Out of these data theremoval of the wall was calculated. Calculations were performed for different particle sizes.Theresultshowedagoodcorrelationfortheerosionpatternwiththeeldobservation.However,theauthorsconcludedthatCFDsimulationdidnotprovideaccurateabsoluteerosion, althoughit couldbeusedtoobtainrelativeerosionintensitiesandtoevaluatedifferentdesignsrelativetoeachother.3. ExperimentalstudiesManyinvestigatorsreportedtheirexperimentalresultsonerosivewearconductedwithdifferentbasematerialsanddifferenttypesoferodent.Chattopadhyay[27] conductedexperimentstodeterminetheslurryerosioncharacter-istics of AISI 316L, 15wt%Cr15wt%Mnstainless steel andStellitepowder alloyappliedasaoverlaytocastferriticstainlesssteel ofCA6NMtype, whichwasusedasanormal turbine runner material. The tests were conducted in specially designed testequipment(Fig. 1). Thedifferentwearratesofthealloyswereexplainedintermsofthemicrostructure, hardness and work hardening rate. The samples were rectangular insectionandofsize65 mm14 mm20 mmandthicksandslurrywastheerodent. TheARTICLEINPRESSM.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1979authorhadconcludedthat 15wt%Cr15wt%Mnstainlesssteel andStellitepowderalloy applied as a overlay showed better erosion resistance properties as compared with thebasematerialCA6NMsteel.KrauseandGrein[20] carriedmodel testswithvaryingparameterswiththeX5CrNi13/4steel, normallyusedinhydroplants. Thetestrigwasdesignedtosimulatetheowconditionsin a turbine.A naturalsand/watermixture takenfroma powerplantreservoirandsandcontaining99%quartzinvariousgrainsizeswereusedforthetests.Theyhaveconcludedthat the abrasion rateis a function of velocity, sandcontentand proportion ofhardcomponentsandsizeofthesandparticles.Themaximumabrasionoccurredwithinanapproximateparticlesizerangeof4070 mm.Roman et al. [28] reported the development of a new erosion resistant coatingNEYRCOa composite coating with ceramic and organic matter base, designed tocombinehardness andductility. Theycarriedout aseries of model tests inaspeciallydesigned test rig (Fig. 2) to nd the effectiveness of the coating against erosion. Four watervelocitieswereused:20,25,36,and48 m/s.Thewaterowratewas2.5 l/s.Theabrasivematerialwashighsilicacontentwiththefollowingchemicalcomposition:SiO2499.5%Al12O 40.2%Fe2O340.2%ARTICLEINPRESSFig.1. Schematicdiagramofslurryerosiontestequipment[27].M.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1980CaCO340.04%Size 200400 mmHardness 7 MohsConcentration 20 g/lSurface examinations of the samples subjected to same erosive condition showedthatthecoatedsamplesgavebetterperformanceascomparedwiththeuncoatedsamples(Fig.3aandb).Mann [6] studied the erosion resistance characteristics of different hard coatings such ashardchromeplating, plasmanitriding, D-gunspraying, andboronisingwithcommonlyusedsteel inhydroturbines.Theweartestfacilitywasdesignedconsideringthelowandhigh impact wear of hydro turbine components. The samples used were cylindrical in shapetosimulatetheangleof impingement from01to901, whichoccurredinhydroturbinebladesandvanes. Sandwasusedastheerodent, theconcentrationvariedfrom1500to10,000 ppm.PerformanceofboridedT410steelwasfoundmuchbetterthanothers.MannandArya[29] studiedthe silt erosioncharacteristics of plasmanitridingandHVOF coatings along with commonly used steel in hydro turbines. The characterization ofARTICLEINPRESSFig.2. Schematicdiagramoflaboratorytestequipment[28].Fig.3. (a)Wearwavesonstainlesssteeland(b)wearwavesonNeyrco[28].M.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1981abrasive wear was carried out as per ASTMG-65. Angle of incidence, velocity andReynolds numbers were maintained similar to those that commonly occur in hydroturbines, simulatinglowaswell ashigh-energyimpingement wear. Thetest parametersadoptedaregivenbelow:1 kgmineralsandofhardness 1100 HVSizeoferodent 180250 mmErodentowrate 5.5 g/sSamplesize 75 mm25 mm6 mmHVOF coating showed superior performance than plasma nitrided steel, but thedemerits of HVOF coating was that it showed micro cracking, debonding and digging outofWCparticles.Forplasmanitridedsteelstherewasductilemodeoferosion.Engelhardt and Oechsle [30] examined different materials and coatings to evaluate theirresistance tothe hydraulic turbine surface. Ahard, HVOF-appliedTC/CoCr coating,named Diaturb 532 and a soft PU-based coating called Softurb 80 were chosen for testing.The samples were tested in a test rig. Later, the San Men Xia hydro power plant in ChinasHenanprovincewaschosenforfull scaletestingoftheimprovementsfoundduringthelaboratoryresearchprogramme. The project includeda monitoring phase of 2years,duringwhichtheturbinepartswereinspectedseveraltimes.Exceptforsomemechanicaldamagetotheprotectionsystems, wearratesontheHVOF-coatedrunnerbladesweredeterminedtobewithintheaccuracyofthethicknessmeasurementgauge(o40 mm).Thewear rate on the PU-coated surface of the runner blades and the guide vanes wasdeterminedtobearound0.15 mmperyear. Theunitwasinoperationduringthe2-yearmonitoring phase and also during the ooding season with an average sand concentrationof2030 kg/m3.Thapa and Brekke [31] carried out laboratory erosion experiments on curved specimensby particles of different size to simulate the ow in Pelton bucket in a high-velocity test rig(Fig.4).Aluminumspecimenswithdifferentcurvature(Fig.5)wereusedforthetesting.Baskarp-15foundrysandwith66%freequartz(nesand) of size174 mmandarticialsilica sand (coarse sand) of size 256 mm were used as the erosive particles. The results wereARTICLEINPRESSSpecimen Pump Sink Flow meter Hopper(Sand andwatermixture) Hopper(Sand freefall)TankCompart-ments Connectionfor maintaining waterlevel at hopperBypassvalueNozzleParticleWaterlevel Fig.4. Highvelocitytestrig.[31].M.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1982presentedin the formof erosionrate for different proles andsurface roughness atdifferent locations of curvedspecimens. Byvisual observationof erodedparticles theauthorsconcludedthatmostofthecoarsegrainsstrikeclosetothesplitter, whereasthenegrainswereobservedfarawayfromthesplitter.Theerosionrateintermsofweightloss per unit striking particle found smaller with ne particles. This was due to low particleimpact energy of smaller particles and might be because some of the ne particles escapedglidingwithoutstrikingthesurface. Anotherobservationtheymadewaserosionrateinmg/kgincreasedwiththeincreaseincurveradius.4. CasestudiesAhluwalia et al. [32] suggesteda number of measures such as providing adequatedesilting arrangements, facilities for quick segregation and easy access to vulnerable parts,judiciouschoiceofthetypeofturbineaswell asoperatingspeeds, useofstainlesssteelmaterials and availability of skilled maintenance staff to evaluate techno-economicconsiderations for hydropower stations inIndia. Theyalsoemphasizedthat theaboverecommendation can only reduce the erosion problem by extending the period between theshotdownsbutcannoteliminatethesameintheunderwaterturbinepartstotally.Singh[33] reportedthecasestudyof theTilothhydropowerstation(3 30MW) onriverBhagirathi.Thethreeunitswerecommissionedin1984.Theturbineswerefoundtobeseriouslydamagedafterabout2600 hofoperation. Theywererepaired,butextensivedamage was observed again within 30005000 h of operation. The sedimentation chamberhadbeendesignedtoarrestsiltparticleslargerthan0.3 mm.Petrographicstudiescarriedoutrevealedthepresenceofhighlyabrasivequartz,havinghardnessof7onMohsscale.Theconcentrationofthesiltparticlesduringtherainyseasonwasmaximumand reachedup to 4000 ppm. Initially, there was a proposal to provide another sedimentation chambertoarrestparticlesupto0.15 mm.Butduetoveryhighcostandasthesettlingchamberscouldnot completelyremovethesilt particles, theproposal wasnot implemented. Theinvestigators, rather, suggestedimprovingthemetallurgyoftheturbineblades. Thenewrunnerwasmanufacturedwithstainlesssteel (13Cr4Ni), whichwassupposedtogivea better performance regarding erosion. However, it was observedthat there was noappreciablereductionintheerosioncomparedwiththeolderrunners.ARTICLEINPRESSFig.5. Specimensofdifferentcurvature[31].M.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1983Yan[34]studiedtheeffectofsiltabrasion indifferenthydropowerplantsofChinaandhaddrawnthefollowingconclusions:i. The abrasion remained moderate for all particles smaller than 0.05 mm and rises sharplyforlargersizes.ii. Theproduct of operatingheadHandcontent of harmful sediment Sd(d40.05 mm)mustbelessthan7,sothattheabrasiveerosionintheturbinewouldbeminimum.Kumaretal. [35] reportedthecasestudyofTilothPowerstationunderManeri BhaliStage-I Project, whichis basically a run-of-river scheme, at river Bhagirathi betweenManeriandUttarkashi(India).DetailedexaminationofrunnerbladesofUnit-Irevealedthat therewas thinningat outlet edges duetoheavyerosionalongwithpittingmarksbefore1yearofcommissioning.Theanalysisindicatedthatthinningofbladewascausedby silt erosion. There was also heavy leakage of water from the guide vane bush housings.Subsequent inspectionof under water parts revealedthat therunner blades havebeenextensively damaged and big pieces of 300 mm350 mm size has just sheared off. Further,visual examination revealed that there were more cracks near the skirt of the runner blades.InUnit-II andUnit-III, theextendof damagetotherunner blades was similartothedamage noticedinUnit-I. Suchanextensive damage is expectedtobe causeddue toexcessive silt, designed/operating condition, metallurgy of blades and inadequate design ofrunnerprole.According to Singh et al. [36], right from the inception of the project, design and layoutneedtobeplannedinsuchawaythatthereisminimumdamageatacceptableeconomiccost. Speed selection of unit, features ensuring easy and fast replacement of effected parts,etc.canhelpreducedrasticallythedowntimeperiod.Theproblemofsilterosionismorepeculiar to Indianconditions and therefore efforts are requiredto generate adequate datatodeal withtheproblem. Inviewofnoguaranteesofferedbythesupplier, purchaserishesitant to experiment with protective coating on underwater parts because of its high cost.Mathur et al. [37] reported the case study of Salal (6 115 MW) hydropower station onriver Chenab and Baira Siul (3 60 MW) hydropower station on river Baira and Siul. Thesilt contents of the water of above rivers indicated the presence of 7598% quartz bearinghardness78onMohsscaleandabout98%siltparticlesareofsize0.25 mmandless.After 4000 h of operation at Baira Siul guide vanes made of 13Cr4Ni stainless steel, the losswas about 1015% whereas at Salal this loss was approximately 1012% of design weight.Stay vanes made of carbon steel get eroded. Pressure side of runner blades, crown and skirtgeterodedveryfast.In2yearsofoperationrunnerprolegetsaltered.LabyrinthgapinBaira Siul increased from the design value of 0.81.2 to 34 mm in only 1 year of operation.At Salal, lower labyrinth ring and sometimes portions of lower ring skirt get washed awayin 810 thousand hours of operation of machines. Such type of erosion caused increase inthevibrationlevelofmachinesatbothpowerstations.Wood[38] reportedaeldstudycarriedoutbyChinaNorthWestElectricPowerbymountingcoatedspecimensindifferent placesinKaplanandFrancisturbinesandleftduring the ood season. Coated region had reduced the worn out thickness between 5 and43 mm, whereasthesurroundinguncoatedmetal hadbeenworndownbybetween1and10 mm. Theresearchprogrammeconcludedthatminimumlossofefciencycanonlybereached by a combination of design optimization, based on erosion prediction andprotectionofsurfaceswithwearsreducingcoatings.ARTICLEINPRESSM.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1984Pradhan[39] observedduring his studies that inrun-of-river power plants insteepsediment loaded rivers the conventional design criteria to trap 0.2 mmsize sedimentparticles didnotseemto functionsatisfactorily.Ingeneral, projectswerehavingdamagestorunnersduetosevereerosioncausedbysilt. IncaseofJhimrukProject, thewearonrunnerwassohighthatitrequiredrepairaftereverymonsoon.Acasestudywascarriedout for JHP, a12-MWrun-of-river typeproject built andcommissionedin1994. Thesettling basins had been designed to trap 90% of 0.2 mm size particles. During the sedimentmonitoring programme, the investigators observedthat average values for suspendedsediment concentration in Jhimruk river during the peak monsoon ranges from about 2000to 6000 ppm with upper values ranging from about 20,000 to as high as 60,000 ppm, whichindicatedthat thesediment transport inJhimrukRiver duringthemonsoonwas quitesignicant. The runners were signicantly worn out after every monsoon. Fromthedamagesobservedintherunnersitwasrevealedthattherunnersinthepowerplantwereexposedtotoomuchhighersedimentloadthanexpectedduringtheplanninganddesignstage. The author developed a correlation between sediment load to which the turbine wasexposedto and correspondingdecreaseinefciencyoftheturbineovera timeperiodandconcluded that the efciency loss was 4% at best efciency point and 8% at 25% load. TheresultsfromthethermodynamicefciencymeasurementsobtainedduringhisstudiesareshowninFig.6.Thapaetal.[40]investigatedtheeffectofsuspendedsedimentsinhydropowerprojectsbasedonacasestudyof60 MWKhimtihydropowerplant. Duetothepresenceofhighamountsofsediments, thehydropowerplantwasdesignedwithsettlingbasinstoscreen85%ofall particleswithafall diameterof0.13 mmand95%ofall particleswithafalldiameterof 0.20 mm. Theplant wascommissionedinJuly2000andthedamagetotheturbinecomponentswasinvestigatedinJuly2003.Theinvestigatorshadobservedthatasignicant amount of erosionhadappearedinthe turbine bucket andneedles. Eventhoughthe settling basinswereperformingsatisfactorily,particlessmaller than the designsize passed through the turbines and caused the damage. The bucket thickness was reducedbyabout1 mmtowardstherootofthebucket,whichiscriticalfromthepointofviewofstrengthandhencethereliabilityofthecomponent. Similarly, thesplitterofthebucketwas eroded to saw tooth form from the original straight edge. The sharp edge of the splitterARTICLEINPRESSFig.6. Variationofthermodynamicefciency[39].M.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1985had bluntedand the width becameapproximately 4 mm dueto which the efciency of theturbine haddecreased. Tominimize the effect of erosionhardceramic coatings wereapplied on the bucket and needle surface at the cost of around US$ 25,000 per runner, butperformanceswerenotpromising.Darling[41] studiedtherefurbishment of theSvartisenhydroelectricplant (Norway).Thehydroturbineat thisproject experiencedmoreerosionthanotherFrancisunitsinNorwayduetotheunusuallyhighlevelsofsiltinthewater.Fromthecommissioningofthe plant in 1993 until 2003, the efciency of the turbine was dropped by 22.5%.Tungsten-carbide thermal spray coating, appliedonsite, was selectedinaneffort toincrease the time between refurbishments, and to increase the lifetime of the turbine coversby as much as 16 years. Two years later, the performance of the turbine continued to meettheirexpectations.5. ConclusionsSilterosioninhydroturbinescannotbeavoidedcompletely,butcanbereducedtoaneconomicallyacceptablelevel. Manyinvestigatorshavestudiedtheprocessoferosioninhydroturbines throughexperimental andanalytical studies. Fewcase studies are alsoreportedintheliterature.Someoftheinvestigatorshavereportedthatinspiteofdesignchangesintheturbinecomponentsandprovidingdifferentmaterialsandcoatingstotheturbineblades, theimprovement inmost casesarenot quitesignicant. Hence, furtherexperimental and theoretical studies are required for studying the effect of hydro abrasiveerosionindifferentowconditionsofturbinesandparametersrelatedtosilterosioninwater.References[1] MinistryofPower,GovernmentofIndia. /http://powermin.nic.in/S.[2] PandeVK.Criteriafordevelopingsilterosionmodeltestfacility.In:Conference.ofCBIP,1987.[3] SinghSC. Operationalproblemsanddevelopmentofanewrunnerforsiltywater. International WP&DC1990;November.[4] TongD.Cavitationandwearonhydraulicmachines.Int.WP&DC1981;April.[5] /http://www.tev.ntnu.no/vk/publikasjoner/pdf/ArneKjolle/chapter14.pdfS.[6] MannBS. High-energyparticle impact wear resistance of hardcoatings andtheir applicationinhydroturbines.Wear2000;237:1406.[7] Hutchings IM, Winter RE. Particle erosionof ductile metals: a mechanics of material removal. Wear1974;27:1218.[8] Winter RE, Hutchings IM. Solid particle erosion studies using single angular particles. Wear 1974;29:18194.[9] DesaleGR, Gandhi BK, JainSC. Effect oferodentpropertiesonerosionwearofductiletypematerials.Wear2006;January.[10] NeilsonJH,GilchristA.Erosionbyastreamofsolidparticles.Wear1968;11:11122.[11] Gupta R, Singh SN, Seshadri V. Prediction of uneven wear in a slurry pipeline on the basis of measurementsinapottester.Wear1995;184:16978.[12] BainAG, BonningtonST. Thehydraulictransportofsolidsbypipeline. 1sted. Oxford: PergamonPress;1970.p.1316.[13] Gandhi BK, Singh SN, Seshadri V. Study of the parametric dependence of erosion wear for the parallel owofsolidliquidmixtures.TribolInt1999;32:27582.[14] de Bree SEM, Rosenbrand WF, deGee AWJ. Onthe erosion resistance in watersand mixtures ofsteels forapplication in slurry pipelines. Hydro transport 8, BHRA uid engineering, Johannesburg S.A., 1982; paperC3.[15] ElkholyA.Predictionofabrasionwearforslurrypumpmaterials.Wear1993;84:3949.ARTICLEINPRESSM.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1986[16] LevyAV,HickeyG.Liquidsolidparticlesslurryerosionofsteels.Wear1987;117:12946.[17] LevyAV.Solidparticleerosionanderosioncorrosionofmaterials.ASMEIntOhio1995;133.[18] Takgi T, OkamuraT, SatoJ. HydraulicperformanceofFrancisturbineforsediment-ladenow. HitachireviewNo-2,1988.[19] Sundararajan G. A comprehensive model for the solid particle erosion of ductile material. Wear1991;149:11127.[20] KrauseM,GreinH.Abrasionresearchandprevention.HydropowerDams1996;4:1720.[21] NaiduBSK.Addressingtheproblemsofsilterosionathydroplants.HydropowerDams1997;3:727.[22] Goel DB. Mechanism of erosion in hydro turbine under water parts and the role of metallurgy in providingremedialmeasures.Seminaron:Siltingproblemsinhydropowerstations.WRDTC,Roorkee,India,May1997.[23] AsthanaBN.Determinationofoptimalsedimentsizetobeexcludedforrun-of-riverprojectacasestudy.Seminaron:Siltingproblemsinhydropowerstations.WRDTC,Roorkee,India,May1997.[24] Schneider Ch, Ka cheleT. Recent researchresults onpredictingandpreventingsilt erosion, ConferenceProceedings1999;Siltingproblemsinhydropowerplants,NewDelhi,India.[25] MackR, DrtinaP, LangE. Numerical predictionof erosiononguide vanes andinlabyrinthseals inhydraulicturbines.Wear1999;233235:68591.[26] Keck H, Dekumbis R, Sick M. Sediment erosion in hydraulic turbine and experiences with advanced coatingtechnologies.In:IndiaHydro2005;Internationalworkshopandconference.[27] ChattopadhyayR.Highsiltwearofhydroturbinerunners.Wear1993;162164:10404.[28] RomanJM,Xin LY,HuiWM,ReginensiJP.Dealing withabrasive erosion in hydroturbine. HydropowerDams1997;3:6771.[29] Mann BS,Arya V.Abrasive and erosive wearcharacteristicsofplasma nitriding and HVOFcoatings: theirapplicationinhydroturbines.Wear2001;249:35460.[30] EngelhardtM,OechsleD.Smoothoperator.InternatWaterPowerDamConstr2004;June:202.[31] ThapaB,BrekkeH.Effectofsandparticlesizeandsurfacecurvatureonerosionofhydraulicturbine.In:22ndIAHRsymposiumonhydraulicmachineryandsystems2004,Stockholm.[32] Ahluwalia MS, Gaur BK. Review of measures for minimising damages in hydro plants from excessive silt inriverwater.In:Workshopon:Siltingproblemsinhydroelectricpowerstations,CBIP,1987.[33] Singh SC. Operational problems and development of a new runner for silty water. Water Power Dam Constr1990;November:2937.[34] YanMZ.Protectinghydroturbinesinsilt-ladenrivers.HydropowerDams1996;4:224.[35] KumarA, Tyagi SK. Problemof erosionof underwaterpartsduetoheavysilt inGangaValleyPowerStation.Selectedpapers:Siltdamagestoequipmentinhydropowerstationsandremedialmeasures,CBIP,NewDelhi,July1996.[36] SinghAK, ChandraS. Hydro-abrasionof water turbines inHimalayas. In: Workshop: Silt damages toequipmentinhydropowerstationsandremedialMeasures.CBIP,1996.[37] PrasadG,MathurGN,AgrawalAB.SilterosionofgeneratingequipmentcasestudyofSalalandBairaSiul power station of NHPC. In: Seminar on: Silting Problems in Hydro Power stations. WRDTC, Roorkee,India,May1997.[38] WoodJ.Ananswertoabrasion.InternationalWaterPowerDamConstr1998;April:367.[39] PradhanPMS,ImprovingsedimenthandlingintheHimalayas.OSHresearch,Nepal,October2004.[40] ThapaB, StresthaR, Dhakal P, ThapaBS. ProblemsofNepalesehydropowerprojectsduetosuspendedsediments.AquaticEcosystemHealthManage2005;8(3):2517.[41] DarlingK.Easingerosion.InterWaterPowerDamConstr2006;February:323.ARTICLEINPRESSM.K.Padhy,R.P.Saini/RenewableandSustainableEnergyReviews12(2008)19741987 1987