*-=W -.- e* I -- :;

NUREG-1 339

Resolution of Generic SafetyIssue 29: Bolting Degradationor Failure in Nuclear PowerPlants

Manuscript C�n:ipe;�e1: March 1990

Manuscript Cwinp;e;sfl: March 1990)D4:c Published:June I99

Richard E. Johnson

Division or Safety Issues ResolutionOffice of Nuclear Regulatory Researchll.S. Nuclear Regulatory CommissionWashington, DC 20555

+4 a It,

hit

I

MIL WINDRA&L81

L --

ABSTRACT

This report descnbes the U.S. Nuclear Regulatory Com-mission's (NRC's) Generic Safety Issue 29, 'Bo!ting Deg-radation or Failure in Nuclear Power Plants," includingthe bases for establishing the issue and its historical high-lights. The report also desribes the activities of theAtomic Industrial Forum (AlF) relevant to this isue,including its cooperation with the Materials PropertiesCouncil (MPC) to organize a task group to help resolvethe issue. The Electric Power Research Institute, up-

ported by the AIF/MPC task group, prepared and issueda two-volume document that provide in part, the ttchni-cal basis for resolVing Generic Safety Issue 29. This re-port presents the NRC's review and evaluation of thetwo-volume document and NRC's conclusion that thisdocument, in conjunction with other information fromboth industry and NRC, proYides the bases for resolvingthis issue.

NUREG-1339

qlkv ^

iii

- t wt-- tb*. . - -. - --- - '7'7--7

CONTENTS

Page

iii

vi

1112

Absact .................................................................................AbbrvAtions .............................................................................

1 Introduction .......................................................................1.1 The Bolting Safety Issue ............................................................1.2 Problem .........................................................................1.3 Plan for Resolution .................................................................

2 Industry Resolution .....................................................................2.1 Planned Program ..................................................................

2.1.1 Gencral Bolting Tasks ........................................................,1.2 Pressure Boundary Bolting Tasks ........... ........ ...........................2.1.3 Tasks Associated with Internals Boltin. .........................................

2.2 Tecl.nical Findings .................................................................2.2.1 Basis for the Rcsolution .......................................................2.2.2 Supporting Data for the Resolution .............................................

3 Conclusions ...........................................................................

4 References ........................................................................

22334

447

11

15

.

NUREG-1339v

ABBREVIATIONS

ACRS Advisory Committee on Reactor SafeguardsAIF Atomic Industrial ForumAISC American Institute of Steel ConstructionANSI American National Standards InstituteASME American Society of Mechanical EngineersASTM American Society for Testing and Materials

BTC Bolting Technology CouncilB&W Babcock and WilcoxBWR boiling-water reactor

CGWT cylindrically guided wave technique

EPRI Electric Power Research Institute

GSI generic safety issue

HSLA hiah-strength, low-alloy (stcel)

IE (Office ot) Inspcction and Enforcement(NRC)

IGSCC intergranular stress-corrosion crackingISI inscrvice inspection

LAQTLOCA

MPC

NDENRCNRR

NSSS

PWR

RCPB

SCCSRPSOER

USIUTJuTrs

WOG

low-alloy quenched and tempered (steel)loss-of-coolant accident

Materials Properties Council

nondestructive examinationNuclear Regulatory Commission(Office of) Nuclear Reactor Regulation(NRC)nuclear steamn supply system

pressurized-water reactor

reactor coolant pressure boundary

stress-corrosion crackingStandard Review PlanSignificant Operating Event Report

unresolved safety issueultrasonic testultimate tensile strength

Westinghouse Owners' Group

NUREG-1339

i

Ii

vii

F -- I- b

1 INTRODUCTION

1.1 The Bolting Safety IssueThe bolting safety issue originally was an integral part ofthe Nuclear Regulatoxy Commission's (NRC's) Unre-solved Safety Issue (USI) A-12, "Fracture Toughness ofSteam Generator and Reactor Coolant Pump Supports.'Recognizing that the types and variety of failure mecha-nisms active in bolting safety problems were distinctly

*diff .ent from those to be addressed in structural steelsupports, the NRC staff separated the bolting safety issuefrom USI A-12 (Ref. 1) and identified it as GenericSafety Issue (GSI) 29, "Bolting Degradation or Failure inNuclear Powe: Plants." The identilittion of bolting ii-tegrity as a separate issue received inipetus from theAdvisory Committee on Reactor Safeguards (ACRS) inOctobzr 1981. The ACRS recommended that the NRCstaff expand its concerns about stress-corrosion cracking(SCC) of high-strength, low-alloy (HSLA) steel bolts toincludc a more comprehensive approach to the degrada-tion and failure of bolting and threaded fasteners. Sepa-rating the bolting issue from USI A-12 created no conlictwith the other USI A-12 goals. No structural materialswere liable to the same kinds of degradation mechanismsas bolting materials, and adequate fracture toughnesscriteria for bolting materials were available within theAmerican Society of Mechanical Engineers Boiler andPressure Vessel Code (ASME Code).

The safety aspects of GSI 29 can be summarized as fol-lows. Critical bolting applications in nuclear power plantsconstitute an integral part of the reactor coolant pressureboundary (RCPB) and include closure studs or bolts onreactor vesscls, reactor coolant pumps, and steam genera-tors. Failure of these bolts or studs could result in the lossof reactor coolant and jeopardize safe operation of theplant. Bolting also is an integral part of the pressureboundary and other safety-related systems, such as com-ponent supports and embedded anchor bolts or studs. InJune 1982, the NRC staff issued Office of Inspection andEnforcement (IE)Bulletin No. 82-02 (Ref. 2).Thebulle-tin required responsive actions by all pressurized-water-reactor (PWR) licensees because threaded fastener fail-ures had shown an increasing frequency of occurrenceand a variety of underlying mechanisms. Motivated by theissuance of NRC requirements regarding fastenLr integ-rity, the Atomic Industrial Forum (AIF) joined with theMaterials Properties Council MPC) to form the JointAIFtMIC Task Group on Bolting, also in June 1982.Thistask group was composed of representatives from AIFmember organizations-utilities, vendors, and architect-engineers-plus representatives from the Electric PoewerResearch Institute (EPRI). The coordinated industry re-sponses to Bulletin 82-02 and, earlier, to the "For Com-ment" version of NUREG-0577 (issued in October 1979)gave added emphasis to the importance of GSI 29.

When the NRC prioritized generic issues in November1982, GSI 29 was assigned a high priority (Ref. 3).

1.2 ProblemThe NRC noted (Ref. 4) that from 1964 to the early 1980sthe incidence of reported failures in high-strength boltingin Class 1 component supports and other safety-relatedequipment increased. Cornmon characterstics among thereported incidents included materials that were subjectedto high sustained tensile stresses out-of-specificationpretorquing, an aqueous environment caused by high hu-midity, primary and borated water leakage, and materialsthat were overly hard and out of specification. The mostfrequently observed failurc mode for the structural bolt-ing was SCC. Both low-alloy quenched and tempered(LAQT) steels and maraging steels were degraded bySCC. A small number of overstress failures was traced toimproper heat treatment or low-strength material. Pres-sure-retaining bolts failed from corrosion wastage*. In-cluded in the RCPB components were steam generatormanway closures, reactor coolant pumps, pressurizermanway closures, reactor vessel elosures, chemical andvolume control system isolation valves, check valves inthe emergency core cooling system that form part of theRCPB, and other check valves. Some reactor vessel inter-nal components, mainly the lower thermal shield boltsand upper core barrel bolts, were degraded, requiringextensive and expensive replacement of bolts in someplants.

Millions of threaded fasteners, including nuts, bolts,studs, and capscrews, are used in a nuclear power plant.The most irnpo-tant application of these fasteners is theiruse as an integral part of the RCPB, such as in pressure-retaining closures on reactor vessels, pressurizers, reactorcoolant pumps, and sicAm generators. The NRC receivedreports of a number of degraded threaded-fastener inci-dents that involved the RCPB and major component sup-ports. Although none of the reported incidents resultedin an accident,they do reflect an undesirable lcvel ofdegradation in operating nuclear power plants by impair-ing the integrity of the RCPB or component supports.

Most of the reported bolting degradations were discov-ered either during refueling outages or scheduled inser-vice inspections (ISIs) or maintenance and repair out-ages.Thus far, bolting degradation has not caused an acci-dent, and has not produced any immediate advcrse effecton public health and safety. However, NRC is somewhatconcerned that inadequate ISI of fasteners, either be-cause of ineffective nondestructive examination (NDE)methods or failure to include fasteners in the ISI pro-gram, could contnbute to the potential for reduction inthe integrity of the RCPB and structural supports. Expe-rience has shown that either wastage from borated-watercorrosion or SCC can go undetected. Furthermore, such*See Ref. 2 for a description of the waseige problem.

NUREG-1339

IiII

iI

I

degradation in bolting important to the RCPB integritycould lead to a loss-f-coolant accident (LOCA) becauseof bolting failures.

1.3 Plan for ResolutionThe NRC considered possible solutions to GSI 29 as partof the process for prioritizing gencric issues. The NRCnoted in Ref. 3 that bolting has a wide range o! applica-tion in nuclear powcr plants and that no single solution tothe problem is known. lherefore, to minimize potentialbolting problems in new power plants, improvements inone or all of the five following areas could be recom-mended: design, naterials, fabrication, installation, andinservice inspection. The NRC sugVcx:rd that the cti-ciency and adequacy of the ISl program bc emphasized.

The NRC action plan for GSI 29, as it finaliy evolved,included four tasks in its scope of work:

(1) Develop the technical bases for bolting applicationrequirements by the NRC staff through a technicalassistance contractor at the Brookhavcn NationalLaboratory.

(2) Review licenseest responses to IE Bulletin No.82-02, "Degradation of Threaded Fasteners in theReactor Coolant Pressure Boundary cf PWRPlants."

(3) Draft staff recommendations for proposcd criteriaand guidelines to be incorporated into the NRCStandard Review Plan (SRP).

(4) Develop a proposed plan for implementing boltingapp!ication requiremerts.

Meanwhile, the Joint AIFIMPC Task Group on Boltingbecame more active. The original charter of the taskgroup was oriented toward a coordinated industry re-sponse toIE BulletinNo.82-02and to the boltingaspectsof the related document, NUREG-0577, issued in 1979as the For Comment" version. The industry responsewas to emphasize (1) a bolting survey, (2) stress-corro-sion-cracking susceptibility criteria, and (3) corrective ac-tions to deal with the problem. However, through meet-ings of the task group, by itself and with NRC staff, a morecomprehensive industry program evolved. The 19-taskGeneric Bolting Program was presented for review andcomment to the parent AIF Subcommittee on MaterialRequirements in February 1983; the program was offi-cially transmitted by the AIF to the NRC in July 1983.The basic objective, which was attractive to the NRC staff,was that the nuclear industry itself would provide thetechnical basis for resolution of GSI 29; therefore, NlRC'sactivities regatding this generic issue were delayed untilthe NRC received industry's findings.

NUREG-1339

2 INDUSTRY RESOLUTION

2.1 Planned Program

The technical program eventually formulated by the AIF/MPC Task Group on Bolting was a comprehensive, 19-task action plan aimed at resolving GSI 29. The ElcetricPower Research Institute (EPRI) organized a matrix-managed Generic Bolted Joint Intcgrity Program to carryout the rcscarch. The results, to the xtent that theyprovide relevant technical findings, are summarizcd inSection 2.2 of this report.

Early in its existence, the task group revicwcd informa-tion on fastenerservice failuresand categorized them intofour basic groups:

Group I -

Group H -

Degradation or Failure of PrcssureBoundary Bolting Caused by GeneralBorated Water Corrosion (Wastage orErosion/Corrosion)

Degradation or Failure of PressureBoundary Bolting Caused by Stress-Corrr -ion Cracking

Group Ill - Dcgradation. or Failure of InternalsBolting Caused by Fatigue and Strcss-Corrosion Cracking

Group IV - Degradation or Failure of Supports andEmbedment Bolting Caused by Stress-Corrosion Cracking, with Two Sub-Classcs Separated at the Minimum Speci-fied Yield Strength Level of 150,000 psi

Three of these groups (Group III was excluded) formcdthe basis for what was called the Generic Bolting Pro-gram. EPRI assumed the lead for technical integrationand research support in this generic program. Work re-lated to Group III (internals bolting) failures was assignedto individual vendor owners' groups. Resolution of thefastener integrity issue involved many disciplines. Inputwas needed from the areas of metallurgy, fracture me-chanics, nondestructive examination, design, specifica-tions and standards, quality assurance, manufacturing orquality control, corrosion engineering, joint design, andtensioning control. The AIFIMPC task group consideredall of these disciplines. Assessment of priorities related tofastenerapplications led to the focus of action on primarypressure boundary components. The action plan designedby the task group encompassed the following 19 tasks(several of which were divided into sub-tasks). These 19tasks were grouped under 3 headings: Gencral BoltingTasks, Pressure Boundary BoltingTasks, and Tasks Asso-ciated with Internals Bolting.

2

� I ..,- .. _.

2.1.1 General Bolting Tasks (Tasks 1Through 9)

Task 1 - Assment of Prloritks and SufetySignifcance

Task 11. - To monitor bolting priority ranking and toassess the failure and success history for each of the fourdegradation or failure groups listed previously. (See Sec-tion 2.1 of this report.)

Task 1.2.- To conduct a pilot Fcoping study, under EPRIdirection, on the use of decibion analysis for bolting aimedat developing a mcthodology for determining thc tchni-cal parameters that influence the likelihood of bolt fail-ure.

Task 2 - Uterature Survey of Fastener Corrosion

To perform a literature survcy of carbon and alloy steelfastener corrosion in PWRs.

Task 3 - Stress-Corrosion Cracking

To study and evaluate the effects of water cnvironmc.-atson quenched and tempercd low-alloy stecl bolting mate-rials.

Task 3.1 - Fracture Mechanics Analysis. To developstress intensity factors for realistic flaw shapes and load-ing conditions in bolts.

Task 32 - Data Review. To obain detailcd descriptionsof failures involving stress-corrosion of HSLA materialfrom prcviously unpublished accounts.

Task 4 - Include Hardness Test k,ata into theBolting Database

To include data obtained by utilities from hardness sur-veys of installed and spare bolting in a bolting databaseand to assess impact of these data on the issue.

Task 5 - Bolting Database

To maintain a database containing hardness data andother properties of bolting mater;als and to update thedatabace as neessaxy to support industry efforts.

Task 6 - Devalopment of Bolting Speciications andStndards for Nuclear Power PlantApplications

Task 6.1 - To develop a general specification for boltingrequirements that eventually could be adopted by nuclearutilities.

Task 6.2 - To initiate action in American Society forTesting and Materials (ASTM) Committee F-16 (respon-

sible for structural bolting) to revise sanpling require-ments ir new and existing specifications to be more con-sistent with end-product expectaticns.

Task 63 - To prepare a draft ASTM standard entitJed,"Standard Test Method for Leeb Hardness Testing ofMetallic Materials," based on Equotfp hardness test ex-perience.

Task 7 - ASME Code Bolting Requirmcnts

Task 7.1 - To prepare a aitique regarding ASMY' Codcbolting requirements, particularly as rlated to preten-sioning of both pressure boundary and structural boltingjoints.

Task 7 - To review AS7 Code Sectiont m boltingrequirements to determine the need for revising or im-proving.

Task 8 - Develop Field NDE Techniques to DetectWastage and Stress-Corrosion Cracking

To focus pilot studies that were under way on the devel-opment of field techniques, utilizing advanced ultrasonictechniques to detect wastage or stress-orrosion crackingin pressure boundary and support fasteners.

Task 9 - Information Exchange

Task 9.1 - To hold bolting workshops to exchange infor-mation on industry efforts rEarding bolting integrity.

Task 9.2 - To produce and distnbute to utilities video-tapes on the behavior and maintenance of flanged pres-sure boundary connections as aids to improving boltingdesign, installation, and maintenance.

Task 93 - To produe a videotape on dcsign, behavior,and tensioning practices as applicable to structural jointsif warranted from the Task 16 results. %

2.1.2 Pressure Boundary Bolting Tasks(Tasks 10 Through 17)

Task 10 - Screening Strategy and Corrective Actionfor Pressure Boundary Bolting

To develop a strategy for identifying bolts in pressureboundary applications that may be susceptible to boricacid corrosion or stress-corrosion cracking and recom-mend corrective actions.

Task 11- Recommend ASME Code Section XlChanges

To review ASME Code Section XI requirements andsend comments and recommendations to the code com-mittees for action, including (1) appropriateness ofSection Xl size limits for inspection requirements;

NUREG-1339

mr- ~-

1

II

3

! ��' -"-�- t I -� �- .n.", -; -,

(2) provision to ensure adequate visual inspection; and(3) assurance that NDE inspections are effective in de-tecting corrosion wastage arid stress-corrosion cracking.

Task 12 - Recommend Research Programs to EPRI onDetgradation of Fasteners

To recommend three projects to EPRI that would in-crease the undcrstanding of (1) accelerated bic acidattack of carbon and alloy steel fastuners, (2) the effect ofMoS2, and (3) sealants for PWR prinary system compo-nents (thie impact of the recommendations on the con-tracted work under Task 3 also was evaluated).

Task 13 - Alternative Materials and Coatings

To recommend altcrnative materials and coatings andprovide guidance regarding sclcction criteria for the pur-pose of eliminating borated watcr corrosion concerns(this task was includcd in the contracted work underTask ').

Task 14 - Component Support Bolting ScreeningCriteria

To devc.op a stratcgy for idcntifying component supportbolts tha: may be susceptible to strcss-corrosion crackingand recommend plant-specific methods for rsolvingftndings regarding mat:rials that require review.

Task 15 - Assess Fastener Integrity Based on FractureMechanics

To developa technique to evaluate the integrity of boltingmaterial in component support fasteners.

Task 16 - Preload Technology Assessment

To evaluate the need for high preloads, to identify poten-tial relief in preload requirements, to investigate preloadapplicaticn techniques and resulting preload variability,and to recommend optimum techniques. Also, to discussmethodologies for estimating cxisting preloads based onknowledge of the tensioning techniques, sampling, orsome combination of information and to discuss risks ofdetensioning existing joints.

Task 17 - Develop UT Procedures for Inspection ofUltra-High-Strength Low-Alloy MaragingSteels

To develop a field procedure for ultrasonic test (UT)inspection of ultra-high-strength bolts in the lower sup-port feet of the Westinghouse-designed steam generator,using Westinghouse Owners' Group (WOG) funding.

2.13 Tasks Associated with Internals Bolting(Tasks 1D aud 9)

Task 18 - Higb-Strength Bolting

To conduct research to improve the strcss-corrosion rc-sistance of high-strength, age-hardenable, Ni-Cr-Fe al-loys and A 453 Gr. 660 (A236) bolt.g rmnarals and oinvestigate the influence of irradiation and stress andstrain on the behavior of structural materials.

Task 19 - Owners' Croups Liaison

To maintain liaison with owners' groups to ensure thatduplication of effort is minimized and that pertincnt in-formation on the efforts of the task group is cxchanged.

The 19 tasks were modified during the progress of theprogram. Redistribution of effort reflecting reassessmentof relative priorities among the tasks occurred.

2.2 Technical FindingsThe results of theJoint AIF1MPCTask Group on BoltingProgram, described in this section, constitute a recom-mended technical basis for rcsolution of GSI 29 by thenuclear industry. The program, outlined in Section 2.1 ofthis report, was presented in detail along with the resultsfrom the executed tasks in a two-volume report, EPRINP-5769, 'Degradation and Failure of Bolting in NuclearPower Plants" (Rcf. 5). The report was published in twovolumes to make the results of the research easy to reviewand to aid utility engineers in addressing plant-specificbolting and fastener problems with a single sourcc docu-ment. Volume included the background information, adescription of the action plan for the AIFIMPC TaskGroup on Bolting, the approach to resolution of the bolt-ing issue and the basis for this resolution, summaries ofthe findings from the 19 action plan tasks (Section 2.1.1),and the conclusions and recommendations.

Volume 2 included more complete supporting referencesand data. Publication of the two-volume report com-pleted a comprehensive, generic review and analysis pro-xa.. A major finding indicated that the design of critical

closure joint bolting involves enough redundancy to en-sure that there is virtually no pressing cause for concernregarding bolting integrity. A brief review of the EPRIreport by volume and section follows.

2.2.1 Basis for the Resolution

EPRI NP-5769, Volume 1.

* Section 1, "Introduction," provided an historical re-view end the planned tasks (i.e., those given tn Sec-tion 2.1 of this report).

* Section 2, uIndustry Resolution of the Bolting Is-sue," presented additional deta- on the Joint Task

NUREG-1339 4

Group approach to the problem, the rsults ob-tained fromn research according to the task actionplan, and brief discussions of two principal contribu-tions of the wcrk-development (1) of a generalizedclosure itegrity model and (2) of joint leak-tight-ncs criteria. The rsulting iaftormation led to theconclusion that the technical basis for resolution ofthe generic bolting issue was reached.

Section 3, Pressure Boundary Bolting." corcludedthat closure integrity can bc ensured througni appli-cation of a leak-before-break criterion and providesthree analyses (primary manway cover, reactor cool-ant pump main flane, and check valve flange) tcillostrate the methods. The work led to a proposedASME Code Sction XI code case (see the discus-sion of Section 6, Volume 1, EPRI NP-5769, in thisrcport) on inspection of bolted closures with ob-served or detected leakage

* Section 4, Structural and Component SupportBolting," prcscntcd rsults from Task 14 of the ac-tion plan. Component support bolting that may besusceptible to SCC was identified, and both genericand plant-spec±'c review procedures were recoim-rr.crded. It was concluded that application by licen-sees of the proposed screening and disposition (ofmaterials which failed to pass the screen) stepswould be an adequate bolting integrity program andwould scrve to rcsolve GSI 29 with regard to compo-nent support bolting.

* Section 5, "Owners' Groups Summary," prcsentedicsults from the Babcock and Wilcox (B&W) andWestinghouse Owners' Group programs on primaryprcssure boundary bolting. The B&W program lentfurter support to fuidings by others (see the discus-sion of Section 3, Volume 1, EPRI NP-5769, in thisreport) that a leak-before-br eak approach and con-servative failure criteria can bc used to ensure theintegrity of bolted closures. Failure mechanisms n-cluded SCC and chemical watage. A steam genera-tor manway closurc with SCC and a reactor maincoolant pump closure with wastage were analyzed tsexamples to demonstrate the adequacy of this ap-proach and these criteria. The WOG program origi-nally included the following 10 tasks (Ref.6) (seeTable 5-1, Vol. 1, EPRI NP-5769):

(1) Determine the bolting material, number ofbolts, bolt dimensions, and gasket materialused for primary boundary closurcs (i.e.,pumps, valves, steam generators, and pres-surizers).

(2) Provide installation procedures for bolting forprimary boundary closures.

(3) Catalogue service experiences for primarybc,.ndary-closures, and identify scrv'icc-scnsi-tive closures based on utility input.

(4) Follow the AIFIMPC bolting programs, pro-vide liaison for the WOG plants, and preventany duplication of effort.

(5) Dcvelop nondcstructive methods fr beltingfor primary boundary closures.

(6) Frepare spccifications for primary boundarybolting, including quality assurance requirc-ments for procurement, receipt, and preinstal-lation irspections.

(7) Evaluate and qualify scalants for primaryboundary closures.

(8) Evaluate and qualify lubricants for primaryboundary closure.

(9) Establish the numbcr of "failcd" bolts in clo-sures rcsulting in one-gallon-pcr-minute Icak-age (or limits set by technical specifications),and detcrminc margins of safcty for bolting inprimary boundary closures.

(10) Establish feasibility of having an inventory ofbolting (considcring Task 3) for primary bound-ary closures.

Of the original 10 tasks, the WOG actually spon-sored the first 4 tasks; the other tasks were judged tobe adequately covcred by separate efforts sponsoredby EPRI. ASTM bolting standards committees,ASME Codc committees, or other MPC activities.The W05 did assume responsibility forTask 17 (seeSection 2.1.2 of this report), the development of anultrasonic field procedure applicable to ultra-hig,h-strength bolts.

A discussion of the first 4 of the 10 tasks for theWOG's program on bolting follows (Ref. 6):

- Task I supported a leak-before-break ap-proach to closure integrity by addressing thecomplete boltci closure rather than individualfasteners and resulted in publication of a user'smanual.

- Task 2 investigations resulted in recommendedinstallation procedures. Tc arrive at these pro-cedures, the WOG compared several preloadmeasurement techniques, including those fortorque wrenches, stud heaters, stud tensioners,and the Bolt Gage (a Raymond Engineering,Inc., development). The Bolt Gage was thepreferred technique.

NUREG-1339

2� � , - - . - . .

II

I

S

- Under Task 3, from available data, the WOGlisted and analyzed primary boundary closureleakage. It was concluded that (1) the avaltableinformation useful in detcrmining service-scnsitive closures 'is limited, (2) the leak-before-break approach rcommcnded by thcAIF/IPC Task Group on Bolting is a soundengineering attack, and (3) thc ASME Codeshould bc changed to address bolt/stud flawlimits based on closure integrity a,d fastenerredundancy.

- Task 4 providcd for liaison with others in theAIF/MPC task group.

An NDE field procedure was developed for boltsundcr AIFIMPC Task 17 that providcd in situ tcch-niques for inspection of Wcstinghouse PWR stenmgenerator support bolts. The procedurc can obviatecosts prcviously bornc by utilities for removal andsurface examination.

Scction 6, 'ASME to ASTM Codes and Standards."included the results of Tasks 6 and 7 of the actionplan (Section 2.1 of this report). Subtask 6.1 wascompleted with the preparation of "Utility Rccom-mcndations and Guidelincs for the Ptrchase Spe--fication and Receipt/Installation Inspectio,. T-e-quirements for ASME Section IIJ. .mericanInstitute of Steel Construction (AIS.,, AmericanNational Standards Institute ANSI/ASME B31.1,and ANSI B31.5 Bolts and Thrcaded Fastencrs."The report is published in its entirety as Sectior. 1 ofVol. 2, EPRI NP-5769. Subtask 6.2 addressed theadvisability of changing the sampling requirementsfor structural bolting specifications under thejurisdiction of ASTM Subcommittee F16.02, and, atthe time of publication of EPRI NP-5769, approvalof the proposed changes vas still pending. Subtask6.3 resulted in preparation of the draft standard,"Standard Test Mcthod for Equotip Hardness Test-ing of Metallic Materials," its submittal to ASTMSubcommittce E28.06, and its publication as Section2, Vol. 2, of EPRI NP-5769. Task 7 resulted in twoproducts: First, the many places in the ASME Codewhere rules are given for the design and construc-tion of bolted joints, scattered among voluminousrules for welding and other fabrication methods,werc listed and explained in Section 9, Vol.2, EPRINP-5769. Until such time as changes in the ASMECode make the Section 9, EPRI NP-5769, listingobsolete, it will serve as a useful single referencesource. Second, the results of a review of ASMECode bolting requirements were published as Scc-tion 10, "Critique of Bolt Preload Aspcc:s of ASMEand AISC Codes," Volume 2, EPRI NP-.5769. Thematerial in this section was limited to ASME Codctreatmcnt of bolt preload with respect to design,

assembly methods, and quality control, even thoughadditional aspects of fastener preload could havebeen considered.

* Section 7, 'NDE of Bolting," presented the resultsof studies designed to att.iin the goal of developingfield techniques as stated in Task 8 of the AIF/MPCbolting action plan (see Section 2.1 of this report).Conventional UT mthods were cvaluatcd and theirlimitations determined. EPRI funded three separatccontracts and asked the contractors to develop andevaluate new techniques that lend themselves toficld application and that are capable of detcctingboth SCC and wastage. One contractor developedacoustic resonance and reverberation techniques fordetection of wastage, aniother contractor further de-veloped the acoustic resonance technique to detectSCC, and the third contractor developed thecylindrically guided wave tcchnique (CGT) for dc-tection of SCC as well as wastage.

The revcrberation technique used a spectrum ana-lyzcr to quantify the frequency content of a pulsc-ccho cnvclope and detcct characteristic time spacingchanges and then compared the reverberation spec-trum of a target bolt to that of an unflawed bolt todetcct degradation.

The CGWTprovided an inspection method applica-ble to most studs orbolts overa range of 16 in. to 112in. (406 mm to 2,844 mm) in length and 1 in. to 4.5 in.(25.4 mm to .114 mm) in diameter. The techniquceould be used to detect cracklike dcfects as small asU.05 in. (1.27 mm) deep in the threaded region of thebolt. In addition, the CGWTcould be used to detectcorrosion wastage greater than 25% of the bolt di-ameter.

* Section 8, 'Lubricants and Sealants," included threesub-tasks as part of Task 12 in the AIF/MPC taskplan (Section 2.1.2 of this report). Several projectsand studies, described in EPRI NP-5769, provideduseful data. The text of Section 8 was adapted from amore detailed report (Ref. 7). The influence of sev-eral lubricants on boric acid wastage vas studiedwith results too varied to review in detail here. How-ever, one result reported was the detrimental influ-ence of MoS2 and the difficulty of removing it fromfasteners that have been exposed to service condi-tions. The studies of leak sealants and concerns de-rived from them led to several rccommendationsprincipal among them was that the responsible de-sign organization (e.g., ASME) should establishstandards for leak sealants.

* Section 9, Alternative Materials," the task identi-fied as number 13 in Section 2.1.2 of this report, wasdrawn from the more detailed reports of Stress-Corrosion Cracking of Alternative Bolting Alloys,"

NUREG-1339 6

EPRI RP-2Q58-12 (Ref. 8) and Section 3, Vol. 2,EPRI NP-5769.This workalso isrelated toTask3ofthe action plan. lhe resulting four conclusions canbe restated briefly as follows (1) whereas low alloysteelz are vulnerable to boric acid corrosion, otheralloy rteeIs generally are' rsistant, (2) galvanic cor-rosion, depending on specifics of the material com-position and environmental chenistry, can occur,but data are needed for each combination if sensibledecisions are to be made; (3) MoS2 lubricant wasshown to be a factor in laboratory corrosion testingwhen conditions favored the liberation of hydrogensulfide, but its role in service-related failures re-mains to be clarified; (4) more Kcc data arenecded if a damage-tolerant methodology is to bcadopted.

Section 10, 'Training Package," consisted of a briefdescription of two EPRI-sponsored actions aimed atinformation xclange. First, a workshop was heldNovember 2 through 4, 1983, at Knoxville, Tennes-see. Participants included represcntatives from theU. S. Nuclear Regulatory Commission, the AtomicIndustrial Forum, the Electric Power Research In-stitute, its contractors and consultants, and the nu-clear power gencrating industry. The stated objec-tives included alerting industry to the NRC's genericbolt;ng safety issue and the regulator's perspectiveof the issuc. Also, speakers reviewed the AIF pro-gram and the EPRI efforts toward resolution, in-cluding bolting design criteria, codes and standards,specifications, fabrication, quality control, tools,procedures, and general bolting problems. Second, aset of three videotapes was produced and madeavailable to any interested party. They are identifiedas Electric Power Research Institute PressureBoundary Bolting Problems; Part I: The Basics; PartII: Engineering Problems; Part m: The Mechanicand Bolting." Although they were aimed at themanager, the engineer, and the mechanic, respec-tively, viewed together, they constitute a rathercom-plete tutorial on bolting.

Scction 11, Conclusions and Recommendations,"summarized the many conclusions derived fromcompletion of the Joint AIF/MPC Task Group onBolting 19-task program. The diverse disciplines andthe many activities were joined and integrated toprovide what the industry believed to be a basis forresolution of the NRC GSI 29.

2.2.2 Supporting Data for the Resolution

EPRI NP-5769, Volume 2.

Section 1, Utility Recommendations and Guide-lines for the Purchase Specification and Receipt/Preinstallation Inspection Requirements for ASME

Section m, AISC, ANSI/ASME B31.1, and ANSIB31.5 Bolts and Threaded Fasteners," presentedrecommended guidelines for utilities constructing oroperating nuclear power plants,.including certifica-tion, identification, NDE, and storage requirementsfor bolting material (bolts, studs, and nuts) to beused in permanent features. It also included reconi-mcnded guidelines for receipt or preinstallation in-spection designed to help ensure fastener integrity.Recommended guidelines were given for tighteningfasteners when neither preloading,' torquing, norboth are specified by other documents. The guide-lines were written specifically for ASME Code Sec-tion m Code-of-Record plants. They were given asadequate for prc-ASME Code Section m Code-of-Record plants (i.e., ANSI B31.1 and B31.7), but, forplants of such vintage, the user was cautioned toconsider the safety class of the system in which thebolting is used and to provide a commensurate levelof quality. For instance, a plant having ANSI B31.1as the code of record may choose to use ASMESection m Class I requirements for systerrs classi-fied American Nuclear Society Safety Class 1.

* Section 2, "Standard Test Method for EquotipHardness Testing of Metallic Materials," coveredthe use of the Equotip Hardness Tester to deter-mine the Leeb hardness of metal components. Thediscussion included definitions, test procedures, in-strument verification, test-block calibration, and atable of hardness conversion. As previously notedunder Section 6, Vol. 1, EPRI NP-5769, the text ofSection 2, Vol. 2, was a draft ASTM standard, sub-mitted to ASTM Subcommittee E 28.06. As will beexplained in Section 3 of this repoit, the NRC un-derstands the need forin situ hardness measurementas part of a program by licensees to ensure confor-mity to codes and standards, and the NRC agreesthat properly conducted Lceb hardness tests can bepart of that program. However, Section 2 of EPRINP-5769 contained what appear to be technical er-rors because tabulated hardness conversion valuesdisagree with,pubiished ASTM Standards. The ap-parent disagreements must be clarified, presumablyby ASTM Subcommittee E 28.06.

* Section 3, Evaluation of Bolting Experiences inPrimary Pressure Boundary Closures," presentedthe results of compiling and analyzing 125 incidentsof bolting failure reported by nuclear utilities. Theprincipal failuie mechanisms, in order of decreasingimportance, were boric acid corrosion (wastage),mainteriance damage, corrosion pitting, and stress-corrosion cracking. Not included in the analysis werea number of flange bolt problems in the control roddrive mechanism that were judged to be relatedmainly to one nuclear steam supply system (NSSS).

NUREG-1339

I

II

I

i

7

Although the fasterjer rejection rate. varied withthe component and generally was small (not morethan about 10%), the rate war suffcient to ustify

ianidng NRC GSI 29 as a high priority task.

Section 4, Sampling Inspection and AcceptanceCriteria for Bolted Connections," provided a statis-tical evaluation of fastener loads. The evaluationaddressed the main concerns in highly stressed high-strength fasteners, that is, failures that were dueeither to the external load exceeding the preloadacross the joint (overload) or to strcss-corrosioncracking. If the preload of a given fastener wereknown, then one could dccide about prcload ade-quacy simply by checking whether the preloar isinside or outside the acccptable design range for thatfastcner. Because unccrtainties cxist about the ac-tual value of prcload, the detcrministic checkingprocedure must be replaced with a probabilistic cri-terion.

Ideally, standards forbolted connections susccptibleto overload and stress-corrosion cracking should beset by establishing maximum acceptable probabili-ties for the occurrencc of each failure mode, basedon the severity of the consequences. To apply suchstandards, uncertainties on the external loads, thestate of preload, and the maximum flaw size shouldbe quantified. Unfortunately, the state of knowledge(e.g., about SCC) and current deterministic practiccmake it impossible to fully implement probabilisticstandards. A sii-probabilistic format, explicitlyrecognizing uncertainties on fastener preload, butavoiding failure probability calculations, was pro-posed. The uncertainty is changed by sampling in-spection, for which a simple method of sample sizedetermination and uncerainty updating was pro-r'oscd, consistent with the format of the acceptancecritcria.

- Section 5, Nuclear Structural Bolting PreloadEvaluation," reported on the results of completingTask 16 (see Section 2.1.2 of this report). The taskconsisted of evaluating the need for high preloads,identifying potential relief in preload requirements,and investigating preload application techniquesand variability. Section 4 of EPRI NP-5769 providedthe statistical nature of the prcloading process, andSection 5 evaluated cisting prcload design require-ments, the relationship of the specified joint preloadto the minimum preload required to carry designloads, and the effect of potential loading relief onminimum preload requirements for one heavy com-ponent support structural ;oint. The report dis-cussed conclusions that were reached about design

*Rejection rate was the relative number of ailures. degradations, in-spection call-outs, etc., as a percent of te otal fastencn inrseriec.

NUREG-1339

I

criteria, load relief, prcload range acceptability,preload estimation, and preload accuracy.

Section 6, The Bolting Database: An Example ofa Numeric Database Application in the NuclearPower Industry," briefly discussed the nature of thedatabase, the stored information and classificationscheme, access, software, and applications.

* Section 7, 'Assessment of Field Hardness Meas-urements o Low-Alloy Quenched and Tem-pered LAQ`') Bolting Materials at Midland," pre-sented the collection (from four nuclear sites) andanalysis of LAQT fastener steels. Significant devia-tion from specification requirements was observed.It was cstimated that the portion or bolts at onc site(Midland) with a hardncss indicating a susccptibilityfor SCC was less that 1% of the total population; nota serious concern.

* Section 8, "Good Bolting Practices," bricfly re-viewed the two reference manuals for nuclear powerplant maintenance personnel that were developcdunder EPRI sponsorship and were intendcd forrapid-access field or office use by utility staff whomust disassemble and assemble bolted joints in nu-clear power plants. This section described boltingpractices that should help staff members identify,understand, and solve (or minimize) bolted jointproblems such as caks, vibration loosening, fatiguc,and stress-corrosion cracking.

Dhe first of the manuals was entitled, "Good BoltingPractices: Large Bolt Manual;" the second was cnti-tled, 'Good Bolting Practices: Small Bolt Manual."The manuals described the problem-reducing stepsin order of increasing complexity and cost. recogniz-ing hat the options available to maintenance per-soniel are generally limited.

The manuals were not intended for use by designers;therefore, the theories behind the recommendedprocedures were not discussed at any length. Theencyclopedia format for the manuals was intendedto make the topics easy to locate. Topics were listedalphabetically and identified by legends printed inbold face. Each topic was described briefly, withtypical data, if pertinent, and with cross-referencesto related topics, also in bold-face type.

* Section 9, "Bolting Rules of the ASME Boiler andPressure Vessel Code," presented a detailed, point-by-point, review of the ASME Code with explana-tions, interpretations, and suggestions for improve-ment. The many scattered bolting requirementswere collected in this one section of the EPRI reportto provide a source document for refcrencc. Thiswas noted in the discussion of Section 6, Vol. 1,EPRI NP-5769.

8

u

* Section 10, "Critique of Bolt Preload Aspectsof ASME and AISC.Codes," is a comnpanion pieecto Section 9, Vol. 2 (and was cited in Section 6, Vol.1) of the EPRI repor.The stated assignment was "tocritique xisting pload sections of the ASMECode." The Joint AIUFMPC Task Group on Boltingposed two quaestions to g'iide this effort:

(I) Do provisions of the ASME Code contnbuteto the types of bolting failure cxperienced inthe last decade or so by the nuclear industry?

(2) Do omtssions In the ASME Code contnbute tothe types of bolting failure experienced by thenuclear industry?

The Investigators were directed to limit the re-sponses to preload aspects of bolting problems. Thereview identiied several provisions of the AISC andASME Codes that could be troublesome. The pointsraised and the rectifications suggested were too nu-merous to be reported here, but are descnbed inEPRI NP-5769. In addition to general observationsconcerning the ASME Code, preload philosophy,prcload and installation method codification, cduca-tion, quality control, and the role of the mecnanic,five specific problems were cit-d and solutions of-fered. Four appendices completed the section; theywent into greater detail on specific problems andgave support to the conclusions and recommenda-tions stated in the text of Section 10. The titles of theappendices follow

(.) Appendix IOA, 'AISC Specification for Struc-tural Joints Using ASTlM A325 or A490 Bolts"

(2.) Appendix lOB, UASMEE Boiler and PressureVessel Code, Section III, Division 1, Subsec-tion NV"

(3.) Appcndix OC, ASME Boiler and PressureVessel Code, Section Vm, Division 1"

(4.) Appendix 10D, "Comparison of ASME Codeson Pressure Boundary Bolting"

. Section 11, "Evaluation Procedure for AssuringIntegrity of Bolting Material in Component Sup-port Applications," is the companion piece on struc-tural bolting to the RCPB discussion, Section 3, Vol.2, EPRI NP-5769. The Section 11 presentation de-scribes an evaluation procedure for general applica-tion to bolting products used in component supportsand fabricated from steels commonly used for sup-port bolting. The evaluation procedure could beused to justify serviceability of questionable materi-als. It was anticipated that this section would beuseful to a utility as part of a plant-specific plan to

, ~ ~ - . - v . '~ '. . . _ _ _ _ _

deal with materials that require some evaluationunder the generic lssue.

The primary objective was to present the proceduralsteps and requLred Information to determine allow-able bolt loads to avoid SCC under steady-mate orlong-term normal operating conditions. Allowablebolt stress as a function of material hardness, boltsize, and thread pitch could be determined with theprocedures. The allowable bolt stress then could becompared with actual bolt stresses ca-lclated for thedesign. A requirement of the procedure was thathardness testing be performed on the population ofbolts so that hardness limits could be statisticallydetermined, Also, as part of the evaluation objec-tive, allowable bolt stresses to prevent fracture u,.der short-term (accident 'ads must be establishedwhen low toughness was in.nlied by the harr' .essdata.

* Section 12, "Alternate Alloys," consisted of briefreports on five separate research projects, all spon-sored by EPRI and dealing with steel corrosion innuclear reactor environments. Each project haJbeen reported in more detail elsewhere; the fivepublished documents were cited in the EPRINP-5769, Volume 2, reference list.

The Combustion Engir;eering Project grew out ofinterest in fastener corrosion and is entitled, "Lit-erature Survey of Carbon and Alloy Steel FastenerCorrosion in the PWR Plants." The objective of thisproject was to determine the extent of low-alloysteel fastener corrosion problems in the domesticPWR industry and to review available data in theliterature on boric acid corrosion and stress-corro-sion cracking of fasteners. Service failures from bothmechanisms were collected and analyzed. A com-mon fact¢r in sbx SCC events involving steam gen-erator prin3ary manway closure studs, which pose apotential for a LOCA, was the use of MoS2 lubri-cant. Decomposition at high temperatures can yieldhydrogen, which can induce SCC in HSLA steelseven at low concentrations.

The Battelle, Columbus, Laboratories Project wasclosely allied to the Combustion Engineering Pro-ject. It involved a review of joint failures in nuclearcomponents from either boric acid wastage or SCC.The primary objective was to deternine if austenitic,age-hardenable matcrials could be used for boltingapplications. A secondary objective was to reviewthe boric acid corrosion and stress-corrosion crack-ing behavior of currently used low-alloy steels andissues relating to lubricants and sealants. Based onthe review, recommendations were made for furtherwork to improve the industry's capability for dealingwith the bolting problem. It was concluded that

9 NUREG-1339

I* -a kThrL. l -- " 0115112 -: -. r .r r .= : * s

austcnitic materials seem to be resistant to boricacidwastage but vulnrable to SCC. The report pre-sented so many qualifications regarding the use ofhigh-strength high-alloy steels that Battelle couldhardly be aocused of recommending them. Mutch ofthe report discussed conditions that can lead to fail-ure of low-alloy stecls without recommending alter-native resistant materials.

The Materials EngIneering Associates Project waspart of a failure analysis of Type 410 stainless steelvalve studs pur:hased to ASTM specification A 193,Grade B6, with a supplemental requirement of 125ksi tensile strcngth specified by the util.ty. The utili-ty's investigation focused on improper heat treat-ment of the studs, resulting in temper embrittle-ment. The embrittlement permitted SCC, with studfailure occurring once the critical flaw size wasachieved. Mechanical property tests confLrmed thatthe material exhibited low fracture toughness. Incombination with the experimentally determincdrather high tensile strcngth and somewhat reducedductility (less than 50% reduction in area in three ofsix specLrnens), the studs would be vulnerable toSCC.

The Westinghouse Electric Project sought a solu-tion to the cracking of age-hard enable Ni-Cr-Fe al-loys in PWRs and boiling-water reactors (BWRs).Several instances of stress corrosion or, in somecascs, corrosion ftigue in bolts, beams, and pinswere observed in reactors using Alloys X-750,I-718, and A 286. nTe object was to examine thethree alloys in different heat-treated conditions. Al-loy X-7S0 with increased amounts of zirconium,which previously had becn shown to be beneficial,also was included. Stress-corrosion cracking studies,involving both crack initiation and crack propagationspecimens in PWR and BWR conditions, were con-ducted on alloy X-750 in 11 conditions, alloy 1-718in 2 conditions, and alloy A 286 in 2 conditions. Thcoperating conditions of BWRs were shown to bemore detrimental to the alloys than operating condi-tions of the PWRs. Alloy X-7S0 exhibited the mostresistance lO cracking (or propagation) in one of theseveral heat treatments that wcre applied, but, in adifferent condition, it was the least resistant. Long-term (more than 10,000 hours) tests are continuing.

The Babcock and Wilcox Project was a companionto the Westinghouse project, using the same high-strength, age-hardenable Ni-Cr-Fe alloys, X-750,1-718, and A 286. Service failures were attnbuted tofatigue, corrosion fatiguc, and intergranular stress-corrosion cracking (IGSCC). Susccptibility to failureby these mechanisms depenced strongly on the met-allurgical condition produced by thermo-mechanical

NURbG-1339

procesosng. The project included detailed microstruCtural characterization and corrosion testing ofthe alloys subjected to 15 different c:nbinations ofmelting practice and thermomechanical processing.-As in the Westinghouse study, preliminary findingsindicated that AlloyX-750 had thebest resistance toSCC when in a particular metallurgical condition.As in the companion study, the conclusions were notsolid and unambiguous except for recommendingfurther studies.

Section 13, 'Standard Specification for Supplemen-tal Requirements for Structural Fasteners for Nu-clear Applications," consisted of some backgroundand introductory material and the proposed ASTMstandard: F XXX-"Standard Specificatio. forQuality Assurance and Inspection Requirements forStructural Fasteners for Nuclear and Other SpecialAppli'-tions." Expericnce with fasteners has cre-ated several conccrns. The draft specification in-cludes requirements for nuclear fasteners as follows:

- Establish sampling and quality levels for allseries of structural fasteners on a uniformbasis.

- Establish mandatory lot control and trace-ability of fastencrs. By maintaining such con-trol, prevent mixing and possible contamina-tion of parts intended for nuclear systems.

- 'Require positive identification and sourcr offasteners intended for nuclear system as evi-dence of adherence to required quality level.

- Require preferential full-scale testing of fin-ished fasteners in lieu of reliance on possiblemachined coupons from fasteners. Actual full-scale testing is designed to confirm integrity offinished fastener not possible by coupon evalu-ation.

- Permit utilization of state-of-the-art technol-ogy and benelcial effects of heading andthread rolling by specific callout. Such othermajor industries as automotive and ae-spaeehave similarly mandated such requirements.

- Recognizing the potential long-term degrada-tion resulting from the presence of discontinui-ties such as cracks and seams, establish specificrequirements to define acceptable and rejec-table criteria for nuclear system use.

This very important standard, now in the hands ofthe cognizant ASFM committee, was supported bythe Joint AIFIMPC Task Group on Bolting.

*Thzi indard is being dcycloped; the numberwill beassigned after itu compkted and apprwed.

10

so- hajollmogi I issiolka -i-m iFa zi 'VI

* Slon 14, e Bolting Technolog, Council"OIL), providd a brief description of the Council,Its activities, and its rnakeup. e BTC Is isIMiatedw the Mate alPropertks Council, I1c, focmerlythe Metals Proerties Council, which providesadnlnistrtie services as requIred. Tbe BTC wasIoed to provide opportunities for threaded ias-.tener and toot use to engage In a variety of coop-erative acbItiltes. As stated in its bylaws, the purposeof the Co.ncl is o sponsor rescarch to recorn-mend pactices; to act as a clearing house for infor-mation, and to proMde education concerning the artand sdence of the Istallation and behavior of me-chanical. fasteners and their interaction with thejoints they are used in. As anyone who has at-tempted to understand bolted joint behavior wi'iealize, the task selctcd by the BTC X not a simpleone, nor will the effort be inexpensive. Because ofthe magnitude of the job, members felt that it wouldbe dcsirable to pool a portion of their technical andfinancial rcsources and attack the problems jointly.Results achicvcd by cooperative efforts, further-more, oftcn have grcatcr crediblity, are more widelyaccepted, and are most economically achieved. TheCouncil expects to provide benefits to industrythrough interaction with recognized experts nbolting technology, opportunities to partkipate inseminars and syrnposia, opportunities to sha,e incooperatively funded research to be planned, moni-tored, and directed by BTC groups, and opportuni-ties to review publications and research results wellbefore gencral rclease. It is anticipated that the BTCwill identify unresolved bolting problems recognizednow and as they arise from experience in the future.Through'its rsources n personal expertise and infinancial assistance, the BTC will be instrumental indevcloping solutions to generic bolting issues.

3 CONCLUSIONSThe NRC staff has reviewed the technical findings devel-oped by the industry and presented in EPRI NP-5769(Ref. 5) as well as other relevant industry-generated in-formation. The staff has concluded that the technicalbasis for resolution of GSI 29 is available. -

The conclusion that GSI 29 can be resolved Is based onthe availability of several relevant documents. Actionstaken by the Joint AIF/MPC Task Group on Boltingresulted in EPRI NP-5769, three video training films (seeSection 10, Vol. 1, Ref. 5) and the Good Bolting Practicesreference manuals, Vols. 1 ar,d 2 (see Section 8, Vol. 2,Ref. 5). Industry representatiyes established the Bo!tingTechnology Council (an MPC affiliatc) to take the lead insponsoring bolting research, recommending practices,gathering and providing information, and promoting edu-cation on installatlon, application, behavior, and interac-

tions of fasteners.The lnstltute of Nuclear PowerOpera-tions ran be pected to act in response to potentiallyunsafe conditions as in the past when Significant Operat-ing Event Report (SOER) No. 84-5 on bolting (Ref. 9)was Issued. Further refinement In codes and tandardswil be provided by the responsible committecs in ASMEand ASTM (e.g., Committee F16 on Fasteners).

During the period in which GSI 29 was being resolved, theNRC took several adtional steps that must be factoredinto the resolution of the issue. Incidents of threadedfastencr degradation and failure from October 1969 toMarch 1982 wtre idcntlficd and analyzed (see Ref. 4)Fhve docunicnts wert prepared based on results of techni-cal assistance contraas in support of the bolting genericissue (Refs. 10 through 14). In addition, action itemsincluded the following NRC notices, bulletins, and ge-neric letters:

* IE Bulletin 74-03, 'Failure of Structural or SeismicSupport Bolts on Class I Components," April 29,1974.

* IE Bulletin 79-02, Pipe Support Base Plate Dc-signs Using Concrete Expansion Anchor Bolts,"March 8, 1979.

* IE Bulletin 79-07, Seismic Analysis of Safety-Related Piping," April 14, 1979.

* XE Bulletin 79-14, "Seismic Analysis for As-BuiltSafeiy-Related Piping Systems," July 2, 1979,(also: Rcvision I of page 2 of 3, July 18, 1979;Supplement 1, August 15, 1979; Supplement 2,September 7, 1979).

* IE Bulletin 82-02 (Ref. 2), which resulted in W.Anderson and P. Sterner, "Evaluation of Responsesto IE Bulletin 82-02," NUREG-1095, May 1985.

* NRC Compliance Bulletin 87-02, "FastenerTestingto Determine Conformance with Applicable Mate-rial Specifications," November 6, 1987 (later Sup-plements 1 and 2).

* NRC Bulletin 89-02, 'Stress Corrosion Cracking ofHigh-Hardness Type 410 Stainless Steel InternalPreloaded Bolting in Anchor Darling Model S350WSwing Che-k Valves or Valves of Similar Design,"July 19, 1989.

* NRC Generic Letter 88-45, 'Boric Acid Corrosionof Carbon Steel Reactor Pressure Boundary Com-ponents in PWR Plants," March 17, 1988.

* NRC Generic Letter 89-02, Actions to Improvethc Dctcction of Counterfeit and Fraudulently Mar-kctcd Products," March 21, 1989.

NUREG-133911

_ ._P, 1.

* IE Clar 78-14, HPCI Turbine RecrtingChanber Hold Down Bolting." Juy 12, 178

* EB Infonnation Notice 80-27, "Degradation of Re-actor Coolant Pump Stbds,' June 11, 1980. -

* mE Information Notice 80-29, "Broken Studs onTerryTrbine Steam Inlet Flange," August 7,1980.

* IE nformation Notice 80-36, Falure of SteanGenerator Support Bolting," October 10, 1980.

* IE Information Notice 82-6, "Failure of SteamGenentor Privacy Side Manway Cosure Studs,"March 12, 1982.

* IE I nformation Notice 86-25, wTraceability and Ma.tcrial Control of Material and Equipment, Particu-larly Fasteners," April 11, 1986.

• IE Information Notice 86-108, "Degradation of Re-actor Coolant System ressure Boundary Resultingfrom Boric Acid Corrosion," November 1986(later Supplements I and 2)

NRC Information Notice 87-56, Improper Hy-draulic Control Unit Installation at BWR Plants,"Novembcr 4, 1987..

* NRC Information Notice 88-11, 'Potential Loss ofMotor Control Centerandlor Switchboard FunctionDue to Faulty Tie Bolts," April 7, 1988.

NRC Information Notice 89-22, QuestionableCertification of Fasteners," March 3, 1989.

* NRC Information Notice 89-59 and Supplement 1,"Suppliers of Potentially Misrepresented Fasten-ers," August 16 and December 6, 1989.

* NRC Information Notice 89-70, Possible Indica-tions of Misrepresented Vendor Producs Octo-ber 11, 1989.

In various ways, these NRC notices, bulletins, letters, andcircular will influence actions that the NRC or licenseeswill need to take in the wake of the rcsolution of GSI 29.Aitho: they do not, indiykhuly or colleciwly, form abasis for the resolution, neither will these documents northe responses made to them change as a result of theresolution of OSI 29.

The NRC staff concludes that all of the available informa-tion that has been discussed in this report (from industryand regulatory sources combined) provide a sufficienttechnical basis to resolve GSI 29.

NUREG-1339

It must be understood that although the NRC staff recog-nizod the value of the sveral products of the industryeffort (the work of the Joint AIFIMPC Task Group onBolting) in helping to resolve GSI 29, that recognitiondoes not constitute unqualified endorsenent of theirtechnicat content. The NRC staff found technical dis-agreement with seral specific discussions in EPRINP-5769, the three videotapes on training, and the G'odBolting Practices reference manuals. The technical dis-agreements, except for the following, however, generallywere not important enough to mention.

First, the staff notes that a general plan for evaluation ofbolting integrity can be derived from Section II, Vol. 2,EPRI NP-5769. Section 11, Evaluation Procedure forAssuring Integrity of Bolting Material in ComponentSupport Applications," was written to fulflll a specificassignment for the Joint AIF(MPC Task Group on Bolt-ing. With appropriate modifications, the procedure couldse for other than component-support bolting. Thescreening process should usc fastencr material propertiesand fracture mechanics analyscs to ensure thatsafety-related fasteners are unlikely to be susceptible tostress-corrosion cracking. Material propcrties should beexperimentally verified rathcr than assumed to be asspecified.

Second, and closely related to the first comment, incon-sistencies found in EPRI NP-5769 regarding the criteriafor categorizing bolting steels according to strength mustbe reconciled. Categorization should be based only on theactual measured yield strength, Sy, of the material(or S. determined by conversion of measured hardnessvalues) and not on the specified minimum yield strength.The justification for this position is that high-strengthsteels are vulnerable to SCC. A bolt made of high-strength steels may be obtained through an order whichspecifies a relatively low-yield strength, but by improperheat treatment, for cxample, the bolt may develop anactual strength far in excess of the minimum specified.Specifically, that high-strength bolts should be thosewith S. 2 150 ksi, medium-strength bolts should bcthose with 120 ksi < Sy < 150 ksi. The following por-tions of EPRI NP-5769 need to be modified in order tomake them consistent with the abovc definitions:

* In Vol. 1, Section 4, page 4-3, bolting steels arecategorized as "high strength" if: i Sy > 150 ki,!.vhere S, is the yield strength' (compare thegreater-than symbol to tib 1rcater-than-or-equal-tosymbol recommended). At thc same locatio;i, mr-dium-strength materials are identified as those with120 ksi < S < 150 ksl which would bc onslstentexcept for the explanafory text that follows on page4-4: 'herefore, it seems appropriate for the indus-try to cxamine the use of materials vith speci-fied minlr .jm yield strengths greater than 150 ksi"

12

A I l . i .i - ;0Z

(erpphasis added The me Wosar e used on page4-S at two places.

* On pagc 4-4, a proposed category is denned by therange of 120 ksi to 150 ks specified rinimum yield

. strength" dowing that the use of S1 on page 4-3was not to be taken literally.

* In Vol. 1, Section 11, 'Conclusions and Recommcn-dations," page 11-5, the words "minimum specified"are used apin.

* On the nex pagc (p. 11-6) one inds 'specifiedyicldstrength." An inconsistency arises in EPRINP-5769, Vol. 2, Section 1, pagc 1-17 because thereader is advised to consider mateials vulnerable toSCC if the minimum sperfied ultimate tensiestrength (UITS) Is greater than 150 ksl or if the actualUrS is greater than 170 ksi.

* Then, in NP-5769 Vol. 2, Scction 7, page 7-2, wefind: . .. the proposed screcning limit of S, • 150ksi (1034 MPa)" although in the preceding parm-graph the words "specified minimum yield strength"were used to describe thc strength range of 120-150ksi (827-1034 MPa).

A more careful reading might rcveal more discrepanciesor inconsistencies. For the reasons previously given, thecriterion of actual yield strength, Sy _ 150 ksi should beused as the level for consideration of SCC vulnerability.

Third, the data listed in Table 2-1, Vol. 2, EPRI NP-5769are questionable. Indexing off the values of RockwellC-scale hardness as given, the corresponding values ofVickers hardness numbers do not agree with those givenin the ASIM Standard E 140. From the same Rc start,the corresponding values of tensile strcngth do not agreewith values given in the ASTM Standard A 370. Suchcrrors (thcre arc typographical mistakes, as well) alsomake the hardness conversions listed in Table I lA-1 ofEPRI NP-5769 suspect; they should be audited. Accept-ing the ASTM stindards as the authority, the hardnessconversions and hardness-tensile conversions in EPRINP-5769 should be treated skeptically. Siilce Table 2-1was to be part of a draft ASTM standard, the responsibleAS`7M committee can be expected to make such correc-tions as may be necessary. Until Leeb hardness values andconversion tables have been incorporated in a standardtest method by the ASTM, they should be used forInformation only' and not be accepted as edence nllcznsing ections or in safety evnluations.

Fourth, the indictment against MoS2 asa lubricant (foundon page 3-S of EPRI NP-5769, Vol. 2) descryus moreemphasis. Facts gleaned from some service failures andfrom laboratory examinations (Ref. 12) clearly show that

MoS 2 s a potential contnbutor to SCC, especially whenapplied to high-strength bolting steels. One of the prob-lems posed by MoS2-difficulty in removing it from partsthat have been in service (see page 8-3, Vol. 1., Ref.5)-may be close to being resolved. Whereas Czajkowski(Ref. 12) found that CS2 will remove MoS2, handling CS2poses some problems. More recently, tests by Czajkowskdof samples of 'citrus-based cleaners" were subjected to acleaning task.dmilar to that eported in Ref. 12, ad it wasevident that the sulfur component (thc active SCC ingre-dient) had been effectively removed (Ref. 15). Providingthat the citrus-based deaners, themselves, are not SCCpromoters, an answer to the MoS2 cleaning problem maybe at hand.

Ffth, although the fracture mechanics analyses byCipolla cited in Setion 9, Vol. , EPRI NP-5769, areuseful and could well be employed in engineering prob-lems where values for the stress intensity factor, K, areneeded, other more recent results are available. In areport published in 1988, 'Review and Synthesis of StressIntensity Factor Solutions Applicable to Cracks in Bolts'(Ref. 6), values for K for cracks in round bars, boththreaded and unthreaded, subject to either tension orbending, were reviewed. Available solutions were synthe-sized into forms appropriate to analyses of bolts andstuds. The K solutions published in Reference 16 shouldbe used In fracture mechanics analyses of threadedfasteners.

The importance of maintaining adequate traceabilityand control of material of fasteners at nuclear powcrplants was set forth in TE Information Notice No. 86-25(Ref. 17). Because plant-specific bolting integrity pro-grams should include steps to ensure bolting traceabilityand material control and to prevent introduction of incor-rect or defective materials or components, the centralideas from this notice follow.

Awareness of 10 CFR Part 50, Appendix B, CriterionVIII, Identification and Control of Materials, Parts, andComponents," and applicable codes and specifications isimportant. Measures have been established and imple-mented by the NRC for identification and control ofmaterials, parts, and components and for traccability bothto the approved design basis and to the source. It is impor-tant that required identification of items be maintained byheat number, part number, serial number, or other ap-propriate means, either on the item itself or on recordstraceabletothe Ietri as required,land that requied =ark-IAgs be en the Item.

'In Attadment 2 to NRC Bulktin No. 8-10, November 22, 1988,verifiabte traceability was defined as (with minor editing for this re-porl Documented evidence such as a certil5te of complince thataublisha trexability of purrhased equlpmcnt tothe manufacurcr. l

the certi e of cotnpliance hs providd byany party ohcr than themau`i"trer. the validity of the cerliflc:le must be verified by the li-censee or permii -t1 -irough an audit or other appropriate mans.

13 NUREG-1339

K

PR _ _ ---

2

7. Z... . - .. . . .

It Ls the licensee's responsibility to use qualified individu-als to muninc markings on material and equipment andto verify that the markings represent material and equip-ment as specfied by the design drawings and specifica-tlinm In the case of fasteners, compliance with the appli-cable =aterial specication (c.g., ASTM or ASME mate-rial'and grade) is verified by required markings on boltsand nuts and certified material test reports or certificatesof conformance as required by procurement drawings andorders and by applicable codes and specifications. When

. vendor-supplied equipment assemblies contain fasteners,it is important to verify compliance with approved vendordrawings and specifications and such other informationas materials used for equipment qualiicati irn tcsts analy-ses. The required markings on material and equipment,including fasteners, not only must exist, but the markingsmust indicate thc correct material and gradc as specified.

The NRC staff rcsolecd GSI 29 based on the findingspresented herein, including the following thrce condi-tions.

First, all carlicr NRC noticcs, bulletins, and gcncric Ict-tcrs that bear on the issucs involved in bolting, dcgrada-tion or failurc, somc of which wcrc notcd earlier in thissection, should remain in cffect.

Second, it was concluded that an cffective means of cn-suring bolting rcliability, as rccommcndcd in Rcf. 5,would be through dcvelopmcnt and implcmcntation ofplant-specific bolting-integrity programs. Thcse pro-grams should be comprehensive and include all relevantNRC requirements and guidance and the recommendedpositions of the industry-sponsored programs.

Third, it is recommended that a new section of the Stan-dard Review Plan (SRP) be prepared to provide guidanceto the staff for the review of future pla nts. The elementsof the review would include all safcty-related joint design,threaded fastener material selection, and programaticaspects dealing with bolting intcgrity during construction,opcration, and maintenance, except for closure studswhich are addressed in SRP Sections 5.3.1 and 5.23.

In light of the facts that many safety-related systems andcomponents rely in large measure on fastener integrityand that there have bcen numerous reported instances ofdegradation or failure of threadcd fasteners, completionof the studies under GSI 29 has led to the conclusion thatfaatcncr intepitv n"ecq 'u k p'ocedurally controlled.'Ihe information reviewed in this report showed that thesafcty issue rclated to fastener integrity involves a verylarge number of parts in each plant, a number of potentialfailure mechanisms (therefore, a corresponding numberof protective or corrective actions), and severdl technicaland engineering disciplines. Although the rcsolution ofGSI 29 was found to bc rather complex, sufficient guid.ancc is available to rcsolve this issuc, mainly from EPRI

NP-5769 (Ref. 5). Specifically, the NRC staff concurrcdwith the recommendations and guidelines provided inSection l' Vol.2, of EPRI NP-5769. Thc recommcnda-tions and guidclines apply to threaded fastencrs with re-gard to certification, identification, nondestructive exami-nation, storage and tightening procedures except whenstorage and tightening pro:cdures are specified in otherdesign documents or drawings. Implementing Section and other technical guidelincs in the EPRI report wouldhelp ensure fastcner intcgrity.

A comprchensive bolting intcgrity program for a nuclcarpower plant would include all safety-rclatcd bolting. es-pecially bolting used to ckisc the primary prcssurc bound-ary and used for component support.

Of particular importancc to safety are component sup-port fastencrs in the onsite power distribution system,including those power sourccs, distribution systems, andvital supporting ystcms provided to supply powcr tosafety-rclated equipment and capable of operating indc-pcndently of the offsite power system. The onsitc powersystem includes an ac distribution system, a dc powcrsystem, an unintcrruptiblc ac powcr system, and thecmergcncy (dicscl gcncrator) power system. Fastcncrs inthe auxiliary feedwatcr systcm and its support systcms arealso important to safe operation of a plant.

The work donc to rcsolvc GSI 29 has shown that (1)existing requirements, (2) the implcmentation of lcak-before-break criteria for RCPB joints (proposed in EPRINP-5769, Volume , Section 3), and (3) the ongoing pro-grams (e.g., implementation of USI A-46 and the devel-opmcnt of individual plant cxaminations for externalevents).should adequalely limit the risk resulting from,and minimize the sevcrity of, the failure of safety-relatedbolting in current plants. However, licensees with operat-ing plants could avoid many of the prob:emF recorded inthe past by developing and implemcnting plant-specificbolting-integrity programs that include current require-ments and reflect the information and recommendationsmade by the industry-sponsored program managcd byEPRI (with NRC staff exceptions as discussed in Scction 3o' this report). New plant licensecs, however, could m cetstringent bolting requirements with only a very small costincrease if established before they begin operating theirplants.

Guidance regarding bolting for staff reviewers in theNRC Office of Nuclear Reactor Regulation (NRR) per-forming safety reviews of all new nuclear power plantscould be provided by a new section in the NRC StandardReview Plan. Such a section, entitled, for cxample,uSafety-Related Bolting," would expand the limited cov-erage on fasteners now included in the SRP and provide a

lhnc Section tile k Utiliy Recommendations and Guidelines forthe Purchntsc Speciflention And Receiptfl'rcinstallatlion Inspection Re-nuirements for ASME Section I1, AISC, ANSI/ASME 331.1, nndANSI 1331.5 Dolls and Thrcnded F.ulenerL.

NUREG-1339

I

I

e-

fI I

14

systematic mcthod for implementation of ihe staff posi-tion rgarding the basis for resolution of GSI 29. As partof the rcsolution of GSI 29, the staff noted the absence ofan SRP scction on general ireviews of bolting and rccom-mendcd that one be prepared and issucd.

41.

REFERENCESU.S. Nuclear Regulatory Commission, "Potentialfor Low Fracture Toughness and Lamellar Tearingin PWR Steam Generator and Reactor CoolantPump Supports," NUREG-0577, October 1983.

2. U.S. Nuclear Regulatory Commission, Office ofInspcction and Enforcement, "Dcgradation ofTbreaded Fastencrs in the Reactor Coolant Pres-sure Boundary of PWR Plants," IE Bulletin No.82-02, June 2, 1982.

3. U.S. Nuclear Regulatory Commission, "A Prioriti-zation of Generic Safety Issues,' NUREG-0933,Rcv. 0, November 10, 1982.

4. U.S. Nuclear Regulatory Commission, "Treaded-Fastencr Expericnce in Nuclcar Power Plants,"NUREG-0943, January 1983.

5. The Elcctric Power Rcsearch Institute, R. E. Nick-ell, Principal Investigator, 'Degradation and Failureof Bolting in Nuclear Power Plants," EPRINP-5769, Vols. 1 and 2, Palo Alto, California,April 19&8.

6. E. 1. Landerman t al., Assurancc of PrimaryBoundary Bolting and Closure Integrity," Westing-house Owners' Group, Material Subcommittee,Program MUHN 1072, December 1985.

7. R. Rungta ind B. S. Majumdar, "Materials BehaviorRelated Issues for Bolting Applications in the Nu-clear Industry," in Improved Technology for CriticalBolting Applications, E. A. Merrick and M. Prager,cds., MPC-Vol. 26, pp. 3948, American Society ofMechanical Engineers, New York, 1986.

8. R. RJngta and B. S. Majumdar, Stress-CorrosionCracking of Altcrnativc Bolting Alloys," Final Re-port RP 2058-12, Electric Power Research Institute,Palo Alto, California, March 1986.

9. The Institute of Nuclear Power Operations, Bolt-ing Degradation or Failure in Nuclear PowerPlants," SOER No. 84-5, September 20, 1984. Pro-prietary Information. Not Publicly Available.

10. U.S. Nuclear Regulatory Commission, "Lower-Bound Klscc Values for Bolting Materials-A Lit-erature Study," NUREG/CR-2467, February 1982.

11. U.S. Nuclear Regulatory Commission, 'Examina-tion of Failed Studs From No.2 Steam Gene-rator atthe Maine Yankee Nuclear Power Station,"NUREGICR-2793, February 1983.

12. U.S. Nuclear Regulatory Commission, Testing ofNuclear Grade Lubricants and Their Effect on A540B24 and A193 7 Bolting Materials," NUREG/CR-3766, March 1984.

13. U.S. Nuclear Regulatory Commission, "Bolting Ap-plications," NUREGICR-3604, May 1984.

14. U.S. Nuclear Regulatory Commission, "Preloadingof Bolted Connections in Nuclear Reactor Compo-nent Supports," NUREG/CR-3853, October 1984.

15. C. J. Czajkowski, Brookhaven National Laboratory,letter to Richard E. Johnson, U.S. Nuclear Regula-tory Commission, August 4, 1988.

16. L A. James and W. J. Mills, 'Review and Synthesisof Stress Intensity Factor Solutions Applicable toCracks in Bolts," Engineering Fractur Mechanics,Vol. 30 (No. 5), pp. 641-654, 1988.

17. U.S. Nuclear Regulatory Commission, Office of In-spection and Enforcement, 'Traceability and Mate-rial Control of Material and Eouipment, ParticularlyFasteners," IE Information Notice No. 86-25,April 11, 1986.

5NUREG-1339is

N NC FORM 336 U.S. NCLEAR REOUATORY tO9Pt88oN 1. REPORT NUMBER| (2't0) -AsI 9y NRC. Add Vol.,

'RC2l11o2. vuW, e..R , and Addendurn NLr-31, BIBUOGRAPHIC DATA SHEET -" if any.)

(See Irjctlor on the rvrse) NUREG-1339

2. TITLE AND SUBTITLE3. DATE REPORtT PUB~USHED

Resolution of Generic Safety Issue 29: Bolting Degradation or Failure in J_DATE__________________Nucicar Powcr Plants MONTH YEAF

June 1990

R4 OR ORAlT lMuAIE

-N/A6. AUTHR(8) S. TYPE OF REPORT

Richard E. Johnson Technical

7. PERIOD COVERED (hctlva Dales)

N/A

t PERFOR5NG ORGANz.rKJ - NAME AND ADORESS (it N.C. provide DSion, Offic or Region, U.S. Nucleaf Rosulatory Cornmlssion, andmailing address If =.ractor, provide name and mallirg addres.)

Division of Safety Issue ResolutionOfficc of Nuclear Regulatory ResearchU.S. Nuclear Regulatory CommissionWashington, DC 20555

9. SPONSORING ORGAIZATION - NAME ANO ADDRESS i NRC, type '8an as aoo If contractor, provide NFC Division. Oltice or Region.U.S uclear Regulatory Carrenlsslon, and mailing address.)

Same as above

10. SUPPLEMENTAAY NOTES

11 ('t' TRACT 200 words or Ss)

>is report describes the U.S. Nuclear Regulatory Ccmmission's (NRC's) Generic Safety Issue 29, "Bolting Degra-dation or Failure in Nuclear Power Plants," including the bases for establishing the issue and its historical highlights.The report also descnbes the activities of th^ Atomic Industrial Forum (AIF) relevant to this issue, including its co-operation with the Materials Properties Council (MPC) to organize a task group to help resolve the issue. The Elec-tric Power Research Institute, supported by the AIF/MPC task group, prepared and issued a two-volume documentthat provides, in part, the technical basis for resolving Generic Safety Issue 29. This report presents thc NRC's rc-view and evaluation of the two-volume document and NRC's conclusion that this document, in conjunction withother information from both industry and NRC, provides the bases for resolving this issue.

. 1~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~f.

12. KEY WOROSlDESCRiPTORS (List words or prtases that will sslst researchrs In locating the report.) 13. AVALABILTY STATEMENTUnlimited

Bolting; Degradation; Failure; Generic safety issue; 14. SECURITY CLASStFICATION

Resolution; Fastener integrity- Fracture; Fracture mechanics (This Page)

Corrosion; Stress-corrosion cracking; Nuclear power plants. Unclassified

(This Report)

Unclassified15. NUMBER OF PAGES

1-. PRiCE

NAC FORM 336 (2-0)

-