Abstract :
Abstract
TheresultsoftheABBAtom3×3-RodBundleRefloodingTestswereusedforassessmentoftherefloodingmodelusedinRELAP5/MOD3.2.2Gammaversion.Theassessmentcalculationswereperformedusingthedefaultcalculationmodeloptionsimplementedinthecode.
Thetestswereperformedtoinvestigatetheeffectsofdifferentspacergriddesignsonheattransferduringtherefloodingperiodofapressurizedwaterreactorloss-of-coolantaccident(LOCA).Thetestswereconductedunderlow-pressureandlow-flow(LPLF)conditionsusingaPWR-type3×3-rodbundlewithfull-lengthindirectlyelectricallyheated,steppedcosineaxialpower-shapedheaterrods.Threedifferentspacergridconfigurationswerestudied:spacergridswithoutmixingvanes,mixingvanespacergrids,andmixingvanespacergridstogetherwithintermediateflowmixers(IFM).
Atotalof36testswithdifferentspacergridconfigurationswerecalculated.Fortwoselectedbasictestswithnon-mixingspacergridsanextendedcomparisonofcalculatedandmeasuredparametersispresented.Thecomparisonofthepredictedandmeasuredmaximalcladdingtemperaturesandquenchtimes,whicharethemostimportantparametersinlicensingcalculations,ispresentedforalltheperformedtests.
Theassessmentcalculationswereprecededbynodalization,timestep,andmovingmeshstudies.
TheRELAP5/MOD3.2.2Gammacodewasfoundtostillhaveseveraldeficienciesintherefloodmodel.Thecalculationresultsshowasatisfactoryagreementwithexperimentalinnerpeakcladdingtemperature,howeverthepredictedtemperatureturn-aroundtimesandquenchtimesaresignificantlytooshort.Theresultsalsoshowasignificantover-predictionoftherefloodheattransferandthevapourtemperatures.Thevoidprofiledownstreamthequenchfrontisnotcorrectlypredictedeither.Finally,thepresentrefloodmodeldoesnotproperlyreflecttheeffectsofspacergridsontherefloodheattransfer.
InspiteofthesedeficienciestheimprovementsincorporatedintoRELAP5/MOD3.2bythePaulScherrerInstitute(PSI)eliminatedtheunphysicalbehaviorssuchascontinuouscoolingwithoutclearturn-aroundtemperatureandnovisiblequenchingphenomena,whichwereshownintherefloodcalculationsbymeansoftheRELAP5/MOD3.1code.
