ML17053C002

Final Deficiency Rept Re Biological Shield Wall Linear Indications.All Shop Weld Joints Evaluated Per Aws D1.1 & PSAR & Were Determined Acceptable or Repaired.Qa Program Modified to Reduce Possibility of weld-related Problems

ML17053C002
Person / Time
Site:	Nine Mile Point
Issue date:	08/01/1980
From:	DISE D P NIAGARA MOHAWK POWER CORP.
To:	CARLSON R T NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I)
Shared Package
ML17053C003	List: ML17053C002 ML18037A293
References
10CFR-050.55E, 10CFR-50.55E, NUDOCS 8010150041
Download: ML17053C002 (166)

Text

LN7NlAOARALNNlAQARAMOHAWICPOWERCORPORATlON/300 ERIEBOVLEVARO WEST,SYRAcuSE.

N.Y.l3202/TELEPHONE (315)474l5llAugustI,1980OfficeofInspection andEnforcement RegionIAttention:

Mr.R.T.Carlson,ChiefReactorConstruction andEngineering

,Support8ranchU.S.NuclearRegulatory Comiission 631ParkAvenue,KingofPrussia,PA19406OearMr.Garison:Re:NineMilePointUnit2OocketNo.50-410'.Enclosedisthefinalreportconcerning theNineMilePointUnit2biological shieldwallinaccordance with10CFR50,Paragraph 50.55(e)(3).Thismaterwasinitially reportedtoyourstaffonMay30,1979asinvolving adefectinthebaseringtooutershellplateweld.Investigations subsequent totheinitialnotification revealedthattherewasapotential deficiency inotherweldsofsimilargeometrical configuration.

Theincreased scopeofthebiological shieldwallreviewwasreportedtoyourstaffonAugust24,1979.Afterextensive engineering evaluation, ithasbeendetermined thatthecondition ofthebiological shieldwallweldscouldnothaveadversely affectedthesafetyofoperations oftheNineMilePointUnit2planthaditremainedundiscovered.

Thebasisforthisconclusion isdiscussed indetail-inSectionVI,AnalysisofSafetyImplications, oftheenclosedreport.Alsoincludedinthereportisadescription of.thedeficiencies andthecorrective actiontaken.Verytrulyyours,NIAGARAMOHAWKPOWERCORPORATION OonaldP.OiseVicePresident Engineering OirectorOfficeofNuclearReactorRegulation U.S.NuclearRegulatory Commission ington,O.C.20555Rash80>Ox500tlPEF:jaEnclosure xc:OirectorofInspection andEnforcement U.S.NuclearRegulatory Commission Washington, O.C.20555

TABLEOFCONTENTS~PaaNo.I~IHTRODUCTXON BCStatement ofProblemandSummaryofCoaclusioas Overvie~ofReportRevisions toInterimReportDESCRIPTXON OFBIOLOGXCAL SHIELDWALLABoCeDiEoF~GoHoPhysicalDescriptioa Fuactioaal Requirements Summaryof'esignCriteri'a aadLiceasiag Commitmeats SummaryofAnalytical Techniques

'ummaryofStressesforDesignLoadiagsSummaryofFabrication andErectioaStrategySpecificationRequirements Qh.ProgramRequirements 2,234566IXI+.DESCRZPTIOH OF'ROBLZH ABiStatement ofProblemMetallurgical Discussion 9~10EHGXHEERZNG EVALUATION AHDCORRECTIVE ACTION12kaBiC+DiIntroduction OverallApproach1StressAnalysis2~FractureMechanics XaaerWalltoStiffener Evaluation 1Approach2~UTTechniques.

Employed3~HapofDefectSizesandLocations 4iSummary.ofEvaluation ResultsaadRepairsRequired5ExampleCalculation CoverPlatetoStiffener Evaluation 121212132Q20202121211Approach.2UTTechniques Employed3iNapofDefectSixesaadLocations

'4.SummaryofEvaluation ResultsandRepairsRequired5ExampleCalculation 2324252526 I*I I~~e~C~PeeNe.EStiffener toStiffener Evaluation IApproach2HapofDefectLocations 3SummaryofEvaluation ResultsandRepairsRequired4iExampleCalculation FConservatisms InEngineeringIEvaluations QJALZTTASSDRANCE/CORRECTXVE ACTION2728282830ABoCoDiEoIntroduction SummaryofWeldProblemsShytheProbZemsVereNotDiscovered intheShopCorrective ActionfortheBSWActionstoReducethePossibility ofRecurrence 3333343434ANALYSISOFSAFETTIMPLICATIONS COHCLUSXOHS APPENDICES 404lARevisions toXnterimReport:BHSZWWeldJointMTDataC~HistoryofEvents424445

~~I0LISTOPTA3LZSTableNumberTLeleInnerVaIItoStiffener UTIndication DataCoverPLatetoStiffener UTIndication DataStiffener toStiffener UTIndication Data J

h~ILISTOFFIGURESFiureNumberTitleReactorBuildingConfiguration 3BaseDetailElevation CoverPlatetoBasePlateVeld8'ertical Stiffener toInnerWallandCoverPlateVeldsHorizontal Stiffener toVerticalStiffener VeldsIHorizontal Stiffener toInnex'ali andCoverPlateWelds.CoverPlatetoBasePlateWeldToeandRootIndications 10Horizontal Stiffener toInnexWallVeldIndic'ation SurfaceandSubsurface Defects12DefectSize13Stiffener toInnex'ali StraightBeamand4~egreeAngleBeamScanningPatternsCalibration BlockUsedtoEstablish PrimaryRefex'enceLevel1516Calibration BlockUsedfortheExamination ofFusionandHeatAffectedZoneinVerticalStiffenex MapofInnex'all toStiffener UTIndications 1718Subsurface DefectCoverPlatetoStiffener WeldZones19CoverPlatetoStiffener WeldZone2BeamA'ngles

'J PieNumber20TLtlaReference BlocksHapofCoverPlatetoStiffener QTIndications 22ExampleofCoverPlatetoStiffener UTIndication-TypesofUTforStiffener toStiffener WeldsHapofStiffener toStiffener UTIndications PipeRestraint WthGussetShearStressDuetoTemperature Differential

'l~\4IINTRODUCTION AStatement ofProblemandSummaryofConclusions

/Duringnondestructive examination ofthebiological shieldwaLL(BSW)attheJobsite,welddefects*werediscovered inanumberofshopwelds.Investigation hasshownthatthewelddefectsexistedprimarily inbackingbarwelds;onlyminordefectswerediscovered indoublebevy;welds.,ALLshopweldpointswereevaluated 1naccox'dance withAWSDLIandoriginalPSARcommitments andeither-wereshowntobeacceptable orwererepaired.

Basedontheengineering evalu-ation,ithasbeendetermined thattheBSWweLddefectscouldnothaveadversely affectedthesafetyofoperations oftheNineKt1ePo1nt2planthadthewelddefectsremainedund1s-coverado*Note:Defect"isusedthroughout thisreporttodescribeawelddiscontinuity andshouldnotbeconstrued tomeanthataweL6discont1nuity isnotacceptable BOverviewofReportThepurposeofthisreportistoprovidethefollowing:

1~Adescription oftheBSW,including itsspecification andQualityAssurance px'ogrsmrequirements 2A,detailedtechnical description oftheweldproblems.

30'presentation oftheresultsofanengineering evaluation andacorrective act1onplan,including adiscussion oftheoverallapproach, summaryoftheevaluation resultsandrepairsrequired, andexamplecalculations

,TheQualityAssurance corxective action.5.Ananalysisofsafetyimplications.

C'evisions toInterimReportPortionsoftheZnterimReportofApril15,L980,requirerevisions torefLectthefinalclosureplanactionwhichwassubsequently adoptedandclarifications tomaintaincon-sistentterminology.

Eatheinterimreport,itwasinadvextentlystatedthattheeng1neering evaluation wasnotinaccordance withAWSDlLEAsstated1nthisreport,allUTevaluations andacceptance ofindications axeinaccordance withAWSDl,l,paxagraph 3.7.6Detailsofinterimreportrevisions arelisted1nAppendixAtothisreport.

ZIDESCRIPTIOH OPBIOLOGICAL SHIELDWALLAPhysicalDescription TheBSWissupported bythereactorpedestal(Figure1)andisattachedtothepedestalbymeansofembeddedanchorbolts(Pigux'e,2)

~TheBSWisanextremely stablestructural systembecauseofthelateralsupportatthedrywellfloorinthereactorbuildingandthestartrusssystemnearthetopofthedryweII.TheBSWconsistsoftwoconcentric steelcylinders connected byhorizontal andverticalstiffeners (Pigures3and4).TheBSWis48ft4incheshighandhasaninnerradiusof14ft3/4inchandanouterradiusof15ft91/4inches.TheBSWwasfabricated inthreerings,eachapproximately 16feethighsEachringwasshopfabricated inthree120degreesectionsi Theinnerandoutershellsandthestiffeners are1I/2inchesthick,A537ClassIsteelplatesconnected byfullpenetration weldsThespacebetweentheshellswillbefilledwithnonstructural highdensityconcreteforneutronradiation shielding purposesi TheBSWispenetrated byairductopenings, inspection

openings, instrumentation lineandpipesleevepenetrations, anddoox'penings forvariouspiping-systemsAttachedtothewallaxepiperestraints, aBSWextension tosupportthestartrussandstabilizer, clipanglesupportsforfloorbeams,andinsulation supportbrackets.

Thefullpenetration weldsusedintheBSWax'ebothsinglebevel(withbackingbars)anddoublebeveltypesThefollow-ingtableliststhevariousweldconfigurations, theirabbre-viations, andthefigureinwhichtheyareshown.WeldJointAbbrevia-tionPigureNo.CoverplatetobaseplateVerticalstiffener toinnerwalICoverplatetoverticalstiffener Verticalstiffener tohorizontal stiffenex Horizontal stiffener toverticalstiffener Coverplatetohorizontal stiffener Horizontal stiffener toinnerwaIlCPBPVSIWCPVSVSHSHSVSCPHSHSXW5667a7b88BPunctional Requirements Thefunctional requirements oftheBSWare:IProvideshielding againstradiation fromthereactorvesselo2Provideanchorage supportforpiperestraints, pipesupports, flooringbeams,andinsulation.

3~Providesupportforthestartruss/stabilizer system.4.Protectthereactorpressurevesselfrompipewhip,getimpingement, andmissileloads J

REACTORBUILDINGCONFIGURATION REACTORREACTORPRE5SUREVESSEL(R.P.V.)STARTRUSSTOPOFBSWEL.314'-1~/g" EL.316'-1I\28'1/l~67'-4"IEL.302'-0BENDLINEPRIMARYCONTAINMENT REACTORSUPPORTSKIRTBIOL0GlCALSHIELDWALL EL.266'-6~/q" EL240'-015'-3"DRYWELLBENDLINEREACTORP'EQESTAL 4l4IMATEL.175'-07FIGURE1 J

BASEDETAIL3'-5>is"-8sic"lNNERWALLGROUTAREAQ'o:<~:o'::1>/a":PASEPLATN'.:>:i'4?o..!

'o::-'>/~"

COYERPLATENASHERPLATEEL.265'-5>lz"3)'ISOLEPLATE'"":0ANCHORBOLTPEDESTAL21'I31$

P 3RDRING17'-6I/2"2NDRING)5'-)S/I6")STRING)5'-63/16")gt03/gllRINNERWALLCONCRETECOVERPLATEHORIZONTAL STIFFENER

'ASEPLATESTALFIGURC3

VERTICALSTIFFENER

)g.01'-81/2".31/2"(TYP.}11/2"(l'YP.}FIGURE$

O.CC COVERPLATETOBASEPLATEWELDCOVERPlATE7/S"l/4"45'llBASEPLATEWASHERPLATEFIGURK5 I

VERTICAL.

STIFFENER TGINNERSALLANDCOVERPLATEWELDS1>/2"lNNERWALLi/8"4$'ERTICALSTIFFENER I1/2"COVERPLATE'I/O"324~(244+84)FIGURE6

/

HORIZONTAL STIFFENER TOVERTICALSTIFFENER WELDS)/Qll4$0VERTICAI.

STIFFENER HORIZONTAl STIFFENER

.-HORIZONTAL SjlFFENERVERTICALSTIFFENER

-,b P

HORIZONTAL STIFFENER TOINNERWALLANDCOVERPLATEWELDS'II/>"80I1/2"/)30o$4oINNERWALLCOVERPLATEHORIZONTAL STIFFENER FIGUREI

CSummaryofDesignCriterfaandLfcensing Commitments TheBSWisdesfgaedfaaccordance withtheAISCManualofSteelConstruction fornormaloperating loadconditions.

Forabnormal/extreme environmental loadcombinations, theallovable stressesarefncreased faaccordance withthefactorsspecified intheNMP2PSAR.Thefollowing loadshavebeeaconsidered intheanalysisaaddesignoftheBSW:1~Deadload.

andseism1cloads2+Accidenttemperature casesconsisting ofthemaximumtemperature dffferentials betweentheinaerandouterwallsoccurring astheresultofa-loss~coolant accideat(LOCA)3Accidentpressuredffferent1al betveentheinnerandouterwallsoccurring astheresultofaLOCAPiperestraint loadsoccurring astheresultofrestraining pipesfolloving apostulated rupture5Jetfmpfngemeat loadsresultiag vhenpressurized flu1dfromarupturedpipestrikestheBSW.D.Summary'f AnalytfcsL Techniques Thestructural analysisoftheBSWvasperformed bythefiniteelementmethodusingthecomputerprogramSTRUDLTheBSWwasmodeledusinga180degreemodelwiththeappropriate boundarycoadft1oas forthesymmetric andantisymmetric loads.Analysesforgeaeralloadingconditions vereconducted usiagpriacfples ofsuperpositioai TheinnerandouterwallsasweLlasthehorfzontal andverticalstiffeners weremodeledusingisoparametrfc elementsAfterthecomputerrunsofthefndfvfdual loadcasesweremade,thestresseswerecombfnedinaccordance withtheappropriate loadcombfnatioa equationfromtheHMP2PSAR,Section12e5e2e8~3a Theeffectofaccidenttemperature conditions vasstudiedtakingintoaccounttheconcreteinsidetheBSW,Itwasconcluded thatthecoacretewouldcrackuadertheseconditions andthattheeffectofincluding concreteonthestressesintheBSWfsinsignificant Forotherloadingcoadftfons, thestressesintheBSWvouldbereducedffcoacretevereincluded; hence,ftvasconsidered conservative nottoincludetheeffectofconcrete k

E'ummaryofStx'esses forDesignLoadingsThevariousloadsaadloadcombinations considered intheBSVanalysisaaddesignmeettherequirements ofPSARSection12.5.2.83.Thestresseswhichvereconsidered includethoseduetodeadload(D),accidenttemperature (T),accidentpressure(P),seismic(E),piperestraint loads(Rr),andjetimp1ngemeat loads(R]).Inallareas:oftheBSQ,,thestressesduetothevariousRoads'were combinedinaccordance withthefollowing governing equations fromthePSAR:I6S<IOD+I~OT+I~OPle8S<I~OD+IOT+loOP+IORr+IOR)+IOEwhereSfstheallovable stressbasedoaAISCMsaualofSteelConstruct1on Thetwoloadcombination equations abovex'eflectabnormal/

eztremeenvironmental coaditfoas andgoveratheBSVdesigaThestressesiatheBSVforloadcombination equations fromPSAR'ection 125283besidesthoselistedabove,suchasfornormaloperating conditions, areverylovTheconditions whichcontrolthedesignoftheBSVaadunderwhichthestressesapproachtheallovables aretheaccidentcond1tfons.

Theaccidenttemperature andp1perestraint loadspxoducethehigheststressesgoverning theBSWdesignTheaccfdeattemperature loadsproducelongitudinal compressive stressesfathefanerandouterwallsfnthe25to30ksirangeandlongitudinal tensilestressesfnverticalstiffeners inthe15to20ksiraageThepiperestraint loadsproducecompressive ortensilein~lanestressesiahorizontal andverticalstiffeners wh1charelocateddirectlybehindthepiperestxaints.

Becauseofthelargenumberofpiperestraints attachedtotheBSV,andbecauseeachxestraint canhaveanumberofload1agdirec-tfoasandmagnitudes, thestressx'angeinthestfffeners behindtherestrafats variesfromapproximately 5to35ksfineithertensionorcompression Onlystfffeners fntheimme-'diatevicinityofapiperestra1nt arestressedneartheallowable stx'essduringapiperestraint loading,Thestressesduetodeadload, seismic,pressure, andgetimp1ngemeat loadfngsaxe,1ageneral,lessthan8ksfwhencombined,

P.SummaryofFabrication andErectionStrategyOnthebasisofashippingandeconomicstudy,itwasdetermined thattheBSWshouldbefabricated andshippedinnine120degreesections, eachapproximately 16feethigh.Thesesectionswouldthenbeassembled toformthree360degreeringswhichwouldbestackedandweldedtogethertoformthewaLL~-Thefabrication sequenceofeachring(consisting ofthree120degreesections) wasasfollows:2%kfabrication jigwaserectedtoalignandholdtheshieldwallplatesduringfabrication Theinnershellplatewaserectedandweldedtothefabrication jig3aThehorizontal andverticalsiffenerswereweldedtotheinnershellplatei4~Thecoverplateswereweldedtothestiffeners 5Thethree12~agreesectionswereremovedfromthefabrication.

jigforshipment.

6-Thesectionsweretransported tothesiteforfieldassembly..

ThefieldassemblyofthenineBSWsectionsoccurredintemp-eraturecontrolled enclosures, separatefromthecontainment buildingi lntheassemblysequence, thethreesectionsofeachringwereerectedandroundedupusingjackingspiders.TheinnerwaLLandhorizontal stiffener weldsateachverticalseamwerewelded,andthecoverplatesattheverticalseamswereweldediPollowing fitupandweldingofthefirstring,itwasinvertedinordertolevelthebaseplatewhichhaddistorted duringfabrication, Thebaseplatewasleveledbyattaching shimplates,depositing weldmetaL,andmachining thesurfaceTwointermediate postweldheattreatments ofthefirstringwereperformed whileitwasintheinvertedposition.

Weldjointreezmdnation byultrasonic andmagneticparticletestingwasperformed ontheshopwelds,andasubstantial amountofrepairwasaccomplished Whileworkwasprogress-ingonringI,rings2and3wereassembled, fittogether, andgirthweldedRingIwasinvertedandrings2and3werestackedonringI~HMandengineering evaluation ofdefectswereaccomplished andrepairsweremadeasrequired.

Uponcompletion ofallfitupandrepairwelding,thecompleted BSWwasheattreatedandpreparedfortransport andplacement onthereactorpedestalinthecontainment building.

'J GSpecification Requirements TheBSWfabricat1on syecificat1on hasthefollowing technical, workmanship, andinspection requirements.'

Allfabrication workwasperformed uaderthefabricator's QAprogram20Allwelding,weldiagyrocedures, andwelderqualifica-tioasshallbeinaccordance withAWSDll.30ZaadditiontotheAWSDlIrequirement forl00percentvisualinsyectioa, allfullpenetration weldswererequiredtoreceiveadditional NDE1naccordance withthefollow1ag oyt1ons:aRadiographic orultrasoaic inspection b~Progressive magaeticparticIeinspection atI/3,2/3,and3/3weld)ointthickness.

(Thefabricatox'hose toemploytheUToptionaswellasMToftherootpass)4Thequalityofworkmanship shallconformtotherequire-mentsoftheAZSCCodeofStandardPracticeforSteelBuildings andBridges,i976.5Nondestructive examination ofweldsshallbeiaaccordance withAWSDlI,Section6Nondestructive testoperators shallbequalified iaaccordance withSNT-TC-LA, Recommended Practices, Nondestructive Testing,Personnel Qualificatioa andCertification, l9756Thesteelshallconformtotheapplicable AS'pecifi-cationasgiven,aadshallbetraceable atalltimestoaspecificheatnumber.AllplatesshallbeUT'edtoAS'IKA578,LevelIStiffener andbaseplatesteelshallaotexceed0OLpercentsuLfur.7Theweldf11lermetalshall.coaformtotherequirements ofAWSDll(1975edition)~HQAProgramRequirements TheQAProgramimposedbythespecif1cation requiredcertainactionsbybothS&WandtheSeller,CivesSteelCompany.Thefabrication specification wasreviewedandapprovedbyS&W'sQADepartment (QualitySystemsDivision) toensuretheinclusion ofayyropxiate QA/QCrequirements.

'S Thespecification andTest,Xnspection, andDocumentation (TID)Reportrequiredthefollowing:

GivesSGW1~Compliance withAppendixB10CFR502..Submittal ofQAprogram,3.Transmission ofQArequirements toidentified subcontractors 4Conformance ofNDTtoAVSDlI19755Submittal ofweldingandNDTprocedux'es.

1l.Qualification bysurveyandauditofCivesasSeller2.Perfoxmance ofinspections (overa26~onthperiod)'overing specificTXDattributes 3Thefollowing attributes addressweldquality:VeldingProcedure Electrode Contx'olProcedure Qualification ofVeldersVeldPreparation VeldInspection RandomCheckofPabrication Completeness Xnspection ofSurfaceDefectsHDTTestOperatorCertificationa.

HDTXnspection Procedures HDTXnspection ofVeldsReportsofNDTTestsInaddition, SSV'sProcurement QualityAssurance Department performed anannualevaluation oftheSeller'squalityhistory'ncluded inthisevaluation wasareviewof:PriorQualityProgxsmAuditsSellerSurveillance Activities Thenatureoffrequency ofhaxdwareunsatisfactory inspection reportsandnonconfoxmances Resultsofauditsbyothersources(i,e.,Client,CASE,ASME,HRC)whenavailable Seller'sresponsiveness andcooperation inresolving questions orpx'oblems relatedtoQualityAssurance.

Theabovedatawasevaluated fortx'endswhichwouldindicateaneedforanaudit,survey,orotherStone6VebsterManagement actionTheabovesummaryconstitutes theinvolvement oftheShVshopinspector, andcomprised ournormalQA/PQAeffortonthistypeofstructure, i'

Belowisasummaryoftheman-hours expendedbyPgAoatheNineMilePointNuclearStation-Unit2BSW(Contract NoNMP2-S204G):

VendorOivesTeskDistrictSurvisoOut~f-Plant RelatedTimeHours159ZofHours91FinalDocumentRev1ew-95Inspection Prep/Report Writing-80UpdatingandMaiataQd.ag Specification

'4189108TravelTime42024.1In-PIaatTimeRecordsVex'if1catioa SystemsVerification StatusDevelopmeat HardwareInspectioa 720*0613TotalIn-PlantTimeCIericalSurt68526539~4.15,2Typing,filing,reproducing inspection reports,updatingofficecopyofspecification, processing veadordocumentation, etc.OthereTotsl251,7431.41000Theabovefiguresdonotiacludeheadquarters supportactivities

'oquaIifysupplier, aud1ts,certifyinspectors, monitoring inspection reportsfornegativetrends,reviewqualityassur-anceprogrammanualx'evisioas, etc.~reeauditswereperformed butaotchaxgedtopurchaseorder.

J IIIDESCRIPTION OFPROBIZHA.Statcmeat

.ofProblemNumerousNDEfndfcatfoas whichwerczejectable toSecefon6.19ofAWSDl~l,andhencetothefabrication specification, werediscovered fnshopveldsafterthenineBSWsectionswereNDE'ediatheshopaadshippedtothejobsite.Theinitialindicatioas wercdiscovered withaaMZiaspectioa, duriagfieldwork,fathetoeazeaofthecoverplatetobaseplateveldjoint.UTresults,fromasampleUTexamination ofthisveld,shovedthatalthoughnoUTindicatioas verapresentfntheveldtoe,reflectors verepresentin'hezootareanearthebickfngbar(Figure9).Asaresultofthis,theweldjoiaewas100perceaeUTexamined, repairsweremade,aadasamplewasremovedformetallurgical analysis.

Approx1mately 20percentofthelengthoftheCPBPveldvasfoundtobezejectable bythereexaminatfon.

Following thediscovery ofthecoverplatetobaseplateweldiadications, visualindicatioas onthehorizontal stiffener toinnerwaII(HSIW)veldswherebackingbarshadbeenremovedwereidentified duringfnspection ofthethreethizdmfng sections(Figuze10).Asaresult,thequalityofbackfngbarwcldsiatheentireBSWwasquestioned.

Anengineering 1nvestfgation wasperformed ontheHSIWwcldsbyexamining therootofthcveldusingprogressive grindingand'agnetic particletesting,to determine thedepthandleagthofthedefects(ThcMTresultsarepresented 1nAppendixB.)Theinitialresultsshowedanapproximately 22percentdefect.rate.Duetosuchahighdefectrateinbackingbarvelds,ftvasconcluded theeallweldconfigu-zat1oasshouldbeinvestigated AsamplingplanapproachwhichemployedUTinspection wasdeveloped forexamination ofthevariousBSWveldconfigu-rations~Inaddition, fourthfzdmfng HSIWspecfmens vereremovedformetalluzgical analysisTheresultsofthcsamplingplanshowedthatllofthe18weldconfiguratfons dfdaotmeetexpectedconfidence levelsSubsequently, thedecisionwasmadeeoperformUTinspection ofallshopweldjointsoatheBSW.Theresultsofthesampl1ng,plan UTdid,however,showthatHSIWweldjointindications occurredwithlessfrequency.

thanthefrequency encountered during.the iaitialMTinvestigatfon Itwasconcluded thatdur1nginvestigation oftheweldindfcatfons, grindingcausedthecrackstopropagate becauseofthejo1atrestraint; hence,amisiaterpretatioa resultedvhichoverestimated eheoriginalcracksize+Adetailedtfm~equenced h1storyofeventsfspresented inAppeadixC.

l COVERPLATETQSASEPLATEWELDTOEANDROOTINDlCATlONS COVERPLATEROOTINDICATIONS INITIALTOEINDICATION'ASE PLATEFIGURE9

HORIZONTAL STIFFENER, TOINNER%ALL,NELDINDICATON

.HORIZONTAISTIFFENERINNERWALLROOTINQICATION BACKINGBARREMOVEDINOPENINGFIGURE10 ll' B~Metallurgical Diecussioa Horizontal Stiffener CoInnerWallandBasePlatetoCoverPlateWeldsAtotaloffivemetallurgical specimens whichrepresented theworstcaseUTindicatioas foundduringtheinitialengineering investigatioa wereremovedfromthesetypewelds(singleyeewithbackingbar)todetermine theaatureofCheiadicatioas Pourspecimens wereremovedfromthethirdringhorizontal stiffeaer toinnerwaII(HSXW)weldaadonefromthesimilax'over platetobase.plate.(CPBP)weld.-Metallurgical evaluation.

ofthreeofthespecimens (twoHSIW,oneCPBP)showedthatalloftheindications wereweldingx'elated(e.g,,slaginclusions orlackoffusion)andwereinsignificant iasizeaadeffect.AcrackwasfoundineachoftheCworemaining HSlWspecimens, andthecrackingcondition wasinvestigated bytwoindependent consultants~

Xtwasdetermined thatthemostprobablecauseofcrackinitiatioa washydrogenandthatthecxackpropagation wasaresultoftheconstraint ofthestructure.

Thehydrogenpossiblyoriginated frommoisturewhichwasaotdrivenoffbythepreheat,butwhichwasretainedonthebackingbars(Thefillermaterialwasnotcoasidered tobea.possiblehydrogensource)Therewasnoevidenceofczackinitiatioa orcrackgrowthsubsequent'o Checompletioa ofweldingintheshop.Since.thelastshopweldswerecompleted iaDecember1978,ithasbeendetermined thattherewasinsignificant riskofcrackinitiaCion orfurthex"crackpropagation betweenthetimeoftheinvestigation (lateHay1980)andthetimeofpostweldheattreatment (PWHT)(mid-July 1980).,Allpossibility offurthercrackingintheunloadedstructure hasbeeneliminated bythePWHT(Thepossibility ofcrackpxopagation intheloadedcondition isaddx'essed inSectionIV.B,2)Pux'thermore, twointermediate PWHT'spreviously pezfoxmed onthefirstringdemonstxated thatthePWHTprocessitselfdidaotinducecrackinitiation orpropagatioa andChatreinspection aftex'WHT wasaotreq'uired Therefore, theengineeriag evaluation andcorrective action,presented inthenextsection,showsthatthiscondition,was adequately

detected, evaluated, aadrepaired.

CoverPlatetoStiffener WeldeThezewasahighrefection rate(17percent,22percent,and12percentonringsI,2,and3,respectively) duringCheinitialreinspection ofthecoverplatetostiffener weldstothestandardAWSDl,lcriteria, Subsequent investigation byexcavation determined thattheindications wereinboththeweldmetalandinthestiffener platebasemetal.Therefection ofthebasemetalindications waslargelydueCodifficulty ininterpreting theAWSDl.lUTresults.Thelaminar-type plateindications wereacceptable totheASTHA20plateUTcriteria, ASTMA578LevelI,aadtotheAWSDl.ledgepreparation criteria.

10 I

However,plateindications acceptable toA578Level1andlatersub)ected toAWSDlIstandardweldUTacceptance criteriawouldberespectable.

SinceCheinitialAWSDl.lUTdoesnotprecisely categorize bytypeorsize,orlocatetheseplateindications, theUTinspectors conservatively assumedthattheindications wereintheweldmetaLand,therefore, respectable.

Asaresult,somereworkoftheseweldswasperformed inaccordance withthestandardAWSDl.lUTacceptance criteriapriorto-establishing appropriate engineering.

criteriaasused,intheengineering evaluation addxessed inSectionIV.UTreexamination ofsomexeweldedexcavations resultedinunexpected stiffener platelamellarseparations whichwereattributed.

Corepairweldshrinkage stresses.

Insummary,itwasfoundthatthefirstappx'oach, usingthestandardAWSDlIUTinspection acceptance

criteria, wasrequiring agreateramountofrepairthanwasnecessary, andChattherepairsinmanycaseswereineffectual inChattherepairitselfcreatedasevereweldrestraint condition andplateLameliar.

Cearingconcern.Therefore, adifferent approachwasdeveLoped tobetterdefinetheindication sizeandlocationwhere-possibleasitrelatestoitsmetallurgical environment, asdiscussed inSectionIV.Anengineering evaluation ofChesizeandlocationdatalimitedtherepairCothatrequiredtomeetdesignrequirements.

Then,toavoidtheconcernoflamellartearingduetorepairweldshrinkage

stxasses, stxictweldingcontrolswereimple-mentedGanera1WeldualitTheweldsChathavebeendiscussed inthisxeportweremadeintheshop,orduringfieLdrewoxk,repairor-replacement, usingeitherthe'floored processwithgasshielding, oxtheshieldedmetal~reweldingprocess.ALIweldsexaminedmetallurgically exhibited overallhighquality.Nocrackinghasbeenfoundinweldmetal;allobservedcrackinghasoccurxedintheheataffectedzone(HAZ),orelseoutsidetheHAZin-thebasemetal.Thestabilization ofthecrackingcondition asaresultoftimeandthePWHT,theevaluation ofthiscondition bystressandfracturemechanics analysis(addressed inSectionIV),andstxictweldingcontrolsafterPWHTensuresadequateweldmentquality.

IV+ENGIHEEEING EVALUATION ANDCORRECTIVE ACTIONAIntroduction Theresolution oftheBSWwelddefectpx'oblemisdividedintotwomainphases-UTofallaccessible shopweldsandengi-neeringevaluation ofUTindications.

IA100percentUTreexamination hasbeenperformed onaccessible shopwelds.(Approximately 90percentofthe.shopweldsareaccessible for'T)Eitherastandard'UT inaccordance withAWSDl1,Section6.19wasperformed whereaccesspermitted oxaspecialUTinaccordance withAWSDl1wasperformed 20Basedonthe,UTdata,anengineering evaluation waspezfoxmed todetermine whichindications wereacceptable asisandwhichrequiredrepaixsTheevaluation con-sistedofbothastressanalysisandafracturemechanics analysistoestablish technical Justification fortheacceptance ofinconsequential defectsAllPSAKcommit-mentsxegazding loads,allowable

stxesses, andothertechnical requirements weremaintained TheUTandengineering evan.uation ax'einaccordance withAWSDl.l,paxagxaph 3.7.6.Inaddition, theresultsfromthemetallurgical analysesofthefiveweldspecimens whichwereremovedfromtheBSQhavebeenfactoredintotheoverallplanforresolution oftheproblemsBOveralLApproachIStressAnalysisTheUToftheinnerwall.tostiffener weldsprovidedwelddefectsizes,locations, andorientations.

Astx'essanalysisconsistent withthefollowing stepswasperformed inthevicinityofeachwelddefect:aThemaximumtensilestressduetothecombination ofindividual loadcasestresses(inaccordance withthePSARloadcombination equations) wasdetezndned b~Theauuehnnatensilestresswasfactoredbytheratiooftheoriginalweldazeatothezeducedweldarea.Thisreducedweldarearesultedfromthewelddefect.12

cTheresulting stzesswascomparedto-thefactoredallowable stressesIfthex'esultfng stresswerelessthanthefactoredallowable stx'ess,thewelddefectvasacceptable fromastressanalysisviewpoint;ffgreater,thewelddefectvasrepaired-Ifadefectvereshowatobeacceytable fromastzessanalysisviewpoint, afracturemechanics analysiswasyerfohaed usfngthestressanddefectinformation todetermine thedefect'sacceptability.

Ifadefectveredetermined to.beeasilyaccessible andrepairable vfthoutunduehardship, itvssrepairedeventhoughdetermined tobeacceptable basedoaanengineering aaalysfs2FractureMechanics Intheindustrytoday,aconservative fzactuxemechanics analysisfsaovaregularmeansofassuringtheintegx'ity of,weldedstructux'es whichrealistically containsomediscontinuities fn,thematerial, suchasslag,yozoefty, orLackoffusfoaSuchananalysisprovidesasoundbasisforestablishing acceptability criteriaforthediscontinuities aadthuscaneliminate unnecessary repairs.Pormoststructural steelsundernormaldesignconditions, linearelastfcfx'acture mechaaics (LEFM)mightnotbeapplicable.

Thex'efore, inthisanalysisweueebothLEAMaadatechaique byDovtfngandTowaley(Reference 1)'hismethod,,knownasthe~rfteziaApproach, haebeenusedbyothers(References 2and7),andcoversthespectrumofconditioas frombrittlefracture(whereIZFHfsapplfcable) tocompletely plasticfailure(wheresomeformoflimitanalysishastobeused)Thestructure febuiltofASSAMA537Class1steel,whichhasexce1.1eat fractuzetoughaess inthelongitudinal direction IncasesWereftfsrequixed, thedirectional properties fnthethrough~hie%ness dfxectfoa axeconsidexed.

Porthepurposeofthisanalysis, thevaxietyofdiscon-tinuities whfchmightbeeacountered inaweldedstzucture canbereducedtotwoma)ortypes:suxfaceandsubsurface defecteo13 l

~'~~Itisassumedthroughout theanalysisthattheappliedstressisatensilestressperpendicular tothedefect,theappliedloadsaredynamicandthereisnocyclicloadingwhichcouldinitiatefatiguecrackingInthisanalysis, surfaceandsubsurface defects(Figurell)aredefinedasinASMEXI,Division1Thefollowing relations describeLEFMapproachasitappliestothisstructure:

IInthecaseofasurfacedefectattherootofaweld,"thestressintensity factorisgivenbylr'V'('/W"~-Inthisequation, thedefectsizea~ha+a(Pigure12),wherehaistheactuaLdefectsizeandaisthestressedportionofthebackingbarThestressSSlied+Sid~Thefre~urface correc-..factorPS~'1+012(I-<);theshapefactorPE1/~Q(seeASMEXI,AppendsA)PactorP>isthecorrection forfinitethickness, t,ofaplates2tcaP~-tan-'g.i~Zt(la)In.thecaseofaweldmentwithabackingbar,t~(thickness

+aFactorP+accountsforthestressfieldgradientcausedbychangesingeometryofthestressedstructure Inthiscasethefactorappliestodefectsemanating fromtheroot'andthetoeofT~LdsAnalysesofsuchpointswereperformed inReferences 8and9.Themostconservative ofthevaluesfromthesepaperswereusedinthecalculation.

FactorP(a/a)describes thestressdistribution inthebackingbar:P(a/a)~---(a/a)-1+</2-sina/a2B2-1'.NThefracturecriterion canbepresented as(Ib)(2)Tobeconservative, thedynamicfracturetoughness, Kld,isusedinthisanalysis.

Kdiscalculated fromtheSai?orsWorten relation(Reference 10):K15873(C)'hereCistheCharpyV~otchimpactenergy.v14 Pl SURFACEANDSUBSURFACE DEFECTSSURFACEDEFECT(B(a).2aSUBSURFACE DEFECT(B>a)FlGUREll

~'

~]1/s~tHORlZONTAL STlFFENERBACKlNGBARlNNERWALl.FlQURE12 P~'

Porsubsurface cracks,thestressintensity factorisdefinedas:K~MFES~caHere,Misthecorrection factorformembranestress(seeASMEXI,AppendixA).PactorPistheEsameasinEquation1~Thedefectsize,a,.isdefinedinFigure12~TheTwoWriteria relationforthecriticalapplied.stress,Sp,is:J2-1S~Sx-cosPu22expxx(S-S)kresS2Theultimatestressis:S+S2-2<1(1-t/t)-~2"(1-t/t),whereSandStaretheyieldstrengthandtensileystsstrength, respectively.

Thecriticalstress,S,iscalculated fromthefollowing LEFMrelation:

Xd1+4.593(a/g)

Sk~Zt[1.12Ftp(a/tg)]

ttn't2tInthecaseswherebendingandmembranestressesareactingonadefect,thestressintensity factorforasurfacecrackisgivenbythefollowing LEPMequation:

ISWbbGE~whereMisthecorrection factorforbendingstress(seeASMEXI,AppendixA),SisthemembranestressandSbisthebendingstress.Theotherparameters aredefinedasabove.TheLEPMequationforasubsurface crackis:K~(MS+MSb)PE+Ca(Thetwo~riteria methodisnotusedincaseswherebendingandmembranestressesareactingsimultaneously becauseofsomeuncertainties inapplyingthismethodtoanonuniform stressfield)15 J

Zaconclusion, itshouldbeemphasized thatthefracturemechanics analysisisveryconservative.

Conservative assumptions areusedforthedefectsizeandalldefectsareassumedtobesharp,whichforthelargema)orityofcasesismoreseverethantheactualcase'lso,itisassumedthatthestressesduetovariousmechanical andthermaLloadsactsimultaneously andthattheappliedten-silestress-is perpendicular tothedefect.Thelowestvalueofthefracturetoughness isusedthroughout thestructure Thus,theanalysisprovidesadditionaL assuranceforthesafetyoftheBSWstructure~

Toillustrate theapproachoutLinedinthereport,atypicalexampleisgivenbelow,whichevaluates theeffectofexclu-sionofI/8inchfromtherootareaofabackingbarweld.Thematerial, A537CLass1steeL,hasminimumtensilestrengthS70ksiandminimumyieldstrengthS50ksi.Thelowestxracturetoughness canbeexpectedintPethrough-tathickness direction Thepostulated defectislocatedintheEC(Figure12),wheretoughness inthethrough-thickness direction, according toSMexperimental data,is'etterthaninthebasemetal;thebase.metaltoughness isusedhereforthesakeofconservatism.

Basedontheexperimental Charpyimpacttestsperformed at90Pandavailable published data(Reference ll),theCharpyenergyat100Pis:C>20fr.IbIZPMAoachThestressintensity factorforthiscaseisgivenbyEquation2(seeSectionIV.B.2)Theparameters intheequationaregivenbelow:a<<LLa+aPorthisparticular examplea<<0.090inch.Thus,inthisexample:a<<0125+0090<<0215inch.ThemaximumaveragetensilestressisgivenasSlid25ksi.0appliedTheresidualstressafterPWHTat1,100Pisassumedtobe10percentoftheyieldstre11gthy thatis+S<<O,lS<<01x50<<5ksi.res's16

Therefore, PS~1.12S~S+S~25+5~30ksiappliedresPEisgiveninAppendixAtoASHEXZ,butitcanbealsoexpressed bythefollowing equation:

11P1+4e593(-)';,Oe212(S/S

'WereEisdefectlengthP>isgivenby'quation 1a(seeSectionIV.B2)P0~1328(inthiscsee)8(a/+)8(-')0~86(see8ecnionZV82,.E,qcanion Ih)0.2150090Pinally,notethatthe'efect isassumedtobeinfinitely longsothata/<~0Thus,X~le12PP(a/a

)S~ca-tan1+4593(-)0212(S/S)21~12x1.328x0.86x30~0.213x 2x1.590.215'an0215am2x1.591+4593x(0)'212(-)le653025033.2ksij/inch Thefracturetoughness ofthematerialiscalculated fromEquation3(seeSectionIViB2):2613~823(20)48.8ka1~1ach0.375Kl~33.2<KZd48.8ThecriticaldefectsIzeevaluated fromEquation1(seeSectionZV.B.2)is:da~0.47inch.crTheevaluation isperformed byaniteration processuntilacalculated stressintensity factorequalsthefracturetoughness ofthematerial.

17

~~TwoMriteria AoachThecriticalappliedstr'essisgivenbyEquation'5 (seeSectionIV.B2)~Theparameters inthisequationaregivenbyEquations 6and7(seeSectionZV.B.2),andS~5ksi.Thecriticaldefectsizeisthenevaluated fBmEquation5(seeSectionLV.B2):da~~0.42inchcrTheactuaLdefectsizeisshowntobemuch,l'ess thanthecriticaLdefectsizeandis,therefore, acceptable.

Furthermore,theratioofthecriticalcracksizetotheactualcracksizeislarge.3.36.cr0.42ha0.125Therefore, aI/Stanchdefectattherootofaweldsub)ected to25ksi.average-tensilestressisacceptable.

18 I

REHRENCES 1~DowlingyAeReandTownley,C~H.A.TheEffeetonDefectsonStructural Failure:ATwoWziteria ApproachTheL'nteznational JournalofPressuxeVesselsandPiping,3,2,1975,p77-107.2Chell,GCACombinedLinearElasticandPost"Yield FractureMechanics TheoryandItsEngineering Applications, FractureMechanics inEngineering

Practice, Editor,PStanLey.AppliedScience,London,England.3~Harrison, R.PAUnifiedApproachtoFailureAssessment ofEngineering Structures FzactureMechanics inEngineering

Practice, Editor,PStanleyAppliedScience,,

London,England.hMuscati,A.andTurner,CEPoet-Yield Fx'acture BehaviorofShaIIMotched AlloySteelBaxsinThree-Point BeadingFxactuxeMechanics inEngineering

Practice, Editor,P.Stanley.AppliedScience,London,England5~Townley,CHATheIntegrity ofCrackedStructures UnderThermaL.LoadingiFractureMechanics inEngineering Px'actice, Editor.,PStanley.AppliedScience,London,England6Roche,RLAnalysisofStructures Containing Cracks-SomeCommentsontheJIntegralCriterion TheZnternational JournalofPressureVesselsandPiping,7,1979,p657'loom,JMPrediction ofDuctileTearingofCompactFxactureSpecimens UsingtheR-6Failux'eAssessment DiagramsTheInternationalJournalofPressureVesselsandPiping,8,1980,p215-231.8~Usami,S.etaL~Transactions oftheJapanWeldingSociety,April1978.9<<Guerney, T.RFiniteElementAnalysisofSomeJointswiththeWeldTransverse totheDirection ofStressWeldinginstitute ResearchReport,E/62/75,Mazch1975.10eSailorsyRoHoandCortenyHeToASMSTP513yp164oLL~Lents,JS~JournalofPressureVesselTechnology, Vol'00,February1978,p77.19 P

CInnerVaLLtoStiffener Evaluation 1ApproachBasedoaUTdata,velddefectsizes,.Locations, aadorientatioas vereobtainedforevaluation.

ThespecialUTperformed fromtheinaerwallwasthesolebasisforevaluation ofindications aadsubsequent acceptorzevorkdispositioas Stressesinthevicinityofthedefectswereevaluated asaresultofreducedveldareaduetothedefectsAllPSARcommitments regarding allowable.

stresseaweremaintained.

Afterit'waeshownthatthestressesvereatanacceptable level,thestressanddefecCinformation vasusedinafracturemechan1cs evaluatioao 2.UTTechniques EmployedAllhorizontal andverticalsciffeners toinnerwalLveldsvere100percentexaminedfromtheinnerwallsurfaceusingspecialultrasonic techniques inadditionCothosedescribed inASSDl.lThesetechniques wereespecially effective indetecting plana~ype discon-tiauities locatediafusioaandheataffectedzonesoftheseweldsTheexamiaatioa employed5MHz1/2&nchdiameter, straightbeamtransducer and4MHz,8x9mm45&egzeeanglebeamtraasducers toscan,theweldsasshowniaFigure13Thetx'ansducer frequencies andsizeswereselectedtoprovideoptimumsensitivity, resolut1on, aadminimaliaterfezeace fzomnea~ieldeffectsTwocalibration blocksasshowniaFigures14and15werepreparedforestablishing primaryreference sensitivity levels.Thecalibration blockinFigure14containsaflatbottomslot1/8inchwide,.vh1chwasusedtoestablish the.reference levelforthestraightbeamexamination Theprimaryreference levelfox'he45&egreeanglebeamexaminat1on vasobtainedbyusingtheslotinthecalibration blockorientedata4~agreeangle.Inaddit1oa, apartofChefusionzoneofverticalstiffener alonga45&egreebevelChatdoesnotgetadequatecoveragebyanglebeamCxaneducer vasexaminedbystraightbeamtransducer Thereference seasitivity levelforthisexamination wasestablished byusingthenotchWamecond calibration blockshowninFigure15.Adistanceamplitude cuzve(DAC)wasestablished foreachexamination aadalLtheindicatioas above20percentDACvererecorded.

Allrecordable indicatioae werefurtherevaluated byadditioaal ultrasonic examination.

Theresponses fromvelddiscontinuities verecomparedtotheresponses fromvarioussizereference reflectors whichsimulated theorientation andlocationofthediscontinuities.

Duriagthiscomparison, corrections veremadeforanydiffex'ences insecondattenuation 20 4'

STIEEENER TOINNERWALLSTRAIGHTBEAMAND45'NGLEBEAMSCANNINGPATTERNSINNERWALLVERTICALSTIFFENER

>>/*"45454rr/r/rr0404INNERWALL304~///rrrHORIZONTAL STIFFENER

/454//rrPOQO45'545'IGUREl3 4

CALIBRATION BLOCKUSEDTOESTABLISH PRIMARYREFERENCE LEVEL4tliSj8118611411tt8FIGURE14 AIP CAI.IBRATION BLOCKUSEDFORTHEEXAMINATION OFFUSIONANDHEATAFFECTEDZONEINVERTICALSTIFFENER VERTICALSTIFFENERO~ey.OOO'NNERWALL,3113)lFIGURE15

characteristics betweenthereference blockandstiffener material+

3.Map.ofDefectSixesandLocations Amapofdefectsixesandtheirrespective locations isincludedasFigure16.(Thedefectsare.shownons,.developed viewoftheBSWfz'omtheperspective of'beingoutsidetheBSW,Lookingtowardtheinside,)~Atotalof23indications (fouxinthefirstring,twointhe.secondring,,and17inthethirdring)wereevaluated andareshownonthemapDetailsoftheinnerwalltostiffener UTindication'ata arepresented inTable14.SummaryofEvaluation ResultsandRepairsRequiredALL23indications wereevaluated bystzessandfracturemechanics analysesandweredetermined tobeacceptable However,tenindications werereworkedbecausetheywereaccessible fromtheinsideofopencompartments anditwasdetezmined thatreworkwouldnotbehazmfultothestructure Theremaining 13indications werenotzeworkedbecauseofoneormoreofthefollowing conditions wherereworkoftheacceptable weldswouldbemoredetrimental thanthepresenceofthedefect:aRemovaLofacoverplatewouldberequiredtogainaccesstotheindication bCuttingofanaccessholeinastiffener wouldberequiredinordertochaseanindication intoan'd)scentcompartment c~Cuttingthroughtheinnerwallwouldberequiredtogainaccessibility totheindication dExcavation ofasignificant amountofweldmetalozbasemetalwouldberequiredtoremovetheindication, thusimposinglargethrough~hickness shrinkage stressontheinnerwall.Therepairstatusofeachindication issummarized inTable1~5ExampleCalculation Subsurface DefectintheInnerWalloftheBSWThefollowing wxampleissustrates anactualcaseofareportedindication intheinnerwallofRing1.Theindication wasinterpreted asbeingasubsurface defectparalleltothesurfaceoftheinnerwall(Figure17)~za 7t sef.lxset@0snto(sersrI~snreo(tn'nceIntr(IffeffelrItt'er(et~loteerI~ISle'Xof-extlr.tXIeo.IXI.OXII~'let0Ser.ll'llee'Irlteerlel'OX'IotIXSle'XterIXsteerneerI-tiWSea'14ler.rsff'0xr-nr.txter.t'lpslr.tx.ter.0tetStI'0~04ItrIXttr4x.styx.nr.ex.0%IlrIXtar.ags4r4+tte'srFIGURE16-MAPOFINNERWAI.t.TOSTIFFENER UTINDICATIONS

~>>setIpllolleoQg.eot.oooe,oeoocloffteetofgfoloeafftalllt OIIOIICOIICttff COIIIIlaffs55.Ioloofftalffo.

Alfferoocelff conenaffltltox

TABLE1-INNERWALLTOSTIFFENER UTINDICATION DATAIndica)f~n No,Length(Inches)Thru-Thick Depth(Inches)Accept/Rework DisositionBasisforDisositionl-l1-21-31-42-12-210(int)(2)13/1615/8281/417/81/81/81/81/81/81/8AcceptAcceptAcceptAcceptAcceptAcceptInaccessible>

Inaccessible, Inaccessible, Inaccessible, CPremovalCPremovalCPremovalCPremovalInaccessible, CPremovalInaccessible, CPremovalrequiredrequiredrequiredrequiredIrequiredrequired3-13-23-33-43-53-63-73-83-93-103-113-123-133-143-153-163-175/8371/2(int)23(int)51/83/42213/8265293/8293/8'3(int)5/8181824451/21/81/81/41/81/81/81/81/81/81/81/81/81/81/81/81/81/8AcceptAcceptAcceptReworkReworkAcceptReworkAcceptReworkAcceptReworkReworkAcceptReworkReworkReworkReworkInaccessible, CPremovalrequiredExcessive weldmetalorbasemetalremovalrequiredExcessive weldmetalorbasemetalremovalrequiredAccessible, onlysmallamountofmetalremovalrequiredAccessible, onlysmallamountofmetalremovalrequiredExcessive weldmetalorbasemetalremovalrequiredAccessible, onlysmellamountofmetalremovalrequiredExcessive weldmetalorbasemetalremovalrequiredPossibleburn-through Accessible, onlysmallamountofmetalremovalrequiredExcessive weldmetalorbasemetalremovalrequiredAccessible, onlysmallamountofmetalremovalrequiredAccessible, onlysmallamountofmetalremovalrequiredInaccessible, CPremovalrequiredAccessible, onlysmallamountofmetalremovalrequiredAccessible, onlysmallamountofmetalremovalrequiredAccessible, onlysmallamountofmetalremovalrequiredAccessible, onlysmallamountofmetalremovalrequiredNotesl.Indication numberscorrespond tonumbersinFigure16.2.(int)Indicates thedefectisintermittent inthegivenlength.3.Allindications wereacceptable baaedonengineering eyaluations.

r SUBSURFACE DEFECT11j2"REPORTEDUTINDICATION 1.32"~2a1I/2HORIZONTAL STIFFENER INNER%ALLFlGURE17 1

LEPMADDroachThestressintensity factorforthiscaseisgivenbyEquation4(seeSectionZV.B.2).Porsubsuxface defects,thedefectsize,a,inEquation4ishalfoftheactualreporteddefectsize(denoted2a).Notethattheactualdefectsize,2a,@asreportedaslessthan1/8inch,butitmsconservatively assumedherethat2a~1/8inchandtherefore a~1/16inch.Thedefectlength,according totheUTreport,wig~3/4inch.Hence,a/<~0.083.Thestress,onceagain,includesbothappliedandresidualstresses.

Theappliedstressinthisexampleis16ksi,andtheresidualstress,asinSectionis5ksi.ThusSSlied+S~16+5~21ksi.appliedresFactorM~1(defectparalleltotheplatesurface)PactorPisgivenbythesameequationasintheexampleinSectionZV.B.2above.Thusp~9.13ksi~in21cx1/16~sn1+4.593(0~083)-0.212(50)

Sincethefracturetoughness ofthematerialinthethrough-thickness direction attheminimumtemperature whenthestxessmightdevelopis:Kd~48.8ksi~n(seeSectionZV.B.2)Thus,QisQ51chlessthanThecriticaldefectsizeevaluated franLEFMEquation4,(seeSectionZV.B.2)assumingthattheratioa/~isconstant, a~13/4inchand2a~31/3inch.crcrTwo>>Criteria AroachThecriticalappliedstressisgivenbyEquation5(seeSectionZV.B.2).Theparameters inthisequationaregivenbyEquations 6and7(seeSectionZV.B.2).S~5ksiresThecriticaldefectsizeisevaluated byaniteration processfranEquation5:a~1.7inch,crsothat:2a~3.4inch.cx'herefore, the1/8inchdefectsub)ected to16ksistressisacceptable.

22 I

DCoverPlatetoStiffener Evaluation 1.ApproachThecoverplatetostiffener weldswereevaluated inasimilarmannertotheinnerwalltostiffener welds;namely,withastressanalysisemploying mappedUTdata,effective weldareareduction, andstresscomparisons, andwithafracturemechanics analysisemploying stressandUTdataTheengineering evaluation limitedtheamountofcoverplatetostiffener weldrepairtoonlythatwhichwasnecessary tomeetthedesignandPSARcomad.tments.

ThisLimitedrepairwasnecessitated bythefollowing events.Therelatively highpercentrefection rate(17percent,22percent,and12percentforRingsL,2,and3,respec-tively)duringtheinitiaLreinspection ofthecoverplatetostiffener weldswasdetermined byexcavation tobedueto.1ndications bothintheweldmetalandinthestiffener platebasemetalTherefection ofthebasemetalindications wasLargelyduetodifficulty ininterpreting theUTresultsTheselaminar'ype indi-cat1onswereacceptable totheplateUTcriteria, ASKSA578LevelI,andtotheAWSDl1edgepreparation criteriaHowever,ifthesameindications werein'theweldmetal,theywererespectable totheAWSDl.lstandardweld.UTacceptance criteriaThestandardAWSDl.lUTdidnotpreciseLy locatetheseindications; therefore, itwasconservatively assumedthattheindications were1nthetheweldmetalSomereworkoftheseweldswasrequiredinaccordance withthestandardAWSDl1UTacceptance criteria.

UTreexamination ofthereworkedcavitiesshowedunexpected stiffener pLatelamellarseparations whichwereduetotheweldshrinkage stressesinducedbyrepa1rThisapproach, usingthestandardAWSDl.lacceptance

.criteria, requiredagreateramountofrepairthanwasnecessary; therepairsinmanycaseswereineffectual sincetherepa1ritselfcreatednewweldindications.

AUTprogram,described 1nSectionIV.D.2,wasdeveloped tobetterdefinetheindication sizeandlocationOnrepairswhichwerenecessary, weldingtechniques wereimplemented tomitigatetheweldshrinkage stressproblemItshouldbenotedthatsomecoverplateweldscanbeUTexaminedonlywhenthead)scentafte~oncrete coverplatesareattached23 r

2~UTTechniques EmployedAllshopcoverplateweldswereexaminedbyultrasonics iaaccordance withtherequirements oftheAmericanWeldingSocietyCode,AWSDl.leIaaddition, allindicatioas re)ectedbythestaadardAWScriteriawerefurtherexaminedbyadditional UTtechniques toestablish their.relevancy aadtobetterquantifytheirnatureandsize~MostoftheAWSrespectable iadicatioas wex'emargina1.

aaddataisaotavailable whichrelatesAWSdefectratingstoactualflaw'izeInordertodevelopsuitableprocedures whichwouldprovide'ufficient iafoxmation forstressaaalysisaadfracturemechanics evaluatioa ofthediscontiauities, the,weldareawasdividedintothreeseparatezonesasshowninPigux'e18,aadthewelddiscontinuities tabulated foreachzone.Indications inZonelwereaotreexamined, sincetheultx'asonic responses fromwelddiscontinuit1es may'avebeeninfluenced by'umerous smallbutacceptable platelaminations contained withinthehor1zoatal aad/orverticalstiffeaers~

Allind1cations inZonelwereremovedandtheweldswarerepaired.

Indications, inZone2wereexaminedusing,inallcases,both45Megreeaad7~egreebeamanglesasshowninPigurel9Refereace blockscontaining artificial reflectors whichsimulatetheorientation, size,andlocationofcrit1calflawsasdefinedbyfracturemechanics analysiswereusedduringtheexaminatioa.

TheyareshowniaPigure20Othertestparameters, suchastransducer frequency, probesize,andinstrument cal1bratioa, wereaddressed toprovidetheresolution necessary forproperinterpretat1on during.theexamination.

Duxingtheexami-nation,allresponses wererecordediadecibelsrelativetotheresponses fromeachreference blockreflector.

Iad1catioas iaZone3wereexam1aedtoestablish thattheirphysicalsinsd1dnotexceeda1/2-inchthrough-walldimension~

Thiscriterioa wasdetermined byfracturemechanics analysisasameaningful threshold forgathering dataTheuseofa48Hz,8x9mm,7~agreeanglebeamprobewithinthehalfskipdistanceprovidedasoundbeamwithinZone3whichwaslessthanthe1/2-inchthrough-wallsize,therefore, otherestimated limitsofreflectoxs largerthan1/2incharemeaningful bythe6dBdropmethodsAlso,theseparation ofprobepositions atthe6dBlimitsfrompointsourcereflectors isfarlessthanfromthe1/2Mnchreflectors Probespacingswerexecordedforthe6dBlimitsfromallreflectors inZone3Theresultsofallexaminations werexeportedforfurtherengineering evaluation asstatedbelow.

Ct ZONE1ZONE3COVERPLATEysiSTtFFENER ZONE'FlGURE18 t

COVERPLATETOSTlFFENER WELDZONE2BEAMANGLESCOVERPLATE7P'084SSTIFFENER FIGURE19 1,P I/8"DEEPI/4"DEEPy11a.45'EFERENCE BLOCK411'I3/+SIMULATED BACKINGBARb.70'EFERENCE BLOCKFIGURE20 l~'

Interetations Theresponses fromallindications inZone2werelessthantheI/8~hdeepnotchinthe45&egreereference blockandalsosubstantially lessthanthe7~agreereference blockreflector.

Theresponsefromallreflectors inZone3didnotindicatediscontinuities withthrough~ll dimensions exceeding.

I/2inchEvaluation Thethrough~11 sizeofallZone2discontinuities maybeconsidered tobelessthanI/Oinchandthethrough~ll sizeofallZone3discontinuities lessthanI/2inchThesevalueswereusedaspartoftheengineering disposition foracceptance orrefection 3.Mapofdefectsizesandlocations A.mapofdefectsizesandtheirres'pective locations isincludedasFigure21.(Thedefectsareshownonadeveloped viewoftheBSVfromtheperspective ofbeingoutsidetheBSW,lookingtowardtheinside)411indications inZones2and3whichrequireanengineering evaluation areshownThefirstringhad71defectsrequiring evaluation; thesecondring,.51defects;andthethirdring,121defects.Additional coverplatetostiffener.

data.aresummarized inTable2~-4SummaryofEvaluation ResultsandRepairsRequiredThetotallengthofcoverplatetostiffener weldsexaminedwasapproximately 48,100inchesOfthetotallengthexamined, 243UTindications totalling approximately 1,470incheswerediscovered inZones2and3,whichwasapproximately 3percentofthetotallengthexamined.

Throughstressandfracturemechanics

analyses, 228indications wereacceptable as-is;15indications totalling approximately 142incheswerereworkedhsummary,including indication sizesanddispositions, oftheindications ispresented inTable2.InZoneIofthecoverplatetostiffener welds,48indi-cationstotalling approximately 450incheswerediscovered andreworked.

25 PI Srvetl-srr-oa-srvteIer'I0IOV1%mr.0%Ier.t)p\sacIIIntIoter.srnl'IIInser~).1).2Irr).2).0-I).I)Iae'retfr'll'll tetlr'ar-Itet'll'10

~)J~2~)~2attruat.10~Ins'IronsreraretIvre.avs.aP.IO'I.10-10lieevIrrleIrrOO'%).SlllIro.$%I~sF110~t.~.Il.11.Sl~I).1st-0I~arIerloaIev11atel4%Ier.lg~vIr'ertule't~IIIter.e'IIlr.IN22nersrneel2.~2.11~lQI.l)~I~I~2~2~2Iel'll'.1 aer.Cl~S/-~~2~'I'Ir.e-lit%aI~I'1tn.ly;nreetttl're-IS'I-2I.)2.Ci.)41~2~2~I'rt'el.syr

-trr.sa-n'lrÃrlffnSar.otasetst'll'ee'O'I~rl0fflntffrn'trnr'Irnatrne~tsa'4'Io'S

~Isoreffa'lnss'utalrnraItr~sl.sv~tre.te.FlGURE21-MAPOFCOVERPLATETOSTlFFENER VTINDICATtONS IloreoQg~rel,oooe.oeoocrorlroeoggrelearlrAnle.

WIOICCIWterrl CalletlAfflgrS.teteerrrurro.

Arrre.coecnff cornreeffs

~-

T48IH2~COVERPLATETOSTIFlEHER UTIHDICATIOH 04Th(kllfadfcatfoas sereacceptable froude411engineering evaluation unlessothersfse noted)ladfcatfon

~!la.1Length~Zona2~?eh"ladfcatfoa

~Ão.ILength~Inch)III2I3.14ISI&I7I&IW1-10III1-12I131-14I15I16I17I18I191-201-211-22I23I24I25I261-271-28I29I30I31I32I33I341-35I36I37(3)1-38I39I~I&I1421<3=1~IASI~1-47l~I&91-50I51I52I53I54.I55I56I571-58I591%022232322323232323233223232.32.2333222322333"23323332223323233'2I262II/430I/OI123/411I/O43/82,3/4II/2745/818I5I/2II/4I3/8I3/8II/42I/2II/810II/84I,I/423/4I3/4II/238I/O4I/28I/4I3/431/44I/O157I/O25/87I/827I/22II/OII/OIII/283/83I/22I/25I/2173/82I/4I3/84I/42I/478I/8I&I1&21&3I~1&51%61&7I&81-69I70I71212~2232&2S2&2~72M292102112122132142152162172182192>>202212~22223(3)2242252262272282292302312~322~332342-352-362-372382-392~2-412&22&32-442-452W62&72&823222222222.22222'322332322222,222323223323222333232332223'I/8II/OII3/8I3/4223/8'73I/22I/?3I/OI3/43/4II/4I2I/4253/8II/2II/43I/24I/2II/2I3/423/8I3/41653/42I/O43I/2843/4I3/4143/82I/OI3/4,87/842I/23I3/8I23/4II/2ISI/O6I/82I5/82I/2II/8I3/432I3/8II/243/41563/4243/4Hotes(I)Indfcatfoa aunberscorrespond toaunbersfn2fgure21.(2)Zona2indicatioas haveathroughthfckcess depthofI/4fnchlzona3fadicatfoas haveI/2inch.(3)Indication uasunacceptable froaanengineering evaluation andrevorked, (4)Indication vasacceptable froaaneagfneerfag evaluatioa butreworkedbecauseftwased)scenttoazoaeIiadicatfoa, rJ

~~Indication

~Ra.12-492502-513-13537(4)k4(4)'(4)3103ll3123-133-14.3-153163-173-LS3193203213223233243-25326327328329330'313-323-34.3353363-373383-39~I(3)~2(3)~3(4)3-41~8~9350.3SI3S23533S43-SS356(4)3573SS359(3)~2~3KocES2.32223332233.2332232222.22222222.3222322222222232,22222322322232I3/8II/2405I/4133/4103/827/83I/2335/810L2'/2II/O18I/24I/4.28'83/421/814I/241/83I/823/4II/2223/42I/225/83I/OII3/833/4I233/8I2L5/83I/83'5/85I/O103/4II/83I/233/8I5I/210I/8I5/88I/8IIII/2L33/447/815I/2I3/4II/25I/432I/2207/883/8I13/45I/4II/4II/2II/2Indication

~Ha.~5~6~7~8~93703-71312373374375376371378(4)3-79~0~L~(3)~4.~(4)~&~7(4)~8(4)~1392393394395~7398~93LOO31013l023'0331043IOS3L063-10731083-10931103lll31123-'11331143ILS31163lll3LLS311931203-12122222333233223323222233'332222'222222222,2'322323322223223Length~ZaehesII/24I/43710I/210I/233/44I/226I/24I/463/42I/O333/423L/4I3/415I/23/47I/43/423/47I/218I/23I/419213/44I/OII/27/83/41/84I/2I3/4ILI/2315/16I3/423/4II1/83/4II323/4I5/8I7/8I1/4I5/89573/42I/2(I)Indication ncncberscorreepond tonccnberainFigccre21~(2)Zone2incU,caticma haveathroccBhthicbneaa depthofI/Oinch,tone3indicationa haveI/2inch.(3)Indication vaeccnacceptable'roe anengineering evalnaticm andreccorked.

(4)Indication vaaacceptable frocaanenBineering evalnation botrevorkedbecanaeitweadfacenttoa!oneIindication.

r4 5ExampleCalculation SurfaceDefectintheCoverPlateThisexampleillustrates anactualUTindication inthecoverplatelocatedinZone2oftheweldment(Pigure22)~TheassumeddefectdepthinZone2isI/4inch(seeSectionIV.D.2)~Thetipofthedefectdoesnotreachthesurface,butsincethedistancebetweenthetipandthesurfaceislessthanhalfoftheassumeddefectdepth,thedefectiscategorixed as,asurfacedefect(seeASMEXI)~Therefore, thetotaldepthis0325inch.Thisweldmentisexposedtocombinedmembraneandbendingstressesduetoapostulated accidentwhentheoutsidewallisatahigher.temperature thantheinnerwallInsuchacaseEquation8(SectionIV.B2)givesthestressintensity factorTheparameters inthisequationaredetermined asfollows:a~4a+a4a~0325inch;a~0181inch(inthiscase)'~0325+0181<0506inchAppliedmembranestress,Sappl.,is-2.7ksiTheresidual.stress,S,isassumedtobe5ksi;therefore, thetotalmembranestressis:res'~Sappl+S~27+5~23ksim'.mresBendingstress,Sb,is33.3ksi,sothetotalmaximumstressis!S'+Sb2o3+33'3356ksiThelengthof,theindication according totheUTreportis4.25inches,soa/g~0119andPS1+012(I~/g)1'932P~1'4(SectionIV.B2,Equationla)8~0.84P~1'04(inthiscase)GP(-)~0885(SectionIV.B.2,EquationIb)E1+4593(a/i)'0.212(S/Sys) 1652 IP EXAMPLEOFCOVERPLATETQSTIFFENER UTINDICATION

.COVERPLATE3f~1.5"aa='/i"0.325"STlFFENERUTINDICATlON

='/+(1.5-1.3-1/8)

=0.325"FlGURE22

1E0,9851.6535.61+4+593(0.119)

-0212(555-}.K~(PPS+MS)PP(-)P~(l093x104x2.3+0.84x33.3)x1.04x0885x0.985x~~3498ksi~inItisconservatively assumedthatthecriticalfracturetoughness isthesameasinthethrough-thickness direction oftheinnerplate,thatis,K~48.8ksi~inTherefore,Q~34.98isLessthanR~488IdStiffener toStiffener Evaluation ApproachInvestigation hasdemonstrated thattheoverallweldqualityofthestiffener tostiffener weldsisverygoodShileallofthestiffener tostiffener weldspreviously receiveda100percentUTexamination intheshop,alloftheaccessible weldswerereexamined byUTinthefield.(Over40percent.ofthestiffener tostiffener weldswereaccessible foraUTexamination

)Onlyminorindications werediscovered duringthereexamination, andtheinci-denceofoccurrence oftheweldindications wasextremely

~LowLessthen-3percentofthelengthofweldsreinspected werere)actable totheAWSDl.lstandardUT(Pigure23),whichiswithintherangeofrepeatability expectedfromaUTreexamination Nodefectswerediscovered asaresult,ofthespecialUTexamination (Pigure23).St&xcturally, thestiffener tostiffener weldsarelesscriticalthanthecoverplateandinnerwalltostiffener weldsTheloadtransfermechanism isprimarily oneofshearalongthestiffnertostiffener weld.Ithasbeendetermined thatama)orfracturemechanics problemdoesnotexistinthisparticular configuration, Evenwithextremely conservative assumptions, suchasneglecting theinherentstructural stability providedbytherestrained, ceLLula~ype configuration ofthestiffener toatiffener welds,acrackofapproximately 3/4inchthroughMhickness depthcanbetolerated fortheentirelengthoftheweldUTdatabasshownthatnowelddefectevenapproaching 3/4inchcanbeexpected27

'

TYPESOFUTFORSTIFFENER TOSTIFFENER WELDSVERTICAl.

STIFFENER

//'//HORIZONTAL STIFFENER STANDARDAWSDl.lUTHORIZONTAL STIFFENERHORIZONTAI.

STIFFENER VERTICALSTIFFENER 00HORIZONTAL STIFFENER

~s~r')i4$iIr00404SPECIALUT(FORWELDSINACCESSISLE TOSTANDARDAWSDl.lUT)FIGURE23

Inordertoabsolutely demonstrate theacceptability oftheBSW,thestructural integrity oftheiaaccessible stiffener tostiffener weldswasaddressed iathefollowingmanner:Sheax'tresses duetotemperature aadpiperuptureloadswereusedtodetermine therequiredveldarea.(Thestressesduetodeadload, seismic,aadpressureloadswezenegligible.)

Considering theunlikelyoccurrence ofalargevelddefect,theweldarea'vailable vasdetezmined basedoathelargestdefectfoundbyUTforthisconfiguration, a1/4inchthrough-thickness indication discovered intheianexvalitostiffener examination.

Bycoasezvatively assumingthatthedefectextendsfortheentirelengthofthestiffener, theveldax'eaavailable isapproximately 84'percent oftheoriginalveldarea;80percentwasusedfortheanalysis.

Foreachinaccessible stiffeaer, itwasshowathattherewasadequateweldareaavailable toreacttothepostulated accidentloads.2.MapofDefectLocations Thelocations oftheindications discovered byUTareshowniaFigure24.(Thedefectsareshownonadeveloped viewoftheBSVfrantheperspective ofbeingoutsidetheBSW,lookiagtowardtheinside.)Additional stiffener tostiffener indication dataaresummarized inTable3.3.SummaryofEvaluation ResultsandRepairsRequiredAlloftheindicatioas originally disc'ovezed bythestaadardAQSDl.lUTverereworked.

TheMTdatawhichvascollected duringthereworkwasiavalidbecauseofthecrackaggravatioa duetogrindingasdescribed iaSectionIII.B.Therefore, themaximumUTindication sizeof1/4inchfzomthestiffener tostiffener innez'ali data(Table1)vasconservatively usedintheanalysis.

Alloftheinaccessible stiffener tostiffener weldsveredetexmined tobeacceptable basedoneagineeriag evaluations sind1artothatpresented inSectionZV.E.4.4.ExampXeCalculation Forveld)oint278/12B)readas"weldjointatelevatioa 278'-S3/4",azimuth1229',belowthehorizontal stiffener"),

theminimumrequiredweldareasaredetermined forthefollowiag tvoloadings:

a.ShearStressduetoPieRuture(Figure25)Considertherestraint withappliedmomentMof24,000in-kandpulloutloadPof744kips.Theequivalent forcecoupleFattheextremeflangesis-~298kMs28 t

trl'0>>trna>>3%tro'1>>ttriWl.~$100I'l~tttitO')tr'll'i'"~0~or~Sr'lI$00~L~lr~'.0>>nrrltortr>>52~~515S-ll$0$'0>>ttr0>>$000>>trr5>>5-5lloydt.~tlirt>>SMrrlit'10't5'4>>1110~>>tttstre510.$%etrl'I>>Inr5-1.-srrI>>.Srt'o>>tIt'ItsyS-t-trl'I>>tI~wrt'IrrI>>5-~trrI>>tII'.IrSSln'll'tr'SV tt~'ll'rtlt'lolno*\ltrolrFIGURE24-MAPOFSTIFFENER TOSTIFFENER UTINDICATIONS

~rrt,0000,oltortorrrroronnoorrrawo.wrorllonclrrl corttnlnl~mtMrlkrrro, Jrrlt<orntrn IOTIItllrll 7r 4TABLE3-STIPPENER TOSTIPPENER UTINDICATION DATAIndict~aNo.l-l1-2(3)1-3(3)1-41-5(3)1-6(3)1-7(3)1-82-12~22~32-42-52W2-73-13~23~33-43-53-63-73M3-93-103-113-123-133-143-153>>1.63-173>>183-193-203-21~3-22(3)3~23Length(Inches)33/810(int)97/8(int)1,1/241/441/4161/4111'/4751/493/8331/4,33/4(3)103/4.73/43%21/433.334-1/247/8651/221/261/273/4841/2,33/4Thtoogly~

Thick.DepthInches)O-I/81/8-1/40-1/81/8-1/41/4-1/20-1/81/4-1/21/4-1/21/2-3/41/4-1/21/23/41/4-1/21/2-3/4O-l'/801/4-1/2'0-1/8 0-1/80-1/8.1/8-1/40-1/81/4-1/201/2-3/4001/4-1/201/8-1/41/8-1/40-1/8Notes1.Indication numberscorrespond tonumbersinPigure24.2.Thelengthandthrough>>thickness depthweredetermined bytheprogressive magneticparticleinvestigation oftheUTindications.

.3-Indi.cati.on was.reworked withoutdatahavingbeenrecorded.

(int)-Indicates thattheUTindi.cationisintermi.ttent alongtheweldlength.

PlPERESTRAlNT V/lYH6ussET.I>~~Ap,III)IIIIIIIiI'~4e-tGUSSET(BOTHENDS)IIPIPERESTRAINT I<IIII)IIIIs=SPACINGBETWEENRESTRAINT FLANGESH=CLEARDISTANCEBETWEENHORIZONTAL STIFFENERS by.=FLANGE'ENGTH a=LENGTHOFUNIFORMLY DISTRIBUTED LOADw=UNIFORMLY DISTRIBUTED LOADt=THICKNESS OFSTIFFENER FIGURE25 4

ThestressQintheflangesisF+PP-18.6Rs1t$Fromsimplebeamtheory,thereactionatpointAiRA>isR~waA-(2H-a)205k2HTherequirednetweldarea.Atisnet'..A~-~5'inA,2neta~whereCSfactoredallovable shearstressThepercentveldarearequiredforRisAx100X~221netweldbShearStressduetoAccidentTemerature(Figure26)The.freeincreaseinlengthofcoverplateAB(4~)isequated.to thysumofreduction inlengthofcoverplateAE(4~)dugtodeveloped forceplustheverticaldisplacement (4~)ofthestiffener.

4where4~4~wst'AH+4AE~Q(4T)H(H/)~THtAE,24~~.THtdd(1.2)+-3EIAGstThisequationreducesto1Q(4T)EHd312dXt2A+-+-'3IAststwhereT<QEad~A~wAstI0st4T~shearstresscoefficient ofthermalexpansion modulusofelasticity distancefrominnerwalltocoverplateareaofwallsectionareaofstiffener momentofinertiaofstiffener increaseintemperature ofouterwall29 r1' SHEARSTRESSDUETOTENlPERA'JURE DlFFERENTlAL:

HORIZONTAL STIFFENER INNERWALLIVERTICALSTIFFENER dCOVER1'LATEINNERWALLo+COVER'ILATE(b)FIGURE26 J;~t Inthiscase,t9,2ksiThetotalforce,P,developed intheweldisF~-~st~24I5k"M'xAd8EThetotalfox'ce,P,isreducedbytheallowable pressureforce,Fall,togivethenetforce,allow'<<F-Fal~615knetallowwherePallallowable foiceduetostiffenex'eaxing ontheinnerorouterwallTherequirednetweldareaisA~netPnet~~19in2SThepercentweldarearequiredisAxIQOX~7.5Xweld.Therefore, thetotalpercentwe'drequiredis22+7'~29.5percentBasedonanetweldareaof80percent(totalweldarealess20percentforindication area),thefactorofsafetyFSagainstshearfortheweldisnetarea2+7requix'ed weldareaF.Consexvatisms inEngineering Evaluation Numerousconservative assumptions whichweremadeinthestressandfricturemechanics analysesarelistedasfollows:StressAnalsis1Inthethermalstressanalysis, itwasassumedthatwhiletheinnex'ndouterwallsarebeingheatedduringanaccidentcondition topx'oducethemaximumtemperature differential, thestiffeners remainatthesame(opexating) temperatuxe; hence,higherthermalstressesresult.30

20Intheheattransfex'nalysis toproducetheaccidenttemperature differential acrosstheshieldwall,theverticalstiffeners wereignored;hence,agreatertemp-eraturedifferential

occurred, indthusthestresseswereoverestimatedi 30ZntheLOCAannuluspressurization

analysis, nocreditwastakenforeffectsofheatsinks;onlyflowdiverterdoorventingoutoftheshieldwallpenetrations wasassumed;theblowdowncalculation includedtheeffectsofinventory andsubcooling; andthebreakopeningtimewas"assumedinstantaneous.

Hence,ahigherannuluspressureresultedInthep1peruptureanalyses, ifadynamicanalysiswasnotperformed, adynamicloadfactorof2indastitic1mpactfactoxof1.3wereused;minimummaterialyieldpropertyfactox'sof13forboltsandstrapsandI1foraluminumhoneycombs wereused.Hence,higherstressesdueCopiperuptureloadsresultedZnChestressanalysis, thepeakvaluesofvariousloads,suchasaccidenttempex'ature, accidentpressure, andpiperupture,werecombinedinaccordance withPSARloadcombination equations resulting inconservatively h1ghstresseslntheeventofanaccident, however,thepeakloadvaluesoccuratdifferent timesPractureMechanics Anals1sTheultrasonic responses fromallweldd1scontinuities werefarlessthentheresponses framknownreflectors incalibration blocks.20Zftherewasanydoubtaboutthelocationandorientation of.awelddiscontinuity, CheworstLocationandorientation wereassumed.30Itwasassumedthatallthedefectsweresharp,whichwas'usuallynotthecaseiXncaseswhereChrough~hickness propex'ties wereapplicable, thelowestknownvalueofthefracturetoughness inthrough-thickness direction wasused.Dynamicratherthanstaticfractuxetoughness wasusedintheanalysisalthoughactualstrainratesareexpectedtobeatleastanorderofmagnitude lowerChanthosecorres-pondingtodynamicfracturetoughness.

tr 6Itwasassumedthatthe.postulated failureconditions 0developat100Falthoughthetemperature forapostulated accident's expectedtobegreater,than135F.Thefracturetoughness corresponding tothe100Ftemperature 0wasusedinthecalculations 70Ztwasassumedthattheappliedtensilestresswasperpendicular totheplaneoftheassumeddefect.PMotethatassumptions I,2,3,and7resultinhighvaluesofstressintens1ty factorsandassumptions 4,5,and6giveaveryconservative estimateoffracturetoughness.

Iaadd1t1ontotheaboveconservatisms anothermarginofsafetyis1ntroduced bylimitingthesizeofallowable defect'to afractionofthecriticaldefectsizeforeveryweldwhichhas'beenevaluated.

32 t(

~~~~VoQUALITYASSURANCE/CORRECTIVE ACTIONAIntroduction Throughtheinspection/rework programdescribed withinthisreportthebiological shieldwall(BSV)willcomplywithallPSARcommitments andcriteria.

However,inordertoreducethepossibility ofrecurrence ofthetypeofproblemsencoun-texedwiththeBSW,aplanofactioahasbeendeveloped Thepurposeofthissectionisto.summarize thenatureaad.possiblecausesoftheweldproblems, todiscussthecondition ofeheBSVwhileitwasiathesellex'sshop,toexplainwhytheproblemswerenoediscovered intheshop,toreviewthecorrective actionwhichwasimplemented, andtopresentaplanofactionaadcertainstepsalreadyimplemented toreducethepossibility ofrecurrence B~SummaryofVeldProblemsThetwotypesofweldswithevidenceofproblemswerethesinglebevelbackingbarwelds,whichincludedhorizontal seiffeaex'o innerwall(HSIV)andcoverplateeobaseplate(CPBP)welds,aadthecover.platetostiffener weldsHorizontal Stiffener toInnerWall(HSIV)andCoverPlatetoBasePlate(CPBP)WeldsTheseweldiexhibited crackiagwhichpropagated fromtheweldroot.Asdiscussed inSectioaIIIC,itwascoacluded thatthemostprobablecauseofcrackinginitiation washydrogenThehydrogenpossiblyoriginated frommoisturewhichwasnotdrivenoffbypreheatIewasalsoconcluded thatthesourceofthehydrogenwasnotfromtheweldfillermaterialoxthehandlingoftheweldfillermaterial.

Thecracksappearedtohavepropagated astheresultof)ointrestraint Inaddition, itisnowbelievedthatduriaginves>>tigationoftheweldiadicatioas, grindingcausedehecrackstopropagate, resulting inmisinterpretation whichover-estimated theoriginalcracksize.Otherweldindications encountered iatheseweldsincludedgeneralworhnanshi~elated indications, suchasslaginclusions andlackoffusioa.Thepresenceofthistypeofindication wasshownthx'oughengiaeering evaluation tobeacceptable

withoutcompromising theweld)ointintegrity.

33 7I CoverPlatetoStiffener Feldh('ac~g~aTheproblemsencountered inthecoverplatetostiffener weedswereinboththeweldmetalandbasemetal:Thecxackingandworkmanshi~elated indications whichwerepx'esent, wex'ecausedbythesamereasonsasthosefoundintheHSIWandCPBPwelds.Howevex~asdiscussed inSectionIIIC,thestiffeners hadplatelaminations whichwereacceptable toaplateUTacceptance criteria(hSTBA578,LevelI)butre)actable taaweldQTacceptance criteria~

Suchlaminations axeinherent.'in platematerialandare,theresultofplatemanufacturing processes Somestiffeners alsoexhibited lamellartearingwhichwastheresultofacombination oftheweldpointcon-figuration, weldshrinkage stressesintheplatethrough-th1ckness direction, andtheplatelaminations 4CWhytheProblemsWereNotDiscovered intheShopAlthoughdefectssuchasslagorlackoffusionoccurredintheshop,itispossiblethatsomeindications didnotmctstintheshopbutoccurx'ed atsomelatertimeduetohydrogeninducement, erection/handling

stresses, orfieldwelding.C1vesinspection systemdididentify1ndications whichwererepaix'ed inthecourseofmanufactuxe; however,itispossiblethat.someindications weremissedItshouldbenotedthatsomeindications mayhavebeennoticedbyCivesbutpassedinthebeliefthattheywereacceptable undertheCode.Investi-gationofthismatteriscontinuing, andanysubstantive newinformation w111besubmitted totheCommission DCorrective Actionfox'heBSWInordertoresolvetheshieldw'allproblems, anextensive programofultxasonic ezamination.

andengineering evaluation, described inSectionIV,wasconducted TheBSWwasanalyzedorreworkedtoassuxethatallsafetyrequirements aremet.EActionstoReducethePossibility ofRecurrence ProamEnhancements Previousl ImlementedTherehavebeencertainlou~angeS&WStandardQAprogrammodif1cations thatoccurredwithintheprogramtomeetcontinuing newindustx'y demands,reguLations, codesandstandards, andmodifications suggested byS&Wprogramaudits,aswellasASME,licensees, andregulatory audits kr Thefollowing describes certainsignificant Lou~angemodi-fications putintoeffectbyS&WsinceAugust,1978.S&WPQAInspector

training, qualification, andcertification inNDEdiscipLines wereintensified in1978Theresultant programmodifications introduce moreextensive trainingtocoverallNDTdisciplines withinthedivision(andPQAdistrictoffices)andmostspecifically certification inaccordance withASNT-TC-1A guidelines WhilePQAInspectors arenotnormalLyperforming hand~n.inspection, thisenchanced trainingshouldassistindetecting supplierNDTproblems.Xmplemented

-August1978S&WPQAhasnowimplemented afozmalPQAinspection planning.

activityThisfunctionprovidesforPQAengineering revie~ofprospectspecifications, codesandstandards, andpastvendorperformance withanoutputconsisting ofaninspection planiThisplanallowstheinspector toconcentrate hiseffortsonsurveillance activities anddocumentation Implemented

-ApriL1979(QAD-T.14, RevisionA)30S&Winspection reporting system,whichallowscentralized dataanalysisandinputintotheQualityXnfozmatLon Center,hasbeenmodifiedtoprovideavarietyofinspection reportfozms~Thesenewlydeveloped formsallowforsimplified computerentzyandgeneraleaseofcompletion sincetheyaregearedtospecificinspection actionssuchashardwareinspection, processsurveiLLance, andsystemsurveillance.

S&W'sPQAeffortcontinues toemphasize seLLersurveillance andconformance evaluation towardeffective implementation bythevendorofhiscontractual qualitycontrolzesponsibilities Thesenewreportformatsaredesignedtofacilitate thisapproachImplemented

-February1980(QS-14.2, Revision0)4TheS&WQAprocedural systemhasbeenrestructured toprovideflexibility inaccommodating uniqueprospectrequirements withoutsacrificing theinherentadvantages ofastandardprogram.Implemented

-June1980(Prospect Pzoceduze No64)

Additional ProramEnchancements NotYetFullImlementedOvex'all, theNineMilePoint-Unit2(NMP2)QualityAssurance (QA)programhasbeeneffective inidentifying deficiencies.

Nevertheless, thefollowing enhancements arebeingaddedtothispxogramsoas,toreducethepossibility ofrecurrence ofmaterialandfabrication-relateddeficiencies:

Xmprovedcommunication andsystematic evaluation ofcurrentQAproblemeventsAdditional engineering invoLvement AdditionaL direction toPQAinspection andprojectengineering personnel Moreextensive processsurveillance inSeller'sfacilityProcedure andProgramimprovements Theseadditions aredetailedbelow:lImproved,Communcations andSystematic Evaluation ofCurrentQAProbLemEventsTheS&WPxojectQAManagerwillestablish aprogramtoreview,on,amonth~o~onth basis,allCategoryXandotherselectedNonconformance andDispositions (N&D),Engineering andDesignCoordination Reports(E&DCR),ProblemReports,Inspection Reports(IR),andauditfindings(HRC,ClientandS&W)foridentifi-cationofpotential genericproblemareasasweLLas"shortandlongtermtxendsthataffectquality.Thisinformation willbeprocessed intoamonthlysummaryreportthatcontainsatreadanalysisandalistingofN&DsandE&DCRstobereviewedatameetingchairedbytheS&WProjectQAManager,Thismeetingwilltakeplacemonthlyunlessthereaxenosubstantial problemax'eastoreviewMandatory attendance forthePotential ProblemReviewMeeting(PPRH)willincludetheS&WProjectEngineer, PQAManager,Superintendent FQC,ortheirdesignees.

HMPCQAandPx'ojectpersonnel willalsobeinatten-danceThiscommittee wiLreviewthelistedN&DsandE&DCRsforgenerictypeproblemsandprovideimmediate evaluation forcorrective actionintheformof,butnotlimitedto,specification changeand/oradditional inspection requirements forthefieL'dsiteorthespecificSeller'sshop.Theresultsofthesemeetingswillbedocumented InitialImplementation

-August1980FuLLImplementation

-October1980

/

2.Additional Engineering Involvement TheS&WPQADivisionhasrevieweditsoperations todetermine wheresystemimprovements mightavoidsellerproblemsAsaresultofthisreview,thefolLowing willbeappliedtoallnewmajorCategoryIordersandselectedordersalreadyawardedSellerSelection ProcessExpandthepresentS&W,QAsurveyprogramtoincludeverification oftheseller'smanufacturing capability.

Thesurveyteamwillconsistofaqualified qualityassurance auditor,theresponsible Engineerontheproject,andtheengineering specialist asrequiredTheteamwillverifythepresenceofadequateengi-neering,design,manufacturing, andqualityassurance capabilities Post-Award Activities bIncludeS&WEngineers onprojectand/orS&Wequipment specialists asmembersofthePQAPost-Award Conference.

IncludeS&WEngineers onprojectandS&Wengi-neeringassurance personneL asrequiredinauditsforrecertification.

InitialImplementation

-August1980FullImplementation

-December19803Additional Direction toPQAInspection andProject"Engineering Personnel S&WPQAandprojectengineering personnel willreviewexistingandfutureCategoryIspecifications forproblemsinherentinheavysectionweldments, including theQA/QCrequirements relatedto.WeldingjointdesignWeldingcontroLandinspection Fabricator NDEprocedure reviewandapprovalFabricator conference tomanufacturing plansVendorpersonnel qualifications Exerciseofstopworkprocedures underappropriate circumstances InitialImplementation

-September 1980FuILImplementation

-November198037

4MoreExtensive ProcessSurveillance inSeller'PacilityAreviewofmd.stingaadfutureCategoryIspecifi-cationswillbeperformed andadetermiaation madeofthoseordexsthatwillrequire,increased S&WPQAsurveillance.

Inorder.toimplement thisincreased surveillance, S&Wwillprepareaplantailoredtoeachpex'tineat orderaadsubmitittoHMPCforconcurrence InitialImplemeatatioa

-August1980InitialSpecificatioa ReviewaadRevisedIaspectioa Planswillbeimplemented byNovember19805iOthexPx'ocedure andProgramImprovemeats aeForfutureandselectedexistingCategoryIorderswhereultrasonic testing(UT)isutilized, thespecification willestablish holdpointsatwhich,priortoperformance ofanyUT,theS&WPQAinspectox',

aidedbyNondestructive Testing,(HDT)Divisioaengineers, willevaluatetheeffectiveness oftheseller'sUTtechniques, operatox's, andgeneralimplementation ofthetestpax'ameters AllUTpx'ocedure qualifications (techniques) willbeappxovedbytheS&WNondes-tructiveTesting(NDT)Divisioaeagiaeers atthepreestablished holdpointasrequiredbyspecificatioa foxeachdiffexent configux'atioa.

bForfutureandselectedadstiagCategoryIspecificatioas S&WQAwille'stablish notification pointsforapplication ofallNDEmethodsotherthanUTTheNDTDivisionEngineerwillverifyNMapplicatioas onarandombasisaftercoordinatiag

'heactivitywiththeS&WPQAiaspector.

Procedure qualificatioas (techniques) willbereviewedonaselectedbasisdependent onobservations atthepreestablished notification point.cRequireS&WPQAattendance atanypre~wardmeetingbetweenS&WEngineers oaprospectaadseller.d.RequiresellerstonotifyS&WPro)cotManagement ofaaychangesintheirqualityassux'ance managements InitialImplementation

-August1980InitialReviewofExistingSpecifications willbeImplemented byNovember198038 t

Increased Surveillance andAuditAdditionally',

theHMPC~lit@Assurance Department wU.1increaseitssurveillance andauditprogramtoverify'mplementation oftheforegoing programenhancements.

InitialImplementation

-September 198039 r

VI~ANALYSISOFSAFETYIMPLICATIONS Basedonanextensive eagineering evaluation; ithasbeendeter-minedthattheBSWwelddefectscouldnothaveadversely affectedthesafetyofoperations oftheNineMilePoiat2planthadthewelddefectsremainedundiscovered.

Inthecontextofthesafetyevaluatioa and10CFR5055(e),theterm"und1scovex'ed" raChex'han "uncorrected" isusedsincetheevaluatioa hasshowathatsomewelddefectsneedaotbecorrected~

Thebasisforthisconclusion iediscussed indeta11belowfor'achweldcoafiguxatioa Asd1scussed'hroughout thisreport,theprimaryweldproblemsinvolvedthefollowing backingbar-welds:(l)horizontal stiffener toinnerwaLLandcoverplateCobaseplaCeand(2)coverplatetostiffener.

Althoughminorweldindications werediscovered inotherveld)ointconfigurations, theiacidence ofoccurrence oftheindicatioas iswithinCherangeexpectedforrepeatability ofareexamiaation byUTTheseinsigaificant welddefectshaveanegligible effectonthestructural integrity oftheBSWandwouldnoChavebeenasafetyhazardifthedefectshadgoneuadiscovered.

Horizontal Stiffener toInnerWallandCoverPlatetoBasePlateWeldsAsdiscussed inSectionIV.Cofthisreport,thehorizontal stiffener toinaerwallweldswereevaluated basedonQTexamination results.Fx'omTable2;allweldindicatioas areacceptable basedoastressandfracturemechanics analysesTherefore, iftheweldindications hadgoneundiscovered, a.safetyhazardwouldnothaveexistedThecoverplatetobasepLateweld,whichwasCheinitialproblemdiscovered, wasrepairedbasedoa100percentUT.Evaluation hasshownthat,evenifthedefectshadbeenundiscovered, agrossstructural failurecouldnothaveresultedand,therefore, asafetyhazardvouldnothaveexistedCoverPlatetoStiffener WeldAsd1scussed inSectionIV.Dofthisreport,AWSDl,lwelddefectsweregroupedaccording tolocationofthedefect1ntheweldjoint.(Refer'to thediscussion ofSect1onIV.D.2andFigure18)~Thoseveldindications locatedinZoneloftheweldpointwererepa1red, aad,those indications locatedinZoaee2and3wereevaluated.

FromTable3itmaybeseenthatallbut17indications inZones2and3ofthecoverplateareacceptable basedonengineering evaluatioa.

By.arealistic consideration ofthemarginswhichexistintheanalysis, ithasbeendetermined thatthese17indications couldnothavepropagated andwouldnothavebeenasafetyhazard.Addressing indications foundandrepa1rediaZonelofCheweld,ithasbeendetermined thattheseindicat1ons wouldnothavebeenasafetyhazardhadthe1ndicatione remainedundiscovered Evaluation oftheworstpossibledefectwhichcouldhaveoccurred1nZonel,takingaccountforrealistic marginswhichexistintheanalytical procedures, showsChatChegrossstructural integrity oftheBSWwouldbemaintained andasafetyhazardwould,therefore, aotexist.40 P

VII+CONCLUSIONS Theexhaustive investigation andcorrective actionoutlinedinthisreportdemonstrates thattheBSWwillprovideradiation shielding andmaintainstructural integrity forallconditions described inthePSARAllshopweldpointswereevaluated inaccordance withbothAWSDl1andtheoriginalPSARcommitments andeitherwereshowntobeacceptable orwererepaired.

Ithasbeendetermined thattheBSWwelddefectscouldnothaveadversely affectedthe'safetyofoperations oftheNineMilePoint2planthadthewelddefects,remained.

uncorrected TheQAprogram'as beenmodifiedtoreducethepossibility ofrecurrence offutureweldmelated problemssuchasthosediscovered intheBSWi41 PP APPENDIXAREVISIONS TO-INTERIM REPORTITEMNO.LOCATIONININTERIMREPORTINTERIMREPORTVERSIONCORRECTED VERSIONp.1,sect.I.C.>2ndparagraph Stiffener tostiffener evaluation usinginnerwallUTdataStiffener tostiffener evaluation usingaccessible stiffener

.tostiffener weldUTaswellasinnerwallUTdataFigure2"innerwallplate"and"outerwallplate""innerwall"and"coverplate"Figure22inchgroutthickness betweensoleplatyandtopofpedestal3inchgroutthickness Figure3"innerwallplate"and"outerwallcoverplate""innerwall"and"coverplate"Pigure5washerplatetocoverplateandwasherplatetobaseplatewelddetailsareshopweldswasherplateswereremovedandreattached usingwelddetailsshowninfinalreportPigure5Figure6title"Vertical Stiffener toInsideandOutsideMallPlates""Vertical Stiffener toInnerMallandCoverPlates"Pigure6"innerwallplate"and"looseoutsidewallcoverplate""innerwall"and"coverplate"Pigure8titleuNorizontal Stiffener toInside'ndOutsideWallPlates"uHorizontal Stiffener toInnerWallandCoverPlates"10Figure8p.5,sect.IV.A,paragraph 2"insidewallplate"and"looseoutsidewallcoverplate"SameasItemNo.1"innerwall"and"coverplate"SameasItemNo.1

LOCATIONININTERIMREPORTINTERIMREPORTVERSIONCORRECTED VERSIONp.6,sect.IV.A,1stparagraph p.6,sect.IV.h,2ndparagraph p.6,sect.IY.h,3rdparagraph 1.defectsizeslargerthan1/Sinch.willbemapped2.the1/8inchcriteriadeviatesfromAWSDl.lSameasI~emHo.1SameasItemHo,11,Ho.1l.alldefects,regardless ofsize,whichunderwent engineering evaluation were.mapped2.defectsareallowedinweldsiftheprovisions ofhWSDl.l,paragraph 3.7.6aremetSameasItemHo.1SameasItemHo.11,Ho.1

APPENDIXBBmmXS~ZpurZZCDam()TotalTotalLengthDefectEeeeieed~heeeh1/4-3/8'3/8-1/21/2-3/43/4-13(2)3(3)2042832172950.3475.1460.4273.1391.631.535320000000006028104101.I/8-1/4isread"1/8inch~defect51/4inch"2.Thisdatadoesnotincludethedatashowaonthenextline.3.Ã2datainthe30ccnupartments madeaccessible bycoverplateremoval.4.A11unitsareininches.44 Jrt

~'Histo'of"EventsThediscovery ofvelddefectsinthebiological shieldwallandsub-sequentactiontakencanbedescribed inthreeseparatephases:dis-coveryofapotential problem,engineering investigation, andsampleplanapproach.

A.PhaseI-DiscoveofaPotential ProbLem(Hay1979)BasedonUTindications MthecoverplatetobaseplateveldsandvisuaLindications discovered inthethirdringhorizontal stiffener toinnerwaLLvelds,thequalityofbackingbarveldsfortheentirebiological shieldwallwasinvestigated.

B.PhaseII-EnineerinInvestation(June1979toAugust1979)Thepurposeoftheinvestigation vastodetexmine ifaveldqualitypxoblemexistedforthebiological shieldwaLLbackingbarwelds.Thesequenceofeventsforthesecondphaseisasfollows:Junel.Aspecimenwasremovedfromthebaseplatetoperforma.metallurgical examination ofcoverplatetobaseplateindications.

June-JulyZ.UZandHTinspections ofthehorizontaL stiff-enertoinnerwallweldsforallthreeringswereperformed.

Theinspections wereperformed onrandomaccessible velds.Basedonunacceptable defectsdiscovered byEVinthethirdringhorizontal stiffener toinnerwalEwelds,additional HTinspction vasperformed on.allaccessible thirdringhorizontal stiffener toinnerwallweldpoints(approximately 2,000inchesoutofatotalof6,000inches).July3~Theccnrerplatetobaseplateveldpointsveredispositioned torequirel00percentUTinspection andrepairedasrequired.

TheUTwasperformed inaccordance withAWSDl.lwithaI/8inchexclusion allowedattherootofthe~veld(basedonengineering evaluation).

ThisUTvascaapleted inSeptember 1979.Byevaluation ofdataobtainedtothattMe,itvasconcluded thatthehorizontal stiffener toinnermallveldsonthefirstandsecondringswereacceptable butthethirdringhorizontal stiffener toinnerwallweldsvererespectable andvouldrequirecaapletereinspection andrework,asrequired.

PClk

~<.C.PhaseIII-SamlePlanAroach(August1979toDecember1979)Sincetheinitialengineering investigation wasprimarily concerned withtherootofthewelds,inspections wereperformed usingEZmethods.Subsequent investigations todetenaine thequalityoftheentirevolumeoftheweldsweremadeusingUTmethodsinaccordance withthesampleplan.Thepurposeofusingasampleplanapproachwas.toverifythequalityofbackingbarweldsbyusingamorerigorous, systematic approach.

Anationally recognized samplingapproachusingconfidence levelsconsistent withlevelspreviously employedwaschosen.4ThesequenceAugust-SeptemberSeptember-Octoberofeventsforthethirdphasewasasfollows:l.Aneffortwasmadetoestablish confidence levelsbasedonthedataavailable fzomtheengineering investigation.

Anadditional 91inchesofHFinspection onthefirstringhorixontal stiffener toinnerwallweldswasperformed tofulfiLLtherootsamplesizerequirements.

BasedontheEVsampleplan,thehorizontal stiffener toinnerwallweldsforthefirst.andsecondringswereacceptable andforthethirdring,regectable.

h2Thirtycoverplateswereremovedfromthethirdzingtoprovideaccessibility forreinspection andreworkofthehorizontal stiffener toinnerwallwelds.October-November3.-Thesampleplanapproachwasextendedtoallweldconfigurations, including weldfointswithoutbackingbars..Theweldpointswerecaapiledinto18weldgroupsbasedonthering(i.e.,first,second,ozthird),thepointconfiguration (i.e.,singlebevelweldwithbackingbazordoublebevelweldwithoutbackingbar),andthethickness oftheplatesbeingconnected (i.e.11/2-inchto11/2-inchplatesor11/2-inchto2-inchplates).November4.Twospecimens wezeremovedfranthethirdzinghorizontal stiffener toinnerwallweldstoperformametallurgical examination.

~(Twoadditional specimens ofthehorizontal stiffener toinnerwall'welds wereremovedformetallurgical examination, oneinEarch1980andoneinApril1980.)46 4r November-December5.UTinspection ofthe18veldgroupsvasperformed byacertified LevelEZinspector.

'Tdatapreviously takenduringtheengineering investigation wasalsoincorporated intoasampleplanapproach.

BasedonthesampleplanUTdataofrecord,llofthe18veldgroupsverere)ected, 4vereaccepted, and3verenotapplicable tothesampleplan.Duetothehighnumberofre)ectedweldgroups,itwasdetermined thatallvelds,vhichvereaccessible forinspection, wouldbeinspected byUT.47 PJl