ML13316B905
ML13316B905 | |
Person / Time | |
---|---|
Site: | South Texas ![]() |
Issue date: | 11/13/2013 |
From: | Taplett K South Texas |
To: | Division of License Renewal |
Daily J | |
References | |
Download: ML13316B905 (45) | |
Text
SOUTH TEXAS PROJECT ALUMINUM BRONZE TESTING UPDATENovember 13, 2013111/13/2013 AgendaIntroductionsPurposeAlBrzDealloying BackgroundTesting ProtocolsPreliminary AlBrzTesting ResultsFuture Testing PlansAlBrzDealloying Program DevelopmentSummaryQuestions211/13/2013 South Texas Project (STP) AttendeesMichael Berg Manager, Design EngineeringMichael Murray Manager, Regulatory AffairsRob Engen Manager, Engineering ProjectsArden Aldridge License Renewal Project ManagerKen Taplett Supervisor, LicensingMatthew Hiatt Aluminum Bronze Project EngineerFred PuleoLicensing EngineerRichard Kersey Supervisor, Civil Design EngineeringCong Pham Supervisor, Mechanical Design EngineeringKevin Regis ECW System EngineerAaron Heinrich Aluminum Bronze Program EngineerContractorRuss Cipolla Contractor, Intertek AIM (Aptech)311/13/2013 PurposeDescribe the progress of testing completed by STP on aluminum bronze components in support of License Renewal activitiesDescribe the future testing scope to be completed by STP in 2014Describe development of program procedure to manage and analyze aluminum bronze dealloying411/13/2013 AlBrzDealloying BackgroundMetallurgyAluminum bronze alloy is composed of alpha, beta and gamma-2 phases depending on aluminum content and heat treatmentGamma-2 phase and beta phase preferentially corrode and leach out aluminum. Result is a porous structure that can result in a through-wall leak if gamma-2 or beta network is continuous through materialHeat treatment for castings determines if network is continuous and whether alpha+gamma-2 eutectoid or alpha+betaeutectoid forms5Typical MicrostructureA copper-rich alpha matrixB alpha-gamma2 eutectoidC -isolated, preferentially attacked gamma-2 (dark regions within the eutectoid and along the grain boundaries)11/13/2013 AlBrzDealloying Background (continued)Dealloying PropagationThere is a discontinuity in microstructure at the boundary between dealloyed regions and undealloyedregions in a component. This discontinuity in microstructure can be explained by the dealloying process which slowly proceeds across the component pressure boundary. As the corrosion process created by the wetted surface extends through the thickness, the Al-Brzcontinues to dealloyalong the wetted path. The remaining wall thickness is not affected by the corrosion process until it contacts the corrosive environment. The boundary between the two material states (dealloyed / undealloyed) is not very wide and defines the depth of the dealloying in the specimen.Dealloying is measured / identified by etching surfaces with silver nitrate darker regions denote dealloyed areasDegree of dealloying (% dealloying) is a geometrical measure= Depth of dealloying / component wall thickness OR= Area of dealloying / total component cross-sectional area11/13/20136 AlBrzDealloying Background (continued)Essential Cooling Water (ECW) System is constructed of aluminum bronze alloys SB169 CA614Wrought, single-phase alloy, 6.0 8.0% Al by weightUsed for pipe, fittings and small-bore valvesNot susceptible to dealloyingSB148/271 CA 952/954Cast, alpha+beta+gamma-2 phaseCA952 -8.5%-9.5% Al by weight, CA954 -10.0-11.5% Al by weightUsed for fittings, large-bore valves, pumpsSusceptible to dealloying711/13/2013 AlBrzDealloying Background (continued)History at STPDealloying initially identified at STP during plant construction/start-up through large number of leaking small-bore cast fittingsAll small-bore castings have been replaced with wrought Established methods to evaluate through-wall leaks in large-bore castings for operability/structural integrity Captured in UFSAR Appendix 9A and supporting calculationsLong-term management by Leak-Before-BreakNumber of Through-Wall Leaks per YearLarge-bore castings 5/yrat startup, 1-2/yrcurrently811/13/2013 AlBrzDealloying Background (continued)Susceptible Component Population251 flanges, 1reducer, 1 cap, 1 elbow and 19 teesBulk of through-wall leaks have been at flanges151 valves and 12 pumpsWelds or weld-repairs with susceptible weld filler materialMostly above and below ground piping butt-weldsSmall number of weld repairs on extruded tees and socket weld metal Non-Susceptible Component PopulationAll pipe is wroughtAll below-grade fittings are wroughtAll small-bore (<3components were replaced with wrought in 1988-1990 timeframeMost large-bore castings that leaked were replaced with wroughtSome leaking valves were replaced with cast material 911/13/2013 AlBrzDealloying Background (continued)STP has procured large number of non-susceptible spare fittings and valves to enable rapid replacement of leaking components when identifiedReplacement flanges and other fittings are wrought AlBrzAvoids cracking problems with welding SS fitting to AlBrzpipeReplacement valves are stainless steelDesign changes for valves are in process; some will be replaced with cast AlBrzuntil stores are exhaustedSTP has pursued NDE techniques to characterize dealloying in-situZero-degree and Phased-array UT can detect and characterize dealloying under optimal conditionsContinued development is ongoing but is not expected to be viable in near future1011/13/2013 AlBrzDealloying Background (continued)Analytical models for evaluating structural integrity based on ASME Section XI and GL 90-05ASME Section XI Appendix H 1989Basic model assumes dealloyed flaw region (identified from through-wall leak) has zero strength or fracture toughness and uses conservative values for material propertiesDuring License Renewal, NRC challenged underlying basis for analytical model as limited test data was used to construct and validate analytical modelMechanical properties, dealloying flaw length correlation based on outside diameter flaw length, and other factors used in analytical model based on small sample sizeOnly one dealloyed component was bend/pressure tested to validate model predictions1111/13/2013 Testing ProtocolsAnalysis Confirmatory Test (ACT)Bend TestComparison of actual stress applied to the component compared to the critical bending stress predicted by the modelInspection of sample for chemistry, mechanical properties, microstructure, degree of dealloying, and/or crackingProfile Examination (PE)Sectioning of component to map dealloying progressionCorrelation of observed outside diameter (OD) flaw length with flaw length at mean radius of componentInspection of sample for chemistry, mechanical properties, microstructure, degree of dealloying, and/or cracking1211/13/2013 Testing Protocols (continued)provide reasonable basis that aluminum bronze components could perform intended function during period of extended operationSTP was credited for 1 ACT and 8 PE exams performed in the different sizesminimum level of dealloying degradation1311/13/2013 Testing Protocols (continued)STP identified 18 cast components for testingOnly 3 of the 18 had been identified as leakersRemaining components were selected based on potential of finding dealloying in those locations (i.e. same component in different train had dealloyed previously, stagnant flow conditions, etc.) and accessibilityComponents were removed during ECW System drain-downs to minimize unavailability impact on system2 components removed in 2012 and 4 removed in early 2013 were part of initial test scopePE and ACT have been or will be performed on all components1411/13/2013 Preliminary AlBrzTesting ResultsAll results are considered Preliminary as Final QA verification has not been completed on reportsTesting performed by Intertek (Aptech) and subcontractors under Appendix B programAptechwas heavily involved with AlBrztesting and analysis during plant start-Aptechis highly experienced in material testing and ASME Section XI flaw evaluations1511/13/2013 2013 Testing Completed To DateACT (bend test + hydro) and PE (sectioning and etching) completed on:2 1 1 2
- Through-wall leakers with no crack** Through-wall leaker with crackMechanical/chemical testing was not performed on every sample due to combinations of:Lack of dealloyingDealloyed area too small to fabricate test specimensMicrostructure examination performed on selected components to ensure dealloying continues to propagate as bimetallic area1611/13/2013 Valve Bend Test1711/13/2013 Valve Bend Test ResultsService Loading ConditionCalculatedMarginASME Section XI AppendixH Required Margin(SF)Level B (Upset)54.22.77Level D (Faulted)44.21.3918Predicted bending stress to fail componentActual test bending stressAnalytical Flaw Length11/13/2013 Valve Pressure Test1911/13/2013 Valve Pressure Test ResultsValveInitialPressure(psi)Hold Time(min)FinalPressure(psi)VisibleLeakageEWFV-69361005100No Leaks1515151No LeaksEWFV-69371005100No Leaks1545154No LeaksValveInitialPressure(psi)Hold Time(min)FinalPressure(psi)VisibleLeakageEWFV-69365001140345No Leak5005500No LeakEWFV-69375001200203No Leak500180487No LeakPneumatic Pressure TestHydrostatic Pressure TestNote: Design Pressure = 120 PSI / Operating Pressure = 80 PSIPressure Margin is approximately 4:12011/13/2013 Flange Bend Testplaced in area of max tensile stress2111/13/2013 Flange Bend TestStable crack tearing during failure. Failure initiated from original crack location-like dealloying around crack location2211/13/2013 Flange Bend Test ResultsNotes: since others did not have OD flaws to evaluate-C-1964-5 Fig. 4-1examination of fracture surface at mid-wall23Service Loading ConditionCalculatedMarginASME Section XI Appendix H Required Margin(SF)Level B (Upset)16.12.77Level D (Faulted)14.81.3910-inch NPS11/13/2013 Flange Pressure Testflange2411/13/2013 Flange Pressure Test ResultsNote: Design Pressure = 120 PSIPressure Margin is approximately 2.3:12511/13/2013 ACT Testing SummaryAll tested components were able to support a bending stress greater than the predicted bending stressAll components were able to hold a pressure without failure of at least 2x design pressureAll components had substantial structural margins (structural capacity greatly exceeded ASME required Structural Factors)3 of the 6 tested components had minimal dealloying but supported loads in excess of critical stress for pre-service components2611/13/2013 Profile Exam Results2711/13/2013 Profile Exam Results2811/13/2013 Profile Exam ResultsID#DescriptionMax % DAAvg%DADealloying Character(Plug /Layer)Crack (Y/N)2cto 2dCorrelation Valid?F-26121.0~5%Plug.Narrow, isolateddealloyed regionsNN/AF-16939.2~5%Plug.Narrow, isolateddealloyed regionsNN/AF-05922.0~5%Plug.Narrow, isolateddealloyed regionsNN/AF-064100.0~40%Plugwith more extensive dealloyed areas. Limited axial YYV-037Valve100.0~40%Plugwith more extensive dealloyed areas. On both inlet/outlet flanges and throughout valve bodyNN/AV-041Valve100.0~40%Plugwith more extensive dealloyed areas. On both inlet/outlet flanges and throughout valve bodyNN/A11/13/201329Notes: DA = dealloyingAvg% Dealloying is estimated, actual average not available yetMax % DA is a local maximum at varying circumferential cutsOD crack angle to through-wall dealloying angle (2c2d) correlation is from AptechAES-C-1964-5 Profile Exam SummaryMinimal dealloying on 2 More extensive dealloying present on both valve Existing correlation for OD crack length to TW flaw Dealloying is plug-like with through-wall leakage occurring before average dealloying % reaches ~60% away from the through-wall flaw3011/13/2013 Mechanical Testing Completed To DateTensile TestYield (Sy) and Ultimate (Su) StrengthYield by 0.2% Offset (OS) and 0.5% Extension Under Load (EUL) methodsTypically ductile materials are measured by 0.5% EULNote some older tests did not calculate yield, only ultimate strengthCrack Tip Opening Displacement (CTOD) TestFracture Toughness (KCTODor KIC)SpecimensMix of CA954/952 material, ~24 tensile and ~25 CTOD specimens-bore fittingsFlanges were wrong geometry/thickness to produce acceptable test specimensSpecimens were all sub-size (but standard)Sub-macroscopic properties of larger specimensvalues for Sy, Su, and KCTODfor dealloyed material are likely higher than reported test values3111/13/2013 Mechanical PropertiesPre-service Material PropertiesSpecified Minimum Strengths Models conservatively use strength of CA-952 for analysisCMTR for as-fabricated material typically reports higher yield and ultimate strengthsFracture ToughnessNot specified as part of material specificationrange of 63.5 -95.1 ksiin1/2Conservatively taken as 65 ksiin1/2in analytical models32CA-952CA-954Sy(ksi)2530Su (ksi)657511/13/2013 Test SpecimensCTOD SpecimenTensile Specimen3311/13/2013 Tensile Test Results Yield Strength3411/13/2013 Tensile Test Results Yield Strength3511/13/2013 Tensile Test Results Ultimate Strength3611/13/2013 CTOD Test Results3711/13/2013 Mechanical Testing SummaryMaterial retains strength and ductility in dealloyed stateUltimate strength asymptotically approaches ~30 ksias % dealloying increasesYield strength asymptotically approaches ~28 ksias % dealloying increasesMaterial retains fracture toughness in dealloyed stateFracture toughness falls into the 25-30 ksiin1/2 as % dealloying approaches 100%Material retains ability to resist crack propagationValues for 100% dealloyed ultimate strength (30 ksi) and pre-service fracture toughness (65 ksiin1/2)are consistent with values used in previous analysesSamples with 20-25 years of aging have same properties as original samples given the same level of dealloying3811/13/2013 Chemical Testing/MicrographyResults11/13/201339~4% Al~11% Al Chemical Testing/MicrographyResults11/13/201340~0% Al~9.5% Al Chemical Testing/MicrographySummaryDealloyed regions are low in aluminum and undealloyedregions have aluminum content consistent with CMTR chemistryReflects corrosion of high Al-content gamma-2 and/or beta phases in dealloyed regionsAlpha phase grains are intact in both regions but have deposited Cu in dealloyed regions from corroded eutectoidChemical testing and SEM examination validates use of visual methods (surface etching) for characterizing depth of dealloying as chemistry corresponds to visual indications11/13/201341 Future Testing PlansSTP intends to continue testing of AlBrzcomponents in early 201411 components are currently planned for removal and testing between October 2013 and February 20143 4 1 2 1 Components will be tested (ACT and PE) regardless of presence of leaksor % dealloyingAs opportunity presents, additional tensile and CTOD specimens will be selected for testing to build out material property curves4211/13/2013 Dealloying Program Procedure DevelopmentSTP is in process of developing AlBrzDealloying Management Program procedureAddresses Aging Management concerns by trending results of destructive testing and specifies destructive testing for future componentsProvides consistent, clear guidance on applying previously developed methods for structural integrity evaluations, operability reviews and relief requestsProvides guidance on selecting NDE methods for examining dealloyed componentsProvides repository and reference for numerous reports, calculations and correspondence that support dealloying licensing basisExpected to be complete in early 20144311/13/2013 SummaryAll testing completed to date indicates that analytical models for managing and dealloying are conservativeLeak-before-break remains valid method for managing dealloyingAll components had substantial structural marginsMaterial properties appear to only be affected by dealloying percentage, not component ageSTP proceeding with test plan in 2014 to obtain requisite number of ACT and PE testsSTP is developing aluminum bronze dealloying management program4411/13/2013 Questions?4511/13/2013