ML20206U699

Forwards Draft Rept Summary Rept on EQ Programmatic Review, for Review & Comment

ML20206U699
Person / Time
Issue date:	04/08/1996
From:	Holahan G NRC (Affiliation Not Assigned)
To:	Hodges M, Shao L, Sheron B NRC (Affiliation Not Assigned), NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
Shared Package
ML20206U672	List: ML20206U667 ML20206U680 ML20206U687 ML20206U694 ML20206U699 ML20206U705 ML20206U711 ML20206U717 ML20206U724 ML20206U729 ML20206U733 ML20206U743
References
FOIA-99-82 NUDOCS 9902170282
Download: ML20206U699 (135)

Text

_ -

<. a

! April it. 1996 MIMORAwilM 10 Brian W. Sheron, Director. DE, NRR M. Wayne llodges. Director. OST..RES 5

LawrenceC.Shao.0lr#Ctor.0Er.l.00 Charles E. Rosst. Director...$P0i, Scott F . Newberr Dir tordPOLR r0 RPM NRR. . . .

. lose A. Calvo. C of LB. DE. N R tedyard B. Marsn Chief. SPtB, DSSA. NRR IROM Gary M liolahan, Director (Oriainal stor >d by G.M. linlahan)

Division of Systems Safety and Anllysts Office of Nuclear Reactor Regulation Stilui ( l DRAfi REPORT ON COUIPHfNT OUAL1iICATION The at: ached draf t report ("Sumary Report on the Equipment Qualification (f0)

Proijr, m tic Review") is being provided for your review and coment This r epoi t .unanart/c' an catensive re assessment of the NRC l.0 proQram in an at t empt in identify the program Strengths and weaknesses and any recommen-ila t i o - for improvement l'o l l on unJ re'.Olution of Comments the report and any POComendation Will be f orwai zi. d t o NRR Senior Management. Please orovide any comments to

. lames ' alum. Sr Reactor Systems Engineer, BOP Systems Section. Plant Systems Branch tiy May lb,1996. 946 A1.1 at hwnt As stated (c A Ihadant 1 /1mmerman l' (. rut Clif lold i Gillef.pte 'j Tat um

&nMR 0%AI I od%

nipnN iile

. IX)CtlMf NI NAMI. G sii dDRAf1[0 em -w .n. 8,a.au.a ... r . e m winen..* - ... v . u. . ,

.2....,.e......a. ...

• w. e OffICL DM.WW NAML (,Mily' ihan pm /%

04/ /96 04/ /96 04/ /96 04/

i>A l t. 041Q oln OFFICIAL. Da.UMLNT COPY 3 y J. - $- 2 '

~

I 9

9902170282 990211 v - ,

PDR FOIA KOHN99-82 PDR , w .. g \ \ ) ' ff f-/ # 3 ~ l bf ~ b 1 x  !

b Attachment c.;

EQ Programatic Re,lew Summary Report (TAC M85648) 1.0 INTRODUC110N i

ln response to issues that were raised b, the Office of the inspector General (010) in a report dated August 12, 1992, the staff completed an assessment of the NRC fire protection program. As a result of this review, the staff-identIfled a number of weaknesses and made spectfic recomendations for ,

programmatir improvements in a report that was issued on february 27, 1993.

i in view of the woaknesses that were identified relative to the NRC (t're protection program, the staff concluded that other programs that are similar

In nature to fire protection, such as environmental qualification (EQ), should i also be reviewed to identify and correct any programmatic weaknesses that may exist.

Indopandent of the staff's reassessment of fire protection, [Q was identiflod An AD Area th.41 r0 Quired further reVlaW as a result of the staff'b activities i related to licenso renewal. As discus'ed in $[CY 93 049, a major concern related to 10 was whether the [Q requironants for older plants were adequate to support 1 cnnte renewal. Consequently, the staff concluded that i

j 4

differences n (0 requirements between older and newer plants constituted a potenttal generic issue which should be evaluated for backfit independent of

license renewal activities, i

in support of the licenso renewal initiative,.[Q testing of electric cables was performed by Sandia Hattonal Laboratories ($NL) under enntract with thn NRC, Some t r<.t s were performed to determine the ef fects of aging on typical clortric cable products used in nuclear power plants and other tests

, (unrelated in license r0neWal) were performed to assess the functionality of f damaged elect ric cables during loss of coolant accident (LOCA) conditions.

Atter acenlerAted aging, some of the environmentally qualified cabics either

' failed or erbibited marginal insulation resistance during accident simulation.

While some nf the SNL tests may have boon more severe than required by NRC' regulations, the test results raised Questions with respect to the J

e nv i ronmen t .i l qualif Icat ton and accident performance capabillt y of cert a'a arttittially aged electric cables [1 5).

I t, order to assess the significance of [0. the NRC staff performed a preliminary iisk scoping analysts on the potential impact of inadequate j equipment quallftcation on core damage frequency. The scope of the analysis wAs limited to core damage prevention Considering internal. events unly with postulated Iailuros of in-containment electrical equipment, with emphasis on electric cables. The major conclusions of the preliminary risk scoping analysis wrina (1) [Q f ailures could have significant risk impact if electrical tomponont rollabilities are reduced in the presence of a harsh environment. (2) the magnitude of the impact on core damage frequency is plant specific; and (3) the lack of reliability data and limitations in current probalillist n risk assessment models result in significant uncertainty. Based on the results of the preliminary risk sco)ing study, the staff concluded that

. a more det ailed LO . tsk assessment should as completed.

Thus, the t urront LQ lssue is ono . hat pertains to opeiating reactors, but I

4

• - , ee. w

1

(

lt's resolution is also important to the plant life extension initiative. The l

j staff issued an Environmental Qualification Task Action Plan (EQ-TAP) on June 16. 19}3, in order to define and coordinate the actions that are j necessary to rrsnive this issue. Action item 3 of the CQ.iAP lists those j

act ions that pert ain to the programmatic review of [0, which include:

3.a Reviow 1iconse Ronowal Back,round Informat100 3.h Review fire Protection keassessment Report 3.c Elicit Opinions from Others (Regions.. [Q Experts) l l 3.d Review Existing EQ Program Requirements i

3.e Review NRC Audit / Inspection Practices ,

3.f Review Licensee implementation Practices i

! 3.g i1 alize Review Results '

1his report is intended to address EQ Tel At.tlan item 3.g " finalize Review '

l i Results," and represents the nverall results of the staff's EQ program review.

i Section 2.0 of this report discusses the review methodology. Section 3.0 is a j summary of the potential issues that have b000 ldentified while completing (Q-1AP Action item 3.4 through 3.f. Section 4.0 provides the staf f's assessment i l and recommendat ions, and the conclusions are contained in Section 5.0.

j 7.0 M VIIW MilH00010GY l

The anal of the L0 program review was to take a ...rsh look at what has been i

done to addre n IQ issues and concerns and to identify whether any significant Issues or contorns currently exist that need to be resolved. The various l elements of the (Q program review are outlined under Action item 3 of the

] [0-TAP Idisruued above).

i

In completing IO-TAP Action items 3.a through 3,f, the goal was simply to identify potenttal issues that may exist, it is'important to roCognilo that the potential issues that were identified are rather speculative, since they

• were generated based on the specific information that was reviewed and no at tempt was made to pursue any of the po:itulated issues to determino if they had somehow been resolved by the staff. Also, in order to assure objectivity, LO-IAP Action items 3.a through 3.f were completed by individuals who wore not previously asso< lated with EQ and who were therefore not familiar with how speriftr f0 problems wore dealt with in the past.

Th- objectives of this review ([0-1AP Action item 3.g) are in: (a) co niidate th. potentia: 1ssues that were identified while completing f0 LAP ,

Ar 6on items 3.a through 3.f; (b) validate the issues through a peer review prra.e s s invnivinq individuils who are experience'l and knowledqcable in the i area of IQ (l e . NRC staff, contrattors, anti inclust ry expert 5); and (r) make  !

recommendattons f or f urther act ton.

.g.

3.0 IMUE

SUMMARY

AND CONSOLIDATION The programmatic review that was outilhed in'the'EQ-TAP was quite extensive and consequently, many potential issues were identified for further consideration. Appendix A is a consolidated listing of the pntential luuns that have been identified while completing (0 TAP Action items 3.a through 3.f. eliminating duplication between l'1e various EQ TAP reports that have been i s s ued l r,-l l) . The potential issues are organized into one of the following section', of the appendix, depending on the general nature of the issue:

A Scope / Applicability (page A 1) 11 [Q Methodology (page A-16)

( Current Status and Implemontation (page A-71)

D Assurance of Continund Qualification (page A-81)

I [quipmart-Rolated issues (page A-93) i NRC Oversight (page A 10.)

6. Miscellaneous Peer Review Comments (pageAll5)

The consolidated listlug of potential (Q'(ssues was reviewed by the NRC staff and others who are experienced and knowledgeable in 10, and the commnnt 5 that were rec eived during this poor review process are int.luded in the appondix along o th the listing of potential issues and problem statements to provide a balanced perspective, Appendix A also includes the staff's assessmont of the potenttal [0 issues that have boon identified, lhe f ollowing sumary of the informat ton contained in Appendix A provides the overall results of the EQ programatic review 1

a. $c opg/Appikabil i ty issues

• lhe [Q programmatic review found that inconsistencies exist relative to lhe scope and applicability of f 0 requirement s, in particui4r:

. single-failure requirements have not been applied consistently (e.g., single failuro criteria was not imposed for qualification of cold shutdown equipment, and the staff's resolution of TAP A-21 )

regarding the " super heat" effects of a main steam line break did not include single failure considerations);

- the need for sinale failure protection is not clear if the purpose of [0 is to protect against the occurrence of " common cause" er

" common mode" failures * ,

- being able to reach hot shutdown was a qualification factor for some plants while being Plc to reach cold shutdown was the l

)

consideratton for other plants; l

. qualification of mechanical equipment has not been addressed in the same fashion as qualification of alectrical equipment! and l

1

)

I electrical equioment located nutsido containmont is subject to the l

same E0 rigor as electrical equipment located insido containment.

b LQJcthodolQ21_ll M .1:

Many of the potential issues that were identified during the E0 program review are related to either: (a) Justification of the EQ methodology that has been imposed, or (b) resolution of technical issues related to )

qualification testing. With regard to (a), many facets of the methodology for establishing the initial qualification of electrical equipment for " harsh environment

• conditions evidently were not justified on a rigorous, technical level. In particular:

a curreni requirements vs.'what is reasonably possible within the state of-the-art capabilities;

- imposition of differont qualification standards depending on plant vintage;

• age conditioning / preconditioning requirements;

• use of generic temperature profiles; and e qualification basoJ on bulk vs. local temperatures:

• test margin requirements With regard to (b), resnlution of many teChn1 Cal issuns relative in qualification testing were identified for further review and follnw-up action, including:

- certifIcatton of testing laborateries: f definition of " worst case" electrical conditions; j

+

. leakage current considerations;

. MStB vs. LOCA qualification requiremonts;

. effects of long-term exposure to moisture:

a combustible gas and chlorinn formation effcLt%;

a dust effects;

• mechanical and flow-induced vibration effects;

. seismic and dynamic offeets; and l

+ fire scenario considerations.

Cutten1. 5.1atuLand.lcphmentatIon Ismi:  ;

c The staf f

• s review under EQ-TAP Actton Item 3.o [10) generally founii )

4 l

that the current status and implementation of CQ requirement: is well understood. While rek utiontof t many of-the administrative and minor EQ issues that were raise' with individual licensees may not have been well documented, licensees /pically recognized these as valid issues during meetings with the NRC staff and agreed to correct discrepancies of this nature. Therefore, for purposes of the [0 programHowever, review, the staff as a result considered this matter to be of min e importance.

of the evolving nature of EQ requirements, the staff may not have been entirely clear on what requirements were being suserseded during development and promulgation of the EQ rule and tsere may be some confusion in the industry on this point. Also, to the extent that

" generic qualification" (as in qualification of a generic type of insulation. for example) was credited, which evidently was the case for some Itcensees, further review and justification may be necessary since this approach was later found to be unacceptable.

d. A n u r a na nLI DAUng cd_Q.W A 1 i f I c a t .la n :

i lhe f 0 programmatic review found that current qualIficalion partice:. l (i.e., initial equipment qualificrtion certification without periodic monitoiing and assessment) may nu. ,rovide assurance of continued qualification of electrical equipment over time, recognizing that:

. substantial uncertainty exists in the qualification process.

especially in the ability to accurately project a *quallfled life;"

a installation. maintenance, and surveillance practices can have a 1 degrading effect on equipment qualification; and e unanticipated conditions and occurrences that take place over the life of the plant can have a negative effect on equipment qualification.

Also.

• reasons-to the contrary" for not upgrading replacement equipmeat to current requirements appear to be without merit since more than enough time has passed to allow licensees to establish In programs particular. " for reasons cualifying equipment to current requirements.

to the untrary" that are suspect include; e the item is part of a piece of equipment that was qualifi I as an assembly:

• the item was on hand as part of the utility's stock prior to Ichruary 22, 1983;

• replacement equipment qualtfled in accordance with the provisions of 10 CFR 50.49 does not exist; and

• the use of replacement equipment qualified in accordance with the provisions of 10 CFR 50.49 would have a significant probability of  ;

creating human factors problems.

5-l

l* , -

l 1

i e, [__outoment-Related Issugi:

During the EQ programmatic review, a number of equipment items were identified that may deserve furtM r review and consideration by the staff to assure that the existing level of qualification is adequate.

In partIcular:

. some electrical components such as penetrations and connector assemblies may be more etitical than electrical cables er;d1;on this b.v is, more attention may be warrar.t ed for these compor.

- moisture transmission through cracks in cable insulation or ir,tn the cable enre through diffusion may pose significant problems; j

1 i - solenoid valves, EQ barrier elements, equipment seals and vapor j

barriers (especially on plants.that are subject to the DCR

' Guidelinn), and epoxy compound used for pntting electrical l

penetrations may deserve further-scrutiny; and

. the use of polylmide Insulation (Kapton). Butyl rubber insulation, l wiewral wool insulation (espetta.ly in wet and humid environments), bonded jackets, coaxial cable, and terminal blocks may need to be better defined and/or justf(ted,

f. NE PerfoCmance and Oversi_cht issues:

hiven the evolving nature of EQ and the uncertainties that are involved, tt appears that NRC efforts to address and resolve this issue have not l

l been altogether sound, for example:

- the ability to determine 4 "qualifted life" by age conditioninr, I technioups seerrs highly questionable: l

- the imposition of different, more rigo ous standard < for the newer plants was not technically justified by the staff;

. the prescriptive regulatory posture that the staff Look relative to E0 seems counterproductive and may hate inhibited progress and innovative approaches for resolving this complex issue;

. the stif f failed to include allowances in the E0 rule f or the temporary removal of E0 barriers to facilitate maintenance, surveillance, and replacement activities; and

- no extension period beyond the end of quallited life (similar to the 25% extension that is typically allowed for completing surveillance requirements) was established to allow flexibility for equipment replacement during the next scheduled outage.

Additionally, based en the information that was reviewed under the T0-TAP. 11 appears that continued NRC oversight and follow throughFor to monitor and assure issue resolution have not been sufficient.

erAmple:

6-

~ __ _ _ __ _ . _ __ _ _ _ . _ _ _ __ _ _ __ _

ww- _

. NRC research activilles have not been very successful in addressing the E0 issues that were inittally identf(ted;

• NRC programs and initiatives have not been established to continually monitor progress and to. restructure, redirect, and

• rrove EQ program requirements as' appropriate; a NRC review and inspection programs have not been maintained in the t area of C0:

- reporting requirements have not been established to assure that emerging E0-related problems are referred to the NRC staff for consideration and appropriate resolution; and

• Generic Letter 88-07 does not require that licensees request an cramption from the E0 rule for equipment found to be vnqualified which appears to be incunsistent with 10 ffR $0.l? requirements.

Fin;11y. based on severe accident considerations and based on difficulties that have been observed in d aling with E0 problems, three additional issues were identifu.d which were not previo '.ly discussed in the LO TAP reports that have been is:,oad. Specifically:

j 9 lifJ1ML LPiflf_ Alt' l The curr ent interface that exists with the Nuclear Energy Institute (NE1) m ten does not allow for the objective and unbiased exchange of I

information on a purely technical level. Licensee representatives.

l industry representatives, and NRC staff are sometimes frustrated by the l

NEl interface when trying to resolve technical issues in a cooperative manner. It is important that this problem be resolved to allow for a cooperative effort in addressing fQ issues and other issues that are l

j important to the staff and the nuclear industry.

h. 1,ng_ller t ew Re s oon s i bil liv:

l Although LO deals primarily with electrical equipment and it is l typt !!y subject to electrical industry standaids, an electrical discip1'ne within the NRC is currently not assigned the lead l

1 responsibility for this area of review. Design considerations specific l

to elet t rical equipment that are important for equipment qualification are mo r apt to be overlooked or misunderstood under the current l

arrangement, and the staff is at a riear disadvantage when interfacing with in' inst ry expert s on impnrtant [0 issues that af fect elec tr ein) equipmer.' lhts assignment of st af f resources appears to be j inefftiient.

i. Iqq.htmeri burvivabilltv:

1 Equipment performance requiremeats for the advanced reactor designs i include survivability criteria ior severe accidents in addition to [0 I '

' requirements for design basis accidents. However equipment l survivability fnr severe accidents has not been addressed fo* operating re ac t on ,

f

n: x n.; .

4.0 [0 PROGRAM ASSESSMENT AND RECOMMENDATIONS To a large extent, the potential issues that were identified during the E0 pronrammatic review deal with limitations and uncertainties that exist in the qualifir st,;n process, and it appears that there are programmatic weaknesses and lingering technical and equipment-related concerns that require further attention. While much has been done overthe past 25 years to try to better understand E0. ;. cicar strategy for the long-term resolution of E0 issues is not readily ar"irent and existing requirements have not evolved over theThe years t o a r c tn:,t f or t h a uncertainties that are inherent in the E0 prncess.

discussion and specific recommendations that follow are intend:d to es'.ablish i a stra'.egy f nr assuring qualification of electrical equipment over its .

insta lled li f etirre. This is an integrated approach and as such, sow of the recormnended actions are desirable program enhancements while others are cnnsidered necessary in order to assure an adequate level of equipment qualif hatton. Those recommendations that are considered to be desirable progran, enhancements are listed in brackets.

In addition to the regulatory perspective that is preaented by this report, it is important to recognize that E0 programmatic improvements may also be possible and of substantial benefit from .,I industry parspective, fnr ern ple. it may be possible to better (ccus E0 requirements by using graded QA rrquirements and PRA techniquest better definition of the role that single hilure plays relative to E0 cnuld result in better-directed ovalification requir ments: and stress testing may prove to be a viable alternativc to preaging. Therefore, in addition to the specific recommendations that are d1;cuued below, the NRC staff should actively support and encourage industry initiatives to iriprove and streamline E0 requirements, methnds, and practices.

4.1 General Considerations in purming [0 programatic improvements and addressing specific equipment concer ns, it is isnportan' to make use of inforrmation that has been developed throwsh research activities, operating plant experience, anr' the advire of ,

l quali f ied experts. The NRC staff should work closely with industry expcets in addressing current and future E0 issues and equipment concerns, taking full 4

1 advantage of other init iatives such as tl.e mainteri...ccvery rule and graded Quality costly for )

auor nce as appropriate. [0 requirements tend to t licen.ees and any changes to existing program requirements or further expe< t itions of licensees should be well justified and properly connunicated.

lhe 'ollowing recomendations ire directed toward satisfying these fundamental concctts: l

[ RECOMMENDATION ))

lit may be possible to ar' dress many of the potential issues that nave been identified by reviewing and better understanding past research ef forts and E0 information that has been aeveloped over the years, and this approach should be pursued before considering nther alternatives. Additional research should be perf rmed only 30 (a) there is a well defined need for additional informatinn; (b) there is a good likelihood that tha desired information will l

j be obtained; and (c) the cost is justified in terms of the evpacted benefit , to public health and safety.)

RECOMMENDATION 2 The NRC .itaff should review the results of past and ongoing E0 research ef forts, qualification tes' results and practices, and other E0 information, and maintain in up-to-date data base containing this in'ormation in order to: (a) better manage, catalogue, and share E0 information and advances in technology:

(b) identify specific issues that may deserve additional research arJ resolution: (c) provide a basis for resolving EQ concerns; and (d) better focus NRC staff and industry resources.

[ RECOMMENDATION 3)

{ A functional interf ace between the NRC and indust ry should be established for addressing E0 issues and concerns t'. a cooperative and tecnnically sound fashion. Since the existing interface 4fth NE! ttnds to inhibit the exchange of information and ideas between indust,y exper ts and the NRC staf ,,1RC 1.:anagement should either resolve this problem or establish other avenues for industry participation. In support of this initiative, this final report on the EQ programmatic review, as well as the ether reports that havc been completed under E0-IAP Ar. tion item 3 [6-!!), should be made available to the general public.)

RECOMMENDATION 4 While it is important to alert the industry of potential generic problems with equipment qualification, NRC expectations of litensee actions should be communicated through issuance of be Bulletins or Generic tetters. Information Notices should in' used as a vehicle for implicitly suggesting that licensees tould take some sort of action.

^:

LLhlt The E0 rule (i.e., 10 CFR 50.49) was established before much of For theexample, research on EQ was completed and the rule is outdated in this respect. '

the rule requires that a qualified life be determined as part of the equipment qualification process, but such a determination is theoretical, the effects of many deurading 4.fluences cannot be accelerated, and 'he accuracy nf sucn a determination is unknown. A rule that is more generai would be better suited to the t heoretic al nature of E0 and the uncertainties that are involved, and would more readily allow the NRC staff and industry experts to pursue other Also, while the E0 approachts and methodologies for addressing EQ concerns.

rule doe', not preciude allowed outage times for E0 barriers and equipment, guiaance has not been established in this area.

[ RECOMMENDATION $}

llhe NRC staff should make changes to 10 LFR 50.49 as appropriate 9-

i 1

in order to facilitate and encourage industry initiatives to improve the EQ process. Specific snethodologies, techniques, and details that are accept:ble to the NRC staff for establishing and maintaining EQ should be provided through Regulatory Guides, the Standard Review Plan, NUREGs, and othcr documents where changes can easily be made as more information becomes available and advances are made in the state of EQ technology.

Guidance should also be establishe ' to address uperational considerations, such as allowed nutage times for EQ equipment and barriers.)

a.3 E0 Procramraatic Weaknesses The potential issues that were identified during the EQ progransnatic review indicate that EQ requirements are not commensurate with limitations and uncertatnties that exist in the qualification process, for example, the assignment or determination of a qualified life is theoretical, the effects of many degrading influences cannot be accelerated, and the accuracy of such a determinatinn is unknown.

There is also uncertainty as to how well the qualification requirements accuunt for such things as equipment installe' ion, normal wear and tear, and periodic maintenance and surveillance activitles.

Another un.'ertainty factor that bears on Fe icqg term qualification of elec t rical equipment is that dif ferent qualification <tandards h ,ve been impo.ed over time depending on plant vintage, for~ea mple, the older plants were not required to preage electrical equipment prior to qualification tasting. Therefore, in order to account for the various limitation: and ur:crtainties that exist and to provice adequate assurance of EQ for eier t rical equil' ment over time, additional measures must be taken, RECOMMENDATION 6 in nrder to compensate for the various limitations and uncertainties that exist relative to equipment qualification, to provide assurance of continued qualification over time, and to identify and correct any EQ deficiencier that may exist.

additional EQ programmatic requirements are nece'.sary, including:

. periodic condition and environmental monitnring of electrical equipment, and

• rigorous identification, assessment, resolution, trending and reporting of equipment qualification problems that occur.

With regard to condition monitoring, over the next several years

.tha NRC staff should develop, in concert with industry representatives, guidance for the application of cnn* **

• monitorirq techniques, RECOMMENDATION 7 The NRC staff should establish a more focused program of EQ j oversight by:

. establishing a NRC Headquarters focal point responsible for 1

. J

l I

identifying, monitoring, trending, ataloguing, and resolving E0 concerns on a continuing basis, and to assure consistency of regulaticn from one licensee to another;

• maintaining (0 guidance documents (including the SRP) up to date based on at'vances that are made through research and Indo;try initiatives; a promulgating information and guidance to licensees and the l' NRC staff as appropriate:

+ establishing specific reporting requirements for equipment deficiencies that indicate qualification expectations have not been satisfied for the given environment so that the staff will be better informed of E0 problems that are being identified and better able to recognize and resolve emerging t0 issues; and

- better anaging and directing E0 research activities.

I With regard to the NRC focal ptint, consideration should be given to assigning the lead respnnsibility for EQ of electrical coulpment to an electrical discipline. Also. In the area of j research, the existing NRC plan for performing EQ research (It}

55ould be adjusted ta incorporate the re'ults of this review. For example, in addition to the need to establish condition monitoring methods and techniques (see Recomendation 6. above), further assessment is needed for a number of specific technical and equipment related E0 concerns (discussed in Section 4.4. below).

Also, recognizing that much more emphasis must be placed on periodic condition monitoring to assure continued equipment coalification, cxtensive efforts and expenditure of resources to correlato artificial aging with natural aging may not be warranted.  !

RECOMMENDA110N 8 Cert ification of E0 testing laboratories in acc.,rdance with gener 411y accepted non nuclear practices (e.g.. ASTM or ASME certification) along with nuclear QA standards is recommended to assure thAt E0 testing is properly and consistently performed l throughout the industry.

[PECOFMENDATION 9) 8 5 truly necessary for licensees to upgrade

[to the extent that it is the more rigoroJs E0 requiremants contained in the EQ Rule.

more appropriate ' reasons to the contrary

• should be established than those that are currently listed 10 Regulatory Guide 1.89.  !

However. resolution of this coe'ern should be coordinated with induttry initiatives to impiave the [Q process.] l I

i

_ . . . . _ u

(

f

)

i  !

4 l  !

j 4.4 Yar.Nus lethrdf,ALautlaul-at ko tiedlonr.saa .

i

! In addition to the programmatic wenitnesset that were identifled during t' e [Q l i,

l progrannatir review, a number of technical and eQuirwnt-related concerns were i; identifled for further consideration. Most of these are not issues in the sense that problems are known to exist but rather, they are speculative l' concerns that sten' from the early evolut' n of [0 requirements and the various uncertaintles that are associated with the qualification process. The 1 l

i following recomendations are for concerns of this nature: '

f RfCOMMENDATION 10 The NRC staff should determine and document to what extent slagle '

I failure considerations are applicable to [Q. ,

1 RECOMMENVATION 11  ;

. l i lhe staff should determine and document to what extent  !

l Qualification of equipment for achievino cold shutdown is truly  :

i nece nary (irrespective o' licensing bases) to assure that a  !

sa'et, concern does not exist for t ,se plants that were not j recut .1 to qualify equipment necessary to achieve a cold shutdown condt, ton

{ l l

l RECOMMCNDAT!bN 12  !

1hc NWC sta' should assure that identification and resolution of i signi' trant L0 concerns have been addressed within the scope of '

l i the i PI initiative. l i

i j RECOMMENDATION 13 The following concerns should be further assessed by quallfled [Q j

cupert s and the NRC staff to determine whether or not and to what g extent additional re<olution is warranted

a, Qualification of coid th9tdowa egolpme.. and resolution of 1 Ar A.2) may not be sufficient if single failure

{ c ons iderat ions apply (see Recommendat ion 10) .

l {

I

b. The use of ' excess margin

• to j'Jstify the short duration ]

l.0CA tests that were allowed for the 00R Culdelines plants j may not be suf ficient to assure equipeat qualif tration. '

l i c r.eneric temperature profiles that were allowed for some PVR$

l and BWRs were not fully ,)ustified and may not provide ,

j suf ficient assurance of qualification.

i j d. Resolutton of IAP A 2) may not have been entirely '

appropriate if resolution of the " velocity profila* is .

dependent on the resolution of MR D qualificattor for DOR l

r.uidelines plants since the MSI.B qualification issue was not ,

i e

h . ll .

l

.,~

fullyaddressed;andthe've)Ec{typrofile*representsadynaale effect that may not have been' addressed in term of fQ.-

c. 1he use of " generic qualification

• may not provide sufficient assurance of equipment qualification in those In'tances where this approach was used, f.

The resolution of other issues that were handled separately froa [Q but that could Impact equipment qualification, such as the issues of mechanical and flow induced vibration, seismic effects, dynamic effects, etc. may have allowed [Q requirements to be comproelsed, o, [quipment survivability for severe accidents (requirement f or advanced reactors) has not been addressed for operating reactors.

h Additlonal resolu'. ton of the following operating and accident considerations may be needed to assure equipmant qualification:

. leakage currents and momentary electrical effects:

. hydrogen burn scenariott

. radiation and temperature stratification effects:

. long-term exposure to moisture:

. continuous submergence prior to the LOCA:

. the effects of fire on EQt a combustible gas and Chlorine formation effec (s;

. use of bulk vs. local temperatures:

. adequacy of M5LB qualification for DOR Cuidelines plants; and

. equipment interface problems, i Additional assurance of qualification may tw needed for the following items:

. elec trical penetrat ion:, and connec t or autobiles:

. solenoid valves:

. [0 barrier elements; e seals and vapor barriers;

. epoxy compounds:

. moisture intrusion through cracks:

. polyimtde insulation (Kapton),

. Butyl rubber insulationi

. mineral wool insulation (especially in wet environments);

e bonded jackets;

. coaxial cable

• and

. terminalblocks.

4.6 01her_.ConsiderAltana A fc= questions were raised as a result of the [Q programmatic review that don relate specifica11y to [Q. but may warrant clarification or further

-D- 1 d

• • ~l assessment hy the NRC staff. The following reconnendations are for concerns 4

of this nature: '

[RECOMMENOATION 14)  ;

lThe process required by GL 88-07 for addressing situations where equipment is determined to be ungt Ilfied does not require that licensees seek an exemption from the EQ rule. The staff should

' determine whether the GL 88-07 process is appropriate given the >

exemption requirements stated by 10 CfR 50.12. and provide

" guidance as deemed necessary.) ,

[ RECOMMENDATION 15}

(there 15 a marked difference in requirements that were imposed for f0 of electrical equipment versus what was required for f0 of

' mechanical equipment, and technical justification should be ,

established for the different standards' and the different

' approaches that were allowed'by the staff.- for example, [Q of machtnical equipment did not involve prescriptive regulation, a  !

l det.11ed program review, ano confica tor) on-site inspection.]

I [ RECOMMENDATION 16) lThe current ronducive interf to the ace that exists cocterative exchange between the NRCand of information andideas Nf! 15 not that is needed for thf resolution of Complex technical issues.

This problem between the NRC and NEl should be corrected or some nther industry interf ace needs to be estabitshed that will allow cooperative efforts to be meaningful and productive.]

5.0 (ONCLU510NS 4

The [0 prograrnat ic review identified many potential issues, ranging from uncertainties associated with the qualification process to potential equipment ~~ '

vulnerabilities. 11 must be emphasized that these are potential issues, some of whir #, may be readily dismissed based on more in-depth review or expert

. judgement. Also, while it is important to recogn12s and appreciate the .

various potential [0 issues that have been identified during the EQ .

programatic review. it is also important to recognize the limitations that exist in the state of technology and in the ability to address and resolve  ;

these issues. Consequently, resolution of EQ issues in general requires a

' good understanding of the overall strategy for addreul.ig EQ on a prograssmatic level, an understanding of what can reasonably be achieved, and the use of ,

good judgement in deciding how to proceed on a given issue. ,

from a program perspective, the results of this review indicate that a strategy.does not currently exist for assuring qualification of electrical i equipment on a long-term basis. Given the uncertainties that exist, the l' current requirement of initial [Q certification must be supplemented with i additional requirements for ongoing assessment, validation, and NRC oversight.

in particular, program enhancements are needed that include: (a) periodic condliion monitoring of [Q equipment; (b) rigorous identification, assessment.

resnlutton, trending, and reporting of equipment aualification problems that l J

J I

i

By including these as occur: and (c) a structured program of NRC oversight.

fundamental elements of EQ program requirements, uncertainties associated with i the initial qualification process and questions about the specific methodologies that were used become much less important.

Many of the specific c.oncerns that have been identified can most likely be addressed by reviewing and better und rstanding past research efforts and (Q information that has been developed over the years, and this approach ahould be pursued before considering other alternatives. Additional research should be performed only if: (a) there is a well definet' need for additional information. (b) there is a good likelihood that the desired information will be obtained, and (c) the cost is justified in terms of the expected benefits to public health and safety. Additionally, the existing NRC plan for '

performing f0 research should be adjusted to incorporate the results of this review for example, l12)in addition to the need to establish condition '

monitoring methods and techniques, further assessment is needed for a number of specific technical and equipment related fQ concerns, Also, recognizing that much more emphasis must be placed on periodic condition monitoring to assure continued eq"ipment qualificatlon, extensive efforts and expenditure of resources t o correlate artificial aging with natural aging may not be warrante. ..

Aside from t he recommendations contained in this report, addltlonal adju s t me nt s in existing [0 requirements may be possible and of substantial benefit to t he nuclear industry. This is especially true reccgnizing that more empha='s is needed on maintaining equipment qualification over time and some " trade offs

• may be appropriate. For example, it may be possible to j

better focu. t0 requirements by using PRA techniques, better definition of the role that ' ingle failure plays relative to EQ could result in better+ directed ovalification requirements, and stress htting may prove to be a viable alternative to preaging. lhe NRC staff should encourage and be supportive of industry initiatives to: (a) improve and streamline EQ requirements, methods.

and prac t c es based on the knowledge that has been developed over the last ,two i

decades: and (b) use PRA and other techniques to better focus EQ requirements

' and to help place fQ issues in proper perspective. Changes in the regulation (10 CFR 50 49) should be initiated by the staff to facilitate this approach.

in pursuing the reconrnendations contained in this repsrt and in addressing !Q l issdes i .: doneral, the NRC staff should work closely with industry experts, ,

t aking f ulb adwant age of other initiatives such as the saintenance rule and graded quality assurance as appropriate, Because the existing interface with N[I seems to inhibit this sort of cooperative effort, it is important that NRC l

, management

.ither resolve this problem with NC! or establish other avenues for industry part.tc1patlon.

The programat it weakeasses and equipment-related concerns that were identifted as a result F the [Q program review do not mean that equipment is

. currently not qualifico. Rather, the results of this review indica 9 that:

(4) some adjustments are needed to better assure continued qualificatton over the prnjeried lifetime of the equipmcat that is qualif ted: (b) some of the past 10 rom erns require further review to assure that resolution is complete, and (c) a st ructured program of on-going NRC involvement and oversight is needed. Unt61 such time that specific equipment qualification deficiencies are identified, existing qualification it assured by the inillal qualification 1 -

l

)

i testing that was performed, twenty-five years of research. and equipment performance and operating experience. None of the concerns that were identified during this review was conside % to be an immediate safety problem.

Principal Contributor: J . Tattn, NRR/SP' '4 i

1 i

1 l

I

- is .

l

R[fER[NCIS NURIG/CR-5772. " Aging, Condition Monitoring, and loss of Coolant 1

Crosslinked Accident (LOCA) Tests of Class il [lectrical Cables ,

Polyolefin Cables.* Volume 1. August 1992 2

NUPIC/CR-5772. " Aging. Condition Mc 6toring. and less of ceolant Acc ident (LOCA) lests of Class 1[ [lectrical Cables -- [thylene Propylene Rubber Cablet.* Volume 2. Nevester 1992 3

NUR[G/CR-5772.

• Aging, Condition MoniterlNg, Sad less of*CoelAnt Av ident (LOCA) Tests of Class if Clottrical Cables --

Misc ellaneous Cat'l* Types.* Volume 3.. Novee6er 1992 4

NRC Information Notice (IM) 92-81,

• Potential Deficiency of f lec t ical Cables with Bonded Hypalon Jackett.* Decem6er 11.1992 NRf Information MollCe 93-33.

• Potential Deficiency of Certain Llan it Instrumentation and Control Cables.* April 18. 1993  :

6 Mc=uandue from M. J. Virglito tw .i C. ahodant.

• Review of to  ;

inf ormation Related to Plant license Aenewal ((0.f AP Act ion Itew 3.41 (I Af M8$648),* March 22. 1994 L

l 7 Memorandue from M. J. Ytrgilio to A. C. Thadant. " Review of fire Prnt ec t ion Reassessment Report ((Q 1 AP Action flee 3. b) (I AC M 544 8 ) .

• Ns> 10 1994 Memr andum f rom M. J. Vir9 ilio to A. C. Thadant. *Results of the ivrver ll 8 -

of f0 (rpert$ (f 0. i AF Ac t non I t en 3. c ) ( f AC NBMas ) .

• June 12. 1994 9 Mt orandum from G. M. Holahan to A. C. thadant. " Review of IristIng I nv t ronmental Qual i f sca t ion Prograe Requi rement s (f Q. T AP Ac t len Itc 3.d; (1 AC M85648).* Decce6*r 21.1994 4

10 Menm aredue f rom G. M. Holahan to A. C. Thadant. *Results of the Review of NRC Audti/Inspectlen Practices 'CQ. TAP ltem 3.e) (TAC 4%44).* dated i l

April 14 1995 le Mecorandum f rom C. M. Holahan to A. C. thadant . *t teensee laplement at ten Prat tices Relative to [Q ([Q.IAP ACtlen llee 3.f) (IAC 4%44).*

Oc t utie r 7. 1994

!? Me mrandum from t. C. Shao to A. C. thadant. *Research Progree Plan for i n v i ronment a l Ov al i f ic at ion of [lec t r ic t ou t paent .

• dat ed Jul y 1. 1994 i

- 17 i

I

l i

AppfhDlX A i potential f0 litual  ;

p .,n ol utat eu l ist inn with Peu Rev tew comment 6 and St af f Alte. . -ont )

O Mupr/ApplKabilitys.oldQ ,-

1. On.)ifleation of componenti ciner!!hnn"cablel has not been riuniounty a+traued. for exam)le, relearch in only jult haginninQ to A uut6 the a.laquary nf (4 for R1 1,91 functions and very little (Q rates, li hat i,ron perintmed on pressure twitchen, AID 6, pressure transmittri., and valvu Opst r Al or t ,

l' n L fte d c W.l o m in11 1 a Ounlification of com1onent6 other than cablen have been vigorously 4 Addronted. During i16 (Q inspections that were conducto,I In Reginn 3, all components required to function in a harsh environment were required to be quellfled, Qualification tanting was nuisd for valve operators, transmittert, prellure switchet, etc, llowever, when tonkiderint Ilcenta renewal and the possibility of excestliny the 40 year qualified life, it ilnel appear thAt cablet hAye recolVed more reVleW/relearch than other com)onents, Al for R0 1.97, only certain components are 'aquired in je f4 quellfled to the halts Jor the concern 11 not ,ia4r, i, why do we need research? Unitt have been quellfled, ,

r depen' n' on the lhe first lentence rigorously":afdrelled.

Interpretationof" can be extremeh proVocativo$e D

late 19H)ot Ifled" enmpnnnnLE in itlicen6ee andardt(Q we In(HRC spections. Oltto-for llaff) vendor law 4 : lotIf' ~ of insptctions. "qWa.

standardt are different now, thil-derument thould clarity Otherwl66 you'll have readert laying that nothing Il qualifluil, .

it I disaures, This il not considered an open (0 titus, L

ilett, the industry has addrelled: qualification in accoiilance with requirements of the Illi llandards and HRC documentl, it they were '

acceptable as rigorcun for cablel, then the other tiem6 are In the tame Isague, and in my opinion, even bettee in 60me ta w . (e.g.,

transmitters, batteriet nome valves). Remember that tha industry

- st andardt On qualifical on f ar Vartout flattet of equipnient have had HRt in>ut, and that NRC has not thus far (with a few uncepttont Indicated any tertout dilagreement s with Ibnus  ;

tiendard6 y linutog Regulatory Outdel, a6 15 the common NRC ,

' prartice, l second, 1f the concern relatal to the extent of NRC sponsnrod research on items other than cablel i.then we must keep in minri that landia. Oakridge, '10. and if I am correct ev performedresearchonmaryitemsotherthancables.enWriehAt

l

. . . tw .z h

Ihled and last, it it time to lik if we have not rea.ned a point uf diminishing returnt in (fl research by performir,1 more agin0 and 10CA tents, lime alal again.relearch conducted to ilate have Dninted to the crying need-for. improvements in the 'ethodt used for equipment surveillance and condition monitorinu. In other wordt. )redictive maintenar e and condition monitoring techniques thould is uled to lupplement-(Q.

I believe that we munt focus our scarte rateurret Therefore,dartlanding towardt un the results of the work alrea.iy portormed and using them beneficially to develop and implomani effective conditton monitorin0 methods,

n. The concern il valid,

f. I am not clear how " rigorously" it defined. What kind of [0 research needl to be performed?

Q.

reletreh performed at landini retuarfu wal DuringtheHRCfQ[Qprotell.

performed on the This included.trtiftrial agleg, radiation ilmulat ons and 1.00A testing, lavoral eq.ipment types were telled including connectorte; penetrations, and sniennld Valve 6.

h. Industry responllblIith, ~df thr'a' bout two ' decades til retsarch' fin ann the expaniliture of many other countrial millions of dolltrl atwewell 48 in far i e 61111 the UIA)from an6werinu all the quellI0ntaboutcableQuaification, it would he luilie to undiertake research on the components Illted in this paragraph as a way of learning how to qualify them. At indicated in comments that follow on other.16 sues, It is time to'take tinik of what we have learned and look for ways to ilmplify qualifica Ion without reducing the alturance of safety.

1. The renuirements for environmental qualification of electrical equipm9nt importaM to 6afety fnr nufl aar Dewer planti li codified in 10 CfR lo.49 (also known as the final rule on eiwironmental qualtfication?, and it dated January 21, 1903. In ar.cordance with areferencedJnotnoteinparagraph'(h)(3)10CfRlin49,andin accordance with the HR( Itaff n findings based on information provided by utilities, componenti provIded in planti are quellfled and adequate to provide Regulatory Guide (RO) 1.97 runctions.

Ilowever, additional research in thin area 16 welconi . Remember, licenleen are responsible 'er qualifying all equipment that il required to be quallfled, ;herefore statementi luch at this one should be directed to licensees.

Sta,f.L Asse16 ment i Rosearch it uood in a point, but there are limitat.an' S. what (an be ACComplithod. The NRC staff thould alture that the retuitt of past research efforts arG underltond and related to spectfl6 equipmunt appilt allont, at appropriale, Any further retsarch thouht; a) De haged A?

s

f 1

on a well defined need for addillonal information, b) bn p.rtund only ll thera is a gnad likelihood that las dellrati in(nemation weit un utilained. and c) lis purtusa only if the tunt of rdtsarth '- pattifled in thu tiaff termt of the evps(tod hansfit to p(ublic health

.tuta n will. lho view sapretted by omment I,h that "...it and s af ety . Ilms in t ake slot'k of what w6 have learnt i and look for wayt to i 'plif y elualifitatinn without reducin0 the alturance of safety

• as... industry ,

init iat iva in this regard thquld be encouraged. The staff should i assure that full advantage 16 taken._of<eperallng plant skenrienes and PRA inf amation, equipment perforritance, condition and env o.inment monlinrino, root cauts atlettment..and trending of informaiinn in order in identif y .ind correct f 0 defielencies on an nncolnu hat-1 fven though the Standard Review Plan tuQQatti that NUR[G'utrtA, RG l.A9.

l And liff ,it) may be applicable for quillfication of mechano al ainsipment, specific guidance ha6 not boon providad in thi regned, I

har Rev.gnggigt

! a, lhe need for addittoral guldsnt' i t not clear unists 'ho issue

( dealt WIth ilfe extentioni h, I dliagree, llow about the AlH( (H( Committe docunirnis (QR and QV I terist)? Thens have been under (syslopment ove the past levens years and were finally isnued in 1994, Tha{ irre in u lfically written to Addrels MsChilnifal (quipment Qua ilication (H( Q) , lt is n.V understanding that NRC has participated in their

.f ave lopment ,

On a different note, ist un recognlle that the industr/ offert in ottablithin0 MLQ in planti IIC8nied to operate since 1980 have  ;

cnnelunively e6tabillhed that the effort holl6 down to tdantf(ying and evaluating (mottly by analyll6 onlyi nonmetallla . Th6 Informallen from this analplit it uned 40 eltablish replacement intervals for the sof t parLt. Further, recently the industry has applied for exemptioni from the requirement to treat Hf4 at 4 1sparate program and integrating them lato thn preventive the NRC hat acquiatcod with maintenance programt.

thste exemption requesti, Apparently,I if so, am not tui's what the ballt la fnr thin titue listing, c, lhin li a valid intuni

1. I believe that there in an AIMC, f 0 document that retort tn i i

mechanteal equipment,

o. isvaral of the H10L plants were required to have mechanical (0 prograat during the 1980't. The basic findings were that the most isnsttive agina components were seals (0 Rings. Gaskett, and dlaphragmn). Leal replacement programt were attablished, Mechanical itemt such as inubbers are required to be periodically i' tested, tome mechanica) actuatort have been CQ qualtfled tuch at pneumatic and hydraulic actuator 6 Motor operators, such at A=3

, , ,,. ,q.3 _ ;g

\ Yr< ' ; ;; [:# ff lMg' 3 +

?

f.h. .

2p sm :i; .c % ' ' 'G.t .

t Limitorque, Rotorque)and lTTl Ware qualified.. Generic Letter 85-03 r$

and 89-10 have required 4 version of qualification by assuring "

that MOV's have their operability demonstrated during worst cas.

d flow, including blowdown conditions,  !

s f.

Merits analytical resolution (i.e., analysis of existing information is warranted to reach. resolution). y As is indicat ed in your statement - the referenced documents 9 e (NUREG 05F0, RG 1.89 anf IEEE 323-1974) may be appilesble Nevertheless, for the NRC j .

qualificacn of mechanical equipment.- .i staff has orovided specific guidance for developing programs A copy for of 4 environme :a1 qualification of mechanical equipment. ,

that guir race followst. _F Although U1ere are nondetailed requirements for mechanical i GDC 1, "Qut11ty Stand &rdsiandl Records," and 4 g equipmen't Environmenal and Hit (jle DesigniBaseffr-and! Appendix B to d? ..e 10 CFx 50, " Quality Alturance~.Criterl&5for; Nuclear Power Plants

., p 4 and fucl bprocessing Plants" (Section.Ill..* Design Control," and ; ..

XVil, "0L ilty Assurance Records"). contain the following I requirome .s relatad:to equipment' qualification:  %

Comp: 1cnts shall be designed to be compatible with the '

1 post ,t sted environmental conditions, including those '

asso lated with.LOCAs.

i I

Measui i shall be established for"the selection and review for sunability of application ~of materials, parts and 4 ,

cQuipment that are essential to safety-related functinns. e Design control measures shall be established for verifying [:

the adequacy of design. 7 t

Equipment qualification records shall be maintained and shall include the results of tests and materials analyses, formechanicalequipment"rLthPstaffUreviewsillconcentrate:on g materials which aretsensitiveito. environmental effects t diaphra ms, etc. A review and evaluation should be performed by the app icant thatLincludes'the following:

Identification of safety related mechanical equipment J (1) located in harsh environment areas, including required operating time, 3.

Identification of non-metallic subcomponents of this 4 (2) equipment. .

(3) Identification of the environmental conditions this equipment must be qualified for. The environments defined.i in the electrical equipment program are also appilcable to mechanical equipment, v

A-4 I

i i

k 1 1 (4) Identification of non metallic. material capabilities, f

• 1 l (5) Evaluation of envie onmejblfeffects; 2

^ j lilt. film *med :

There is .i marked dif ference in tne'$taff's handling of (Q for f electrical equipment as compared to mechanical equipment. While the /

peer revi"w comments indicate that specific guidance has been 3 establishiW for C0 of machanical equipment, it was'not doa.e in the same prescript ive manner as was thought to be necessary for electrical l ij equipment and it appears that the'same level of' effort has not been &

placed on. .nsuring that EQ of mechanical ~ equipment-is adequate for operatinq reactors. It appears that the NRC staff has established a  ;'

l less rignious qualification standard for mechanical equipment and focused staff attention is necessary to assure that mechanical equipment is adequat"ly qualtflod. Specific guidance in this regard should be

.Stabli',h. I and promulgatud ta the industry. An approach dtfferent fr w what was e equired for EQ of electrical equipment (e.g.. prescrtptive regulatior detailed program rr 'ew. ard confirmatory on-site inspa tino) should be fully justiteed. ,

3. Under th" surrent requirements, active and passive EQ equipment are y lumped t" o'ther in the development of performance requirements, design requirement s. maintenance programs _and. safety priorities, which may not be entirt'r appropriate. ,/4 ,

Peer peview Comment s:

a. I am not sure of the basis for this . issue,

b. This was done for conservatism.

c. I agree. A few utilities make '.he distinction. This is a good example of where the utility must address required service functlon, not just point at the vendor,

d. No opinion () am not sure 5,-nat the issue is here). Safety funttional performance of an equipment is defined taking into account its interfaces with passive items such as cables, '

te m nattons etc. During the qualification process, this functional capability is demor.strated either in a single test program or through multiple tests and analysis programs. The concerr. regarding safety priorities, and the maintenance area needs to be better defined. The safety priority of the protective / safety function performed by an equipment item governs the safety priorities for its interfaces, be they passive or active. As for maintenance, to the: should extent-there are maintenance have been, (and I know attributes for passive items, they' they generally are) addressed in a utility's maintenance program, Active and passive equipmentjare important " lumping" issue is e.

not.

b

h

^ -

y . . Mpg 2 .. , . .

f. I am not clear how safety priorities are an issue.

9 Successful operation of passive equipment ( cables, terminal

' blocks. breakers, etc.) is often necessary, permitting the operation of active equipment. Thus, it seems appropriate for qualification and document:*lon of its performance, maintenance.

and safety priorities.

l

b. Merits analytical resolution (i.e., analysis of existing ii.iormation is warranted to reach resolution).

i. TheEQrequirementisth'atcomponents.(i.e.componentswithinthe scope of 10 CFR 50,49);must'be able;to perform their required functions when called u>on for as.long as' required. This requirement applies to )oth active'and; passive components.

Licensees have found'.(thd<the NRC,hasiaccepted) various ways to l

demonstrate that equipment in'their EQ lirograms meets this requiremont, There is no specific requirement to lump or ont lump together performance requirements design requirements, maintenance programs crisafo'v ?r)crities. If there are inappropriate lumping of requirements, specific identification and j eIabcratico on such requirements are welcome, l

113 i l.Alltisflul:

from a (afety perspective, full advantage should be taken of operating dition l ,oporience and PRA information, equipment performance, con

> p l .m t and environment monitoring, root ca;,ise assessment, and trending of

' int '.. mat inn in order to identify and correct any EQ def tciencies, that Implementation of the maintenance rule may ,*xist on an ongoing basis.

' wi nelp to resolve this problet tor active components, and the staff should init l Ate action to include hisCtrical equ'pment Within Also,the scope given the of the maintenance rule to better address this concern.

i advances that have been made.;in our understanding of E0 over the past 25 years, cost-effective improvements may be possible In the application of 5 fnrequirementsandtheNRC}staffshouldbe,receptivetoproposed (a)' teveloped as an industry y channes in the EQ methodologyrthat ares Initiattve, and (b) demonstrated ~tosbe technically justified.

4

[0 requirements seem unreasonable for equipment located outside cont ainment anJ exposed to short-term steam condit tons and/or radiation-nnly harsh environments when.considering the low core damage .

cm.t ribution of this equipment compared to critical components that are PRA implications /EQ screening criteria may lot ated inside containment.

be helpful in establishing whether or not and to what extent EQ is truly ne(essary for a given component.

Peer Review Comments:

a. E0 requirements for equipment located outside containment that is relled upon to function during and/or following a DDE are reasonable since the qualification requirements for those components do not include harsher environments than they will see

'A6-

$~

(

1 l

n .

l g 'J ?. y 1 durin; the event. If the equipment is.not relied upon, then it should not be in the EQ program.~ PRA~can:be used to assess the

level of reliance out this should be done with caution and should

~

i include a review to ensure that the plant's safety analysis

remains valid.

4

b. Maybe the criteria for master-listing should be revisited, in ]

i 1 general, I don't feel that-qualification for outside containment 1

! is unerous.

i i c. I agree. This is long overdue and a concerted effort has the '

potential to narrow the scope of equipment included in EO l ~

programs. This is true for both in-containment ano out-of-i cnntainment equipment populations. Also, note that just because

! an ec,uipment item is located in an area outside the containment, and that it may only be exposed to short duration environmental

{ extremes, it cannot be automatically concluded that it doesn't. _

l

- have to be qualified. Some equipment / components have the l

potential to .xperience common cause failures even under these conditions. - .

l g; .- -

rurthor. I believe that We'do now have sufficient real world experien' data to permit a meaningful assessment of equipment l fallure ..tes in nuclear p'lant environments. This should be used when performing the PRA. '

i

d. this is a valid issue; PRA is a good screening tool,

e. ihe philosophy utilized to date has been for equipment needing to i

' unction or fall safe in~ harsh environments, that it be qualified to its harsh environment- The qualification almost always includes some testing to assure its operability in the harsh l environment, even if this environment is relatively low steam i

ennditions or radiation only. Many types of equipment do have problems with these less severe DBAs. For instance, the steam iauses condensatim and many types of equipment have experienced l

problems such as switchgear, MCCs. and hydrogen recombiners. The -

increased radiation levels on the order of IE4 RAQs Some and above IC's, suchdo as

~

rause problems for integrated circuits (ICs).

l 6.M05 devices are susceptible to radiation levels of IE3 to IE4 RADS. Some newer devices, such as HMOS devices-(found in computer l

j chips) have experienced failures in the 100's to 3000 RAQs rango, i r

! f. Merits analytical resolution. 1!t will be'useful for the NRC to establish its position on the applicability of PRAs to equipment l

qualification. Also, there is a need to evaluate the use of PRAs to justify short I.0CA tests. )

l 9

first of a 1, a properly developed EQ program includes 9.oh 1

those items of equipment.i(1.e. equipment withir the scope of 10 CFR 50.49) that are relied upon to remain' functional during and following design basis events. Therefore, since the program i

includes only those items that (nad function, PRA does not seem to A-7 i

, c_y a

' aAn have a role at this point. The'E0 requirement for these items of equipment is that they must he able to perform their required ['

functions when called upon for as long as required, it is not unreasonable, for example, for applications where a component is ,

required to be qualified for service inside containment and a similar component ist requirs ' to be:quali(led;for service c , .f"

~

conditions outside containment:(where the'.l. potential harsh

. environment is significantly less ~ severe), if the licensee chooses to use the same component outside: as iti used inside containment. .

Under these circumstances.Elicenseestartinot: required to use the '

same componentt however, using the same' component may provide desired flexibility.: (Again...this is not unreasonable, it is the ,

licensee's choice and(it provides desired. flexibility. .

5t df Asses smen'.: y Given the advances that h4Ve been made in our understanding of E0 over '

the past 25 years, and based on the TMI-2 experience. PRA tc<tchts. and plant experience. some improvements C in the E0 Therequirements staff shouldmay ne be receptive possible and of benefit to the ' stry to proposed changes in the E0 methodology that are: (a) developed as an industry init iative. and (b) demonstrated to be technically just ified.

5. 14A studies indicate that E0 Master Lists may need to be updated to na.lude additional equipment.

Pyrr Review Comments:

4. If true, the E0 lists should be updated.

I b 1his is valid, but additional equipment'to be added is probably non-safety related..

t Merits analytical resolution (i.e., analyze existing informatlen to reach resolution).; Also see comment 4.f (above).

d. What PRA studies? Please identify the indicated studie . anc . "

provide copies for NRC staff information. It is the licensees' 4

responsibility to update E0 master ilsts as new information ,

becomes available.

',.t a f f As smment : '

Additions to the E0 Master List would only be appropriate if the benefit

'n safety is significant as defined by the IP[ initiative. In this regard, the st af f should assure that E0 shortcomings would in f act be addressed under the IPE program. Beyond this. the use of PRA for E0 applications may be of significant value to the industry in better The staff focusing E0 reouirements and the expenditure of resources.

should be receptive to proposed uses of PRA with regard to f0 that are:

(a) developed as an industry initiative, and (b) demonstrated to be 2

technically justified and in keeping with the Commisticn's policy on '.he m.e of PRA.

.A-8

>--t --

l l

n ,

g.

Y)h. < jf c

,,;> ~ - .

6. The single failure cr.lterl&Lhas not been appiled consistently relativa to C0 For example:;

Peer Review Comments!

a. I don't think that plan

• design bases are consistent.

b. May be true, but so what! Each istua should be evaluated separately.

• 1he single failure criteria was not imposed for qualtfication of cold shutdown equipment (i.e., lEB 79 01. Supplement 3, only required one train of sold shutdown equipment to be qualified).

Peer Review Commentst i

a. The concern regardingilEB 79 018 supplement 3 ciarificatIon is h- perhans a legitimate one in that'there is no documented basis that 1 am aware ofitt to why this exception is acceptable for. older

~

l plants. As oneiof the ,arties to the many HRC and Industry discussions ondhis subject during t.he early 1980s, my l

recollection {tithatt

- for somefolder plants, requirin the environmental j

quellfication of the total popu allon of the cold shutdown equipment would have been impractical to implomont with the '

! as licensed system configurations, and f

- The intent was to limit the environmental qualification I requirements to the system / equipment complement in one (complete) path required to achieve a cold shutdown condition rather than applying it to all equipmont called l

fer in the emergency procedures.

Perhaps, a better documentation of the basis for this will help clarify why this is considered consistent application of the single failure design criteria,

b. Qualifying one train of cold shutdown equipment will insure that a single f ailure will:not occur. liowever, 10 CFR 50.49. superseded lEB 79-01 an(dts- supplementsl5f t requires qualifying a' path to safe shutdownis Please note that qualification in the context of 10 CFR 50.49 assures that a single failure will not occur, and by j doing so the single failure criteria is imposed. l I

e' With regard to the '*superheat ef fects" of a MStB. the st af f's resnlution of TAP A-?) failed to include single failure considerations.

Peer Review Comments: l l

a. Single failure was considered; it was factored into the mass and energy releases from MSLB. There is no " additional" single failure considerations from "superheat effects."

A9 I

I

.,w. _ _ ___- _ _ - __s _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ __

te .

The issue regarding H$1.B superheat effects analysis not including single failure considatations.. if trua.:Is an acclJant analysis issue.NOTan[0issutip

,a q A- .tated in Regulatory Guide 1.89. the purpose rf environmental qualification is to avoid "commor :1use" failures. Given this. It is not c lear why it is necessary to qualify equipment to protect against sinqle failures.

Pear Review Cowents:

a. I agree. However, what's the dif ference in terms of tvoe testing?

The tests show that A component can perform.

b. The ability to withstand a single failure is one of tie elements of defense-in depth applied in the design of safety sy'tems. The purpose of EQ may be sumed up as preserving the defense-in-depth.

In other words, it is to identify and eliminate the potential for common cause falluras.that may challenge the defanse-in-depth, That means EQ may not even specifically in accident anst*onments.

focus on (i.e.. we do not intentionally 10 looking for) the potential for common cause failures if tie equipment is operated only under normal environments.at all times...As such.. single failure falls into the category ofcrandom failures. Therefore; it is not clear why consistent application of single failure is an EQ issue. In my opinion, if there is a concern regarding single failure application. cit should be treated as a design basis issue.

( The purpose of E0 is-to demonstrate that the safety related coutp ont does not have a " common cause failure" which could take out v.Jundant sets of'the same equipment. Qualification A randomdoes not failure, assure that a random failure Would not occur.

however.shouldhavefverylowprobabilityoftakingoutredundant equipment. if the probability it'high,' then it is probably a "comon cause." In order to demonstrate that equipment duas not i have a " common cau:e f ailure", then all f ailures during qualification and in service should be analyzed to determine the root cause. Once the root cause is known, then the judgment as to the cause being random or common mode can be made.

Any f ailure in qualification. or in service, which is common cause. would render the qualification suspect untti it can be corrected. The area of assuring that failures seen in service are not common cause would seem to provide the best payback for safety.

d Merits analytical resolution (i.e.. analyze existing information to reach resolution)D Although single failures are assumed to occur independently.of any other failures, it is possible that the failure is the same as a failure that'can ro wli from a common cause. Unless I misunderstand the statement. its logical implication might be that qualification is not necessary at all.

A-10

~ .-

qf:g * 'L' . ._.;;.m _

h ? s,? -

. . qQg:

Q,'

e. Perhaps a better choice of words for RG 1.89 would have been "to insure that common-cause (111ures do not occur *.

H 4_Aues',M n.J:

l'a 'taf' should determine to wt ~ t extent single failure tonsiderations v r- applicable to ID. and assure that Qualification of cold shutdown e q u i pme rit a nti resolution of 1AP A*2111s consistentfwith the position t hat is estatiltshed.

J

7. t'e adecuacy of safe shutdown capabliity with regard to to has not been v'qnrously Addressed (e.g.ijhot shutdown vs4 Cold shusdoW6 Capanility s:ngle failure reQuirementi, etc.).

Ecer Review Comments:

! a. I believe that this.watDadd.re:

s ,f , sed.. 3 ,

b 1 disagree. Given That'the't0 master-litt called for in lo CFR 50.49 is established t a tr) into account the various accidents.theresultingenvirpnments,andthecomplementof systems and eculpment required far each of those accidents, I (all to tenderstand the balls for this statement, e Non issue. Safe shutdown is evaluated through many paths:

i Appendix R PRA. f5AR, etc. (Q master lists take this into I "onsideration Merits analytical resolution (i.e., analyze existing information

o l t o reach resolut lon).

lhe f0 rule (i.e. 10'CFR $0.49) requires quellfication nf safe 1

' e shutdown equipment. If this statement in suggesting that this l

requirement is not adequate, than this issue should be discussed

' in more detati with the NRC staff and management.

d.L.ADe.sP.cM:

LsentheadvancesthathavIbienmadeinourLinderstandingiff0over tne past 25 years and based on the IMI 2 experience and PRA *nsights.

e c adjustments may be warranted. The staf f should determin- to what

r. tent cual1ftcatton of equipment for achieving cold shutdown is truly yessary irrespective of licensing basts, to assure that a saf ety ncern does not esist (for those plants that are not requireil to

.;..alif y eauipment necessary to achieve a cold shutdown condit ton) and

.l.a ument the basis for tne postiton that is established. Ine NRC staff

" auld also be recepttve to proposed changes in this regard that are:

.i )

developed as an industry initiative, and (b) demonstrated to be .

'n hnically justified. l l

I l

This includes equipment reQutred to remain functional and those l i

whose failure Can affect the safety functional capability of other safety-related equipment A ll  ;

f ,

N

/

. . g.

' 8. Bet t er definition of which instruMhts are required to be quell'ind is 4 ner< led. with suppo.rting basij,.

Peer Review Comments: .(

a. I agree to the extentoit etatet to the.need for a PRA based nide f inition. If the current determinittic evaluation is what~is intended in this stateMnt. I must question why? Are we.then

},

saying:

{

I - that RG 1.97 is inadequate? If so. it is not an [Q issue.

but may be an accident ans'/ sis and emergency operating

[ procedures istue.:-

1 '

f;j that the criteria enumerated in 10 CFR 50.49 If so, for under what

. , iw establishing an [0 list is incomplete.

condition or accident-scenarlo(s)?

b. Non-issue. Criticality of instruments is factored into selection l

' for MEL.

i

c. CQ list should be adequate to determine which equipment should be i M qualifled. ,

3

p -

457 ,

j [%q ;A y d. Inresponseto79-01BIutt11tleswererequiredtodncumentthe safety related functioni for safe shutdown.; Once the equipment was identified and itcwas determined,that it-was-located in a.

c harsh environment.~in order to assure its-function,'that equipment was required to be qunlified. 'I witnessed the NRC EQ audit team M review of these analyset during many audits. Thus, I'm confident

$y that the. utilities have this documentation and supporting balls.

The design basis reconstitution process that many utilities 9 undertook was partly involved at upgrading these analysts,

e. Merits analytical resolution (i.e.. analyte existing information

} to reach resolution).

f. All equipment. including instruments, required to be qualified is '

(b2). and (b3). This regul.ement

[1 discussed in 10 CFR 50.49 (bl)

• 1 is written in a manner th.t includes the flexibility that is I necessary to t.ognize the differences between the many plants that is required to comply witii all the Commission's regulations.

It is the responsibility of each licensee to determine exactly what equipment fitt into each of the'thre4' categories, s SIM L.hnen.mtn.t:

fla.donthestaff'sreviewunderCQTAPActionitem3.e[10),it j app.4rs that the instrucentati" that was required to be qualifted was -

well defined. However. given tite advances that have been made in our. I une rstanding of [Q over the past 15 years. And based on the TMl 2 en orient e and PRA insights.< some adjustments in the requirement s may be pn a.ible and beneftetal to the industry. The staff should be roceptive A-12

to p."pescit changen in the instrumentation requitements that are: (a) deve T*N A% an industry int t lat ly8 . and (b) demon %ffated to De t f( h' tally justifled,

9. Safetv ielated equipment located:in mild. environments that experience sevem environmental t inditions'du . to the operating condition, such as snif ',nating f rom being continu&lly energized. may not he adequately qual ed.

P_ttr Re y_Lt.w C emme n t1 :

a. May be true, but probablyL& rare occurrence. Design requirements for safety-related equipment include requirements for assessing environmental effects such as self heatC

b. Such coutpment should be covered undt (Q design range testing.

c. I disagree. 10 CFR 50.49 does not i iuire qualification of mtid-nvironment equipment. At , esent. *nvironmental qualificatton is focussed only on demonstrating the inherent functlonal capability of harsh environment equipment, particularly in accident  !

l

-nvtronments. for mild environment equipment implementing ippropriate surveillance maintenance, and fallure analysis to address degradation due to normal operational and service environments is believed to be.. adequate. :This position has been-accepted by the NRC in the past in several meetings with the industry. What has changed now? If equipment operating and fallure experience suggests that maintenance programs are .

inadequate, than it should be pursued' as an oversight issue related to the maintenance and surveillance of mild-environment equipment. Perhaps, a systematic study of ali equipment failures in the indurtry can shed light on this st.bject, in any event, !

amnotclearwhythisislanEQ-issue.

tiowever, it should be noted that one can make a seismic 1 qualification issue of this, if indeed it can be shown that cperating/f ailure erperiente shows a potential for reduced seismit functtonal capability.

d. This is a valid luut. Forus has been on harsh environment because of 50.49. l

c. 1 am not clear whether this is an [Q issue. l

f. Safety selated equipment in mild but not benign environments, do l not now have to be qualified. Rigorous qualification and maintenance programs to identify and document qualification are l l

probably not cost effective because much time would be spent on reviewing paper work. A more cost ef fective approach, which would increase safety asturances, would be to encourage the utilization of more modern, sophisticated non-intrusive test ecuipment to menitor the condition of equipment. All electrical

! A-13 l

7-l 7 '

f i -,. 5 euuipment has heat as a by. product'. In the generation of NPAR j

i Report NUREG/CR 5762, it was noted that infrared thermography was 7

a new. modern tool for measuring'non-intrusively, the temperature -

l of equipment, this tool and others. such as Vibration signatures.

[

have been shown to be sensitive to age related degradation,

{

j The sensitivity to age related degradation and the non-intrusive i attribute provide a mucn more economical method of detect ng d mradation before equiprnent failure. It is condition monitoring.

! which focuses on looking at the hardware in its normal state. The l proper focus should be at~looking at hardware instead of paper. .

i

' Thus. the encouragement of equipment condition monitoring would

.ean to have significant cost and safety impact.

g. More NRC oversight may be needed.to assure compliance with E0 l

iequirements.

l I

j h, in accordance with the Statement ~of Consideration for the final EQ  ;

rule dated January 21,198h- The final rule doet not cover the electric equipment located ,i.. a mild environment. The Commission has concluded that the general quality and surveillance g:

requirements applicable to electric equipment as a result of other inmmission regulations including'10 CFR part 50. Appendix B (see f or example Regulatory Guide 1.33. " Quality Assurance Program l

Pequtrements (Operation)," Revision 3) are sufficient to ensure i

I adcouate performance of electric equipment ireportant to safety l

' ac at ed in mild environments. ~

i ,

llill ' &.McD.I .

Trom a ,afety perspective. full advantage should be taken of operating l

! plant o portence and PRA information, equipment performance, condition and c". - onwnt nnnttoring, root cause assessenent. and trending of i l

l irif orr a s on in orce to ident ify Irnplement and correct any (0 deficienc ies thit at ion o f the maintenance i.;c may ei .t oa an ongoing basis.

j i # 11 h.:p to resolve this problem for active components, and the staff should nit iate act ion to int iude alect rical soutpment within the scope such as this one j of thi taintenance rule to better adderst concernt e

l SE!N D Ba sed on t h.-

staf f's review of scope /appilcability issues, the followina

) recorrrnandat 'ons were made :

j 1

a. Althou e f0 research on some compnnents may not have been as exten.ive as (att e research. additional research should not be performed unless:

j

<> based on a well defined need for further research, (b) there j (al aitcaod likelthood that tne desired information will be obtained, ano 15 J

(c) tr e cost of the research is justified in terms of the expected benefii> to public health and safety.

l j b. f ull a1.antage should be taken of operating plant experience and PRA inf orM 'en . coutpment performance, condition and environment j

i 4

A-la

~ ' ' ' " - + ~ -. . . . , _ , .__ ._,

^ ^ ^ ' ' '- - - - - _ _ _ _ _ _

i l

monitoring. root cause assessment, and trending of information in ord?r to identify and correct:any [Q deficientiss that may exist en an ongoing 1 basis, in order to f actistate this effort.?the staff should initiate artton to include electrical equipment within the scope of t$e  !

maintenance rule. .,

There ($ a marked difference in requirements that were imposed for (0 of j c.

electrical equipment versus what was required for LO of' mechanical coulpment, and technica) justification is required for the different j standard. and the different approaches that were allowed by the staff.

for erample. [0 of mechanical equipment did not involve prescriptive replat ion, a detailed program review, and confirmatory on-s te inspectton.

o The NRC staff should assure that significant [0 shortcomings ,<ill be addressed under the IPE initiative. \

e the NRC *.taff should determine to what extent single failur, l i

c onsiderat ions are applicable to [0, and assure that qualif '. at ion Sf old shutdown equipment and resol"' inn of LAP A 21 is const'.'ent with tne posit ton that is established,

f. The staff should determine to what extent Qualification of equipment for i ac hieving cold shutdown is truly necessary irrespective of licensing basis to assure that a safety concern does not exist for those plants tnat were not required..t0 ~ 3qualify equipment:Decessary tn-achieve a cold shutdown condition.

a Given the advances that, have bern made in our' understanding ( (Q e.*r the past 25 years, and based on the THl 2 experience and PRt. Insights.

adjustments in existtrg E0 reautremi.r.ts may be possible and of

.ubstantial benefit to the inaustry. For example, it may be po'.stble to )

in t t er f ocu'. [0 reauterment s by u'.ing PRA techniques, and twt ter l def init ton of the rcle that sinole f ailure plays relative to f(; could j result in tiet ter directed Qualif ication requirements. The 'WC staff.

should be recept t"e to proposed change; in this regard that are: (C

$ .cicped as an irdustry initiative, and (b) demonstrated in be t e.r hr it a l l y acceptable.

1 NR: staff did not cors.otr any of the sccpejapplicability 's>ues to be l ar '"med i a t e .afety problem.

1 l

A lb l l

h 8-  !.Q_0.CM19.!1%

l.

Many facets of the methodology for establishing initial cualification_of a ,

"outpment have not been adeouately addressed and Justified on

• rigornes, it>rhnical level, includtrq (for example,i:

P.st.r leview Comments: l

a. The bullets that follow are too general for comment. I thought t i

that these. points (other than cost) were coversd for properly qualified equipment 1

b. The main consider 5 Lion is'that'the' methodology-selLcted will l

increase plant safety, P

c. Issues such as dose' rate effects, oxygen diffusion considerations, mandrel bend test 2 requirements .etc., are (Q issues that are now, r and have been for some time, the subject of 93ny discutsions and ,

require engineering solutions accepttDie to both industry and the To date, these issu 6 We wt prevented EQ pregrams from r

NRC. We should being developed by . industry and approved ty the NRC continue to work to resolve these issues in a way acceptable to

~

all involved. a a use of test data ver tus other method; f.uch a operat ing expe r tence ind operating analytical techniques (for example, extrapol.ito of data o' espei tenc e. 1t adequately justified, may be wet' suited for r0 4pplir at mm),

Peer Review Con g )is:  :

a. I disagree. Why '! this an issus $1nce the NRC has already .

outruled (defacto)ification of harsh-environment equipment demonstrating qual Delete this from.furthera-considerat'on because: '

- For all practical ~ p'Jrposes, establishing qualification is l complete for, plants that are operating.

We have vary.llmited experience on the performance of equipmee under accident conditions to draw from. t

- lt is impractical if not impossible, to analyze equipment .

performance capaDility under accident' conditions Thus, preference for test over other methods fcr environmental qualification is justifiable.

for seismic qualification, it is prutient to examine a greater i Irvel of use of operating experience, and this is being addressed by lEEE. '

b. The key words are ' adequately justified." A number of licensees l unsuccessfully attempted to use data extrapolation for i A-16

k qualification %here the ditt presented did not represent the formulation el the component'to be 'ested or the source of the i data did not identify that it was obtained in a manner that l duplicated the e Mected DBE environment. For example. cables l

manufactured byj. tie same manufacturer using generic terms for its materials (e. gig buty))~can use different formulations of the material and data for matarla1s used in taped splices can look acceptable unless you recognize that the adhesive can If fall in a )

] harsh environment that incivdet submergence in water.

adequatelf justified, the use of such data is acceptable,

c. This is a valid issue,

d. The DOR Guidelines had some good words requiring tests for harsh pressure-temperature-steam environments. How do you extrapolate normal operation to conditions that by. definition are more severe

] (e.g., you can test water forever ai ?ll*f and be ignorant about what haopen* 41 213'f)?

e, Some physical. data will alway'. t.e required. Operating eeperience and analysis woulde t'y themselves, not be acceptable, f.

IEEL 323-74 allows qualifirition by type testing, opera':ng experience, and. analysis.L NRC made it clear via workshops in the 1980's, auditsi.and correspondence that the preferred method was type testing .To.cualify by operating experience, equipment would save to'haveldotumentation showing that it has experienced and operated properly in a DBA. Since only TM! hn experienced a DBA. very few equipment items have been qualified for harsh environment 4 based on operating experience. Analysis only is also very difficult to have demonstrated since there are no formulas to reliably predict operation of any type of equipment in a harsh environment. Thus, little effort should be placed on addressing technique:, which are not normally used. A stroncer policy statement than.the one in RG 1.89 may be considered to dis c urage the other less utilized methods of qualification.

Merits analytical resolution; industry responsibility. All g.

three methods and combinations of them are acceptable with justification. IEEE has initiated an effort to constder preparation of a standard that will address the use of operating expertence in qualification.

I think type tests should be the basis for qualification. Other h.

methods of quallficatior. such as analysis and operating experience l are extremely,llaited in application and have little practical value,

l. UndBP Whlt thtlmillntti il (Mt ditt not ipproprtitti lithQ intent of this question is to suggest that test data need not always be required, than I refer you to 10 CFR 50.49 (f)(3) where experience with identical or similar equipment under similar conditions with a supporting analysis to show that the equipment A-17

c 4

to be qualified is acceptable. In addition, NRC staff experience over the years indicates that there is no basis to support analysis only or extrapolation of data, and what can operating experience tell you about the ability of a component to survive a LOCA7 Finally, what is adequate justifie.ation?

t ' # f_ _A uts Eltd :

Gi<en the advances that have been made i.n our understanding of IQ over tnr past 25 years, and based on the TM).2. experience, PRA insiuhts, and operating plant experience 4-some ad,justments The staff in the requirements should be receptive may be pc nible and beneficial to the industry. (a) developed as an to proposed changes in this regard that are:

nJustry initiative, and (b) demonstrated to be technically justifle '.

+

. environmental conditions (e.g. temperature, pressure, radletion,  :- ,

e ) that are postulatad for both inside and outside contain Ic:r Review Cemmetui:

Haven't we a 1 disagree. I am not clear what the issue is here.

already established the framawork for a consistent set of methods for deriving the environments in both DOR and NUREG documents?

What is the inconsistency that still needs better techn1ral If the concern relates to minimitira the level of

.)ustification?

c on s ervat i sm, then it should be a Regulatory Recuirement, Reduction issue rather than an [Q lssue.

l' This is a valid issue, t

This is not a methodology issue.

c The environmental conditions postulated for inside andPlants outside have containment during a DBA probably are conservative. and mass used computeritcc modeling techniques to predict the heat transfer. There have been a few tests to aid in developing these models, such as the FTIR tests and CE's full scale torus testing.

Some utilities have done some special tests to aid in predicting th environment.

In many cases, the postulated DBA is in conflict For. instance, most DBA curvas show with natural steam phenomenon.

superheated steam -i.e. 340 'F and 60 psig (at saturated Then the curves conditions the pressure would be over 100 psig).

show chemical spray coming on and the temperature and pressure would be the same as before the spray.

In actual testing, the spray causes the steam conditions to goTypical into saturaston, which immediately drops the temperature.

postulated curves don't show this phenomenon.

Additionally, most postulated DBA's outside of containment show instantaneous temperatures above 212 *F being distributed to?!atural many areas in the reactor building which are not pressurited.

condensation, deflection of steam off of the walls, around corriers A-18 1

,- y. M .; ' }, .'

. .m.. .

and on equipment would cause co'ndensation and a reduction in steam l

! I temperatures. Additionally, cabinets with' louvers or few openings would contain trapped air, which would have to be displaced by the steam and thus a natural thermal' delay:results. Research into tho steam phenomenon wonid most Ilkelf reduce the predicted severity l of steam line breaks outside containment and in compartmentalized containments, Radiation is also predicted to' occur instantaneously. The levels i

i n( radiation And the instantaneous release are probably overly j

predicted.

e. Industry responsibility; this is best resolved by the industry.

l 53ifi f ~ us" tmell.t :

In ce.w41. the environmental conditions that were required for E0 were '

based on accident analyses that included additir,nal margin and are l

i belic.ed in be conserva*.'ve. However, given the advances that i. ave Deen j made n our understandity of EQ o .c the past 25 years, and based on the

TMI.) rpesience and PRA insights, some adjustments in the requirements j may b. necessary. The staff should assure that informatton in this r e g a i .' is eaialogued *nd well an4 rstood, and made available to the s ersees' *esponsibility to make use of new l

mdu c lt is the

{

gencei. information, as well af platt stetific information, and to make adju.* ment 5 to thair (Q programs when such action.is warranted. This j i does not necessarily mean that (Q programs must always become more 'i

en t r tive; new information might very well. support a relaxat'on in pron' .c

oqu i rement s ,

\

. age < mdit inning.

Etti *" (31w C ommen t s :

a. I agree. it is well known and accepted that the methods

] established were the "best practical" given tne (then) state of ,

j the technology. Holes have been identifled in both the technology

' and its applicatlon. I am also of the opinion that eventually we

need to get out of 1.he frUbe of mind wherein we assign a j

"cualified lifc' based on simulated aging tests and then treat it j

as sacrosanct, and pencil whip it to incraasingly higher levels of precision depending upon who does the main using what information.

I believe that the true answer lies in verifying the correlation between real world aging vs. simulated aging, in other words.

j condition monitoring.

! b. This is a valid issue.

c. tonsiderable research has been edlcated to age conditioning.

Prior to Sandla's research~. NRC was presented with information ,

J on how Arrhenius theory formed the basis of the Underwriter i

Laboratory Specification UL 746B and IEEE Stds. 99 and 101.

A-19 I

I Additionally, Arrhenius theor.v is the basis for reliability calcula' ions in military, NA'n, and the semiconductor industry.

d. Merits analytical resolution, The level of past and ongoing experimental research on age conditioning is consistent with the importance of this issue, However, careful evaluation is needed to assure that future research is directed to investigations of the highest priority that have.the prospect of success in a reasonable time.. More effort should be d rected toward applying the lessons learned.during.the last;two decades to modif f qualification requirements to resolve the problem of demonstrating I

a meaningful qualified ~ life.

.taff Assessment y t' reconditioning of equipment is necessary as a f undamental partthought of to mitial qualification testing, and the current me'hodology i- 1 tm conserve *1ve.

The amcunt of pretonditioning inat is truly necessary l

to establish qualification is subject to debate, however; and given the undt standing of E0 ove the past 25 41 van.c> that have been made in

. ears and based on the TMl ? experience and PRA insights, some changes ny be appropriate lhe staff should be receptive to proposed changes in'ti4tive.

n the methodology that are: (a) developed as an industry

.ind (b) demonstrated to be technically jest f(f ed

. -a rg , (in general) and use of " excess margin" to justify w ortened post - ac c ident test duration during ' TOR LOC A t est ing;  !

) Peer Periew Coments:

a. Mary.ns in general: I disagree, I don't think this iceds work  !

unless one wants to.go after establishing a basis for possible relaxation of current'IEEE. requirements. The current practice of ,

using IEEE-323 type margins for test parameters is justiflable and l represents a practical engineering approach to accomodate-some uncertainties such as manufacturing variat195, and should be C on t 1 #1Ued .

.r xcess margin" used for-justifying short duration tests- this issue deserves some attention by performing a set of very focused LOCA tests to detertnine if the results support the method in which margins or conservatism in the test parameters were used Manyto types justify shorter test durations in some older plants.

of analysis techniques have been used. One should examine the

} validity of the extrapolation of Arrhenius parameters to temper,t'.ure ranges far beyond where they were experimentally J estabitshed. In my opinton, this can only be resolved by testing.

b.

This is a valid Issue, but it is being addres'ed via the current j NRf research program plans ralative to EQ.

e. Determice whether the test is modeled adequately.

A-20 O;

. . . :3

%f}l:%..,,i%$q;Q.e: y i d. Margins were required to be doctimentedifor all plants, regardless of E0 licensing basisi ,

..The c Merits analytical resolutioni merits experimental research.

correlation between margin and its contribution to safety assurance is not likely to be established quantitatively with any p a

reasonable amount of research) and it will therefore remain largely a matter of engineering judgment. However, if there is a serious question concerni:'g the justification for the use of ,

excess margin to compensate for.short DOR LOCA tests, I doubt the question can be resolved simply by engineering judement. It might j i be necessary to repeat some tests using the current LOCA testing i

i practice and to compare the outcome with that of the DOR tests.

stdL.AUJt.5 5me n t . ,

However. given'the Margins (in oeneral) are thought to be conservative.

a vances that have been made in our understanding of E0 over fho past 25 years, and based on the TMI 2 experience and.PRA insights. some to.the! i-adju'iments in the requirementsemly..Je posd ble and beneficial ~

industij. Also, to the extent that focused altention is placed on' werating plant experience and PRA.inferr.ation, equipment oerformance, iondition and environment monitoring, root cause assessment, a"d  ;

trending of information. it may be appropriate to relax some wouirements. The staff should (a) be receptive to proposed chanu m in the developed as an industry in'tiative.

r.irgin recuirements that are:

ed (b) demonstrated to be technically justifled.

Wit' regard to the use of excess margin" to justif) the short duration

( M A tests of the DOR Guidelines plants, this may be a valid concern and l

't should be reviewed further by the staff.

- "e test woonnce and int duratton. inc lud ino pou -acc ident t,.orat ino j

t mes.

. I's t r Review Comments:

a This should not be an issue after all the research that has been

, conducted to date. The current literature review [being completed under contract to the NRC).should put an end to this once and.for

~

all.

However, one can make a case for a shortened post-accident duration for test purposes, and consistency in this regard. A study focused on this narrow question that systematicaily evaluates each accident scenario for the time required to achieve cold shutdown (or some other acceptable intermediate plant condition) is desirable.

b. This is a valid issue.

c. Merits analytical resolution and experimental research. As for test duratio',, there is a need to evaluate the use of PRAs to justify short LOCA testse A-22 s . .

i

l l

'd 4

• f h)fd Ee0.1: ..

Ro arc h i <, good t o a p0 :n t . (N! there ao' lim't at ions to what c an be ai, wplished. The results of past research ef fort s should be catalogued ar: the informattun should bc wall understood and used in addressing is.aes such as this one. An) further research should; a) be based on a w,> 1 1 defined need for additional information, b) be pursued only if there is a good likelihood that the desired information will be ont ained, and c) be pursued only lf the cost of research is just ified in terms of the expected benefit to.pubilC-health and safety.

Wi t h regard to post-accident operating times, there appears to be confusion and inconsistency. Given the advances that have been made in our understanding of E0 over the past 25 years, and based on the THI-2 e,perience and PRA insights, some adjustment to the requirements may be .

possible and beneficial to the industry. The staff should be receptive to proposed changes in the r.ethodology that. are: (a) developed s an industry initiative, and (b) demonstrated to be technically justified.

. ..edrel hend t est rt-Quirement, I

l hsg h yiew Comment 5:

I agree. The Question is Whether or not they are required. My understanding is that the writers of IEEE-383 intended it to 4.

assure that vibratory and shock forces from a seismic event are properly accourted for in the environmental qualification process.

Many in the industry are of the op{nion that these requirements may be too conservative. One should reserve judgement on this issue because, con'rary to what many believe. there are many installations (e.g..; free hanging cable' at~ termination points such as those with connectors, and cables in ficx conduits at termination points) in plants wherein cables will be subject to such vibratory and shock forces during a seismic event. In an

' aged condition, these-installations may be more vulnerable to a J common cause failure. There are a few examples of failures whercin motor lead wire: grounded.to short caused by wire insulation (in aged and brittle condition) falling off from motor-starting forces. But, such forces are much greater than those from a seismic event.

Perhaps, a study and some tests to determine if aged cable jackets and insulations can withstand these forces, can answer these questtons. This should be a low priority item. Pending completion of such a study, it is prudent to continue current practices, b ibis is a valid issue; needs to be addresseJ.

( lh15 test bounds the expected worst case dynamic / static forces that a cable may be exposed to in the field yet which are not reproducible in a test chamber. If another more realistic method is available, it should l,e preposed by IEEE.

.A-22

. s . ' - ~.

-__~_-'

e 92;.. 8.% . age

• b , w l

L

d. Merits analytical resolution; merits experimental research.

There is substantial agreement that post-LOCA mandrel bend tests are too severe; and in so:ne recent cable EQ pengrams, these testr ,

i have been omitted. - However, if they are omit ted, part nf their original purpose, i.e., to account for vibration and seismic effects, woult have to be addressed.

e. Please provide an acceptable alternative.

i 4 11){f Assessment:

f.iven tne advances that-have been made in our understanding of r.Q over the past 20 years, and based on.the THI-2 experience and PRA insights.

• nme adjustments in the requirements may be possible and benefscial to tbe industry. The staff should be receptive to proposeo changes in this regard th3t re: (4) developed as an industry initiative, and (b) demonstrated to be technically juttified.

. .ynergistic effects; Peer Review Comments:

4. After all the tests by the Industry, and by Sandia and o'thers, it is hard to believe that this is still an issue.

4

b. This is not a valid. issue; sufficient research has been done.

c. The synergistic effects of sequence have been addressed for many years ..th must test programs using the sequence o' radiation exposure prior to thermal aging.

J. Merits analytical resolution; merits e m mental research. It  ;

i would be useful for the NRC to update its position to account for the research conducted durie) it$ last decade. The research results are diffic'It to generalize and are somewhat inconclusive.

There is some evit...ce that the degradation of equipment during i LOCA conditions maytovershadow aging degradation to such an extent that synergistic aging-effects shodld not be a major concern.

$t af f Assessraent:

10 the extent that focused attentton is placed on operating plant experience and PRA information, equipment performance, condit ton and invironment monitoring, root cause assess:nent, and trending of information. this issue becomes one of minor importance. However, given  :

t he advant es that have been made in our understanding of E0 over the past 25 years. and based on the TMI-2 experience and PRA in.ights. some l i

adjus'ments in the requirements may be possible and beneficial to the industry. The NRC staf f should be t eceptive to proposed (%nges in this i regar. that are: (a) develuped as an 'ndustry initiative. .ind (b) ,

l 1emnn' t ra t ed t o be techr.ic ally Ju st i fied.

. to e rate ef fects:  ;

l i

<A-23

l I

a - s

~

,a

" I ' "/R ^- C %.'"

}

Peer Review Comments:

l

a. After all the tests by the industry,'and by Sandia and others, it  !

l i

is hard to believe that this is still an issue l I

b. Not valid; sufficient research has been done.

c. Dose rate effects are cUnsidered~to be second order effects, more j

than adequately accounted for by using the very high radiation j

total doses typical of qualification.

d. Merits analytical resolution: merits experimental research, it l i

would be useful for the NRC to update its position to account for l the research conducted during the last decade. The research

' results are difficult to generalize and are somewhat inconclusive.

There is some evidence that the aegradation of eautpment during j

LOCA conditions may overshadow aging degradation to such an extent i

that iose rate cffects should not be a major concern.

llaff AUf.D M.0.1 :

f j

See the :,taff assessment re: synergistic effects (abuve).

• oxyg m diffusion consideration!*. and j Peer Review Ccmments:

a. This is a valid issue.

b. In my opinicn oxygen diffusion effects are second order ef fects because 1) many plants use inerted containment, for which E0 credit has not generally'been given, 2) oxygen'is used in temperature and LOCA chambers, and 3) root cause analyses have not shown oxygen diffusion to be a source of differentes in dnticipated and expected results. I

c. Merits analytical resolution; merits experimental research. It '

would be useful for the NRC to update its position to account for research Conducted during the last decade. The research  !

th't l

results are dif ficult to gNwralize and are somewhat inconclusive. i 1h<.re is some evidence that the dagradation of equipment during LOCA conditions may overshadow aging degradation to such an extent that oxygen dif fusion effects should not be a major concern.

Also the oxygen diffusion effects predicted by research on insulation materials were not always evident in later cable research.

d. What realistic assumption are being referred to? It is realistic assumptions that recuires the consideration of oxygen diffusion.

The assumptions being alluded to should be provided for review and discussion.  :

-g -4 A-24

.m: ~

• r"

. the staff assessment ie: synergistic effects (above).

a

• i"st.

Enyr Review Comments:

a. This is truly an issue,'especially in?the current competitive utility industi, environment. This should be addressed perhaps by keeping an.open mind'to. review and where practical accept innovative' alternatives proposed by the~ industry to address EQ.

An example of this already >elng. implemented is -the MEO exemption request. Another area worth pursuing from a regulatory aspect is to remove the requirements to establish and track " Qualified Life," and substituting it with greater reliance on equipment operational reliability: analysis.' reliability centered maintenance and condition monitoring / assessment.) from the' regulator's perspective,thiswillrequireastudi'toestablish'aframework for implementation.. Such an approach will also go a long way toward addressing E0 related license renewal considerations. I might also note that this apprJar.h will be consistent with those of Germany and France.

Alsv. the industry needs to do its part by doing more joint group wor' on qualifying new equipment. EQ problem resolution, nd standardizing systems / equipment for use iri the nuclear industry at least at plant vintage levels. This is what the French do and we can learn from their experience, b This is not an E0 issue; do value impact.

c. The concern is not clear;.the' specific. cost problem (s) should be identified.

Staff _651ejimRal: ,

fast is definitely an issue for the industry when it comes to E0, and . ,

there ray be accPDlable cost-cutting measures that can be taken. While Ihe staff should be receptive to reducing costs that are imposed on the i industry. the onus is on the industry to propose and justify lower cost aitornatives.

The 11owing problem statements expand on certain aspects of this issue:

- l'ifferent E0 standards were imposed (i.e., DDR Gaidelines. NURIG-0588 tat'egory 1. and NUREG-0588 Category 11) without supporting teinnical j at if ic a t ion as to.: (a) why more rigorous standards were warranted, and (b) why " progressively less strict standards" were adequate.for the older plants (e.g., older plant equipment qualification is not as ,

r'qnrous- as NUREG-0588 since the components have been qualif ta ! without ,

aoing margins or considering syr,ergistic effects).  :

i

[ecr Review Comments:

a. There may be some instances where equipment quallfled under D0R ,

A-25

a

{

l l

.. ~. >>.7 l

This requirements would not have passed' Category I requirements.

is an issue that should be evaluated by.RES in regards to life  !

extension $lntt the efflCli Of in*$lhl 89 n9 f palt 40 year plant operation must be addressed (particularly for cablet),

b. Somehow. I thought that there was " supporting technical justification."

c. This is being addressed by the current HRC research plan.

d. Different E0 standards were applied because of licensing differences between vintages of plants. The attempt was made to assure that technically each was effcctively the same. When 10 CfR 50.49 was issued and licensees were required to meet the rule. the ma,)or impact was to add some items to be qualified.

lne EQ contentions at Shoreham (NUREG-0588 Category !! plant) were largely based on the differences in E0 requirements for different

~

vintage plants. The contention was that the EQ program at Snoreham was deficient tocause of.severalittems including the concern that equipment' Was cuallfled by> grandf athering to older.

less stringent standards anc that there Was inadequate demonstration that all safety' related equipment was properly qualified to meet aging and other life requirements.

The testimony of NRC Staff James E. Xenney and Vincent S. Noonan concluded that "The new legal requirements (10 CFR 50.49) 're based in large part on the previous requirements and are act The expected to significantly modify the existing [E0) program."

l A5LB found in favor of titCO.

l At the time of 00R Guideltres and 10 CFR 50.49 promulgation. It l was the consensus that 1) all plants had equal technical .

requirr ant s to domonstrate by testing that equipment could operate .aroperly during and following DBA's and that methods allowed to address aging were the main differences. 2) certain equipment such as motors cables, and HO" actuators were qualified using pre-aging. regardiess of the plant's 00R. NUREG-0588 Cat I or Cat 11 licensing basis.

My opin*on is that the known synergisms.,such as dose rate and

' sequence are second order effects. wThe' changes in properties caused by these second order effects are insignificant when compared to the degradation caused by using conservative testing l conditions. Sandia's Mark Jacobus agreed with this and so stated it in NUREG/CP-0135, p. 2-16.

e Merits analytical resolution; this is a valid concern and one that has been recognized by the Commission. Since the NRC is already investigating this issue, no further coment is offered.

f. lhere are basically twn standards, simply because the 00R Guidelines and NUREG-0588 Category 11 are quite similar.

A-26

~

i i

i Therefore, we have NUREG-0588 Categories I and 11. NUREG-0588- >

Category I is the standard which all plants are supposed to-eventually reach. The NRC staff was supposed to develop technical 4

justification for the progressiilely less strict standards for j

_j older plant s. However, to -date' that technical ' Justification has not been developedia qv l I

' I 51Af f Antil!!!t.!d: i t

4 Although different E0 standards were imposed (i.e., 00R Guidelines.

NUR[G-0588. Cat. 1, and NUREG-0588 Cat. !!), eac6 was intended to j establish a reasonable level of assurance that equipment'woul<1 function a when needer' during a postulated event, glVen that some plant. were [

already oi ting while others were in various stages of construction. t

!! has br rgued that one method is more rigorous than another..but l this beco irrelevant if one accepts that each method is sufficient to- t establish qualification for some initial, prolonged oeriod of time. The i tritical question becomes one of how long the cualification i good for. l j

None of the qualification methods has been successful in establishing a "cual'fied life" with any degree o' mtainty and all rtust be i

.upplemented with operating plant experience and PRA informatton,

.guipmNt performance, condition and environment monituring. root cause {

a u e s smen t . and trending of information on an ongoing basis 1i order to j

rovide assurance of continued qualification over the life of the plant i

this is discussed more fully in the steff assessment of the 'ourth state of the art capabilities" and l problem f.tatement totermination of a *concerning Qualified " life, below).

j

- Ine current version of IEEE-323 may be better suited for demonstrating  !

l 10 than the 1974 version since .much more information and exper ience are f wailable now than there was when-!EEE 323-74 was endorsed by the staff.

Peer Review G mments:

a. Do the 115 or so operating plants, and their vendors, meet.the new j IEEE-323 standard?

l

b. This is a valid issue.

-t

c. The current version of IEEE-323 is not different than the 1974 version with respect to qualification practices and there is not' )

any further knowledge provided in the standard. Thus endorsement of newer versions of IEEE-323 is unnecessary.  !

d. Merits enalytical resolution; IEEE claims that the two versions i are equivalent, but not all parties agree that such is the case.

On various occasions, the NRC has communicated its position on ,

this matter orally: it would be helpful'if the position were j documented. i e -. This is totally false and would'only be made by someone totally unfamiliar with'EQ. What-is the basis-for such a statement 7 ~How is the current version better suited to accomplish.the intended  ;

i i

1 A-27  !

l L

'~

p Has the person who posed goal of environmental ~ qualification?if so, what is the basis for this question read both vers'u.m this statement? If not, please read the two documents.

f 11aff Aggng, g ;

Given the advances that have been made in our understan< ting of f0 over the past 25 years and based on the TMl.2 experience anr1 PRA insights,  ;

some adjustment in the requirements may be possible and tieneficial tc j the industry. Also, to the extent that focused attention is placed on performance.  !

operating plant experience and PRA information, eautpment and condition and environment monitoring, root cause assessment trending of information, some relaxation The staffofshould the more rionrous to be receptive ,

requirements may be warranted.

proposed changes in this regard that are: (a) developed as an industry initiative, and (b) demonstrated to be technically justied.

m NRC staff Given the R 7u13 tory Rtquirements Review Committee and view that backfitting the IEEE 321-74 requirements would provide ". .a small, unquantifiable increase i. e ie.el of assurance that equipment is qualified as compared to the significant costs that would be '

involved.. IEEE 323-74 may not be warranted or surfa sently justified regarcless of ,

as a necessary qualification standard for power reactor*.n issued.

when the Construction Permit Safety Evaluation Report Peer Review Comments

a. I agree, except in the case of life extension, Cost impact is

b. Maybe a new cost-benefit analysis is needed.

certainly greater fnr existing plants,

c. Non-technical issue; outside the scope of researt h.

d. Backfitting to IEEE 323-1974 in my opinion is unnecessary because

1) there is little equipment that has not already been upgraded by new testing including aging, 2) as equipment is replaced, new qualification programs are being performed to 10 CFR E0.49, 3)

Even DDR Guidelines plants have substantial percentages of their safety related equipment qualified using aging. Increased concentration on knowing equipment's condition and analyzing compnnont failure root causes will lead to necessary feed back and corre. iw actior

e. Th" tr C*lon P'y.n* SE4 date was simply a way to separate ope- . , plants e ..tose that had already purchased signif' ant amour..a ,.,l. a . .' that time. The standard ured to judge me capabili'yofaV.smenttoperformitsfunctionwhevitiscall ), and is upon to e so is a safety issue, cost notwithstandi 7'is why the supposed to represent minimal safety standards. Th il Commission directe.d the NRC staff to provide technt justification for the progressively less strict star fard for older ,

plants.  ;

, l E

A-28 l

P g

S Rff Assessment: [; -

See the staff assessment of bthe' previous'two problem statements.

- Current " state of the art capabilities" may not be sufficiently dev' loped to support existing 'CQ requirements, such as determination of a qualif tv Ilfe;"

! Eger Review Comments:

a. Unless the issue deals with life extension, I believe that ,

qualified life has been adequately determined using existing technology.

b. ..and then again, in many cases, they may be. The concern is not specific enough to comment on. As written, it sounds shallow and superficial, as well as negative,

c. This is being addressed bv the current NRC research plan.

d. Merits analytical resolution; the concern is valid and a concerted effort should be made to resolve the issue. The experience of the last two decades has'demonstr(ted the difficulty of establishing qualified life without Ih ge Uncertainty, One alternative t o the requirement for qualified life is the use of stress testing prior to LOCA testing and enhanced: condition monitoring in servn e.

in the early

c. The state of the art "is' the state of the art.

1980s when the current E0 rule was codified, it was concluded that in order to b1 assured that a specific piece of equipment can be relied upon ',o perform a required function at a given time.

equipment had to have a' qualified life. The method that was acceptable to both industry and the NRC was the Arrhenius Methodology of calculating qualified life.  % date, a more acceptable method has not been develope".

f. Current " state of the art capabilities" are suf ficiently developed to support the existing qualification requirements, including qualified life. This is really a two part answer. The first part addresses whether safety related equipment can operate properly during the DBA. The second part addresses the ability to simulate  ;

4 aging.

j Operation of safety related; equipment during the DBA,1) is the most important part of environmental qualification, 2) was l required of all plants (00R and NUREG-05BB), and 3) is the documented proof, through LOCA simulation tests that safety related equipment can perform its safety related function '_efore, l during and following a LOCA, The NRC's and the nuclear industry's reliance on type testing for environmental qualification, rather than the other qualification options, such as aralysis or operating experience, provides A-29

i l

l substantial assurance that saf ety related equipment is qualified l 1

to the harsh environment!. .

Additionally, the vast majority of equipment was tested to enveloping, worst case accident profiles, The significance of this is that equipment i,,,1y be subject to a line break during its installed life. That line break'could be a small..line break or a major double ended guillotine type break. It does not have to survive more than one.1The: Qualification practice was to identify all of the potential 41tne breaks.for equipment in all locations in the plant. An envel'opihg' temperature / pressure. profile was then developed. Additionally, when equipment was being qualiffed for multiple plants, the: worst case enveloping profiles were again enveloped. This practice assured that the significance of any actual DBA was less in sever.ity to the enveloping requiren nt.

Thus, additional conservatism was added to the temperature, The pressure, chemical spray, and radiation requirements.

successful completion of a DBA simulation test, of a representative sample, to.ted to this worst case enveloped profile, provides significant assurance that the equipment will operate as needed.

Addit tonally, for items such as cables, more than one test was mest likely performed. For instance, in NUREG/CP-0135.

page F-?), it was noted The that one manufacturer's cable was tested state of the art's ability to gener4te DBA to over 11 DBAs.

Temperature / Pressure / Steam profiles is adequate since many test facilities in the US and throughout the world have this capability. Dependingcon the plant type, typical temperature and pressure maximums during DBA" simulations are shown in Table 1.

Table 1 Typical MLximum Typical Maximum Plant Type. 3 Pressure of DBA

Temperature of DBA 308 'I 50 psig 00R Guidelines BWR 275 'F 35 psig D0R Guidelines PWR 340 *f 50 psig NUREG 0588 Cat i BWR 400 to 500 *f 60 psig  ;

NUREG 0588 Cat i PWR Additionally, fort St. Vrain, a HTGR, had much of its equipment qualified by type tests for temperatures of 900 'f. Its cable was the same as many other plants.

All plants were regilired to document the qualific3 tion and NRC '

reviewed and audited these results.

, A-30

, s.

Thus, the capability of; safstprelated equipment to uperate in a DBA has been adequatel,v demonstrated using the state of the art of type testing. Strong support'for testing comes from the theorem thit one test is worth a thousand' expert opinions.

The second part of tha issue, simulating tte deleterious affects of aging, has been performeo using the state of the ort. In the late 70's several studies were performed to determine the methods '

to be used to simulate aging. There were many who cited the lack of complete understanding of the aging process as an excuse to not move forward. However, many.ilnothe. nuclear Industry and the NRC saw the need to consider what was known about aging and how other industries had addressed the problem. .

Three major industries had been' utilizing techniques for many years to address aging.and all three had basically been using Arrhenius Theory. Thine were:- 1) Underwriters Laboratories for electrical insulation and plastics,.using lit 746B; 2) the cable industry using IEEE 99 and 101', formerly ANSI A57; and 3) the military, NASA, and the sem' .anductor industry using reliability theory for life testing. which uses the Arrhenius equation to calculate life and~ failure rates.

Other aging theories had been proposed. The main reason that the nuclear industry accepted Arrhenius theory as an acceptable form of accelerated aging was 1) Arrhenius Theory had the most data behind it, 2) Arrhenius parameters for the most part had been developed by testing, and 3) Arrhenius Theory htd been succM sfully used in military and NASA reliability efforts.

Arrhenius theory was the best theory available at the time when it was found to be acceptable in NUREG-0588. It continues to be the best and state of the art.

The accuracy of the qualified life determined by the Arrhenius equatton has been a histori: argument. In order to ,atisfy industry concerns on accuracy, several conservatisms are used in qualified life calculation assumptions. The most important are:

1) assumed operating temperature, 2) assumed material function, and 3) asse~;d interaction of multiple materials.

~

These conservatismsland assumptions were addressed as follows.

The calculation for qualified life usually assumed that the materials of the safety-related equipment were at the ...aximum of the assumed temperature range and then heat rise and hot spot temperatures were added to this maximum temperature. Over the years, lessons learned. like determining the actual temperature lika solenoid rises of items containing significant heat sources valves, motors and transformers, were incorporated ir.to the qualified 'ife calculations.

Materials typically showed different rates of dettrioration as to whether electrical or mechanical properties were being evaluated.

A-31

- w.

(( n g-f.yg:

.a g b '

Thus the assumption as..to;whether the materials had an electrical .k function. mechanicalefunction, ur'both gwas an important j- ,

.y assumption. .: 5

.,7 . , .

'?

Lastly, few safety relhted p'ieces of equipment are simple enough Jw to be comprised of only one. material,_ such as most terminal blocks. When multiple, materials 9are present, the material with M.

the lowest activatioWlenergylWastchosen ?for the Qualified life 1 calculation. This assure 1<thatLall: materials 'with a higher activationenergyarVaged?.toiationgedqualifiedlife.- 3 '

Additionally, with ~ multiple'materialsW an implicit' assumption is that the materials arp compatible withleach other:and that new, t

synergisticreactions, don!.,pformspjyp . '![

Sincemuchsafetyreihedequipmentwas1qualifiedbyaccelerated f ,

aging, material incompatibility was discovered when it wasRedesign [? and present, because theseiltems f ailed aging . tests.  ;

rete >'ing were requiredLin order to achieve qualification.

r in th. 1980's has generated a lot of 3, The results of much resear evidence that qualified lives may be. conservative and little g

y evidence exists that they may be overstated. k In my opinion, the ac' curacy at the qualified life calculation ^$,

g hinges dramatically on the underlying assumptions and that,the -

state of the art is capable of acceptable accuracy, given that 1 there is not an inherent flaw in the assumptions. g The most dramatic impact on qualified [ life would arise 'if the f i

underlying assumptions were grossly different in actual service. y $

Thus. new and continued focusYshouldtb'>on'theisafety;related e

l' cwhich would; provide.the indication that_an M

' l}sg.

hardware's condition,iwas:

underlying assumption m y =va.g.m flawed.l p'.:p g[ 9 ,

E L 9 Equipment's conditionfcan beh scer'tained with intrusive periodic i testing, but this may cause more failures th*n leaving the t equipment installed.2,Non-intrusive condition monitoring isN preferred.to monitor the temperature of individual pieces of equipment,  ?,. no intrusively. Thus, tin additionito ambient temperatures, it is ~

possible to obtain temperatures actually at' and on each device. y(

Knowing the temperature of each device allows the assumption of

' temperature exposure.to be verified. Data taken to date indicaten i an equipment population which for.the most part is operating at significantly lower temperatures than were originally assumed. { .

For those devices for which the temperature was found to be higher.

than assumed, corrective action can be taken. This corrective action includes recalculation of qualified life and lowering ,y temperatures when possible, it is much more preferable to know y the condition of each safety related device and therefore have W evidence of the conservatism in the qualified life. -

In NUREG/CR-5762ninfrared; thermography was shown to be p sensitivej to age related degpdation,4As;electribal-equipment deteriorates g

QW

.; A-32 1 i

t

.Q.

y

• 2 c .'  % , dr eh ( .

overheating results from poorfconnections and less efficient heat f L.

transfer occurs.

The assumptions that form,the basis of Qualified life that involve g J

material functions and material interactions can be addressed with 5 a few refinements of existing. practice, ,first, existing e

operation, maintenance and surveillance practices provide information on equipment operational state, . Anytime safety A related equipment failseto operate pro)erly, a component root k.

should .)e performed. CRCFA's ^t cause failure analysis (CRCFA)f the root cause finds evidence of should be scrutinized'to see i material function capabilities and/or material interaction or ,

4 otherwise indicates a mechanism which may not have been accounted ~

for in the original qualification. This direct feedback to the CQ process based on the knowledge gained in CRCFA's is very important and necessary to assure that the state of the art in the original 4 aging program was adequate.

Several examples exist where the CRCFA identified a flaw in the assumed conditions. 'NUREG/CP C'34 pages'C5 to C-19 provide some )e.

examples. Additionally, a few recent, failures of penetrations #

showed that the environment was more humid than originally assumed, g[,.6

. j. ("

.)

St9 t [ .A13 m m en_U T if 7.

Based on the information that has been reviewed under E0-tap Action V lt.m 3. the term " qualified life"tappears to be a misnomer. There are {

simply ton many unknowns and uncertainties related to the qual.fication {-

met hodology (e.g. , formulation' of compoundst assembly, installatiun, *$

operation .:nd upkeep of equipment)' to be able to make a specific c determination of qualified life with any degree of accuracy. While it i not possible to establish a specific qualified life, the initial j.

is 1 qualification testing methodology (i.e., 00R Guidelines, NUREG-0588, Cat 1. and NUREG-0588 Cat. 11) does establish equipment qualification i for sorte prolonged, but indeterminate, period of time.

In order to address concerns such as this one, the NRC staff should )

e .ure that full advantage is taken of operating plant experience and PM information, equipment performance, condition and environment T neoitoring, root cause assessment, and trending of information, 50 that '

LO deficiencies can be identified and resolved on an ongoing basis.

implementation of the maintenance rule will help to resolve this issue y' fu active components, and the staff should initiate action to include y nier'trical equirment within the scope of the maintenance rule to fully address this concern. {

ny

~

it. staff's assessment does not mean that equipment is no longer .

Qualified: itsimplyrecognizessomelimitationsthatexistinthestate.{f, fj of E0 technology that need to'be acdressed in a more focused fashion.

Until such time that specific equipment qualification deficiencies are  ;

idantified by enhanced monitoring methods, existing qualification is '

assured by the initial EQ testing that was performed, twenty-five years of research, and equipment performance and operating experience.

A-33

J

y ,

d;;T

'IY '

3 i

ftrally, given the advances that have been midt in our understand'.nq of it wer the past 25 years. .and based on the TMI 2 experience. PRA ,

in qhts, and operating plant experience, some adjustments in the ~p[

re wirament; may be possible and-beneficial to the Industry, the staff shiold be receptive to preposed changes in the C0 methodology that are:

(a @veloped as an industry nitiative, and (b) demonstrated in be a e, 'nically j u '. t i f i c a .

A , r e c 1 4 t + r. h a t, not been estatlighed between art if icial and natural aging P.nr Arvjew Co>nment s : ,

True. In son,a instances'/;paturaicaging can be more severe than .

a.

artificial, Tb!s hasEbeen demonstratedifor cables. Y)

b. A correlation mtv never be established for all components and materials,

.p i This is a valid issue.

d Several attempts have been made and are in progress towards correlating artificial aging with accelerated aging. UL and ,

reliability testing nave established that the same failure $

mechanisms were developed in artificial and' accelerated tests. i The EPRl/ University of Connecticut project on cables and other materials is a long term attempt at answering this questinn. The use of Lomponent Root Cause failure Analyses of actual plant compunent f ailures provides direct feed back and an opportunity to -

corre' ate the artifidtal and natural aging. .

Merits analyttcal resolution: correlation,has been established e

between accelersted thermal and radiation aging and natural $s t hern;al and radiatt on: aging:under restricted conditions and for f a  ;

selected materials. ".However.;it is true that the correlation '

hotween accelerated?Md.naturtlftging of' equipment assemblies is $ l l

modest at best. This' concern it one of..no factors to be taken into account in a re evaluation'of the qualified life requirement ', l The 'ollowing additional views were also expressed in response to other related issues and problem statements: ,

The experience of the last two decades has demonstrated the difficulty of establishing qualified life without large uncertainty. .One alternative ~to the requirement for qualified life is the use of stress testing prior to LOCA . .b testing and enhanced condition monitoring in service (re: 1 l Itate of the art capabilities).  !

i More effort should be directed toward applying the lessons learned during the last two decades to modify qualification ,

requirements to resolve the problem of demonstratino a meaningful qualified life (re: age conditioning].

I

%^

A-34

. . n w

i' f Theideabehindartif.lblalagingistop~utequl>mentinanendof life condition before~ accident testing. Both tie NRC and Industry {

have agreed that artificial aging i:, the best way to accomplish &

+

this. If there is a,better.: Way;of'l accomplishing this goal. we would all like to ht"Jil t .'

cc J. '

Lijff Assessment: 1, 4 p T h i '. is one of the uncertainties inherent in'the C0 methodology that was -(

ref erred to in the staf f's _ assessment of the previous problem statement. '

lo the extent that focused attention is placed on operating plant experience ud PRA information, equipment perfcrmance, condition and t-environment monitoring, root -cause assessment, and trending of information. this issue becomes one.of minor importance. .j 5(roe significant aging mechanisms cannot be accelerated. Some aging meil.anisms required to be simulated may not be sigrificant for all s,imples.

Es pr 1eview Comments:

a. Probably true.

b. This is a valid issue,

c. Some aging mechanisms diay not.bs' adequately simulated because 1) the environment was not adequately identified and 2) daughter &,'

standards have attempted to define the' aging simulation. This is an issue in two areasic First, failures within the last 6 years ,C have identified higher'.zthan anticipated humidity conditions at several plants. Assumedilow.humiditytrequirements would allow dry $

heat aging. Also, IEEE-317 for penetrations, specifies dry heat aging, Given recent experience that Sumidity is higher than

- assumed. *t.is standard.should probably be modified and the state of the art in aging under higher humidity conditions may need to be improved. Other industrics, for instance Navy and other military, do have standard humidity, salt spray, fungus, etc.,

test methods which could be employed.

d. Industry responsibility; best resolved by the industry. The first j' sentence identifies a valid concern; the fact that some significant aging mechanisms cannot be accelerated is one of the factors to be considered in re-evaluating the qualified life requirement. The second sentence appears to be inconsistent, because there is no requirement to simulate aging mechanisms that are not significant. The following additional views were also expressed in response to other related issues and problem _

statenents:

The experience:o)Jtheilai two decades has demonstrated the difficulty of establishing qualified life without large 's 1 uncertainty. -One alternative to the requirement for qualified life is the use of stress testing pricr to LOCA  ;

1 l

A-35

i

q. ..

. ?:4 :. . . _

j %l

, y h$ .%f:hG :h %

7 +h.

~

l testing and, enhanced ccendition m nitoring in service (re: 4 state of the: art capabliities).

- More effortsshould be directed toward applying the lessons-learned during _the lidt two decades to modify qualification.A requirements'.to resolve the oroblem nf demonstrating a.

meaningful qtlal(fiedellfe.(re: age conditioning). +

J f iu.

If one aspect of

e. Ifsomething'canndt"bedone,itcannotbedone.

a policy cannot be' accomplished, that is not necessarily a reason:

to scrap the policyit in cases such as this, examples should be '

provided. Anyone'can make negative statements about something that they disagree with.

M ff Asif Em_ tot:

See the staf f assessmentJf the previous problem st atement. 3

' w. m .

Excessiverelianceisplacedonanalytical.agingcalculationsthatmayy P not i.e as reliable as testing. e,,:clal,y'in older plants. ,

w.

Peer Review Comments: f '3- F

a. Couldbetrue,bdtthesignificancemaybesmall.

?

b. Maybe. Costisa{ factor, y e,, '

c. Industry responsibHity; more NRC oversight may be needed.

Neither accelerated, aging nor. aging analyses are very reliable as methods of estimating qualified life. In the case of l abora toryy. J aging the problems include the uncertainties introdu'.ed b the Arrhenius method to thermal aging, uncertainties introduced by extensive contraction of the life simulated to the laboratory aging time, and the fact that itInisthe notcase practical of aging to simulate analyses,some it-significant aging mechanisms.

is difficult to establish a mathematical model o' the equisment!

Basically, it is not feasible to account for aging in a techni-cally rigorous way; regulatory oversight can at best rule out a'ny excesses in the7 aging component of. qualification, ^ The following taking into a additional account the limitations of the pr cess.

views were also : expressed .in response to other related issues _and problem statements:

- 4 The experience of the'last two decades has demonstrated the difficulty of establishing qualified life without large uncertainty, One alternative to the requirement for qualified > life is the use of stress testing prior to t0CA.

testing and enhanced condition monitoring in service (rei-state of the art capabilities).

More effortushould be directed toward applying the lessons learned during .the last two decades to modify qualification A-36 r - - - - - - - - - - - - - - - - - - - - - _

requirements to resolve the problem of demonstrating a meaningful qualified life [re: age condi'.ioning).

- The fact that some significant aging mechanisms cannot be accelerated is one~of the factort to be considered in re-evaluating the qualified life requirement (re: acceleration of aging mechanisms).

It is true that the correlation between accelerated and natural aging of equipment assemblies is modest at best.

This concern is<one of the. factors to be taken into account in a re-evaluation of the qualified life requirement (re: .

correlation between_natura.1 and. artificial aging).

1  :- .

o

d. Theintentofthisst'6{ementMs'n6t"' clear. Analytical aging calculations are (or should always~be)-based on testing. The problem in older plants is that there were neither artificial aging nor testing. Note thatuin this context, artificial aging ,

and testing is synonymous. ln addition, if this statement is suggesting that equipment in o'eder plants should be tested to <

determine its capability after having been in service for some period of time, than 1.say that a test such as that is fine for However, such a getting some appreciation for past performance.

test does not, and cannot, give any information about future performance of the equipment in question. On the other hand, if equipment in older plants is artificially age to some end of life condition and then tested, then some useful information may be obtained.

$'3'LfsM'S5'M31 5"r the staff assessment of the g three previous problem statements.

loo pment ' aging" has typically not been performed in the same F

f om t ional state as it is used .in the plant _ (i.e., energized or de-erw eg u ed ) . f y f. ,

Peer P.eview Coments:

a. If true, the significance may be small.

b. Hay be true in some cases, c Typically!

d Non-issue; addressed by IEEE standard.

e in most cases in the plant, safety related equipment is not continuously energized and thus aging simulations utilizing unenergized specimens is app.opriate. However when devices are energized, such as solenoid valves and motors, the aging in many cases did utilize energized devices.

i A-37

l 1'

' l

,, l

f. Industry responsib11ltylLbestresolvedbytheIndustry. When it l 15 feasible to energite equipment during accelerated aging, doing so may contribute?to the timulation of some significant aging effects better than would:be.the case without energizing.

H'; wever, it is notth1waysEfessible to energi2e the equipment, e.g., energizing 4t'aluated temperature may introduce aging mechanisms that do not exist'in real service. In such cases, an ef fort to account.for the effects of energization can be made by 1 treating the temperature rise:as part of the service cnnditions, i.e., by calculating;the' operating tiemperature of. a component by adding the temperature rise due to energizing to the environmental temperature. Research on this topic is not likely to be J productive. The concern emphasizes the point made in earlier paragraphs that an effort is needed to replace qualifted life as a major element of.EQ.

9 It shou 1: have been.

5.11Lf..fmnr.enl :

See the staff assessment of the four previnus problem statements.

In lieu of attempting to define a

• qualified life." it may be more appropriate to develop. methods for addressing and/or monitoring in-service degradation.

Peer Review CommentJ:

~

a. I disagree. Qualified'1tfe means the time that a component can function in its installed environment and still be expected to withstand the effects of 4 OBE. Without understanding the failure modes induced in the test specimen by the DBE environment, how can periodic degradation monitoring accurately reflect how the component would react in the harsher environment created by the DBE? for that matter, how would we know that the component would functton in a CBE on the first day it was installed?

b. Both are probably needed.

I c. This is a valid issue.

d. Qualified life as currently used is normally interpreted as time, it is more appropriate to consider qualified life as a condition.

As long as the condition of the equipment has not degraded to a condition impacting its performance during a DBA, it should still be considered within its qualified life. The current concept of a qualified life time does establish good controls on maintenance and replacement,' but it is possible that an equipment's condition could deteriorate prior to the attainment of its qualified life.

When this is the case; only the concept of condition monitoring would allow the:1dentification of this degradation.

e. Good point: merits analytical resolution. The following A-38

A mw additionalveenswere$1soexpressedin:responsetootherrelated issues and problem statements:

The experience of the last two decades has demonstrated the dif ficulty of *stablishing qualified life without large uncertainty. Ore ~ alternative to the recuiree:nt for qualified life is the use of stress testing prior to LOCA testing and ennanced condition monitoring in service [re:

state of the art capabilities).

More effort should be directed toward applying the lessons learned during the last two decades to modify qualification requirements to resolve the problem of demonstrating a meaningful qualified life [re, age conditioning).

- IEEE has initistY; w ud7sn'effoWkMhhiider prep'aration of'a standard that will' address' the' ult'of rperating experience inqualification[ ret use of test data).

f. It may very well be: th{$tulettionhasbeenmadepreviouslybut no one seemed to want to or was willing to, develop an acceptable way of monitoring in service degradation.

Sta e s s uspnt . -

- (. ,.

See the staff auessment of the fourth problem (tAtement (above)

Mar. n requirement s f or demonst rat ing (Q (e.g. , one hour minimum op ' at ing iIme, thermal aging,. et;.) fray be too severe and witbout suf' .'ent just if 'r at it a; overali margin requirement 5 need t n tw Detter de' od w't h support ing tec hnit a' just ification.

i P.T 3.ty_iff $QEtB1.5

a. 1 cannot respond without knowing the basis fu the term "too '

severe.*

b. This was done for conservatism.

c. Be careful -- margin can cover a lot of' sins. Specifically.

margin can address unknown and unquantified concerns.

d. This is a valid issue,

e. I haven't experienced.the tituation where the margins have been too severe and thus' jeopardized the safety related ee,.ipment's performance. The use of margins does simplify the concerns abou; test equipment accuracy and seems appropriate considering the uncertainties in predicting DBA environments.

f Industry responsibility; I am not aware of any specific. required values of margin. There are only " suggested values.* Therefore, E0 requirements do permit the adjustment of margins to avoid any l

A-39 l

? P.

The values that are too severe or that cannot be justifled.

problem is tnat it hat become common practice to adopt the

" suggested values;* and anyone who chooses otherwise faces the burden of Justifying the cargins chosen. Incidentally WRfG-0588 states that iargin does not apply to the aging component of CO.

The following additional? view was also expressed in response to another related issue or problea statement

- The correlation # between' margin and its contribution to safety assurante is not likely to be established quantitatively" wit ore h:any_reAlonable amount of research; a it will there remaln largely a matter of engineering judgment (ret ~ margin $ in general).

9 fication is the lack of preciseness in the The technical accident justi.#tetting techniques, the number of specimens scenarios',

tested, and variations in manufacturing techniques. What is the justificallon for not having margin? If a good technical justification can be preserted for not having the current margin But requirements, perhaps marg.a. equi'ements can be changed.

just to state that they are too severe without further elaboration is not justification.

f a f f, ADef)@e,nt. '

aven the advances that have been mede in our understanding of f0 nver '

'hc past ?$ years, and basted on the THl.2 expertence and PRA insights.

.<>me adjustment in the requirements may be possible and beneficial to l 'he industry. Also, to the extent ~that focused attention is placed on I

perating plant experience and"PRAlinformation, equipment performance. and

< ondit irin and envit onment/ monitoring; root cause assessmer t I

t rending of informationksomei.felaxation of the more rigorous ",'

requirements may be warrantedi aThe it'aff thould be receptive to proposed changes in this regard that are: (a) developed as an Industry '

initiativa, and (b) damonstrated to be technically .tustified.

i the need and/or ability to establish post accident qualification beyond l , two tc iour .<eek period is questionable.

P_eer Review Comments:

l

a. I dor.*t understand the issue.

I

b. Beyond [0 scope.

I interpret this c.

This is not clear; merits analytical resolution.

statement to mean that operability does not need to be demonstrated for periods exceeding two to four weeks, i.e.,  !

instead of periods of 100 days to a year. (1 assume it does not l l

l' mean that the LOCA test need not be longer than two to four  !

weeks.) If the question is based on PRA studies that show LOCAs can be controlled with.very little equipment operating after a few days, the NRC should document its position.

As to tF.e " ability" i

A-40 f

t C- __

. . m ~ Q;!gQQVGQQ . . a.Q

- y.ggf.f?yp;. ;];:

j h;fi )@  ;; . ,

• o establish pott-ncident- Qualification for periods exceeding two i

to four weeks, the ability esists; out the cost incr. rases with the duration of the period.. The following additional views were also expressed in responte to'other related issues and problem statements:  ; '

• There is a[htek to evaluate the use of PRAs to justify short LOCAtests(re: lettduration).

• 11 will be useful for the NRC to establish its position on the appilcability of PRAs to equipment quellfication (re:

PRA implication 51

d. A significant number of' plants. and all of the most recently licensed plants, have establish post-accident qualification for 100 days. Some plants claim to have estabitshed post-accident Qualification for.one year (Seabrook. for example, established a post accid?nt qualification timerof one year). One year I:. not a NRC staff requirementi IF fact it'is not even a staff suggestion.

Mi is an exampleswhere post awcident qualification and monitoring provided information far beyond-the 100 days that the NRC staff requires. What.tupport 1 1s there for the :upposition that two to four weeks is sufficient?

kt A!_f 3De.U. mfd.1' Given the advances thetIhave been made:in our understanding of C0 vver the past 25 years. and: based.on the THl 2 expertence and PRA insights, snme adjustment to the reoutrements may be possible and benaficial to the industry. The staff shot.ld be receptive to proposed changes in tne methndolngy that are: l(a) developed as an industry initialive, and (b) demonstrated to be technically juttified, lime and dose rate testing requirement s suggest that equit"*ent qualif it at ton extends beyond the DBA and into severa at:c ident spare.

wnich may not be entirely appropriate.

Peer Review Commenti:

a. [quipment needed to mitigate an ace.ident. such as a LO(.A. must be f demonstrated to be capable of perforning its function, in some cases, this function occurs after the initiating event, this l equipment has to withstand the environment created by the LOCA until called upon to-perform. Examples could be a valve needed to align RHR to theLeontainment sump or RG 1.97 indication needed to assist operators in assuring the plant isfremains stable.

b. Done for conservatism.

I i c. I don't understand ' Time and dose rate requirements suggest that...." Documents.like NUREG-0588 address only DBAs and the required post-accident serv he, and component qualification test reports are very specific about test conditions. What's the A-41

i problem? If severt accidents must be addressed, I think theraf sj no doubt that much presently qualified coutpment cannot be qualified for much service, i

d. I aorce. It is true tl'at the doses u eo in E0 thus far may be /

very conservative, but' i arn convince 0 that this area does not merit additional research work becaust:

Thecurrentaccumulateddose,doserateandtimearebaedf N r

on releases. calculated using.f!D 13444; Through the severe Y:

accident studlespit?has4een shown that this model may(bel [ I too conservative'and) result in higher overall dose estimat~es 1 for (Q purpotesi

} g But, given?thattthe qualification affort is mostly-complete N and it has.been shown that the materials used can withstand these levels, what ($ the issue here? .

f >

it that by recogniting the levels of conservatism, we can relax the requiren ents? If so, it should be taken up under 15 j

' the systematicLregulatory requirements review program.....l i Also, this lisue does not merit any additional research effort under the [Q TAP because plants operating today are not likely to derive significant benefits from such ,

relaxation. The cost of cleaning up the paperwork to use the new reduced dose will be more than any potential bPnefit. >

The ongoing literature review [being conducted under contract for the NRC) should provide confirmatlon that the materf els usedn . qualifiedsculpment'doindeedhavemore'thansufficientcapabilAty towithstandthedotes' currently-used,andthatthelifelimts;

               -                  are gnverned by thermal degradation.

e. Most equipment ~i[Qutilfied for some period of time after a DBA.

This varies depending or the operational funttlon of the safety related equipment and typically varies from I hour post DBA to'2 years post DBA.'

f. Theissueisnotklear;whattimeisintended? I am not aware of any DBA qualification requirement that goes beyond the defin.ition of LOCAs, MSLBs'.cand other HELBr.

All requirements within the scope of 10 CFR 50.49, including time 9 and dose rate testing, are limited to design basis accidents. kl.LLf_ A)McA1: been made in our understandi'ig of EQ over Ges n the advances that have the t35t 25 years, and based on the IMl-2 experience and PRA insights. some relaxation in time and dose rate requirements may be possible.. The staff should be receptive to proposed changes in the methodology that 4 A-42

e. _ . _ _ _ _ _ . __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __-

f i J.) ne: (a) developed as an industryvinitletive and (b) demonstrated to be.g ter*.nicitly justiftad. 7 - 44y: h " double peak

• requiremet)L' (i.e,, exposure to two cycles of 'n& Alatum g j 1"mperature and pressure) 31'not~ eepresentative of design basis s 4 W ittons and may be too p vere. Jp 7 >

3 0 l [cgr Review Comments:

a. This was done for conservatism.  ;

b Ara -a planningtodeVelopnewequipment(andmanufacturers)thatI will benefit from a single peak requirement? c Valid, but the NRC has adopted a conservative opinicn. d I have not experienced that double peak tasting during DBA l simulations was too severe nn equipment. It was changed 'o one - peak testing because no credit was given for the assumed margin that it was supposed to rep .asent. t- Merits analytical resolution (i.e., analyze existing information 4 tu reach resolution). Introductionofthe"doublepeak"inLOCAf, testing was based on achieving reasonable. engineering Judgment, with the objective process. For example.vit introduces conservatism that helps countertheconcern.thst'onlyone'specimenistestedtoestablish{a -j oualification.. WhileMi Caay not be representative of design q. - conditions, no such claim was intended. The question of whether * , ' it is too tevere, and whether alternative LOCA testing profiles - are adequately conservative, may merit more engineering Judgment  ; hased on the extensive accumulated data base of 1.0CA testing (in 4 the USA and in other Countries); however, experimental research I1 , not a pru rity.  ; ,

                                                                                               ]

f Ibc bounding qualificatinn ;,tofiles in Appendix C (NURfG-0588) were generated based on a wide spectrum of postulated accidents, t in some cases, these profiles can be considered to be overly j ' l conservative; however,. in the absence of an approved plant-specific profile, this profile may be used and is considered the m'nimum bnunding profile. In general, this profile may represent 6 hours of superheat conditions followed by 18 hours of saturated

• condition <. The actual degree of superheat is left as an open parameter for, as a minimum, the test temperature is to be 340*F' for the time s)ecified and the test pressure is to be eqtal to or >

oreater than tie containment design pressure. Obviously, the 1 higher the pressure the less superheat that will exist for a fixed ' tempera'ure. See NUREG-0588 Rev..).~Part 11, comment and - Onegj resolution no. 97 for~iadditional'-discussion on this issue. ' should recognite that the curve in Figure C-1 of'NVREG-0588 is . provided for those BWR and I'VR ice condenser facilities which do ' not have plant-specific accident profiles available for use in f their equipment-qualification program. I must be clear here, the A 43 i

3- . y 4b .

                                     "doublepeak'lijnot>anablo)nterequirement,butrather.Itfiito be used in lieu.of usthg a piesit specific containment temper 1ture and pressure design profile (see WREG-0588 Rev.1, Sections 4 1.l(3) and 1.2(2).

l 5taff Asits sSt(tt * , f.1ven the advantes that have been made in our understanding of I0 over the pa'.t ?S fears. and based on the TMI.? esperience and PRA insights. be possible and t>*neft(la).to Some adjustments in the Also, to requirements the extent may focused attention is placed on that inc intustry. nparating plant experience and PRA information. foulpment performance. condition and environment monitorifig.. root cause assessment, and I' trending of inforteation, some relaxation of the more rignrous requirements may be warranted. The staff should De receptive to proposed changes in this regard that are: (a) developed as an industry c-l initiative, and (b) demonstrated to be technically Justified. the generic temperature profile that was allowed by the DOR Guidelines and NtREG 0588 for equipment qu. 'c at i,n ( i .e . . I ,,g f o. PUR s and j i 1,, + 20* f for BWRs) was not fully justified. 5

                                                                                                                                    ;. )

Peer Review Comments!; It was my understanding that the i it.toi: nation was used as a a. toolforscreeningwhenjidginginff'vidutilplantDBAanalyses.' Each plant was required it. ave specific pipe break analyses of all potential line breaks and these were used in the qualification protes .

b. Merits analytical' resolution (l.a., analyze existing information to reach resolution). It would be helpful if the NRC document,ed its rationale.

c. You are stating that the 00R Guidelines and NUREG-05BB allows a generir. temperature profilet the key word here is " allowed.'!

These documents' alto &llow a plant specific analysis. In this case, a plant specific analysis means a plant specific profile. For additional; discussion on this issue see NUREG 0588 Rev.ill, ~ Part II. comments and resolutions nos. 57 and 97. 1 Etaff Asses 3 ment: lo the extent that the generic criteria are reasonably representative of the postulated accident environment this simply becomes part of the uncertainty that is discussed in the staff's assessment of the fourt' problem statement (above). However, in order to properly judge the significance of this issue. further actian is warranted by the NRC staff to determine why the generic profiles could not be justified and how this relates to the plants where the generic profiles were P.ed. Licensees typically do not evaluate the amtient temperatures around 60 equipment, basing equipment qualification en aver.1e bulk temperatures instead of local ambient temperatur(s. A-44 ,

                                                                                                                                           )

WM 3,4; m

[ a 5 i f.u L Birias Commenis: 4 i

4. Probably true, but in most cases this should be accept 4blo and .

normal plant monitorin should ldentify instances where j* s itin i f ic ant .d lic ripgny e s: er.ij t p . p[ b EPRI held a works 6op on "Honitoring Equipment Environments During , Nuclear Plant Operation" in April 10. 11, 1990. Many plants discussed their monitorir1 nethods and many plants continue to add: monitoring, t *ndustry responsibility; more NRC ovarsight may be needed. It is the industry's responsibility to account for significant  ; deviations from bulk 7 temperatures where equipment is in:.1alled..'., andmoreNRCoversig$mighthelpassurethatthisisdone. i d This is ,,.inerally t rue, especially for equipment inside , containment; although liter sees should (and do) take into

onsideration local hot T Wts, in addition, most E0 programs ,

separate the plants into [Q .:ones, and Qualification is often based on the bulk 'emperature in each Zone. Zcne temperature data 1, used in aging calculations to determine the calculated life of , equipmer.. 5,4. .

                                                'E,s
         ' ' aI.[.3D.ed Mtat ?                    [;.                                            .I r.).,r ^                                           g
          '  c cnncern is two fold: (4)?theamountofthermalpreconditiu'iing                      57
         ; tor t o f 0 t est tr.g may nel Save beeti suf ficient given the lor.il ambient                  '
            .nperature. and likewite . (b) the peak anident temperature inat the a
         .suspment was cualified for tday not be tufficient. To the artent that,                   '
         ' , used at tention is placed on operating plant experience and PRA

• M ormation, coutument performance, condition and environment  ;

         -"<.itoring, ront cause nstessment, ar.d trending of informatton. the i.pect of this cancern that deali with the adequacy of thermal aging                   ,.

M om.is one of minor irrpertanc6. However, the aspect of this cnncern " ' ' iat questtons the adequacy of the assumed peak accident it.perature T (as: sampared to the local ambient temperature) should be addressed by the , W staff

         ' ' . regulations do not state the a:ceptance c riter ta f or qualif ying                         !
              .ainmnnt tiased on operating experience.                                                     i N tr Review Comments:

4- Probably true, although it is difficult to demonstrate qualification of components that will see a harsh environment , using operating experience since that experience will not dupittate a LOCA or HELB. l ti . This is a valid point

c. To my knowledge, the use of operating experience to qualify equipment has not been practiced and is frowned upon. Thus, it C is A-45 i

                                                                 ;.s                                           .

_q-

a. .

4 5 probauly unnecessary to try to;tstablishiacceptance criteria for its Me as an [Q method. g Ji Mertt' analytical resolution; existing standards and RGs permit 4. the ".e of operating s Nerience essentially as a way of estat'ishing qualified sife. Hcwever, the rettrictions are such that it is rarely a practical approach. The ultimate acceptance -- criterinn is the provision of reasonable assurance that the spei 'f led safety f unction can be performed under applicable f.er. e conditions, including accidents. The following additional vien were also expressed in response to other related issues and  ; prob'em statements: -

                  -           IEEE has initiated an effort to consider prepar.tton of a standard that will address the usn of operating experience                                           ;

in qualification (re: use of.. test data 1vs. other methods).

                                                                                                                                 ;t .

More ef fort should be directed toward applying the lessons E learned during the last two decades to modify cualification j v. reautrements [ro: geconditioning). l

                   -          1he experience of the lattitWFdecades hat demonstrat.ed the 6

dif ficuity of establishing qualified life without large uncertainty. One attornative to the requirenent (nr qualified life is the use of' stress testing prior to LOCA f testing and enhanced conditlen monitoring in service (re: stateoftheartcapabilities). .

    $taf,f,_Au                ner11-                                                                                                ,

Given ibe it antes that ha., been made in our understanding of [Q over .

    'he past " years, and based on the IMl.2 experience and PRA insights.

u te a d.) o ' wr.t s in the eoutrements may be possible and beneficial to the indus' .. The st af f shnuld be recept ive in prnpnsed changes in this j

segard t' ' ve (a) developed as an industry initiative, and (b) ,

f itemons t r a ' to :;r technically justified. Given noic realistic assumpitons for the release fractions, the timing of the rr " v. e . ti.e chemical form of the release, and accident mi t Nat io- 'fects resulting from equipment response an immediate and 3 1arge .no, c term (llD) may be overly conservative and inappropriate.

     & er pg.rt.ew rp m enu:                                                                                                         '

l 3. Coul'1 de tru .

b. Irw C. lhis is a valid issue.

f The large radiation dose used in E0 programs in the US does seem

d.

to be overly conservative. ,,

j Modern sophisticated electronics are more susceptitle te lower i A 46 f l [ . - - - . - ~ . - - .  !

3:  ; radiation doses and thus the u$e of inli equipment is jeopardized by having overly conservative radiation requirements. The alternative to new toutpment is reliance on old less afficient and less reliable technology,

e. Another source term (9 raft NUR[G 1465 source term) has now been '

approve by the HRC staff for use in the CC-System 804 plant desion. There is no requirement to'lWitCh to the new app.oved r r ', source term; however, those plants-interested in switching should j In addition, contact the NRC staff and for further discussions. the staff is also in the process of review ng an additional sourr.e term for the AP-600 design. (Consecuently, there are currently two staff approved source terus available (11D 14844 and Oraft NUREG-1465 source terms) (or use in E0t and upon completion of the , NRC staff review of the source term proposed by the AP-600 design. there could be a third . kt> ' ' lufnatn1: 24 G i v. the advances that have baan ade in our undarstanding cf f0 over the past ?$ years, and based on the TMl-? expertence and PRA instuhts, so" adjustment s in the requirements may be pntsible and benef ic ial to Ihe a<1u s t r v . The itaf( should be rer apt ivc Io proposed (hanqcs in this rec J that are (a) de. eloped as an tejustry init iat ive , ,e n d it t dne at rat ed to be t echnMaily juttif'ed. A q.....ral e emptinn for radiat ion quai t ficatico t e st ing of equiptront exp s.ed to low level radiatton may be well suited 'or (Q purpose. under y..* ,,,

                                                                                                                 ~

c er. . in def ined circumst ames. Ete7 RtilDLCgymg.aL' .

                                                                                                          ~

a. Could be true. '

                                                                                    ~

b. I thought that we had a general ex9mption at 10' (or 10') Rads except for solid state electronics, if not. I agree there should l

j be one. 1

c. This is t valid. We still do not understand long-term. low-rime effects

d. Merits lytical resolutten (i.e., analyre existing inform. tion to reac esolution). This noint has been discussed at length by the [3 anity; and relevi '. information exists in the 4 literat .

It should be feasible using existing informatton for the NRC document its position,

e. The NRC aff position on this issue is that electronic equipment

' within e scope of the E0 rule that will pe exposed to total integra d doses of radia'.lon less than 10 rads (10' for other ecutome ) is considereri to be in a mild environment. Therefore. nviron ntal qualification in accordance with 10 CfR 50.49 is not D require This position is based on literature searches. comments A-47 s I' C i

w , from industry, and HRC experience. If new or additional data 11 available to support-a position different from this, that data should be presented to the NRC for review,

                                             'l                                                 g M d L.A11111Et.qt :                                                                                '

See the staff assesseent of the previous problem statement.

   -   I r. areas designated as radiation harsh:only environments or high energy 1o e break (HflB) areas, the conservative assumptions used in                                      i c a.i ulat ing radiation levels make it dif ficult (if not. i.mpo'.ilble) to                       '
 '     opgrade to the more modern and efficient digital.equtpment.

feer Review Comments: can be adversely affected by

a. Digital rquipment (and analog) issue should remain on ensuring safe radiation so the focus of the If licensees can justify a lower ,

oper:. tion of such equiptnent. level of radiation exposure than originally determined, this new v.;ue should be ellowed. If this comment is directed at the methods used to calculate the b. potential radiation environment, it should be discussed with the Nevertheless, it is known and widely Radiation Proter:lon Branch. , accepted in industrysthat electronic (equipment (which includes a ' -o significant amount of the more' modern end efficient digital ' 4 equipment) is more susceptible to radiation damage at lower thresholds than other equipment. If the current methods of is calculating the amount of radiation in a given environment J inatcurate or incorrect 'then perhaps a fresh icok at this situation is wsrranted. However, evidence to support a new investigation stiwuld be presented* to the Radiation Protection Branch. St4'f_.A>}p u me11: See t t'e it af I assessment 5 of Ihe two previous problem 51 atomer.t ;

     -                 not t>e appropriate f or the NRC to rego.re licensees to eit ner l',may' ; tomant the new source term "across-tha-board" or not at all . Instead c' allowing licensees to use the new so"rce term 'nitially for              dist. rete Allowing Ic at tom (such as in E0 radiation-narsh enviivnments).

we (lcathility in applying the new source term wou ld allow lic.cnsees

• rec alculate the exposure levels of some E0 components without

' e.;ending sifyificant resources to update all of the post-accident I"ocedures

             .arce term could anohelp calculations to eliminaterelative    to E0. cflexible some components      from theuse cottlyof the new
          . muirements of 10 CTR 50.49.

f_CE.lrview Cetdi: a. The focus of the issue should remain on ensuring safe operation of equipment. If Itcensees can justify .a lower level of radiation A-48

i

 !                                                                        vi
;                                                                         j}.

1 vw ... i esposurethanoriginA1)y. determined.thisnewvalue.shouldbe e

$                                        alinwed.                         #Q *

{

                                                                          'i f j               b.                       This is not valid. You can not have it both ways: you either use the new snurce tern orfyou do not,

c. Tre intent of this c~omment is not clear. It seem as if it is ga b.aing sucgested that,the NRC should permit partial use of the old and partial use of the:new source terms for EQ. If this is the case the.i wewouldbe!creatingathirdsourceterm. If someone j

15 proposing a new source term for.[Q only, this proposal should j a ne presented to the HRC for review and comment. Otherwise this i usment seems technically diseriented. SLAI.f_P elintal:  ; If See the t a ' f a '. s e s sinen t s ' the'three previous problem statements. ,

                                                                          ?2                                                                  ,
              - The cost of qualification ; test 899 is a barrier to the introduction or adap: at                       i cf new productt',ihto tht nuclear industry.

ti j Deff R vL >w Comments

1-]"

[Q i i t. 1MI d ronstrated the need for EQ. While cost is a factor, safety mm.tbemaintained,f}

b. No doubt about it. I

                                                                          *1

c. 1h* current NRC research program may result in reduction of some reoutrements. ,

! d. I do not agree. Cost her se should not be a basis for modifying - j [0 requirements; reducing the cost can be justified only by ! c'inges that do not compromise reasonable assurance of safety. Vnile the statement is a reasonable observation, it does not by l j - ' . elf sugge:t a course of action. j t { l e. Qualificationtestinghsnecessarybecauseitprovidesinformation about equipment that is unattainable in any other manner. at I l in. ores quality, provides ccnfidence, traccability, and most of I all it provides and: demonstrates safety; . r te .= .

                                                                                                                                 ,4 l                                                                              e Su ff 4 , w m.fnJ:

l

                ;).cn u.. advarces that havn been made in our understanding of [Q over l

l l the past 2'. jears, and based on the IMI-2 exper tence4 PRA insights, i plant oid ating experience, some adjustments in the requirements may be l p o n i b ! -- cnd beneficial to the industry. The staff should be receptive l to prar .ed changes in this regard that ar;: (a) developed as an industr3 initiative, and (b) demonstrated to be techni" ally justifu l.

2. Addittoral technical issues and other considerations that were ident it ud as ' potent ial issues

• related to the E0 methodology include:

l A-49

   '-8* -"            - . - . - , , , _ ,

(dagI.Al .Put Review ConcenL1; 4 The 0,i'owing technical issues

• highlight the need for the margin that 'ias been built into the procets.

b. While several of these issues (e.g., the effects of hot, humid I

l condt'lons; the effects of steam leaks and inadvertent spray actuat ton during normal-operation;J deformation:of cable jacket and . l insulation at high stress points! and'others) ;must be. considered during EQ testing, they are not issues that can be (or shoJ1d be) resolve within the context of.the requirements of the CQ rule. Rather. these are normal' operating conditions that'need attention

• l i

and should be addressed by maintenance programs,..[These Conditions existed before there were EO requirements and they would still exist even if there ware no E0 requirements. When they are .' resolved for normal 0,erating1 conditions, they will'no longer be a i l Concern for [Q. l

c. The following additional' technical issues should;be added to the list l

                .         iire rotardant insulation and jacket materials contain halogen compounds that are released during irradiation and thermal enosure. Halogens, in particular chlorine. trapped between the jacket and insulation. can enhance the logradation of the insulatinn. Aging cables without a tacket can provide a nonconservative estimate of insulation life.                      v       -

g .- l 11 has been established'that low dose rete.hns a greater impact on degradation than high dose rate. A dose rate j salue should be established that will provide a conservative i estimate of degi Nation. .rihis value would then be used to irradiate materials to $1mulate field service aging. rer the m>jority of instrumentation cables, the leakage currents have not been measured during a DBA. Leakage currents should be determined for the whole cable system l including spilces, penetrations and end connections. I 9 !!U.).11 M!)1 : . 10 t h e e x t <> n t that focused attention is placed on operating plant c.rerianc e a nd PRA infornation, equipment performance. condition and environmer' monitoring, root cause assessment, and trending of

     $nformat1.                  the cable degradation concerns are (,( minor 1mportance.

With rega

• to the third concerit pertatning to leakage currents, action is requir. Dy the NRC staff to better understand and resolve this issue.

 . qa l i f ic a t        in. c ert i ficat ion of testing laboratories:

Peer Review Comments: A-50' , RAAE

                                                     -           ~ - - - - - - _     . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ _______   l

a. Thiswasabiglssuearound.the1983timeframeattheC00 level.

Itfall.th, rough;bocnuteit,wasimpractical. >

b. I disagree. Ah 6ne of!thote' involved in the long and arduous' effort on this topia during the early 1980s. It is still fresh in my memory, and every one involved agreed that this is ant desirable. Iffit was consioered then to be unfeasibi-economically, and undesirable technically, what has changed now to warrant its reconsideration? Do we really want to be lulled into a false sense of security provided by this " accreditation" process. Even its big croponent, IEEE. wh;ch published IEEE-650 has withdrawn the standard for lack of interest and acceptance by the labs and the industry,

c. This is a valid issue. )

1

d. Qualification and certification of testing laboratories should be i l

performed usin[ common non-nuclear industry processes and imposing ' ruclear QA requirements. e, 1 do not agree. Considerable effort was expended on this topic about ten years'ago. it.was concluded that, while there might be some benefitsito .ertification ~the level of (Q business was too m small for:testilaboratories to. absorb its costs. Since the. level 4 of EQ business is much: smaller now than it was ten years ago the prospect of implementing a certification program nnw 1ppears to be , negligible,

f. Examples should be cited when .tatements such as this are made.

51df_ 6ilt.ument : The siaff atirei., with the view (stated above) that queiifIcation and certification M tetting laboratories should be establi>ned using commonu non-nucl i industry processes and imposing nuclear 0A raquirements. W1l'e it the responsibility of the industry to ensure that testing laboratories are adequately qualifted to perform C0 testing the NRC staf f should assure that Qualification teing is betrg pmperly and c o n s i s '. e n t l y oerformed through develop 1,ent and implemes.tation of ongoing [0 audit and inspection activit ies.

     . determination and resolution of worst-case electrical carlitionu f.gn_ Review Commentsi.

a. I disagree. To the best of my knowledge, to the extent practical, such conditions have been included in the qualification programs, it is true there have been a few lapses. This is an education problem but not a problem of lack of, or clarity of requirements, i0 was a rapidly evolvint technology when much of the work was done. Surveillance, maintenance and condition monitoring, if properly implemented, should take care of any past lapses.

Training and education of engineers involved in specificatie1 of EQ requirements-should minimize, if not prevent recurrence. A 51

                                      -f,.               ,

O & ym. *

                                         ' s           _

e,' Jgp _ q.gy;gg, j . _

                                ;. 8m .         e L.Q       ,

b The statement it too'non specific.-

c. The electrical' extremes of most equipment is known and '

demonstrated during qualification programs. A valid

d. Industry responsibility: merits analytical resolution. overlooked in i concern since worst case conditions However, thisare does sometimee not absolve the nuclear currrnt E0 practice.1 . o industry from identifying and resolving worst case conditions asA tightening-part of its responsibility to operate plants safely.

of E0. surveillance,;and maintenance requirements can help assure that adequate attention is given to; worst case conditfors. c Staff Assenment: ' While the staff agrees withithe view (stated above) that ific the tonditions. licensees - 4.e responuble for identifying and resolving plant'-spec tr.at were i.st preperly accounted for during E0 testing, the NRC staff- . should: (a) assess the informat8 00 gained.over the past 25 years.

                                                            ~

portaining to E0 and wortt-case ...ditions.ard Ib) assess past exist. concerns 04alificatton The NRC staff practices, to determine if any specificshould assure. atcounted for through development and implementation of ongoing E0 audit l a'd inspertion attivities.

          .olution of radiation and temperature stratification effects and
     '.".1rugen F.            buen scenarios; l

Pyer Revie d omment_i: r a i agree that such stratification effects may not have been Ind ed, they  ; addressed in the qualification establishment phase.  ! could not have been because these are site and configuration specific problems that could not be addressed in any generically developed program such as E0. Only. review of operating and failure exparience W and root causeJanalysl> could address these > effectively. I am familiar 1with some instances of such conditions having been identifi6d and corrected through vigilance in maintenance and root 2cause analysis programs. "herefore. I must conclude that valid as this may be.as an EQ issue, it merits research attention only to period 8911y review failures and root This function is effectively being causc: and notify licensees. performed through the existing NRC generic communication program. My recollection is that the EPRI test program performed during the mid-1980s. did show that hydrogen burn is not a significant concern for harsh environment qualified equipment.

b. Tnis is a valid issue, How will you address -

c. They orobably do (at least hydrogen burn). l hydrogen burns? l A-52

                                                                                   .                                                                      i l

d. Hydrogen burn testing was dond by IPRI in the 1980's to estabilsh 3 that the impact on.Otherwise harsi, qualified equipmeat was insignificant. e l'

                                                                -@         , .    .5,,

3 3-

e. Industry responilliil'ilff meritian11yLica11 resolution. The industry can-contribute tc resolutieft.cfctemperature and radiation stratificaticn effects in.part by monitoring environmental conditiont in.apppppriate parts of the plants. ,

The NRC conducted # an extensive .investigattor, of hydrogen burn scenarios about a; decade ago, it is not clear whether the concern is due to 1ack of knowledge.of-the'NRC's resolution or disagreement with the resolution. If existing informatton on this to;,ic is' inadequate or not adequately promulgated, the NRC can 4 taKe corrective action.

f. Hydrocen burn scenarios are outside the scope of the E0 rule.

Temperature and radiation stratification that result from a oesign basis accident are willi ni the scope of the E0 rule, and equipment i within the scope of the rult s'ould be quallfled to withstand these conditions.

                                     ).O_f_f Ass e m (:

i The NRC staff should assets.the information gained over:the past 25 jt m rs pertaining tu [Q.: hydrogen burn, and stratification effects, and

                                     .f ot ornine if a significant safety problem exists.
                                   - resolution nf MSLB vs..LOCA environmental. conditions; Peer Review Comments:       .

I

a. This is not.an issue. ,

j

b. I don't appreciatesthe MSLB'vs -LOCA! consideration because.E0 'l testing uses envelo' ping techniques to establish worst case requirements. When an item is required to be qualif ted to both M5tB and LOCA conditions. both proUles are enveloped by one test.

i 1 which is inherently more severe than either.one alone. The issue is not clear. - \1aff A51csiman,1: this concern was identified by the staff Meause M5tB conditions (which could be more severe than LOCA conditions' were not recognized during , initial cualification efforts, further. Pe subsequent practice of enmparing the M5tB equipment surface temperature to the bulk LOCA temperature as a basis-for MSLB anlification was questioned. Action by the NRC staf f is neces3ary to de'6 mine, based on the additional

  '                                   information that has been obtained over the past 25 years and based on past Qualificat'.on practices, whether this concern represents a significant safety problem, and whether additional measures are warranted.

t A-53 1

l I

                                                            -i4                      ~w;

( . :% . *

                                                                                                                                      -p y

o selfheatingef.fectsofcableslj"I 's-  % y..ne g4g,q'. '

3 ", J P_eer Review Coceents: This has been. addressed by assuming a cable operating

s. I disagree. .

If i femperature is at rate: 96ditions of very close thereto. ! cything, the industry practice is probably more conservative than l *t needs to be given the derating of cables in their applications.

b. This is a valid issue.

f . I

c. Self heating effects in cables.has been typic' ally considered.

This normally would effect only power cables and cables are normally powered to worst case condillons during DBA testing. I I do not agree. Cable qualification practices do account for the

d. .

self heating effect of cable energi2ation; in fact, the industry l

laims that this was done too conservatively in some E0 programs.

          ;14Lf. 4 ku lm_tal.

Based on the peer review commenO:(above).sjt,. appears.that self heating Also, to"the - effect( thathave been Cons (dered at$liatt' to some ' degree. focused at;ention Is'pl. iced on operating plant eyperience extent information, ( 1 PRA equipment' performance, condition and environment i l n.or,,toring. root cause assessment, and trending.of information, th s { issue becomes one of mince importance. diffo olties in simulating accident ~ conditions (e.g., containment spray. hum 1ri t t y , etc.); l i - l Peer Review Comments:

a. This is not a valid issue,

b. Properly equipped commercial-laboratories have no problem f simu'.ating accident conriitions. 1 j

c. The issue is not clear.

l l I 11 Alt h_LUJEtal o

                                                            ~                                                               ditions   n .'

4 There may be some vari n tont add'uncertaintiet"in the' actual con l that are established in the (01 test chamber, but these should be minima Recognizing that the ] tf the chamber is properly instrumented. J j postulated accident conditions were determined through analytical to 3 methnh that also ccntained inherent uncertainties, the concern seem be of little significence. an onqeing audit and inspection program for E0 testing laboratories will J Also, see the staff assessment regarding help to address this issue. l certification of E0 testing laboratories (above). j the effects of hot, humid conditions;- f . i i

 !                                                                   A-54 9

i _ t- - r

                                         $,1               y Peer Review Coments:                                                                   h

a. I agree, humidity canrot be accelerated. Heat in combination with '!

humidity may apply to specific locations in some plants. Generic prugrams such as E0 nnnot address this. effectively. But properly implemented surveillarce, inspection. maintenance, and root cause analysis programs can be effective. Additional research cannot satisfactorily' address this.

b. This is not a valid issue.

c. Thelangtermexposur7ofmaterialstobeatandhumidity, if not Identified as part of the assumed environment, Will cause differences in performance...Someseffort Jhould be made to identify the environment properly and hesearch. simulation p

techniques. The military style timulations may need modification..

d. Merits analytical resolution' (i.e.. . analyze zekisting information to reach resolution). -This is a valid. concern in accounting for aging degradation, because i ere is no practical 'nethod of accelerating humidity effects that can be correlated to ., '

qualified life. However, there are humidity stress tests that can at least provide some indication of the endurance of equipment I under humid conditions. Humidity stress testing has been ' incorporated into at least one standard, IEEE Std 650 for battery chargers and inverters, h!4'!._ h.Le ,_g.a1: _ 10 the estent that focused attention is placed on operating plant e.p.rience anc PRA information. equipm .t performance, condition and environment monitoring, root cause assessment, and trending of n.'ormation, this issue becomes one of mino importance. However. given. the AJvances thu have been made in Our under$tanding Of.EQ over the pa .t 25 years and based on-the THI-2 experience and PRA insights, some ad mtments in the requirements ma) be possible and beneficial to.the indu st ry. The NRC staff should be receptive to proposed changes in this. regard that are: -(a) developed.as an. industry initiative, and (b)~ d ronstrated to be technically justifled.

   -   IFo effects c steam leaks and inadvertent spray actuations during nwnal operation;
       ,ner Re.iew Comments:

P

a. I agree, this is a valid concern. But, it is one where only experience in the plant can guide the determination of areas of vulnerability. 'It may apply to specific locations'in some plants..

Generic programs such as E0 cannot address this effectively. But, properly implemented survcillance, inspection, maintenance, and root cause analysis programs can be effective. Additional research cannot satisfactorily address this.

                                          .s ,

A-55 X

b. This is a valid issue; what about operating experience to address this?

c. Equipment qualified, to ha'rsh environments should not.erperience.

problems when exposed;to' steam:leakstanotinadvertent spray actuations. 51f they co; then the qualification is suspect until the issue is resolved'. ~

d. Industry responsibility; merits.an'al'/tical > resolution. This is a valid concern. Whure experience predicts that certain equipment in certain locations is expected:to be subject to steam leaks and l tr.:dvertent spray;actuations, the EQ program should incorporate testing (such as humidity stress tests) to simulate these service l

' conditions. Where such events were.not anticipated in the E0 program..the existing program _should:be. supplemented by additional testing or enhanced condition monitoring to assure that equipment is refurbished or replaced when it is no longer able to operate as requ' ed during a DBA. S t J [! _ Auclimf1Lt

• Ihe staff agrees with the view (stated above) that properly implemented

                          'urvetilance. insp .tlon, maintenance, and root cause analysis prngrams can be effective in addressing this problem. Therefore. to the extent that focused attention is placed on operating. plant experiente and PRA                          . !

i n f o rma t i on eoui pmen t . pe r f orma nc e , ,: cond i t i onia nd e fiv t ronme n t monitnr %g. root cause assessmenti and' trending ofnformatton, this- j issue becomes one of minor importance. . J

                     . deformation cf cable jacket and insulationLat high stress pnints; l

Peer Review Comments: I j

a. I disagree.. l recall that.some work was done by Sandia and TVA to /

answer this concern. That"should.be sufficient to' address the concern generically. Still this' concern v.111 remain because generic programs'such as E0 cannot address this completely for all variations. But, properly implemented surveillance, inspection. maintenance, and root cause analysis programs can be effective. Additional research cannot satisf actorily address this issue fully.

b. 1his is a valid concern; what about operating experience to address this issue?

c. Sandia did perform some tests of cables hung over sharp edges and concluded that cut through did not occur. Given standard cable pulling practices and cable-routing practices, sharp edges is most likely a rare event and not considered a comon. mode f ailure.

d. Industry responsibility; best resolved by the industry.

11Mf Assh5Fmf.nl: A-56 g  ;

95

                                                                 ,    . s . 4.

kW . . . . ,

                                                                         , :v.

See the staff assessmentire the effects of steam leaks (above).

• dust effects; Pfefr Review Comtents:

a. I agree. Generic programs such as E0 cannot address this offcctively. But properly implernented surveillance enspection, maintenance, and root causa analysis programs can be effective.

Additional research cannot satisfactorily address this.

b. This is a valid is:ue; what about operating experience to address this issue?

c. Most harsh ' qualified items are qualified assuming exposure to sprays and steam,-or otherwise are protected.by seals and sealants, which'have documented effectiveness-in.their own EQ tests. Thus, dust. particles..whichLare assumed to be larger, l would have littla effect. Additionally, most LOCA' simulation chambers would contain cons.derabloocontamination as sprays are recirculated,-thus. simulating. dust and,other contamination conditions.

d. Industry responsibility; best' resolved'by the industry.

j Elaff_AE n W nl: See the staff assessment re: the effects of steam leaks (above).

         -      long-term esposure to moisture; peer Re iew Comment 5:

a. lhts may be the most significant concern,

b. I agree. Generic programs such as EQ cannot address this effectively. But properly implemented surveillance, inspection, !

maintenance, and root cause analysis programs can be effe'tive. l Additional research cannot satisfactorily address this. l

c. This is a validilssue; what about operating cd erience to address this concern?

The

d. Industry responsibility; merits analytical resolution. ,

i following additional views were also expressec in response to other related issues and probicm statements:

                         -      Where experience predicts that' equipment in certain locations'is expected to be subject to steam leaks and inadvertent spray actuations, the EQ program 55ould incorporate. testing (such as humidity stress tests) to simulate these scivice conditions. Where such events were not anticipated in the E0 program, the existing program 4-57
                           - -=

should be supplemented by additional testing or enhanced condition monitoring to assure that equipment is refurbished or replaced when it is.no longer able to operate as required during a DBA [r.e:athe , ap, - effect,s of. ge. steam leaks]. .

                                                                                     -        Humidity represehlsja val {d' concern.in accounting for aging degradation,'because there is no practical method of accelerating humidity. effects'that can be correlated to a qualified life.a Howeveriethere.are humidity stress tests that can at~1 east' provide'some' indication of the endurance of equipment underchumid' conditions. Humidity stress testing has beeniincorporated into at least one starJard:

IEEE Std 650 for battery chargers and inverters (re: the ef fects of hot',yh,umid conditionsji

                                                                                                                         <      +                    .

5U.lf A51R11B D1: foiused attention on operating plant experience and PRA informatton, equtoment performance,~ condition and environment monitoring, root cause a"( sment , and trending of n 'ormation, will help to address this con 4ern. However, NRC staff action is necessary to determine, based on th" additional long information that has been obtained over the past 25 p .n . whether term exposure to moisture rr.> resents a significant sa'ety problem and whether additional mansures are warranted.

                                                                       -    the   impact of fire scenarios.. protective f.eatures, and fire fighting act'vities, including (for example) smoke, fire-retardant                                     coatings and broken         and or leaking Iie. barriers,     fire     suppression system     actuation           ,

flammable gas lines: -

                                                                                                                                                           .c .
                                                                                                                    ,   ?fyl Prer Review Comments:

a. I agree. 'his is probably a v'alid concern and may not have been fully addressed in existing qualifications. A study of the expected effects,. the. methods for factoring them in establishing aging parameters, and determination ~of the levels of margins that may exist in current'a'ging programs ~is'sppropriate. '

b. This is a valid issue; needs to be addressed.

c. Industry rernonsibility; merits analytical resolution.

Ltaff Asif.11mfE1: Aitton is required by the NRC staff to assess the impact of fire u"narios on E0 and to determine if this represents a significant safety p oblem and whether additional measures are required,

                                                                         -    the consequences of combustible gas and chlorine formation; Peer Review Comments:

a. I d'.sagree, this is plant design and physical facility 7 q..

A-58

                                  .            W $ 'l&               '

5 M4k. .x configurationtontr6)T',$1ssueMNt?should

                                                          'N' "d*as not ~be.; treated issue.                 .

b. This ts a valid concerni'needs~to be addressed.

St4_ff A,331mpn,1: J n Attton is required by the NRC staff to assess the consequences of combustible gas and chlorine formation and to determine if this reprasents a significant. safety problem and'whether additional measures

                                        'n              - -

are reautred. .

• interface effects between components that are tested separately; Eter Review Commenti:

1 agree; with some exceptions, the test p-ograms completed thus far have addressed interiaces mostly separately. But, it should a be noted that they have'been addressed in the context of the overall qualification of thi equipment. item to assure that the saf ety functional .capabil.itysofit.he..e,quipment. in, harsh . _ c" environments' are notfdompromised.WWetmay' argue Jaboutwthe acceptability of some'>of the evaluation practices employed. Barring a detailed evaluation.of~the various configuratinns of interf aces, and determining,an acceptable method for each configuration-(of course, we may need tests-to validate), it will be impractical'to determine whether or not there is an EQ issue here that could potentially challenge what has been done to date. From my knowledge uf_'what has been done in qualification, and the design installation,3 maintenance and testing aspects of several classes of equipment Jmy opinion is'that such an effort may not be. cost effective.mindiv.idual. evaluations-on a case basis, particularly based ontf ailure expericnce review (which incidenthlly is-the current NRC and industry practice) may be more effective. b This is a valid concern; needs to be addressed, c interfaces of equipment are qualified along with the equipment. The test laboratories have to interface to the equipment inside chambers and thus this knowledge was passed along to the industry in EPRI NP-5000, " Handbook on Electrical interface Sealing,? 1988...

d. Industryresponsibil'((y;'moreNRCOVersightmaybeneeded.

5tjff Assessment: y ine staff agrees with the view (stated above) that licensees are rnponsible to ensure that equloment interfaces are adequately j qualitied. Focused attention on nperating plant experience and PRA information, equipment performance, condition and environment monitoring, root cause assessment,-and-trending of information, will . help to address this concern.. However, NRC staff action is necessary to.  ! A-59 i I I

                                  'y: -          . v .u        . . w ,, , . y                              *-' -*

.; (:7 w :p.fvL

                                      .g     a.
                                                           ..: _ J g c

n i, {

determine. based on the additional-taformation that has been obtained, o.cr the past 25 years, whether specific.equipmentsinterface problems ' have Deen identified that should be addres' sed.

 *   (') boundary restrictions (e.g., the compensatory actions that must be taken for a short-term breacn of' mild-to harsh area E0 boundaries areduring W re n ive when compared with the probability of.a LOCA,HELB event I     a maintenance evolution):

Peer Review Comments: m ..

a. Iagree,this'is'avalid36nEeff{b'ut)*lt.is'not?onesthatcouldbe, NJ '4 resolved by additional resefechhtitilshouldabetaddressed throughN plant configuration'controlfprogramsh'

                                                .+                 3

b. This is a valid issue;.needs.~to.be~ addressed. -

                                                                      /

c. The assae is not clear.

                                    ^

d. Engineering judgemenE and conson'/sekse sh'ould resolve this issue.

                                      , . .m             j 2 Et.aff Assessment:                                                                                 7
                                                      . wgg.: -                        -

t.Icensees occasionally find it desirable to remove or in disable orderC0to boundaries (e.g., flood walls, concrete plugs, etc.) facilitate maintenance activities. However, removing the barrier may

       'ause otherwise operable safety related eouipment to be exposed tolified i

F rsh-environment conditions that this equipment ibis is ahas notproblem valid been cuathat is

        *nr until the barrier is reestablished, tmst addressed by industry. and the NRC staf f should be receptive to l                                                                                   (a) developed as an prr; posed rcdolutions of this problem that are:

ndustry initiative, and (b,)' demonstrated to be technically justified. v

                                                                ~
   -    scheduling constraints, ont. performing periodic maintenance (i .e.-..E0 components are considered ~to:be inoperable at the end of their qualified i

life, even though the method'useo to. determine

• qualified life" contains j many assumptions and conservatisms)p s t e -

Peer Riview Comments:

d. } 39/ee. ThisconcernarisesbeCi.Jseofourtreab$lentof qualified life as* cast-in-concrete number, despite all of'its uncertainties. This: concern,will be. moot.if we can agree that a given the nature of the uncertaintles'behind the qualified life estimates, it is prudent to treat it more as a yard stick'than as an absolute do-or-die number. As an interim solution, the NRC might consider allowing the utilities to use the 25% grace period that is now allowed for lech. Spec, surveillance, for EQ maintenance also. A justification can be developed for this,

b. This is a non-EQ issue.

c. Great care should be taken in considering the granting of I

A-60:

                                    -        Mcq.gjys. - 4 l

flexibility for end oft: life determinations. -The reason l assumptions and conservatisms are used .is to deal with a complex l process for determining.end of.zlife,4.These assumptions and I conservatisms must belteplacediwithsequally, valid . inputs before they are removed or altered,

d. Merits analytical resolution. Quallfled life is at best an estimate of the pert _ed for which the required operability is reasonably assured. TTh6!0ncertalnties31n;. qualified life determinations are noticonsistent% ithsconsidering equipment inoperable at the end of qualified. life. There is now a need to evaluate a'ternatives to the qualified life requirement.

Operating (xperience and enhanced condition monitoring are among the alternativos that should'be considered. ~The following additional /lews were also expressed in response to other related issues and troblem statements: Nei*her accelerated aging nor aging analyses are very reliable as methodt of estimating qualified life. Basically, it is not leasible to account for aging in a technically rigorous way.. Regulatory oversight can at best rule out any excesses in the-aging compnnent of qualification, taking into account the limitations of the process (re: reliance?.on< analytical aging calculations).

             -      The experience of the'lastJtwo decades has demonstrated the difficulty of establishing. qualified life without large uncertainty. One alternative to the requirement for qualified liferis the use of stress testing prior to LOCA testing and enhanced, condition. monitoring in service [re:

state of the artscapabilities].1

e. Industry responsibility; best resolved by the industry,

f. At the end of qualified life E0 components are/or should be capable of withstanding a design basis accident. Therefore, it may appear to be capable of providing additional service. If a plant wants to continue to use a component additional testing is required to demonstrate that the " family of components" is c able of provioing additional service in addition to being capabl f withstanding a design basis accident. The method used to determine qualified life is not perfect; but until a better; ethod is developed, it is the best available.

hidf Assenment: lhe staf f agrees with the view (stated above) that this problem f ' est addressed by industry, and the NRC staff should be receptive to a proposed resolution of this problem that is: (a) developed as an industry initiative, and (b)zdemoristrated to be technically just ' ,.

   . the accuracy of test instrumentation; A-61

                                         < s ?n         ,

c Ertr Review Comments:

a. This has been sufficiently dealt with through industry review and NRC audits. It is true that an occasional lapse here and there has been discovered, and will probably continue to be discovered in the future as well. We must recognize that in any human endeavor, it is impractical to eliminate such lapses, especially when it involves exacting attention te numerous details. The answer lies in our QA/QC systemidnot additional research.

b. This is not an EQ iss.ue; th.f.sdisyQAt ,

c. IndustryresponsibilitytilamhnotS& Ware'ofanydeficiencyinthe requirements for-testsinstrumentiaccutacy,and calibration. If there is any deficienef.With.'homol. lance, cit is priraarily the industry's responsibility to correctithe situation.

lLL4Lf_.6Dtusal: l sed on the staff's review ua .ir EQ-TAP Action Item 3.e [10]. this does not appear to be a valid concern. Instrumtnt accuracy was spectfically revicwed by the staff during the on-site [0 inspections that were nerformed at each plant. . doc umentat ion requirements; and Eter Review Comments:

a. No further work should be reautred on this. The standards and HRC requirements are sufficiently clear. This is an education problem, not a research issue,

b. This is not an issue; established by lEEE Standards.

c. Theconcernisnotc.lgat;.thei4pecjficidocumentationproblem(s) should be identifiedD --

d. Based on my experience in reviewing EQ documentation, I often wished for more relevant information.

e. Documentation is required'by the' Code of Federal ~ Regulations (i.e. 10 CFR 50.49 (J)). One of the primary purposes for documentation is to insure quality and traceability of components.

Changing the Code of federal Regulations requires rule making. The NRC has no basis nor inclination to pursue such a change.

  $_t_.i f_f A s s g1 s men t :

A c ertain level of documentation must be maintained in order to be able to demonstrate that qualification does in fact exist. The staff's rr.tew under [0-TAP Action item 3.e (10] indicated that specific documentation requirements were imposed. but perhaps it could be argued Ihat the requirements are too strict and unfounded. The staff should be

                                                                                                                                )
                                         .s   .

ni , , r A-62 I

 *

• v.s -

                                    $;N,         ag}hipve         a. ,.                ,
                                'iNSN&$;$0khhhf5                    ..

f% c. . *,; receptive to proposed changesiinithe" requirements-that are: (a) de. eloped as an industry initiative.'and (b)' demonstrated to be ' tec hnically justified..

• other considerations rela'. Ivc to'EQl such as mechanical and flow-Induced

        <toration, seismic effects, dynamic effects, etc.                                          l; Peer Review Comments:           .                                                            ,

a. I disagree. lhese areinot within the: scope of environmental l' Qualification. They should be addressed separately.

b. This is a valid issue.

c. The issue is not clear.

htoff_A Jsessment: The resolution of other lisues that were handled separately from EQ but that could have a degradingginflu ocer.on equipment. Qualification, such , as the issues of mechanicalIand. flow-inducedevibration, seismic effects, ' dynamic effects, etc., should be reviewed to.. assure that EQ:has not been rompromised by the-resolution offthesefother issues; The (Mlowing problem statements expand on certain aspects of this issue:

     -  (nntinuous submergence prior to harsh exposure has not been addressed.

Peer Review Comments:

a. I am not aware of any areas %hereT E0 equipment is normally submerged.

b. Need it L addressed? If it's part of the normal environment, aging should address it. But does-it really happen for electrical.

equipment? i (. This is a valid issue. 7

d. The NRC staff position on submergence is straight forward and clear. If an item of equipment ~ls submerged during normal operation and/or during accident conditions, then that item should be tested for E0 purposes;in:the submerged condition. The staff has always stated that the: test-conditions should always,Totodate, the >

extent practicable, represent ~ the' installed conditions. no one has presented an~ acceptable' alternative to the NRC staff position on submergence. . i.aff A_JE ssmen_1: i j focused attention on operating" plant experience and PRA information,  : 4 equipment performance, condition and eavironment monitoring, root cause l assessment, and trending of information, on an on-going basis will help , I A-63 I . l

                                                                                              .__j

I However. NRC staff action is necessary to to address this concern. I d:*. ermine to what extent this is.a valid concern; especially for plants that came under the E0 criteria of the DDR Guidelines and NUREC-0588. Lategory !!. , the momentary electrical effects of the lostulated. initial peak temperature and radiation stresses have not been addre, sed. Peer Review Comments: [h

a. Why not? Test specimens should have been energized at the time of the event's simulation.so why.is this.an issue?

b. This is not a valid. . ssue; i .thit,is checked during LOCA. t

c. The issue is not. clear..

I d. This coarent needs clarification, i l I #L4[1.S1M1L101

2. *

! Based on the peer review comments. this does not appear to be a valid contern. However, the NRC staff should review the information that has been gained over the past 25 years relative to E0 and electrical phenomena that are likely.lo occur d'u ring event scenarios to determine whether further action is. Warranted to address this concern. Comparisen of the calculated MSLB surface temperature to the LOCA bulk ' temperature may not assure that the eQutpment will survive the MSLB-

                                                   ')

environment. P,3fr Review Comments: ,

a. I am not aware of!this issue, n , . 3, .

b. Needs to be considered..

Although

c. Industry responsibility; merits analytical resolution.

it is the industry's responsibility to account for MSLB effects, additional guidance from the-NRC would be helpful. I

d. Equip.nent qualified in accordance with the requirements of 10 CFR 50.49 will withstand the worst-case design basis acciden conditions. A LOCA will likely produce the peak LOCA and MSLB scenarios.

conditions and a HSLB will likely produce harsh conditions for a i longer period of time. Consequently, plants often develop a composite profile for testing purposes that includes This peak approach LOCA assures conditions with the duration of a MSLB. equipment survival for both LOCA and MSLB environments. S_taff Assessment: The concern is that the bulk'LOCA temperature is an averaged value, and

                                             -e,-.,.,.....          _ 23 ,

A-64 f

                         .e a m ..

                                                                        .O N." . , 9
                                                            .b                       ,
                                                            . . .s. , .

temperatures near specific components may be more or less than the bulk

                           <alue, depending on plant-specift: conditions such as tereperature stratification effects. So..the concern is two fold:                         (a) analification M couipent for LOCA conditions based strictly on the containment bulk LO' A temperature may not be suf fi'. tent, and (b) use of the bulk LCCA tenperature as a basis for qualification of equipment for HSLB conditions may not be sufficient. Action by the NRC staff is necessary to determir.e. based on the additional information that has been ottained over tne past 25 years and basec' on past qualification practices, whether the temperatures that were required for LOCA and MSLB qualification were appropriate and whether additional measures are warranted.                                                                                      .

l fnvironmental conditions for accidents other than for LOCA (such as for MSLB) were not defined for at least 65. power reactors. The staff failedtorecognizethis;fac,tpf,,lrQts,resolutionofTaskActionPlan Item A 21 Peer Review Comments:

a. In the early '80s. HRk formed an' EQ Branch to resolve issues such as this one. The Containment Systems Branch provided input of containment temperature profiles of both HSLB and LOCA to the EQ l I

Branch. The issue should have been resolved more than ten years ago by the EQ Branch, but the resolution may not have been recognized and/or characterized, properly by RES. J

b. I am not aware of this issue.

c. Based on a value/ impact assessment, the NRC staff concluded that Task Action Plan item A-21 has a low priorit/ ranking. However the accident analyses for plants do consider a HSLB accident. And licensees develop pressure and temperature profiles based on those considerations. EQ test orofiles are subsequently developed from those accident analyses.

                                                                                                                       ~

Staff _A_1ginin1: M'NM* ** R ~

                                                                                                                           \

See the staff assessment'for the previous problerr statement. f

                       -     The staff's " final position *'regarding the velocity profile in con'ainment during blowdown was pending completion of Task A-21.

However, the staff *5 resolution of Task A-21 was incomplete (see the previous problem statement) and this issue may need to be revisitert. Peer Review Comments:

a. A " final position" on this issue is not apparent, but there is no need for such a generic resolution . The velocity profile is only needed in very few cases (such as AP600) whe'e plant-specific detailed analysis should be performed.

b. I am not aware of this issue.

A- fi5 C- _ _ _ _ _ _ _ _ _ _ _ _ _ __

) i

                                       .   .n.       ,

Y - 4

                                .LlN*4l &         3. q qp yg
                                  '~ -

Staff Assessment: Theconcernistwo-foldk.(A).resolutionbfthe"velocityprofile" asper'. of TAP A-21 may not be arpropriate if the velocity profile is l dependent on the resolution af M5tB vs. LOEA conditions (see the l prPVious problem Statement), and (b) the " velocity profile *' represents a dynamic ef fect that may not have been adequately Actton addressed is required in terms by theof E0 (see the last bullet'under B.2, above). NRC staff to address this concern.

   - For plants cualified under the 00R Guidelines and up to the mid 1970s.

vendor specifications for E0 equipment contained few performance requirements describing the acceptable performance of cable systems under harsh conditions. Peer Review Comments:

a. Is there a problem? Separating the shortcomings of the DOR l approach, each cable qualification test demonstrated that cable's l ability or inability to functionrinsthe originating plant's. harsh )

environment.  ! l

b. Given the changegrate of equipment / cables, many of the old components may n6t"brintthe plant? The first step would be to perform an inventory of f the materials in the plant.

It is 1

c. Industry responsibility; more NRC oversight may be needed.

the industry's responsibility to verify that the cualification documentation provides reasonable. assurance that cable safety functions can be performed as. required. :!( the original qualification is'found to be deficient, aJditional testing orHora other approaches to assuring operability may be necessary. oversight by the NRC may be necessary to verify that cable qualification programs are adequate in this regard. f

d. This may be true, however in accordance with 10 CFR 50.49 (k) (

l "Appitcants for and holders of operating Ilcense are not required j to requalify electric equipment important to safety in accordance with the provisions of this section if the Commission has previously required qualification of the equipment in accordance ) , with " Guidelines for Evaluating Environmental Qualification of Class lE Electrical Equipment in operating Reactors," November 1979 (DOR Guidelines), or HUREG-0588 (For Comment version),

                " Interim Staff position on Environmental-Qualification of Safety-Related Electrical Equipment," However, replacement equipment must be qualified:in accordance with the requirements of 10 CFR 40.49.

11 1ff As_sessment: The performance requirements of equipment are determined by the results j of accident analyses,'not vendor specifications. However. vendor ] speci'tcations (to the degree that they exist) can provide addit onal l t A-65 l I - - - - . - - - - _ _ . _ _ _ _ _ _ _ _ _ j

assuranto that installed equipment are suited f or their spm f fic applicalinns. Therefore, the staff agrees with the view (stated above) that it is the licensee's responsibility.in ensure that the cualification document & tion provides reasonable assurance that eculpment saf ety functions will be ;;rformd given plant specific considerations.

                            - The 5eismic Qualification Utility. Group guidelines do not rn ognite the performance recuirements of toulpment:during a design basis avent (DBI).

only the damage to equipment that results from a DBC. More specific seismk qualification requirements may be needed. Peer Revie Comments:

a. I disagree. In the area oferelay performance (contact chatter / bounce).TSQVG does consider equipment performance during the event. .

   -p

b. Merits analytical resolution. This issue merits study because of 1 the lack of consensus concerning the applicability of earthquake ,

damage data to seismic Qual .1 cation. Lack of damage during an earthquake does not, per so, assure operability during an carthquake. Therefore, it seems that earthquake data needs to be supplemented with analysis and some testing (less than a r.omplete seismic test) to provide an adequate technical basis for seismic qualification. The conclusions of such a study should not be inconsistent with what are acceptable procedures for seismic iualification of large equipment where testing is not feasible.

c. Seismic qualification is not within the scope of the CQ rule.

ELdL.Auwar.1: See the staff assessment'of'the last! bullet'under B.2. above. i

                            - 'holimitingundervoltageandunderfreQuencyconditions(i.e.,

postulated electrical condition $ .ar seismic as compared to other hostile environmental conditions) may not have been assumed for est ablishing qualif ication, especially f or plants sub,)ect to tho 00R Guidelines. P3er Review Com inu:

a. I am not aware of this issue.

i

b. Seismic is not included in 10 (FR 50.49 and was not addressed in the [NRC staff's) operating reactor EQ inspections.

c. Ihis s#atement is not exactly clear. If the limiting undervoltage and underf requency condluons result from a seismic event. then qualification for these canditions should be covered under the seismic qualification criteria (i.e.. Section 3.10 of the standard review plan). The DOR Guidelines does not cover seismic

          -                            qualification.

g. ,

3. l A-67 \ I.

                                                              .                                        ;y 4>         -                      -
                                                                                                       .i                                                           e     *'

n. x - Jgg; y .

                                                          .."   u.

.+ :; ,

i

                     $141f _ Aucument:                      .

l See the staf f assessment of the last bullet under B.2. Above. l 5F an Based on the staf f's review of [0 methooology issues, the following recommendations were made: l

4. In order to account for the numerous uncertainties that exist relative to equipment qualification, to provide assurance of continued l

qualification erist. over time, and to identify any E0 deficiencies that may oper ting plant experience and PRA informatfon. eculpment tperforman:c, d tondition and environment monitoring, root cause assessmen anTo racilitate this effort, trending of information on an ongoing basis.the staff should also ie'tlat within the scope of the:mtintertence rule, ,

                                                                           ;ults of past and ongoing E0 research

b. The hRC staf f should review the 4 efforts, qualification test results and practices, and other E0 nto date data. base containing this information, and maintain an u setter manage, catalogue, and share E0 information in order to: (a) information and advancetiin technology; (b) Identify specific' issues that may deserve addillonal research and resolution * (c) provide a basis

' for resolvingFor E0 concerns; and (d) better focus NRC :taff and industry example l this data base should be used as an initial resources. step ir. addressing the $pecific [Q issues Itsted in (, g, h, and 1  ; (below). t.n y further C0 research by the NRC staff should only be performed if: (a) it isisabased on a well defined need for spesific information (b) i there good likelihood that the desired information will be l ' entained, and (c) the cost of the proposed research activity islic health and safety.

                        ,;ostified in terms of the expected benefits to pub
                                                                ,aboratories in accordance wit. generally a       'eti fication of E0 tes'.inu accepted non nuclear pradices along with nuclear 0A standards                                                                               is needed Also, to assure that E0 testing is properly and consistently performed, the NRC staff should periodically ~ monitor the performance of [0 testing laboratories through development and.ir;1ementation of E0 audit and l

inspectton acttvitles. l t c The NRC staff should assure that plant-specific conditions are properly l accounted for by licensee E0 programs through development and implementation'of ongoing CO. audit and inspection activities and through issuance of generic communications when appropriate. f. The use of " excess marginto justify'short-duration LOCA i tests d to that assure < were allowed for the 00R Guidelines plants should be rev ewe f that a significant safety problem does not exist. rurtner review is needed to determine why the generic temperature 9 A-68 1

I profiles f or PWRs and WRs.wareinotafuky.Qustified and how this relates to those -Ituitions where the/ges,3ric' temperature profiles were used.

h. further review is needed in assure that the " velocity proftie" aspect of TAP A 2' was adequatelf addressed. First, the staff's resolution may l

not have been entirely apprcpriate il resolution of the

• velocity profile" is dependent on the resolution of M$tB qualificatton for DDR Guidelines plants (tince the MSLB quellfication issue was not fully addressed); and second.tthe
• vel) city profile
• represents a dynamic offect that may not have.been addressed in term of CO (see J. below).

1. further review is needed to better m,derstand and more fullr address the f uliowing (.oncerns relative to E0:

• leakage currents and momentary electrical effects;

                                                           .       hydengen burn scenarios; a        rarti at 'on and tel*perat tir st rat t f 4 c at ion e f f ec t s;
                                                           -        long -t. rm exposure to moist ure:                                                                   l

• rentinuous tubmergenre prior to the LOCA; f
• the effects of fire on [Q l

                                                           -        crebustible gas.and. chlorine formatton effettt;                                                     l i

use of bulk vs.. local temperatures; l a .

                                                                                                                                                                         )

l

                                                            -       adcQuacy of MSLB qualification.for D0R Guidelines plants; and
                                                            .       aquipment interface problems.

J. lhe resolut ion of other issues tha'. .were handli ! sep4ratel,. f rom (Q but that could have an effect on equipment Qualific tion, such as the i% sues of mechanwal and flow induced vibration. 'etsmic ef f ec t s. dynan it effect'.. t: . should be re.u wed to anur e that 10 has nM been 1 rterromsed by resolution of thee other issues 6 A large number of the concerns. suggested that by using the additiona) ~ information that has been obtained over the past 25 years effectively, and based on the TMl-2 experience and PRA insights. some adjustments in

                                                             !ne [0 n..thodology may be possible and beneficial tn the i ndu e.t r y . This ts especially true recognt2 ng that more emphasis is needed on 6

maintaining equipment qualification over time and some " trade off* may t'a appropriate. The NRC staff should be receptive to proposed changes and improvements in the E0 methodology that are: (a) developed 'ified. as an industry initiative and (b) demonstrated to be techr.ically y For exarrple, the f ollowing elements of the qualification metnon -logy were cited as potential candidates for improvement:

• use of analysis and operating experience versus test data; A-69

                                              ?

M*AEdwrW3e>f%Qyn -

                                                                       ',' l '; '>
                                                                                 .. s          .

4

                                                                             .~
                                                . amount of preconditirnint,       s                                          required and possible use of stress t <;s t ir g as an al t err.at t s e t o this requirement ;
                                                . test margin requirements.

' . post-accident operating .ime requirements:

                                                . mandral bend tett requirements;
                                                 . treatment of synergistic effects, dose rate effects and oxygen diffusion effects!

^

                                                 . double peak reQdirementi
                                                 . radiation testing.and source term conside:Jtions:
                                                 . E0 boundary rettric'tions that'make-it difficult if not impossible

} to perform maintenance,and replacement activities; f

                                                  . schedular constraints f0:' performing maintenance and replstement of LQ equipment (i.e., no grace period allowed); and
                                                  . cocumentatton requirement 5.
                                            'he  NRC staff did not consider any of the [0 methodology issues to be immediate safety problems.

209:' ,s . ..O ,, , L A-70 a _ _ _ _ _ _ - _ _ _ _ _ -__ _ _ _____________________-____ _ _____ - __ - -

3 /, ~; : jp -  ! A &S .1~ .-e . C. brtenL.it.a.t#5 ADL)mnlemeniation o( E0 Reoub'agg ' i

l. 11 is dif f icult to determine What mir.imum f 0 st andard as imposed on licensees because a clear record of exceptions that wera allowed by the

                .taff is not readily avaliable. The mat ter is fut t her eneplic ated by tne evolving nature of reQuiremen'* that were being promulga.ted by ((

Bulletins. Supplements, and ultimately by the C0 rule. The (nllowing imob!'n statementi relate to this issue: Parr._Erview Comments:

a. In eeviewing the concern and the following problem statements presented in support this issue. I would be remiss if I did not point out i'.at thi reflect a compilation by one or many who were not involved in .n the early days, and thus are unfamiliar with the rationale for why somethings are the way they are. It is c..darstandable that there is a sense of f,*ustration when one does not know the ba:is for a decisione and cannot readily find it.

I agree that in looking-back, one-can come up with a need for more and better documentation ~of;the bases'for many of the decisions that were made. 11.this not true of any human activity? This will be all the more true if that activity involves an attempt af. backfitting operating plants, and plants under various stages of construction, to meet requirements based on an evolving > technolcgy. I think that we should be'able to say enough is ' enough and move on to the more important items. If one wants to reconstruct events and bases for the many ceCisions that were made by the NRC headquarters staff and the It will be a time

                                                                     ~

field inspection teams. it can be~done. It will require the consuming project and a costly effort. participation t' people f.'om the NRC and the industry who were involved in those days. Luckily, we still have about a dozer. of them actively involved in the industry or in the NRC, although they may not be in the same technical areas. , 1here has been clarification but there are no exceptions to the DDR Guidelines. IE Bulletins and their Supplements require cesponses to specific issues. Therefore, if one has all the Bulletins and their Supplements, there should not be any confusion as to the requirements'of the Bulletin, However, if any such confusion exists, the NRC staff is available to discuss specific issues. 9

                 't (Lf Cpecerls :

E_ sea on the st af f's review under E0-TAP Action item 3.e [10). it appe a

• that the requirements and implementation of [Q requirements were consistently established and weil assured. Resolution of the.more trivial, administrative-deficiencies that were discovered during the NRC staff re tew and inspection process was accomplished through meetings Given the l that were conducted between the NRC staff and each licensee l 1

A 71 I

                                                                                                         , o
           . cry ! arce atabt'e o f de f it t ent it <. t       a typisai af this proc.ss. only 1N.se considered 1o be of $ignif scance mere os.enented in the weting ainutet.         The less significant i t em<. were While rearouni/ed by the applicable the less significact [0 l': ensee as needing to be correttec.

deficsenties may not have been docu?ented, it is the staff's view that , t b d does not represent a -significant- C0fnprom15e- of- [Q requirements.

         -  it     is not clear to what extent the various clarifications                  and staffB
                                                                         ![ Bulletins, Appendix pnsitions that were stated in Generic letters of NUREG-0737. etc.. were fully implemented..and wt.ich ones are r urrently apolicable since they are not specjfically referred to by 10 ( F R $0.49.

Peer Review Commenti:

a. I agree.

Merits analytical resolution. It seems reasonable to expect the b.

                       'lRC to clear this up.

( All c'arifications, staff positions. Generic letters. IE Bulletins, and Appendix B of NUREG-0737 were fully impicmented. The ones that are currently applicable requires a reading of the document in question. 5 ' d! A 5.5 e il. ment : E.ised on t ht: staff's revirw under (0 TAP Actlon item 3.e [10). what ttywas-required ta De implemented in the way of E0 requirements was pre

              < lear.      In geni tal, requirements that were est*,blished Dy Gener.c ietters and Bulletins are'ttill~ applicable unless  However   they    have given   thebeen rapid
               .uperseded by more recent feQuirements.                                            have development        and transition of [Q. requirements, the staff may not imen ertirely clear as to what was being " superseded

• and there Therefore, maystaff the fiRC be

                . cme confusion in thc industry on this point.

st.ould pursue this matter. with industry representatives to determine whether clarification of'the existing reQulrements'is necessary. [moraer.cy shutdown systems " ..used to bring the plant to a in celd a breach

                ,nutdown condition following accidents which do not result cf the reactor coolant pressure boundary together with a rapid 9pressur12ation of the reactor coolant system" wereItrequir*d                  to be is not clear.

soalified by plants subject to the 00R Guideltnes. Also. >ince

                >s ever, that this was the (ase 'or NURfG 0588 plants.         required by the LQ o salific t ton of cold shutdown eogipment was not                     is currently cole, i' is not clear to what extent tnis requirement
                /41id.

f aer Revjfw Comment 5:

a. I am not aware of this issue.

b. Plant-specific design bases.

A-72

(- -- _ g3 }h,. ,.- [ , f 7, ,

c. Merits analytical resolution.3 Although.thls reviewer is not  :

knowledgeable about thit 1$$ue, thw statement liself seems to Justify a resonn:e. .

d. Currently, plants are required to Qualify a path to safe shutdown.  ;

Safe shutdown is not defined the same way for all plants (i.e.. some plants define safe shutdown as hot shutdown). Plants must oualif y a path to cold shutdown only if their license defines safe shutdown as " cold shutdown " This situation applies to all plants , (i.e.. DOR Guidelines, NURCG-0588 Categories I and !!). kt> f 8 His)ime.n.t:

            's e e the taff as'.essment of the previous problem st atement .

)

     -      fvidently,       the NRC staff consWered certain parts of the DOR Guidelines and NUR[".-0588 to be " optional" and consequently, the minimum standards that were found to be acceptable to the staff are not well defin N.

Exceptions that were allowed te L(4 rAquirements may not have bcan proper and ( insistant in all cases. Peer Review Cowents: , 1 4. I am not aware of this issue. Merits analytical resolution. It seems reasonable to expect the b. NPC to clear up this itsue.

c. None of the requirements were optional. However. there may be lhere was opt tonal ways to comply. with a particular reautrement.,

no NRL policy that allbwed for exceptior.s dthin a review The category without approval by the NRC Commissioners themselves. intentional differences in the NRC staff policy are those that places plants into the three categorie$ to be reviewed under the DDR Guidelines. HURCG-0588 category II and NUREG 0588 category 1. d Nevertheless. NRC inspectors may differ somewhat in their detarmination of what is acceptable to meet a particular reautrement. However, these differences should be within the ' range of the " slight differences"'that we all experience in the way we see things, or thcre may be some differences that were unintentional and can be attributed to.a mistake by a particular inspector. , ktML.We WAD 1; Pa'ed cn the staf f's review under E0-1Ap Action item 3.e [10), inis does The process that was used by the staff not appear tu be a valid issue. for evaluating licensee compliance with EQ reoutrements was quite While resolution or many of the less rigorous and well defined. stonificant, plant-specific 1$5ues may not have been well documented. but the procu s was consistent and involved the ssme principal NRC participants which would tend to minimize inconsistencies. l A-73

                                                                                 < =

During the 14RC's [i inspection activities of the mid 198vs. 1- *sistencies cris.ed in the st af f's interpretation of [0 requirements f a- test r e e.a l t s . i ffectffriew Commentt:

a. Region 3 attempted,toaimplementaCons[stentInterpretatinnofEQ

                                                                                     ~

requirements during the inspections'in quattion. This was accomplished through the use of a common ' pool of contractors, frequent discussions with NRR on evniving E0 issues and the attendance of all available E0 training opportunities and meetings. Merits analytical resolutlon. It seems reasonable to expect the b. NRC to clear this up.

c. Inconsistencies that may exist are not the result of NRC policy; but rather, are unintentional errors on the part of tne inspector (s).

M d ' Als_tM m.t0.l' Ib "i or the st af f's review unrjoe [Q.IAp Action !!Pm 3.A (jf]. ff.fs does ri n ' appear to l'e a valid issue. While some inconsistanc ies were ir. it aDie due to the different knowledge and experience levels of the w i o u , PA inspecturs involved, work < hops were held with the R"aions The NRC and with the licensees t o minimite the extent of thi, problem. Headquarters Office was also actively-inve'ved in training the inspectors, providing guidance and addressing issues that were identified, which also helped to achieve' a coiisistent applicat ion of the r eca. i rement s . - D r.11t ing E0 tests of commercial or generically named component s may not h,e e been ent irely appropriate when these tests covered variou', manuf ac t urars, vintages, or designs of cables and interf aring u .pnnent s . Peer Review Comments:

a. I believe hat the generic tests were proven to be acceptable for specific components by carefully reviewing the testr and assuring the results bounJed the component in ouestion.

b 1 don't think tnat this was done. ( l agree, but the statement is too mild.

d. I agree that E0 tests should not be extrapolated to commercial or generically named components. The regulations and IEEE 323-1974 require that the link between the qualified test specimen and the items in the plant, including manufacturer, model and vintage be established. If similarity of the test specimen to the plant l installed equipment was not establi:hed, then installed equipment-was judged to be not qualified. l A-74 ,

i

s. MoreNRCoversightmaybeneededfthisisavalidconcerntothe extent that any qualification program does not satisfy qualification rt auirements concerning the choice of test specimen and the extrapolation of tests to striillat components.

I I am usiaware of such crediting of C0 tests. When did this happen? It is entirely inappropriate. 5 i .41 f A '. 0 v>cN. ' flare may have been circumstances where this practice was considered to be appropriate. For example, this may have been allowed for equipment o.,alification under the DOR Guidelines. To the extent that focused a*ient un is placed on operating plant' experience and PRA iniormatlon, equipment performance, condition and environment monitorinp. root-cause asses went, and trending oftinformationi.this issue'becomes one of minor portance with regard to aging considerationt. However, irrespective of aging considerations, there could'be some Question as to whether inneric qualificatinn" was sufficient to demonstrate that equipment ill ' unction during an event. TF HRC staf f should determin'. whether ey additional actton is warranted to address this issue based on *.he r f nemation that has been ' accumulated over the past ?$ years. t'! t e r. t fatlures (for all attempted E0 tes's- were no: spm if wally

      .. quired 1o be documented, evaluated. And saved as part of the equipment
       ; ulificat ion record and consequently, cualificat ion may not t .e br en tal1< ob jec t i ve (i c. E0 may have been Dased on ~'.elec t ie
           + c cm a : r..i i .

b t.tJJ.Ltta._kWfJ10 ; a This may be true to some extent, but it is believed that the vast. amount c' test data that was reviewed satisfactorily demonstrated qualiiica' inn including explanations of test f ailures.

b. :n EQ testing, most test prearams were performed with a minimum of anomaltesandallanomaliesincludingequipmentcorrectiveactions and retesting ara documented. Since it is necessary to demonstrate a clear path of successful operation, den test failures resulted in substantial redesigns, some vendors chose to redesign and then start tlie . qualification process' on new designed equipment. Thus, not~ all test failures.Were passed along to the licensees. The process was objective in that a clear path of proper periormance was necessary to be documented,

t. More NRC oversight may be needed. Although trie NRC did addre:s this issue in some of its [0 inspections, it is possible that the proulem was not corrected in all cases. The statement of tha issue is consistent with this reviewer's experience with valification testing and Jocumentation and his participation in EQ inspections; and he agrees therefore that it is a valid concern, n 15

                                                                                                         "'A - as y ,

l l i

  \                                                                       Howeverbequirementsfor                          '

d. Perhaps this statement is true.

reporting test failurc, are governed by.10.CFR Part 21. " Reporting i of Defects and Noncompliance." .Any test?fallures that f all within sne scrpe of 10 CFR PWt 11 should have been reporte d in

l l accordance with its provisions. Otherwise, the testing entity J

would be in vion tten of'the Code of federal Ragulattuns,. l

         $33f f **:tstmfJil:

The r e .ults of past research 6' forts should be catalogued and the j inf or 't 'en sho.sid be well ut 1erstood and related to specific Theequit e.taffent appl ,atton; in order to address isu.es such as this one, shoul<3 also assure that full advantage is taken of operating olant

 )

ripe. "m n nd PRA information. *Qutpment performance, condition and j monitoring. root cause assessment, and trending of environment information in order to identify and correct any EQ deficiencies t' may c orrent ly exist. The NRC staff should provide additional guidat..< if nr .m a . , *; ensure 194' ruerent (Q testing practice:, are j ' j approt' ate. operability of l i Non ..if ety-related ins trument s that could impact t % instruments were nt,L i l i s a f e t . . r- l a t ert l equit'c,t that was regt * ' <f tosbe quallfled..and implementatton c' this requi' ament n.4y not be us.iform among all plantsi Pter Revitw,Covent1:

                                                             .y                    -

j a. I am not aware that this is an issue. t j j b. Point Beach w u the pilot plant; all plants were similarly f addressed by the E0 Branch. I c. The inclusion of non-safety equipment that could impact safety All cy ipment is a requirement of all E0 programs per 10CfRSO.49.Thu>. this l plants had to have safety evalustions for 10CfR 50.49. l should have been addressed for all plants. This type of equipment i

                       =45     addressed at all of the EQ audits of which I was aware.

l d Mnre NRC oversight may be needed. J This statement may be true. Howe"er, it is not consistent with e. j the intent of goalification criteria initially set forth by the NRC staff. for e' ample. IE Bulletin 79-018 usted January 14 1980, stated in the second p.ragraph of item 1 under " Action To Ba l .1sken By 1.icensees Of All Power Reactor Facilities With A Operating License (Except those 11 SEP Plants Listed on Enclosure i 1)* that " Electrical equipment items which are components'of I systems listed in Appendix A'of Enclosure 4, which are assumed to operate in the FSAR safety ant.'./ sis and are relied on to mitigate

                                                                          .ed within the scope of this design basis events are con >a Bulletin, regardless whether or not they were classified as part of the engineered safety features when the plant was originally licensed to operate. The necessity for further up gra n g of non-A-76

.. o ? safety-reinted plant systems will ha dependent of the outcome of the licensees and NRC reviews subsequent to TM*/2.' As a resalt of the EQ rule, olants are requiNJ f o address this issue under 10 CFR 50.49(b)(2). As indicated above, the rts irement to address

• u this issue may not have been as clear at the time of issuance of If Bulletin 79-01B as it was when the E4 rule was issued.

Neverit.eless, this is a safety issue and all plants should address this issue as required to insure safety;and safe shutdown in the > event of a design basis .act.ldent,

                                                # '# */ ~

Staff Assessuni: F4;cd on the staff's review under'EQ TAP Action item 3.e [10). this does wt apoear to be a valid lisue. Instrumentation was included in ti.c NRC

      'mview and inspection of licensee implementation of E0 requiraments for all plants.         The process appeared to be rigorot.s and compre ,ansive. It appaared to be consistent for all plants.
    - cG ).9,      instruments wer: not adJreued in the initial qualif f. atinn cautrements and it is not clear tu what estent (and to what (riteria) nstruments were required to be qualified.

Pen. Bey _ Ley,,lomme n t s :

a. It should be clear now.

      !.       Mer'ts analytical resolution (i.e., analyze existing data to reach resolution).

The guidance provided in RG 1.97 Revision 2, dated December 1980, can be traced to NUREG 0737, published November 1980, entitled

                *clai i rication of TH1 Action. >lan. Requirements, and to NUREG-0737 NOREG-0737 and its
                %pplement No. 1 dated Decembel' 17,21982.

(u,;'ionent contained letters that issued these documents as requirements, ratherdthan just HUREGs. Subsequent to the issuance of NUREG-0737 NUREG 0737 Supplement't, and RG 1.97, the' NRC staf.f metwithalllicenseesanddiscussedqualificationrequirements and implementation dates as related to-RG 1.97. The NRC staff subsequently wrote safety evaluation reports documenting the results of the meetings and the RG 1.97 qualification requirements for all plants. Those requirements apply to all plants and are the same as the requirements of 10 CFR 50.49(b)(3) which rapresents the criteria for all plants. There were however, plant The exceptions were based

                 .pe:lfic exceptions to this criteria.

primarily on differences in plant design which could affect the recuirement(s) for a pa-ticular instrument. 9 Lil 4 ^ tiim_fAl:

        'ae the .taff assessment of the previous proolem statement.

Requirements for protection ' rom contrni system interactions" may not ne uniform fur all plants. A-77

pe g , s w > ,, e .g ..

                                                                                    *

• Nj%%@t N'mK -

t .

                                                                                                  ;f u w l                          f,,eer _ Review Connentit .
                                                                                                 @[$7YNi sigi:
                                                                                                                                 "~<

a. I am not aware of this issue. ,

b. Pierit s analytical resolution 4 Llie., analyze existing data to reach resoluticn). g, ,

c lhis statement is not~ completely clear. 'However, if " control system interactions" is the same .5 fasA Action Plan Item A-17,.

                                                       " Systems Interactions'in Nuclear Power riants." see the staff

! resolution of item A 11 contained in NUREG-0933.

                           >*.*L}_s>LsSyrg:

b. the staff assensment of the two arevious problem statements.

b. 10 inspections of the mid-1980s found that many programs lacked the l

2.

                           + .montat ton nece<sary to support E0 inspection activities. especially e                      the older plants. Vithout. sufficient ~ documentation, it is doubtful mat if n ation was truly;c tabilihed at these f acilities.
                                                                                                     , c.

P. .r_8tview Comments: , , 4  ! disagree, sinceithese' findings were not dropped and licensee ar.t wn to demonstrate qualificillon were . required. t Idisagrec, sounds 1.th:&yhddestlotoproblem. Are not the NRC audits and the findings 1 therefrom usunlly closed out by follow-up l ' inspections or written iesponses from'the licensees? Itwasmyunderstandingthat'E0open-items,suchasE0 deficiencies found during the E0 audits of the mid-1980s. were required fnllow-up items by NRC Regional Offices. J More NRC oversight may be needed.

                                 >                      'he N~C staff audited tha. EQ programs at all nuclear power plants.

and issued violations and, when appropriate, fines whon it was determined that qualification could not be demonstrated at a given facility. However, the qualification standards that were applied to older plants are different from those that were applied to newer plants, for example, the 00R Guidelines are somewhat different from the NUREG-0588 Category I requirements. Differerces in qualification requirements other than those that result from differences in regulations should not exist, and if such differences do' exist, it is a mistake. If documentation at any plant is insufficient to establuh. qualification - that is also a mistake. All operating plants should have sufficient documentation to demonstrate that all' equipment req'uired to be qualified is qualif.ied. .lf such docun.entation does not exist for a given plant, then'that plant .is violating both the law and it's intent. In addition,'such plants should also be considered to be u" safe. A-78

gg L 5;, j 5.LAli.b 1C ME2D.1: . l I Ja,ed on the staff's review under EO TAP Action Item 3.'e (10), this does not appear to be a valid issue. ~Ihe staff agrees with the vfew (stated atsove) that the findings'..tre not . dropped and 1icensee act son was required to resolve documentation:oroblems. sr l Given the evolving naldre of {Q acd the confusion that existed in the l 3. industry, licensee QAipiograms'mAy not have been well structured and incused on implement 109'EQ requirements. I Peer Rev 4: Comments:.

a. I agree, t To some extent, this may have been true during the early stages of f b.

E0 implementation. The situation has lonej since been corrected, j and to the best of my knowledge, the industry QA/QC programs do reflect sensitivity to EQ requirements. In fact, the industry now l j performs self assessments, .- 4/or periodic EQ audits of their own ' programs, and vendor surveillances to verify continued attention

• to quality in this area. This combined with ongoing staff i training performed by the industry should alleviate this concern.

i

c. Licensee 0A personnel were integral team' members in the E0 process and judging that QA' audits were regularly performed at the laboratories, l.believe that they were fully integrated into the EQ process allfalong.

gm . . &. .t .

d. Thisstatementmayfbe;true.PHow.ever, EQ requirements were '

i established in the early 1980s (approximately 3 years ago), and any confusion that existed early in this tiroe frame should have j been cleared up:by now.--However, should confusion continue to i exist, tlin NRC staff is available for discussion and clarification as needed. Environmental qualification programs at plants are. l and should be, living proge ains (i.e., programs should be updated l l as new information becomes available) and when licensees find i mistakes, the NRC expects licensees to correct those mistakes. By the way, licensees can update EQ programs and correct mistakes  ; without being issued violations nr fines. Staff Ass m my d: f I 1he staff acrees with the view (stated abo'e) that to the extent 'that this problem did exist during the early stages, it has Inng since been. turrected, in order to ar. count for any lapses that may have occurred during the initial. implementation of E0 requirements, the NRC st4ff  ! should assure that ful1 advantage is taken of operating plant experience and PRA information, equipment performance, condition and environment l monituring, root cause. assessmenti- and trending of information in orc'er to identify and correctiany EQ deficiencies that'may currently exist. A-79 l 1

                                                           $N hy Wraa Based on the staf f's review of current >tatus and implementation issues, the' following recommendations were made:

a. sven the rapid development and transition of [0 requirements. the staf f

                   '*i d y              not have been entirely clear as to what was being ",urarseded" and
                    'here thay be some Confusion in the industry on this point, lherefore.
                    'he NRC staff should pursue this matter with industry representatives to letern'ne whether Clarification of the existing r eQuirements is
                    ...nessary.

D. ?here could be some. Question.as:to:whether

• generic Qualificatlon" was

                       .uf ficient to demonstrate:thitispecific equipment Will-function during an event.                 The NRC staff shouid determine whether any additional action
                      's currently warranted to address this issue based on the information that has been accumulated over'the past 25 years.

i c. *he result s of past research eff tin should be catalogued and the ,

                      'nformation should be well understuud ano related to sDecif1C equipment
                     .ipplicaltons in order to address issues such as: (a) implic.ations of [0
                     'est failures, and (b) lapses that may have. occurred during the initial mplementation of [Q program requirements. The staff sanuld also assure
                     ' hat full advantage is t& ken of ~ operating plant experiern and PRA nformation, equipment performance,' condition and environmc..t monoring root cause assessment, and trending of information in order o identify and correct any [0 related deficiencies that may ex.st on ar.

rgoing ba ts. The NRC staff should provide add'tional guidante if

                         .ecessary to assure that current [Q testing anu QA practices are
                        .ppropriate.

T he 'JRL s t a f f did not consider any of the current status /imples.. ntation

           . i p o..           . to t,e imediate safety pipblems.

I I h 4 A-80 r

l - i D. enurance of Continued ddalif f hat l6.E h isting programs have not b2en i developed'and implemented to a:sure l. continued coulpment quallfication, incliiding oualification beyond the established

• qualified life." The followleg penblem statemants relate to this issue:

- . s s 4

P_e e r Review Comments:

a. I agree that continued qualification has been fumbled by the NRC.

I remember a spring;1MO meeting where the EDO expressed great concern on this very point, but the EDO and NRC management's interests were diverted to other more imediate issues.

b. I disagree. The industry is keenly aware of the need to assure the continued validity of the qualification that has been established, and have implemented programs accordingly. Again, this is an area where one can find varying methnds and levels of

 '                  details. Many of the prvblem (tatements listed to support this contention are just not valid. I would be remiss if I did riot point out that they do reflect.a level of unfamiliarity with the industry programstin this area,.

If there is anything that the NRC.could do in~this-area, it is.the development of a' guide based on a survey of the current practices. Such a guide will go a long way toward promoting uniformity of industry practices.- It may be argued that developing such a guide is inconsistent with'the NRC Hission. Perhaps a justification can be found under the need to improve consistency in inspection practices, and to_ reduce the cost burden imposed by the EQ regulation. ,

t. lhe awareness of the significance to E0 of fomponent Root cause failure Analyses, information provided in NRC Notices and flulletins, and realization that unexpected degradation impacts EQ 1s grnerally present at licensees. The re-emphasis by the NRC of E0 awareness would probably be appropriate.

This statement is incorrect. In all plants there is equipment d. that is not qualified for the forty (40) year life of the plant. 11 is well known throughout industry that when a piece of equipment reaches the end of its qualified life, it should be replaced or requallfled. If eoutpment is not be'ng replaced as it reaches the end of its qualified life, then the E0 program reviewed and approved by the'NRC'is not being followed and the plant should be issued a violation for being unsafe and not in compliance with the requirements.of 10 rFR 50.49. v Elaff Assessment: for whatever reason, the NRC staff never really got around to establishing guidance for maintaining Continued (ongoing) Qualification of equipment. Other than the " replace or reoualify" aspect discussed in A-P'

I th- peer review comments (above). [$sentially by default, the staff ar'"rted conventtonal maintenante and surveillance pract "as and l e sm roer t s f or serving this f unct ion, and no init iat ives were insesued ' and fc. F o'oping further guidance in this area. Shortcominc ori..taintics that exist in 140 (Q oethodology (see Section B of this , ap; cod i t) indicate tnat additional measures are needed to assureWhile the tentinued eg'apment qualification over the. life of the plant. st af f agrees with the views-($tated above) ~ that a re-emphasis of EQ a m eness and promulgating guidance'. based on current industry practices wou'd be helpful, a more extentive' effort'is needed. in particular, the stalf should assu e that operating. plant experience.and PRA'information, cause equipment performance, condition'and environment monitoring, rcot as'assment, and trending of'information are used in a comprehensive F urt her. the

         .arner in order to maintain [Q over the life of the plant.

staff's guidance should not necessarily be limited to " current . prutices." Implementation ot1the maintenance rule should help to I address this concern for active components, and the staff should ' ir.* t iate action to include passive electrical equipment within the scope ei *he ma:ntenance rule for E0 considerations.

   -    The    DDR Guidelines state that ongoing progiams should exist to review su.weillance and maintenance records to assure that equipment that

• eo itn t s degradat ion (e.g. cables) will be ident i f ind and addre* 'o<t a .

nr. . , u r y . Programs such as these are generally not in place. Py; _pfview Comments a i believe se accepted normal surveillance and maintenance programs These programs as being .cceptable to ne ' this requirement. should 51: 11 be acceptable.- ~

                                             .'      n,,._    .

b. They should be. [;[. .. Ml' *

c. Merits analytical resolution (i.e., analyze existing information to reach resolution).. The following additional views were also expressed in response to other related issues and problem s t a bwent s :

l

                   -       A tightenir.g of EQ, surveillance, and maintenance requirements can' help assure that adequa'e attention is given to WC,rSt case conditions.

Where experience predicts that certain equipment in cestain locations is expected to be subject to steam leaks and inadvertent spray actuations, the EQ program should incorporate testing (such as humidity stress tests) to i l

simulate these service conditions. Where such events were l not anticipated in the EQ program, the existing program l shnold be supplemented by additional testing or enhanced I

                            '   Jition monitoring to assure that equipment is refurbished or replaced when it 15 no longer able to operate as required            ,

during a DBA.  ! 1 l 1 I A-82 , l

Y k h l f h Y$ ?W ' ' - iS* "& l

                                                   .e(

There is now a need"to evaluate alternatives to the qualified life requirement. Operating experience and enhanced condition monitoring are among the alternatives that should be considered. t" Ha ' f Asses g - Soc t he pre vious staf f anessment .

   -       tn isee 0A programs may r.ot be well focused on maintaining f0 reaoirements. For example, analyses performed by licensees (e.g., 50.%
           <e.iows. .            t cause. coreictive action, etc.) may not be well foteed in th-     regard.

Peer Review Comments:

a.  :.,.:

a. There is no evidence ofsthis ate.this.. time,

b. If licensees are not maintal6ing EQ programs as required by the Code of Federal Regulations, then those licensees are in violation of reautrements and suchl Violations should be reported to the NRC.

L'._J L_4.1m 5 me n L : See t! . staff assessment following 0,1 (above).

                                                   ?,"'W ~          .
   -       Changes in manufacturing techf$l Que,s an.:d use of materials when re'urbtshing equipment may not be adequately addressed bv the nr minal eo r, ment qualificatton documentation.

hf E.ht.n_fo!p i vni n :

a. Thi> could be true.

b. Industry responsibility; more NRC oversight my be needed. It is the industry's responsibility to account for Significant differences between the materials and parts used in the refurbished equipment and the materials and parts in the equipment that was qualified. More oversight may be needed to assure shat qualification programs account for. such differences.

c. If not adequately addressed, it should be. Again, EQ should be a living program, if licensees discover that an EQ program or some parts of an EQ program is inadequate, that licensee should take corrective action without being forced to do so by the NRC.

51.:!f Asse!*, ment: ,' The stsff agrees with the view Istated above) that it is the it(ensee's responsibility to assure that r' placement and refurbished or repaired eat.toment is adeauately qualified, and guidance for addressing this rocreen would be best addressed as an industry initiative. A-83 i

f L _ n.g , l l Pun hase specification requiremetit' for replacemer,t parts have not been addressed relative t, E0 Pee Review Comment.s:

                                                       ~

a. Iseemtoremembe'r'the~yIWer&Addres!6dhin'thatreplacement equipment was required.-to meet' Category.I requirements,

b. Have not been addressed;by:whom? This issue is addressed in RG j 1.89 paragraph C.6,.andstn.10 CFR $0.49(1).

WJ f A5 s a.i.Lin.1: [ While the requirements are relatively clear for instances where complete components are being replaced, the requirements are notSee so clear when the staff l1 p m e parts of quallfled components must.be replaced. as',tusment of the previous problem statement, Criteria for atteptable versus unacceptable aging degradation have not d be+ r .'s 1 abl i shed . ' Peer Review Comments: l

a. If this concerns life extension, this may be true.

l I do not agree.: ,, 3 b.

c. This statement is truelonly for equipment that was not.preaged.

for this equipmentclicenseessinust' determine.the limit to which a i piece of equipment can'bejdegraded and yet perform its required action when called upon h .To date, the NRC has left this determination to theslicensees.. For preaged equipment,.the equipment is preaged to..the' end 'of qualified before design basis accident testing. If unacceptable' aging occurs (i.e., aging not in accordance with preaging testing and analysis), In then the this event, preaging ervironment'was incorrectly deterhi;ned.the prea qualified life should be adjusted in accordance with the resultc of the revisit. P d_f_AfgpigqtL1: The staff has relied primarily on initial ih's is a valid concern. qualification testing as a means to establish FO. but focused attention ha. not been provided on maintaining equipment qualification nver the See the staff assessment following D.] (abase). 1 'o of the plant. N effects of installation, maintenance and surveillance practices on equipment qualification have-not been addressed, brr Reviftw .(S3tments: .[ a, Iseenoevidenceof.kthisatthis' time. A-84

{l

                                                   .. : .          . ~                   .
                                                                                            *   ' -~ , ~ '

M 947Q:is:p 3 e

x. _,,

                                           + . v.s b,          Industry responsibility; merits. analytical resolution. The effects of installation, maintenance, and surveillance practices have been addressad only cursorily in. industry standards and~

regulatory guidance. While?it is the industry'.s responsibility to account for these effects iit' would be helpful if the NRC. provided more detailed guidance-than-is currently available.

c. The installation, surveillance, and maintenance of equipment is the sole responsibility of licensees. To the extent these practices need to be addressed, they must be-addressed by licensees.

t

          'tt a.f.f A s s e s sment :         .y               ,

52, lhestaffagreeswiththe.hV'ew( t d aboVe) that if it has been addressed at all,:lt hat"lbten d.one?.in sta e a very cursory manner. See the - staff assessment following D.1 (above),z

        -  Identification and treatment of hot soots, long overhangs, insulation and .,acket embrittlement.cunintencid Icng-ter, submergence. <por:re to
           .hemical attack (e.g. ;borie, acid leakage, decontan.1 nation ac t ivities),

and localized anomalies have not been addressed. Peer Review Comments:

a. I see no evidence of this~at this time. Hot spots, when identified,' are addressed by licensees. So are the other conditions.

b. Industry re:.ponsibility, merits analytical resolution. The issue raised by this statement is similar to the one in the preceding statement. The industry is aware of the need to account for these ef fects to assure ' plant safety, but prescriptive methods of doing so are not available. ,The itvs named in this statement can be.

regarded as weak linksMn safety systems, as 'they may be more ' likely to cause:fai ure.ithankthe degradation of equipment in normal environments {& Consequently. .their-importance cannot be overemphasized However,'since they are deviations frcm the conditions planned,to exist in nuclear plants and can have many individual variations,'itrir not feasible to develop detailed procedures for dealing with them. Nonetheless, it may be possible to provide some broad goldelines. c .. E0 programs as envision by the NRC and industry should be set-up to be continuously.: Updated so as to be able to address these issues if and when'.they occur. That is what is meant when EQ programs are referred to as living programs. It is the respons.bility of licensees to update and maintain EQ programs at required to ensure qualification; this includes identification and treatment of hot spots, long overhangs, insulation ard jacket ernbrittlement, unintended long-term submergence, exposure to chemical attack, etc. A-85

s 1

                 $Jaif ,qng:                        9MCa [c.,"- h'e$. .yk' jen        ,
                                                                      ~

the z',iff agrees with the view' (stated above) that the importance of this ;ncern cannot be overemphasized, and guidance is needed. See the staff as;essment (cIlowing 0.' (above).

               - Maintenance and inspection activities'have not beeri developed and impir-nnted to assist in the aging management of E0 components.

Peer PeyD,wlomment s :

4. I see no evidence of this'at this time.

b. This statement will certainly apply to plants that do not have an acceptable surveillance and maintenance program. One of the primary parts of an acceptable E0 program is a comprehensive surveillance and maintenance program. Plants without acceptable surveillance and maintenance programs do not have acceptable EQ programs. I would like to emphasize here that surveillance and m.iintenance is a major port of E0 and must be performed by the 1 licensees throughout the' life of- the . plant; l

c. 'ndustry responsibility. The following additional. views were also expressed in response to,other related issues and problem statements:

A tightenir.g of EQ, surveillance, and maintenance requirements can help assure that adequate attention is given to worst case. conditions. Whera experience predictsithat certain Equipment in certain locations is expected to be subject to steam leaks and inadvertent spray actuations, the E0 program should incorporate testing'(such as humidity stress tests) to simulate these service' conditions. Where suc' events were not anticipated in the EQ program, the existing program

                                 !,hould be supplemer. Led by adoitional testing or enhanced condition monitoring to assure that enuipment is refurbished nr replaced when it is no longer able to operate at required during a DBA.

There is now a need to evaluate alternatives to the qualified life requirement. Operating experience and enhanced condition monitoring are among the alternatives that should be considered. LIM' 41f s s*ent : This is a valid concern. The staff has relied primarily on initial f qualification testing as a mear,s to establish FO. but focused attention has nnt been provided on maintaining equipment quailfication over '.he life of the plant. See the staff assessment following D.1 Iabove). A-86 j

                                                 +
                                                    . MQg .e

• s  : P W e, u b

                                                      < ;Vg-g      ,
                                                                        .y. s . sg y 7.w    -
                                                  *     #g.m         o.       .

2. .andition monitoring'technicues have not been sufficiently developed to

                  ;roject remaining service.Itfe.

i Mer Review Comi..ents:

                                                                                           ~

I am not aware.that conditi m monitoring is used to determine

                                                  ~
                      .                                                                                   i remaining service life.

D. I agree. This is well recognized.byfthe NRC and the industry. Ongoing research can be effective'in-this area if it is better  ; By: this-limean: laying :lt'out:as .a long range (5 to 10 focused. yr:) progr:m and progressingoinismall manageable chunks. A PSA l based equiptent prior tization would~ be of great help in this i j

regard. . .y j c. It is already. accepted /practicebto, calculate qualified life from

' assumed ambient temp.eratures'and to estabiish actual ambient > I temperatures. Thus7 the use 'of Condition Monitoring techniques.  ! such as infrared Tiermographv and vibration signaturas. are a natural extension.to.establiining remaining lives for most -' equipment. The condition of equipment temperature has been shown to be one of the most significant parameters at detecting age related degradation'in NURIC/CR-5762. 1 i i 1. Merits experimental'research. The conditian monitoring component " of the NRC research, plan has thezgreatest potential for benefiting

                                                                                ~
                                                                                                    ~

! the nuclear industry.

                   ~         This is torrect.                                                             i i                       taff A13p nmgnt:
                    'ne stoff agrees with the views (stated abcVe) that research in this area can be effective ifsbetter. focused:and that this-aspect of the NRC 4

esearch plan has the greatestipotential~for. benefiting the nuclear

                                                                            ~

ndustry. It would.be unr&alistiC to'believe that condition monitoring. -

                    's hniques can be developed::to 4roject% remaining service life with any..
                    .l agree of accuracy, but condition monitoring techniques can be developed

~ ind use3 to provide assurance that equipment has not degraded beyond

                     .ame pre-defined acceptable. level. Also, condition monitoring programs an best be developed and: implemented as a cooperative effort with full ndustry participation. :Over the next several years. the NRC staff nou!d develop, in concert with industry, guidance for the mandatory spolication of condition monitoring techniques in order to anure                     ',

ontinued equipment qualification over its installed lifetime. This is , c a part of the more comprehensive effurt that is needed to maintain ' quipment Qualification (see the staff assessment following 0.l. above). 3'  :

                    )0recoirementsforreplacementequipmentshouldbebetterdefir.e ustitied.       The following prcblem statements relate to this issue-              1 l

Peer Review Comments: 1

                                                        .kh + ; g.87 O r                 .
                                                                                                             )

4 s

• v

a. I disagree. Much has been d6na inthi'sareabotYbytheNRCand by the industry, it is time to let the industry fine tune their programs. This area., requires continued. vigilance on the part of the NRC and industry.j-That :issthennatdre of; beast,. It should also be pointed out that:some ofethe problem statements' listed ' '

below reflect a.levePpf?unfamlliarity31th Ahe2 issues. , 7 -

                                               ; y q,7

p. . . ".

(g g. c,t.ye , . , , _ . z. .

                                                                                             '. .u g.. .m;my&' lowed for initial equipment Since several different standards: Were'al qualification, depending on plant' vintage,.10 CFR 50.49 (the f0 rule) included provisions for upgrading replacemeatiequipment'to the more rigorous requirements of'the rule. Since a' transition period was necessary for upgrading replacement equipment to minimize the impact on operating reactors, " reasons to the contrary" appropriate for the transition process were established by the NRC staf f and int Jed in Regulatory Guide 1.89.       Unfortunately, guidance was not provided for how Inng this .ransition pericJ should be and more appropriate " reasons'to the contrary" have not been est hlished given that ample time has been allowed for the " transition proce n* to be completed. To the extent that it is truly necessary to upgrade to the more rigorous E0 requiroments. more appropriate " reasons to the contrary" should be es'ablished. However, as discu: sed in Section B of this appendix, some chanqes in the methodology for establishing initial equipment qualification may be possible that could be of benefit to the industry, this is especially true recognizing -that ongoing measures must t)e
 @. eloped and implemented to: assure continued, qualification over the installedlifetimeoftheequipmentj(seet.helitaffasses!mentfollowing.,                        ,

0.1. above). Therefore. resolutiontof:this. concern should be c _ rnardinat ed wi t h ~ industry" initiatives to* imp ~ rove?theE0+ process. 'J r W ' t'r4 - " Reasons to the contrary" for not upgrading replacement equipmant to the r e& rement , stated by 10,CFR 50.49 appear;to .be without merit and should be ,)ustified. E gr Revicw (ommc-tS:

a. I am not aware of this issue.

b Several " sound reasons" were listed in a 1982 (?) Generic Letter signed by Eisenhut. Licensees we.e encouraged to develop others where necessary. c

         " Sound reasons to the Contrary" are probably antiquated and should be dropped. The more common practice of upgrading replacement equipment has been in effect since 1983.

d. Perhaps, but a reading of the reasons to the contrary as outlined in RG 1.89 seems to suggest that economi.cs.is involved and technical justification may be somewhat elusive. .

b ff D iepsment. See the staff assessment of the previous problem statement (above). A-88

m--,

                                            -  vv ,    ,e c-     g;. ,;% -
         -  [ouipment that is qualified to the 00R Guidelines and is well suited for its application must be replaced with NUFG-05BB Category I eauipment regardless of whether the upgraded equipment can' perform the desired funciion as well as the older equipment.

Pee' Review Comments:

a. I agree.

I thought " sound reasons" applied. It certainly

b. Is this true?

should -- replacing equipment with a different type often creates problems that we don't want or need.

c. This statement is not correct.(seetRegulatory Guide'li89 Rev 1, Regulatory Position C.6); 'In addition @since replacement equipment is more rigorously tested thaft.. DOR Guidelines equipment, how can it be shown that 00R Guidelf.1es: equipment.is'better suited for a particular application? If a particular item is performing a given function and it 'is replaced withgan. item -that:cannot _

reliably perform that function, then this is not considered to be an upgrade. Liltf Assessme.n_1: wz , . , . -. y 3, The eal problem here is that..to the-' extent that this situation exists, licunsees have not planned for the eventual replacement of plant equi.l ment and have not taken the necessary steps to qualify equipment to the more rigorous requirements. This' problem is best resolved Dy the indo try Also, see the staff assessment following 0.3 (above).

          -  ine

• ouirement to upgrade equipment from Category 11 to the Category 1 (eit.rta c' NURfG-0588 is prohibitive and provides no safety benefit, espe tally if the installed equipment has been preaged and all that is mis ing is the Category I documentation.

Peer Review Comments:

a. If this is true, the requirement.should be relaxed.

b. Apart from my concern about " sound reasons," I don't understa

                                                        ~

the last line -- does it mean that-00R11evel documentation is available? , i

c. For equipment to be certified ~as NUREG-0588 Category I or 10 CFR 50.49 qualified,~ required more than just pre-aging. l 1

                    ' Category I and 10 CFR 50.49 certification requires better and more complete documentation of performance characteristics during the harsh environment, Thus, equipment'which was pre-aged, but was lacking in current practices of documentation, has been certified to meeting Category II or 00R Guidelines requirements.

d. Industry responsibility; merits analytical resolution. If "all i that is missing is the Category I documentation," it is not A 80

                                                                                                           "i obvious why the existing documentation cannot be upgraded.

However, if the existing qualification program does not meet the C more demanding requirements of NUREG-0588, Category J. It cannot i be concluded that upgrading from Category.;ll to Category 1 is - always " prohibitive anii provides~noisafety benefit." On the other s ily-the only Way , hand, to provide upgrading.thequd)jficationy!40@pmentYoperability. reasonable assurance;of )potinacessar In certain cases, increased surveillance.:. condition' monitoring, and - operating experience may be acceptable alternatives.

e. Exactly what documentation is missing from the file of this equipment? Has this item been preaged only, or has it been preaged and LOCA tested?' If it has.been preaged only then it is quite obvious that required performance in a DBA has not been

• demonstrated, and this condition is and'should be unacceptable.

If an item is not tested, how can reliable performance in a DBA he assured? SlitLAJ1U216 ' s The .iaff agrees with the views (stated above) that Category I coaii itti. tion recuires more than just preaging, anri that upgrading , Qua'ification ts not necessarily the only way to crovide reasonable anurance IN; equipment will function during an avent. As discussed in Section 8 of this appendix, some changes in the methodology for l estabi t shing init ial equipmc.it qualification nity.be possible -that could be of benefit, to the industryi:'Thisd @ especiallyitrue recognt2ing:that { onacing treasures must be developed and[iiginented to assure continued Qualification over the instalied lifetime of the equipment (see the staf' assessment following 0.1, above),' Therefore, to the extent that l

th- concern represents a significant problem for licensees, resolution should be pursued as an irdustry. initiative.

l j - There is decreastnn support and cooperation from vendors of qualified cou:pment. Sorre original E0 equipment suppliers are no longer available Third-party vendors will to provide qualified replacement equipment. j supply quaiified equipment. bLL costs tend to be excessive. Pf ar Review rommen,M :

a. Third party dedicators? Quality may be suspect. also. I

b. Industry responsibility,

c. 1 believe this statement is correct, but that is the reality of the market place. It is not acceptable to compromise the safety of a plant by using equipment not qualified to perform required functions when called upon.

1.t.J ff AssessanJ: The staff agrees with the view (stated above) that resolution of this However, as discussed in Section B concern is industry responsibility. A-90' f

                                                                                           .            ,4                  .
                                                                     . . . -                                   n.-...

cf this appendix. changes inay be possibleJin .the (Q methodology that may help to alleviate this concern. The NRCi staff should be receptive to j suci proposed changes that are (a) developed at an industry j initiative, and (b) technicallf justified. j a. Inght budgetiry continues 10 be a challenge to any advancements in the j area r.' [0 J l f.it.r..A v.11w_.(u mme n t s :

a. I agree, and that is the real world. l know many cases where l

at.r.:cments have been or are being made despite this constraint. I bel te.e that time and marht forces will take care of this. 1 b. Merits analytical resolution. The suggestions (made in response 3 to other issues and problem statements)-that EQ requirements be i reviewed in light of the experience ofAtht latt two decades has- i i t he prospect of increasing the asstfrance of safety and reducing CO 1 j costs, for example, if the adalified. life requirement were i replacedbystandardizedst'essteitjrig,itcould~.redut.ecosts l l significantly. y' pOf f Nnsm?M: f 1 The fF- staff agrees with the view (stated above) that a review and ! adjus'"-nt nf the E0 requirements in light of the experience and i inforN: 100 that has been gained over the last two decades has the ! prospe t of increasing the assurance uf safety and reducing f0 costs. Howese' resolution of this concern is industry responsibility and should '>e pursued as an industry initiative With full NRC staff tonce a', ion. l Sumgry l Based on staff's review relative to assurance of continued l avalifico: :an. the following recommenuations were made: While 'he NRC. st af f agrees with ti,e peer review comments that a re-l 3. emphas's of E0 awareness and promulgation of guidance based on current l industry prac. ces would be helpful, a more extensive effort is needed. ! In particular, the staff shouldiassure thatsoperating plant ~er.perience and PPA inf ormation, equipment: performance, condition.and environment monitoring. root cause assessment, and trending of information are used l in a i eprehensiv3 manner in order to maintain E0 over the life of the ! p l a n '. Further, the staff's guidance should not necessarily be limited i to "corrent practices." This approach would be useful in addressing i current concerns such as installation, maintenance, and surveillance effects, hot spots; long overhangs; aging degradation; etc.

b. Implen ntation of the maintenance rule should help to assure continued equipment qualification over the useful life of each active item that is qualif ied. and the staf f should initiate action to include passive electrical equipment within the scope of the maintenance rule to better assur.. continued qualification of electrical equipment A 91

_ _ _ _ _ . _ , . _ - . . - , _ . m ---4

                                                   *taff should develop, in concert

c. Over the next several years, the NCf w ; . r. industry, guidance for the mandatory application of condition

' moottoring techn1 Ques in ordor to assure continued equipment qual ification over its installed lifellme. This is a part r,f the more ' coeprehensive effort discussed ~in (a) above that is needed to maintain eautoment Qualification. ,; ,

d. To th" extent that it is trulf necessarfsto Upgrade to the more rigorous E0 requirements, more appropriate " reasons to the contrary" should some be established. However, as discussed in Section B.of this appendtx.

thanges in the methodology for ettablishing; initial equipment qualification may be possible that could be of benefit to the industry. 1his is especially true recognizing that ongoing measures must be de. eloped a d implemented to assure' continued' Qualification over the Therefore. installed lifetime of the equipment (see (c) above). ' resolution of this concern should be coordinated with industry initiatives to improve the-[Q process,

e. The NRC staff agrees with the po r review comments that a review and adjustment of the E0 requirements

                                                 .n light of the experience anil information that has been gained over the last two decades has the pi .;mrt of increasing the assurance of safs.t.y and reducing      E0 tosts.

lity and H r . .. v e r . resciutton of this concern is industry responsibi st old be pursued as an industry initiative with full NRC staff i mperat ion. l The Nk' staff did not consideriany of the issues pertisining to assurance of contiraec qualification to be immediate safety problens. i

                                               .         c f

A-92

   *                                                                                        .l '

3 77%N. . .. ' , E. Eqtoprent-Related Issues . gj , [ . l 1, f a t ' a rt- of other electrical components such as penetrations and contector assemblies may be more important than the failure of ele t rical cables, and more att(ntion may be warranted for these torr,mnent s . Effr_Heview Comments:

4. Both f ailures could be extremely detrimental for plant response to 1 a DRE. so the issue of more significance / consideration for one i

failure versus the other escapes me,

b. Ooerating reactor inspections concentrated on non-cables.

c. Depending upon one's perspective, a case can be made for one or more component (s) as more important than others. I do believe that the locus on cables is correct and should be continued.

Decisions relatir,; to the importance can probably be best addressed if we complete a F5A of EQ priorities as discussed elsewhere in this document.- $ few guiding factors in establishing 4 such importance rankingt'::includetF - maintenance and surveillance being performed on the items of interest; potential for as yet unaddressed or unmanifested common i cause failure mechanisms; I cost to replace, particularly during an extended license i term; potential for causing multiple system and component f ailures simultaneously; and ' last, but not least, failure experience. i

d. What does "more important* mean? Risk impact, higher failure rate?

e, Judging from the failures noted in BHL's review of LER's and NRC's Notices and Bulletins,; connections and.. penetrations may be .l i experiencing more aging degradation than cables and thus would" warrant additional attention,

f. Merits analytical resolution. A valid concern because it is l

possible that the~ f ailure rates of connectors and~ penetration I

                       .stemblies exceed that of cables.

g. first of all, all eauipment within the scope of the E0 rule is j

mportant, as is indicated by its title, " Environmental l Qualification of Electric Equipment important to Safety for Nuclear Power Plants.": The NRC expects licensees to use good 2

A-9' i

                                                                                               \

i l

engineering judnement when making equipment operability decistoas. As and decisions involving the operation of plants in general. for this particular situatloe although electrical penetrations 4nd connector assemblies are important, they w ill not function a'thout the cables tha' transmits power to them. ?

    ' ' dL!.d.cJiO9J11:                                                                                                  )

Based on the peer review comments (above), the spect(tc concern jli u regard'ny electrical penetrations and connector assemblies is a valid li one and further action by the NRC staff is warranted.to resolve this l l issue. This concern should be reviewed in. light of ~the of going l literatore survey that is being done under contract for the NRC to Corrective l deterr'ine to what extent a significant problem may exist. 3-action should be taken depending.on the nature of the defitiencies that 4 W! are identified (should any exist). Beyond this. f ,used attention on I operating plant experience and PRA_information, eQulpment performance. condit ion and environment monitoring, root cause attessment, and l trending of information in order to identify and correct [0 deficiencies (,, this appendix)..will help to address t *e 4ction A and Section B this tuntern on an ongoing basis. The staff agrees with the peer review i commeots that an importance ratiking may be appropriate given operating l: plant .,perience (i.e.. coutpment failure data) and PRA information. \! Mn i s t ', e transmission through cracks in cable insulation or into 8he l 2

                          *hrnugh dif f usior may compromise adjacent connect ors or l

table we This termiral equipment not designed to withstand moisture. j i vulneribt'.ity has not been addressed. ( I Peer Qvtew Comments: I disagrae. Cable qualification testing addressed this. a. I agree partially. It deserves some attentinn in the ongoing

b. is being done under contract for the NRC, l

literature survey that ! If the scores of cabin testing performed to date indicate that moisture intrusion throtyh cable insulation is a high probability event, we should perhapswinitiate additional research on cable connections. I am familiar with a couple of instances of such l occurrences, but need more confirmation, I have also heard of others, e.g., committee members in the IEEE-383 working group and 1 qualification specialists, mention that cable ;onnections may be ' the weakest link in the cable systems. i j

c. This is a valid issue.

, into equipment has

d. The information about moisture transmission I

been addressed and was passed along to the industry in [PRI NP-5000 "Han d book on Electrical Interface Sealing," 1988. e industry responsibility; merits analytical resolutionthe It is primarily and perhaps more NRC oversight may be needed, responsibility of industry to uncover the conditions described in l l A-94

this statement, but-theiprocell:cou}C-be alded.by guidance end-increased oversig;it'from:the u.

                                    ; g;     .NRC..
                                                 ,psm p..             ,-                  i

f. This is the type of'iltliation~that;prenging'is su1 posed to address. If equipment:Js properly.preaged prior to LOCA testing, cracking should occursddring the accelerated aging process that  !

simulates the cracking,that occurs during the installed life of t'1 cable, thereby exposing the vulnerability. On the other hand, i you are referring to equipment qualified urder the requiresdnts of the DOR Guidelines;Where.preaging did not take pla.:e, th;a you

                                                                                          \

are correct. _

                                        ,,-                                               ]

Staff Assessmtal: The fJRC staff agrees with the view.(stated above) that this concern should be reviewed in light of the ongoing literature survey that is im ng done under contract for the NRC to deter 'ne to what extent a s i..a t f ic ant problem may exist. Also see the previous st af f asseument i.'cse).

3. 5a noid valves may not be sufficiently qualified for certain at: ' 'catinns

[f p r Review (omments:

a. There is no evidence 'of this at this' time,

w. .

b. I disa I am not sware of.any'such inadequacy. Having said that, gree.let me also mention'that there is certainly room for improvement in future qualificatton tests. -Ongoing industry group qualification programs'.for.certain SOVs do address some.of them '

(applications). ,

c. Considerauie effort has been dedicated to solenoid valve qualification. The qualification of solenoid upgrades is currentiv being accomplisheo in industry,

d. Industry responsibility: more NRC oversight may be needed.

e. 1his may be correct but we need to discuss specific applications.

   >' a f A.Ud.t.5Ud :

D concern should be reviewed in light of the ongoing literature

    ,~  .*y that is being done under contract for the NRC to determine to
   #.' extent a significant problem may exist. Corrective action should in Paken depending on the nature of the deficiencies that are identified

( teold any esist). Beyond this, focused attention on operating plant av;...innce and PRA information, equipment performance, condition and er..ironment monitoring, root cause assessment, and trending of i n ' e rma t i o'1 in order to identify and correct (Q deficiencies (see Se<. tion A and Section B of this appendix) will help assure that problems. of this nature are identifjed and; Corrected. A-95 < j

                                                                   .9

a. LQ b4ri in element s may not be &de';u4te.

s

                                                                                 . q. .

i EeerR*L_tyLiq'tLTents' r I j

a. I am not aware of this issue, i f l i

b.  ! agree. This is a valiti concern, but 15 not ont that could be  !

resolved by additional researth, it should be addressed through ! Jl plant configuration control programs. < {  ! 1

c. This statement is not clear. i l

d. The information about moisture transmission into equipment Das

! been addressed and was passed along to the industry in - j (PPl NP-5000,

• Handbook on Electrical Interf ace Sealing." 1988. I!

i

e. This may be correct, Eut Wh' ether-t' h ey: ire'qualtfled or not depends l ont.especificapplication(s).iihis.commentseemsto-bemore plant specific than generfl andds representative of appitc.* tion-problems that must be ado %ed by the' users. E0 barrier elements

{ are capable of being qualified for some application (s); but it l should nrt be assumed that they are qualified for all L applications, if these barrier elements are being used in i  ; i applications for which they are not qualified, it is a violation of the (0 rule and shuuld be addressed by the user and the NRC. l I l Ltaf.' N m men.t: i j I0 ta<rier element s consist of flood barriers, walls. enclosures, pene' rat ens seals, etc. . that provide protection against adverse sts are l environ-notal consaquent.es To the extent that f0 barrier .1e g credi*od. the euulpment thatWhile is betnu protected by t he se cl ament s t '. not j this concern .'aa:s primarily with leak-rega cd to Da qualified. I tigt.: an:lo.o r.,. +t is no meant to be exclusive of other EQ barriers. See tna staff assessment of C.1, E.2, and E.3 (above). 5 Qual st ion of equipment seals and vapor barriers on plants, j espe. ally those that are subject to.the 00R Culdelines and NUREG-0588.' may nit ta sufficient. . f ,. i j P,g,gr, Review Commenti:

a. There is no evidence of this at this time.

3

b. I know of no basis for this concern.

C. .This statement is not clear, l i

d. EPRI NP-5000, " Handbook on Electrical interf ace Sealing," 1988; EPRI NP-673), " Guide to Optimized Replacement of EWment Seals,"

i

                                                                                                           ..),

March. 1990; and EPRI NP-6408, " Guidelines for Establishi i l Maintaining and Extending the Shelf t.ife Capability of 1.ielted t.ide items,(HCIG-13)," May,1992 have been made available to the j

nuclaar industry.

4 F A-96

l r' , i e Industry responsibility; more NRC 6"ersight may be needed. l Thismaybecorrectjbut"whethertheyarequalifiedornotdepends r.

                                                                                        .      This comment seems to be more on plant thespecific   specific application than  general an    (s)d is representative of application problems that must be addretted by the users. (quipment seals and                                i vapor barriers are capable.of being qualified for some application (s); butfitishould not.be assumed that they are                                  , ~

qualified for all. applicatienti < !f equipment seals and vapor  ; barriers are being used in applications for which they are not i qualified, it is a vioittien of the E0 rule and should be i - addressed by the user-of these materials and the NRC. M dll...A a si1 < m,gg : we the ;.rc.1ous staf f asse!'. ment; this is a subset of the concern m pie m d hf L.a (above). l f rosy (o pound used for potting electrical penetrations mAy not be 6 i qualitte.1 to the temperature condi' ions that are experientert post LOCA l m1. o r dur;ng a M5lR Peer Reviey iggnini,1: 4. There is no evidence of this at this time,

b. I know of no basis for this concern,

c. Potting compounds in' penetrations should have b*en qualified as part of the penetration. if not, it is most likely a vendor  :

specific or model specific problec.. Industry responsibility: more NRC oversight may be needed.

e. This may be correct, but whether it is cualifted or not depends on the specific application (s). This comment seems to be more plant specific than general and 15 representative of application problems that must be addressed by the users. Epoxy compound is capeble of being qualified for some application (s); but it should not be assumed that it is qualified for all applications, if epoxy compound is being used in applicattons for which it is not qualified, it is a violation of the [Q rule and should be addressed by the user and the NRC.

M AILhiriEm_fM: ( This concern should be reviewed in Itght of the ongoing literature i survey that is being done'under contract for the NRC to determine to what extent a significar)t prol'lem any exist. Corrective actions should be taken depending on.the nature of the deficiencies that are identified { (should any exist).

7. Use of the following products in E0 applications may need to be better defined and justified:

l 1

I *

.9h+ch ;.u: n.nc fDe(f;[ j.h , -

poly 1mide insulation lKipton) Butyl rubber insulation: i ents) mineral wnol insulation (especially_lh wet env ronm

• bonded jackets

    -            coaxial cable               *'

a terminal blocks Peer Review Comenti: life a. True - especially for butyl if used at a plant entertaining extension. t> . I know of no basis for this concern. c. I am not clear what the problems are with the listed m Define the problem. d in its d. No material should be tisdd unlessNhas been qualifieite application. All of tha qualification tests periormed on:them. merits analytical resolution.

e. Industry responsibility: lified or This may be correct, but whether these materials are quaThis co f.

not depends on the specific application (s). i of be more plant specifit, than generalJ and is representat All of ve - application problemt"that' must'beiddressed d for some by the users. these materials are' capable of beIng qualifie being used application (s); but it should not be assumed l that they are qualified for all applications. :lf these mater

                                                      ~

violkiton of the EQ rule and should be a these materials and the NRC. i.tdf _!dlelFind: literature This concern should be review d in light of the ong survay that extent a significant problem may exist with any f the what Beyond this, nature Corrective actions should be taken depending on the o deficiencies that are identified (should any PRA exist), inf orma'. ion. itoring.androot cause focused attention on operating plant experience equipment performance, condition ar.d environt.,ent mon identify an assessment. and trending of information in or<fer to" ' this appendix LO deficiencies (see Section A and oa :.orrected. Sectice assure that E0 deficiencies are identifi> Efluence on the rate of 8, The color of insulation material may have c. its degradatinn. Peer Rey.itw Conmenti:

a. I am not aware of this issue.

A-98

b, ' know of no basis for this concern. I c. This is a valid :oncern.

d. Insulation color ditterences in' rate of degradation is probably another second order effect '41ch is overwhelmed.by the ' severity of tFe DBA testing.

l e. Merits analytical resol't{on. u Staff Assessment: , . ,, ,,i , 10 t h.t astent that focused attention is placed on operating plant experience and PRA information, eQulpment performance, Londition and env*ronment monitoring, root cause assessment, and trending of-information in order to identify and correct C0 deficiencies (.ee Sect nr, A and Section B nf this appendix), contarns such a> this one are

of t nor importance.

LY2iMJ 4 4 [Quipmer related issues and concerns should be ascessed primarily through , i review o' existino information and by taking full idvantage uf operating l plant e.; or ienc e and PRA information, equipment performance, condition and ' e stronm nt monitoring, root cause assessment, and trending of information on an ' e ; .: i n g b a s i s . forrective actions should be taken as appropriate for < ,ignifi .ot issues that are identifled. [Quipment items and concerns that require orther review and assessment by the staff in this regard incl.ide:

             . ele *.ical penetratic1s and connector assemblies
                                                                  ~         '

e solrnoid valves l' [0 b4rrier elements i e sea , and vapor barriers

• po., pntting compound

             . mo i . ' o re intrusion through cracks l

• pol v imide insulation (Kapt:m) -

l l

              -   b t. : ,' rubbar insulation
              . minoral wool insulation (especially in wet environmtnts)
             +

bo.id.d Jackets e coas al cable , a terminal blocks The NRL staff did not consider any of the equipment-related issues to be ' immediate safety problems. A-99

 ~__                                  _
                                             ~ ,: .

1 F, N3 0vq,r1Lqh,1 .

J. The ::rescriptive regulatory approath'that has been taken relative to E0 is rnunterproductive. Inhibiting progrest'and innovative approaches in resc ving this complex issue. Peet Review Commerth:

a. The regulatory approach taken was driven. to a large extent, by the lack of attention ty the industry on this issue. Therefore. I disagree,

b. Can those of us who kept' plants operating while we performed EQ inspections hope to be." paroled" someday? This comment reflects how the NRC operates.

c. I agree. There is some validity to this concern. It should be noteu that despite this, there have been several innovations on CQ problem resolutions. Hy:n'tess:is that the. Industry will not, and generally has not, sat idle if there are cost savings to be had through innovations. No.research: Work is needed on this topic.

d. Merits analytical resolution. HRC regulations do allow devictions from prescribed requirements.providedsthey are , justified.

Ilowever, the industry' usual 17 finds it ' easier to follow an approach known to be acceptable ~instead of undertaking the risk of justifying an innovative approach. 'Perhaps, the NRC should facilitatt he introduction of innovative approaches by encouragir , hem and providing a more efficient process for their review.

e. Perhaps, but the prescriptive regulatory approach resulted f rom a lack of initiative by the industry, and the out-right resistance by industry to the NRC initiative to address EQ. This can be.seen

                   .n the industry response to IE Circular 70 08. IE-Bulletins 79-01, 79-01A, and 79-01B and its supplements. Ultimately, the NRC was sent a ' Petition for Eme:gency and Renedial Relief

• by the Union

                                                      . The petition sought action in two 4                  of Concerned       Scientists  (UCS)lectrical fire protection'for e             cables, and environmental

areas

qualifit.ation of electrical components. The petition asked the  ! Comission to immediately shut do..n all operating plants, and to ) i halt sonstruction of new plants. As a result, the Commission issued a Memorandum and. Order-(Ct.I-80-21) dated May 27. 1980. l which ultimately lead.to the prescriptive regulatory approach i (through documents such*as'the" DOR" Guidelines, NUREG-0588, IEEE I~would like to emphasize Standard 323-1974 and 10 CFR 50.49). here that the NRC's approach resulted from industry's refusal to be cooperative after repeated requests.from the NRC. Ma'.f A s s ej s me n t. In ::eneral. prescriptive regulations do not allow for innovation and l advances in th state of technology and tend to be counterproductive. i A-100 I

i ! ) I l Ihe NRt il4f f should encourage industrv initialives to 'mprove the state of aouipment oualification. and cnanges in the regulatton (10 cfR 50.49) should be mde to f acilitate this approach. Specific methodologies. oproaches. technioues. and details that are acceptable to the NPC uaii for establishing and maintaininq [0 should be provided through the f i w ance of Regulatory Guides. the Etandard Review Plan, flVREGs, and ot'er documents where changes can easily-be made as more information bec omes avai' *1* and advances are made.in.the state of EQ technology. e

2. The safety-significance of EQ issues is not differenti.ted and recognized in the regulatorf process'. For example, the E0 inspections of the mid-1980s emphasized documentation .without a corresponding emphasis on the equipment's safety significance.

Peer Review Comments:

a. I disagree. The documentation findings were associated with equipment identified bylthe licensee as being relied upon to function during/foll6 wing a DBC. This'put E0 equipment in a separate safety classification from that normally identified as safety related.

D. I disagree strongly. The Itcensees set functional performance criteria reflecting necessary safety functions. The numerous complaints about documentation invite the rebuttal: what do you want instead?

c. Human endeavors are subject to human failings. The emphasis on I

documentation cited in the example is :a poor one. One should , recognize that the NRC' inspections were conducted when the work on implementing E0 was moving full' speed ahead. At that junctura, , the only form of objective" evidence available was documentation. Even then, the NRC did. conduct facility walkdowns to review ' equipment installations, and identified deficiencies. Subsequent inspections at the plant sites and vendor ~ f acilities have focused or hardware installation and maintenance, and test equipment. problems. One will never know, and most likely will not want to find out, if the real safety benefits of EQ are achieved. With that said, let us noW11ook at'how.to: improve attention to safety significance. The answer (f PSA. ' Absent'a systematic effort of that type, we can only have conjectures, and honest professionai disagreements on this issue,

c. ine issue is not clear; it is not clear what is meant by the safety-significance of E0 issues. Does it refer to the different ,

contributions to risk of different equipment items? Does the

tatement imply that there should be different levels of safety instead of just Class 1E? (Note: The emphasis on documentation in the mid-1980s was necessary, because it was not feasible at that time to review EQ programs without at. least adequate documentation.)

e. It has been determined by the Commission that because all EQ issues are important-to safety, for regulatory purposes all EQ A 1^) l l

1

                                                      - - -          ~_________-

Issues have simlier safety;si is reflected in the issuance;of'gnificance...This. Generic Letter 88-07 where in approach its enforcement the staff. wat'n6t required to determine the safety significance of each violation of the EQ rule, but rather, was required to aggregate' the' number;0fWiolations and base its findings on an assessment of the. significance of the aggregate. ltLLLf_6110AE9.nl: The svolving enforcement policy relative to E0 indicates that there is aorr.e degree of truth to this concern. The Q staff agrees with the view issues in proper (stated above) that PRA may help to place perspective. The NRC staff should be supportive of industry initiatives for u:,ing PRA and other techniques to better focus E0 requirements and for ;. lacing E0 issues in proper perspective.

3. biven the state of the art that was in tristence at the 'ime IECE 323 7a was developed and the limitaticns that existed. it would seem that EQ

                 -rogram recutrements may have been misdirected (especially with regard to 'he recuired determination of ' Qualified life" and the absence of wrveillan

• requiremant s f or obt aining advance warning of significant inuradatton,. ]

)

Peer Review Comments:

                                                             '                                                   l

a. I disagree. 1 i

should be eliminated

b. I agree. The concept of

• qualified life" altogether, and emphasis should be placed on surveillance, l maintenance and Londition monitoring.

c. The concept of qualified life has merit in many respects, not the least uf which, is the impact on knowing which safety systems are being relied upon for plant operation. The attainment of qualified life is highly probable unless there is a flaw in the '

i original assumptions. By concentrating affort to monitor i equipment's condition, incsr . flaws, if they exist, could be d e t e r.t e d . The monitoring of equipment condition provides continued assurance that the qualified life is cbtainal.le when no deterioration is found and provides necessary feedback and corrective action opportunities when unanticipated degradation is found. Non-intrusive surveillance for obtaining advanced warning of significant degradation and the performance of component root cause failure analyses for; obtaining the information on actual equipment degradation..are activities which promise both safety and economic payback.

d. The NRC staff does now and always has insisted that licensees and applicants include' surveillance and maintenance as par

• of their a without 3

E0 programs. In fact, the staff considers an EQ progra surveillance and maintenance component to be unacceptaole. However, the surveillar.ce and maintenance insisted upon by the staff are minimum levels .of acceptance, licensees are not A-10? __' T

restrictedtothesemSimum$evelsland'araencouragedtodevelop morecomprehensiveprograms. As~you have implied, the Commission has adopted IEE[ 323-1974, and for the most part, considers it to be an acceptable method of complying with the requirements of 10 CFR 50.49. IEEE 3Li-1974 also provides a minimum acceptable level of compliance, and licenset; are not forbidden to go beyond its limits when developing EQ programs. As for the requirement of determining a qualified life, to date neither industry nor the NRC has found a more acceptable substitute for determining the reliability of electrical equipment.important to safety installed or to be installed in~ Nuclear Power Plants.

• Merits analytical resolution. The reaufrements of IEEE Std 323-74 were not. necessarily misdirected; they were the best that the industry consensus could producatatitnat; time. However, we have v

gained much experience in the two decades.since then, and it 'is important to incorpo.* ate current knowledge in a revision of qualification requirements. The following additional views were also expressed in response ** other related issues and problem statements:

      -       Here effort should be directed toward applying the lessons learned during the last two decades to modify qualification requirements to resolve the problem of demonstrating a meaningful gealified life (ret age conditioning).

One altarnative to the requirement toi qualified life is the , use of stress testing prior to IOCA testing and enhanced l condition monitoring in service [re; state of art I capabilities). The issue emphasizes the point that has been made by several other issues and problem statements that an effort is needed to replace coa.ified life as a major element of EQ [re: equipment aging). Operating experience and enhanced condition monitoring are among the alternatives that should be considered in lieu of a qualified life tre: scheduling EQ repetitive maintenance). The effects of installation, maintenance, and surveillance practices have been addressed only cursorily in industry standards and reaulatory guides and it would be helpful if the NRC provided more detailed guidance than is currently available [re: effects of installation, maintenance and surveillance practices). The suggestions (made in response to other issues and problem statements) that E0 reautrements be reviewed in light of the experience of the last two decades has the prospect cf increasing the assurance of safety and reducing E0 costs (re: tightbudgeting). A-103

i l l i

l i Re f.L6s uament : i i

j Given i*e shr.rt c oming s ani. .ncertainties that exist relative to

              "  cualifiad Infe.' failure cf the NRC staff to recognize the need for i              focused attention nn maintaining. continued qualification over the i

l installed Infe of the equipment.-and-fallure,ol the staff to justify

multiple [0 standards supports tha'contentionithat NRC efforts were somewhat misdirected. However. a lot Of gD0ditechnic41 work has been '

done bra

• by the nuclear industry and by1the NRC, and the NRC staff l

.' agrees ith the view (stated above) that it is.important to incorporatc 1 j currant knowledge that has been Q&ined over the last two decades in a The NRC staff i revis un of the existing qualification requirements. should he kupportive of industry initiatives in this rega d. l  !

a. A lot of research has been completed and much more experience has been obtaincu in the area of E0 since 10 CFR 50.49 was issued, but focused NRC programs ano initiatives apparently do not erist to Continually mon i t or progress in this area and to make use of thic information for j

1 festt14 ? .4 r 4 09. directinc. and i"irovino E0 program requirement <. Peer R_cylew Comments: 6 ) j

a. I agree. The ongoinq literature review that is currently beltg l

i ec formed under can'.ract to the NRC should take care of this ! c nncern. i

b. Merits analytical resolution' (see comment "d? of the previous l cnncern).

i - I disagree with 151s statement. The NRC' sponsored a sigulficant j c. port ton of the rese.'rch conductea 1n;this area,. and uses both the i ' + !sults of that renarch and'alFapp1(cable 6xperience obtainable [_ m an effort to improve E0 program requirements while maintaining nunimum acceptable levels of compliance that will insure high 1.vels of safety, i Sla f f.. **"O.t' 5)EfILI : l However, except for situations where current , This '- a valid concern. f0 pro tices are perceived to be lacking or inadequate, it is primar.ly the re,ponsibility of the nuclear industry to identify improvements and altern4' Ige, that are dasirable and technically justified. Nonetheless. i the NE' < taf f 'hould cat alogue and be f amiliar wit h the advancas that l are p" q made. and -aintain cogntlance and expei'ise in the area of E0 7 j j iorther. !he NW st af f should encourage and be sua; 3rtIve of industry l initta iyes in improve I0 methods and practices based on neu j i deve:opments and advances that are being made. 1 5 NRf. research att iv >t ies have not been entirely successful in resolving the "aoc old' [0 issues that were initially identified (see the summary j of NURf3/CR-0 01 in Appendix L of the staff's report that addresses E0-TAP Action item 3.d for specific examples). Ta-A-104

 )                                                                                                                -

h

 ;                   211r Review 1ments:

4 l a. I am not aware of this {ttua. , I i I anree. The ongoing literature review that is currently being i

b.

perfnrmed under contract-to the NRC should take care of this  ! concern, at least partly.

c. Merits analytical resolution. The correctness of this statement emphasizes the point that'research advances slowly within the limitations of funding. facilities. and qualified researchers.

This rMognition should be taken into account in the planning of new research programs and the allotment'of resources thereto. The i statement also supports this reviewer's sparino use of the "Herits Experimental Rescarch" comment. ! d. If success is measured only in terms of resolving old or existing problems, then you are correct. However, research that doe.t not provide solutions to old problems can.and often does provide new information about existing problems'or: products. The NRC research activities have been more successful in providing new information and in identifying previously unknown limits in component c a p a b il i t i r.s . 'l' -. 113f ! S MS.D mtal:

!                    Research is good to a point, but there are limitations to what can be accctvitshed. The results of past retearch efforts should be Catalogued and th inf ormat ion should be well understood and related to specific coutument applications, as appropriate. Any further research should:

4 I a) be based c>n a well di'ineri need for additional informatinn b) be ! pursuad .nly if there is a good likelihood that the desired information will be ot tained, and c) be pursued only if the cost of research is justi' icd in terms of the expected benefit to public health and safety. l 6 NRf r ..po r t r ; i cquirement s f or E0 related problems (i .e , problems t hat occur daring analif ication testing as well as problems thet occur during l j p; ant aparatinn) ma not be sufficient for establishing an appropriate leve. ,' over.tqht.

Peer P.eview C r .cnn

a. I am not aware of this issue.

I I disagree. If the 10 CFR 50.72. 73 and Part 21, requirements are

b. '

i.nadequate for this, I don't know what else can be, i f niy understanding is correct, there are ongoing efforts by the NRC and I the industry to streamline these reporting requirements either l because they are unnecessary, o" duplicative, or burdensome.

c. Merits analytical resolution; more NRC~ oversight may be needed.

This strtement is consistent with the Common observation that information in industry and NRC. data bases (such as LERs) is not only deficient but can be misleading, 1 A-105

 -.                                                                                                     7

y 4, If

d. Reporting requirements are in accordance with 10 6fR Part 21.

                                                                                                                                                                      ; inadequate.

this statement is suggesting that 10 CFR Part 21  ; then specific examples should be cited. The NRC <taff is l svallable to discuss any shortcomings, including additional  ; oversight. l St((f Assessmgnl: ,, '

                                                                                                                                                                                                                             )

l There is a large degree of uncertainty inherent in the EQ process that sust be recognized and compensated'for. For example, qualification  ; test".9 typically does not involve:a .large enough sample size to bethe statistically significant (in fact.' qualification may be based on l tast results of a single unit), and how equipment will actually age inBya l tion. r aclear power plant environment is largely a matter of specu a onderstanding and resolving equipment problems that ariso during plant l l operation, compensated for. much nf the uncertainty inherent in the E0 process can be reporting requirements: a' low threshold is~necessary for reporting of E0 l

                                                                                                                                                                           '                                                  i that are l                                                    *eficiencies so that the staff will~ be cognizant of E0 prot ems
                                                  . ming identified and better ac e to recognize and resolve emerging E0 i$ sues.

centinued operatica (JC03) allowed under Generic

                               /                   Just i'it at ions for ietter 88-07 may not be appropriate.

P_eer Review Comments: 4. There is no evidence of a problem here. 6 I disagree. This is not an EQ issue that reouires further research focus. It is an industry house keeping issue. It l t That is correct. Sli ff A5se.11matl' The intent of '.be JC0 process 'ias to provide an interim periodduring l fication of relief , l when appropriate to allow-Ilconsees-to establish qua i - The nst ances when eculpment qualification is found to be deficient. ICO process outlined by GL 88-07 allows licensees to 'make a finding of operabii1ty using analysis and partial test data to provide reasonable asiurance that the equipment will perform its safety functic, when

                                                             .alled upon." even thnugh the equipment is supposedly                                                    not qualified However,     the JC0and         is a l                                                                1-es not satisfy tne peovisions of 10 CFR 50.49.

l temporary measure and a long-term solution This aspect that complies of the JC0 with appears the EQ to rule must ultimately be impler.ented. be reasonable given the circumstances and uncertainties relative to EQ in general. The one issue that still needs to be addressed is that the proces; does not recognize the need for an exemption from the E0 rule nei li CFA 50.11.

8. Differences in individual NRr Inspector's E0 knowledge level and inconsistent interpretation of E0 requirements have a severe impact on licensees

• E0 programs. -

A-106 ,

                                                                                                            .e

__._____ A _m_m___________.__ _ . . _ _ _ __________-__--________________._____m_m _ _ _ . _ _ . _ _ _ _ m

                                                                                                                                                                                . - . .               i....is.. a    w.mL I

[11 was suggested that the NRC should. perform technical reviews and t uue saf ety evaluation reports (3[R$ for each Qualification test rrport issued by a vendor'Or!. original)eQUlpmenl~ supplier,.so that lir.ensees would know which"testireporteis: considered valid and Araptable to the NRC for equipment qualification.) Eter Review Comenti:

a. I agree regarding impact of NRC inspectors that have varied experience levels. ,

b. I Ageee. Allhumanlendeavorsare[sublecttohumanfallings. Perhaps the i Maybe there is room' for improvementsvin'this area.

NRC should consider additional training for inspectors, preforably by someone from outside the NRC'with knowledge about industry programs. The second item regarding technical evaluation of test reports. it's a bad idea. Will this not lead to a false sense of security ' amongst the licensees? Nev.r let it be forgotten that the licensee is re nansible for the safety of the plant, and will pay the price in the end, not only for the NRC review, but also for any undesirable consequence therefrom,

c. The same could be said of all' inspections, not just EQ.

J. During the 1980's'significantly more.NRC And industry personnel were regularly attending EQ training courses. These training courses did provide the basics for a minimum underst anding and provided a forum for discussion of latest developments.

e. Merits analytical resolution.. There have been instantes in which '

NRC staff members have been assigned'to E0 activities, although they had no prior familiarity with'the subject. The NRC should assure that inexperienced staff are not'put in the position of making safety decisions before they are adequately trained. The suggestion of technical review of qualification test repnets (presumably by qualified staf f) merits consideration. While it might not be feastble to review all test reports, it might bc 4 feasible to do so on a selective basis: perhaps a few reports in g each equipment category. Since a test report might be used for several different applications, it would still be the owners' responsibility to confirm that the test report demonstrates that the acceptance criteria of each application are met, f. First of all, it was the intent of .the NRC to be as consistent as possible in the application of the EQ rule. However, we realize that in spite of our best efforts, differences in the application of the rule will sc*metimes occur. It is our goal however, to restrict these diffetunres to the differences in individual personalities that we all share in our perception of the world we live in. A-107 o _ _ _ _ _ _ _

   ~_                     '          -                         -. .- . -_ -                   _ _ _ _ _ , _ _ _     _
                                                      . h :,y.s                                                          l 4                                        ,
                                              'u~           .- @

Second, the idea of reviewi" test ~ reports issued by vendors or original equipment' suppliers is impractical because EQ is a bit more complicated than that~. For example / a' test report used to demonstrate acceptable qualification'for a particular : tem to be used in a specific application at site A, may not demenstrate that same item is qualified for a similar application at site B. ' ' Consequently, if the NRC issues an SER stating that the item is qualified, both site A and site B will think the litm can be used l because it was found qualified by?the NRCi".If on the other !)and, < the NRC's SER stated that this item isfnot qualified, site A would  ; be deprived of its.use. Therefore,'we have the situation that exists today, i.e., qualification is site dependent and licensees are responsible for qualification at their respective sites. SDff Assessment: E0 n a very specialized are.' and Whilerequires it is nota~certain level necessary (norofdesired) understancing and expertise. for all NRC field inspectors.tr ha experts,in.this area, snme leve' of training and qualification is ne:essaryP7he NRC Headquarters (HO) Of f ur should establish and maintain expertise'in this area, and resolution of specific E0 problems that are controversial or beyond the . l knowledge level .1f field inspectorsi should be discussed with the HQ staff. With regard to review of test'reMrts, see the staf f assessment  ! of I.9 (below).

9. NM review and inspection programs relative to E0 have not been adequately maintained. For example:

] i l Peer Review Comments: 1

a. There is probably some truth to this on all counts cited (bi: low). H' I definitely agree to the first bullet (below). Not aware of the b.

other issues. However, some of , e The statements that follow are basically true. the current EQ inspectors did take part,in the inspections of the 1980s. In addition,.the SRP (including Section 3.11) is in the process of being updated;.~ l f 5_taff tw.(_.sment: At the time that the EQ rule was publisb..d l fne concern is a valid one. and implemented. there was a lot of contrnversy and uncertainty  ; as.nriated with the qualification methodology that had been approv2d by  : the staff. mijltiple standards existed without technical justification. The situation and information in this ar.ea was rapidly evolving.

!              warrantee continued focused attention by the NRC staff to cata'ogue and                                  j better understand tne information that was being obtained through                                         l i'              research activities. and to closely monitor operating plant experienc4                                   J and E0 problems that were being identified so that continuing problems could be recognized and resolved.         The NRC staff should establish a more l

A-108 J

                                                                                         ?p T "y

focused program of LO oversight b.t: A; establishing and % ';.taining a high level of expertise in the area o E0: (b) maintaining E0 guidance incuments (including the SRP) up to date based on advar:es that are made fhrough re'earch and industry initiatives; (L) developing ar.d - implementing ongoing E0 audit'ard-Inspection programs; (d) setting a low inreshold for reporting of;to deficiencies.so that the sttff will be better informed of E0 problems that are'being identified and better able to recognize and resolve emergirq EQ issues;.and (e) better managing ard directing research activities.

     . Training /oualification of NRC inspectors and reviewers relative to E0 has not been maintaineu. The E0 inspections were conducted on a one-time basis and a periodic NRC inspection program has not been estab'ished and implemented. Current NRC. inspectors are not sensitive to E0 issues, they do not receive-training'on EQ issues and standards, and they did not participate in .the EQ' inspections of the 1980s.

Peer Review Comments:

4. I don't doubt it.

b. This reviewer does not know to what extent the statements in this

] paragraph are true; however, to the extent that they may be true, action by the NRC would be indicated. Also, in response to an earlier issue, the following view was expressed: ) There have been instances.in which NRC-staff members have been assigned to EQ activities, although they had no prior f amiliarity with the subject.a The .NRC .should. assure that inexperienced' staff' ate not2 put~in the position of making safety decisions"before'they: ara adequately trained (re: NRC inspectors' EQ-knowledge). i , Haff Assessment. See the staff assessment of f.8 (above)i.

      . The SRP (Section 3.11) is very much out of date and needs to be made            l turrent    (e.g., the Environmental Qualification Branch is listad as the
          !ead revicw group there is no reference to the'E0 rule; a "cantral              l J
          '11e" is referred to contrary to what was ultimately required by 10 CFR 50.a9; and RG 1.89 and lEEE 323-74 are not recogrized as the             1 appropriate staff guidance documents for satisfying E0 requirements).

Feer Review Coregg,Ui: I a. I agree.

          >M f_f. A u encefl -

1 This is a valia concern. See the staff assessment of F.9 (above).

       . IFEE Standards 381, 535, 627, 649, and 650 (and perhaps others)                l l

pertainir .o E0 have not been endorsed by the NRC. l I A-109

                                                                                                     ~y-Qy MW -
                                                                                                 / MN(4NiQO,$hb;ga. d,: 46                                                          m    ,.      &h M r Review Commen u:

a. Merits analytical resolution.

5Bif. As s e s sgrtt: See the staff assessment of F.9 (above). }

10. The NRC has not provided guidance on how plant data can be used to h modify 'he projected qualified life of E0 components.

Peer Review Coments:

a. I agree,

b. True, but do we;really,waiilthe.NRC to issue such guidance? The a preferably; collectively):and licenseesshouldtdevelopyoneg(The;NRC's.'inputwouldn'.thurt.

implementthemiconsistentih$ Maybe IEEE should beJealldd1ip6n t to do this expeditiously. ' Also, see peer review comments elsewhere on totally getting rid of the clualified lifeiconceptr Merits analytical reso'lution. IEEE has initiated an effort to consider preparation of-a standard that will address the use of ] operating experiencelin qualification.

d. If the qualified life ofJcomponents has been determined by methods acceptable' to both"the!NRC' and -Industry, and the componuts ~ in question have reached the end of their qualified lives, there are only two acceptable possibilities: (a) replace the components, and (b)' retest the components.

EDI!_Auts.tmani : The NRC staff agrees with the view (stated above) that guidance on the use of plant dita to modify the projec.ted qualified life of EQ components would best be pursucJ as an industry initiative. However, to. the extent that focused attention is placed on operating plant' l D experit.rce and FRA information,-equipment performance, condition and environment monitoring, root cause assessment, and trending of information in order to identify and correct any EQ deficiencies that may exist on an ongoing basis, this concern becomes cne of minor importance, ll. Some licensees expressed the view that Informat ion Notice 92-81.

                                                                           " potential Ocficiency of Certa.n Instrumentation and Control Cables."

was not well focused and. created confusion. Peer Review Coments; l

a. There is some truth to this. Again, we should never forget +. hat human activities ire subject to human failings. Hopefully, we have learned from this.

A-110

I

  ' o                                       .

b. Merits analytical resolution. if confusion exists, there is a need for clarification. Perhaps more importantly, there is a need to answer the questions raised as a consequence of the fact that the Sandia National Laboratorics' testing' applied all irradiation (aging plus accident doses)' prior to' thermal aging. It has been claimed that this procedure T*oduced much more cable degradation than either tne sequence of thermal aging followed by aging-and accidert-irradiation.or'the sequence:of aging irradiatfon, thermal aging, and accident irradiation.

c. InformationNotices'(includingithfohmationf92-81 are provided for the edification of licensees,1and.doJnot.yequire)a. response.

However, 'f the information contained.in.an Information Notice is applicable te rparticular21(censee's structures. . systems. and components, then that licensee-isvexpectedsto take corrective actions as necessary.to maintain safety in accordance with the requirements of the Code of. Federal Regult.tions. Information Notices are often written in a general' form in order to provide af fected licensees with the 'lexibility that nia/ be necessary to minimize the impact.of a potentialsresolution. 5)ecifically, . Information Notice 92-81 discussed failu'es r of ca)le jackets and insulation manufactured by two specific vendors. Plants having these cables in environmentr represented by the test environments in which the cables failed, should-conduct an investigation to . determine if (and what) corrective action is.necessary.and proceed with that action. Finally, Information Notices always provide an NRC technical contact (s). Licensees should use the applicable technical contact to discuss clarification.as necessary. M a f f Ant 1Lmtnl: ) inis is a valid concern. While 'it is~1mportant to alert the industry of potential problems with equipmer.t qual.ification, the NRC staff should 1ake c4re not to act prematurely. before enough is known'and underst00d ~ arout tha spacific problem that has been identified, In the case of IN 91-8]. there was some controversy about the aging sequence that was used ' by Sandia. Al m , irrespective of failures, under the accepted methodology a single successful test result provices sufficient basis i j for qualification and the NRC staff was not clear as to what specific j ( ,nditions the suspect cables were not qualified for. D Cf R 50.49 does not define the terms "similar" and "significant" and l 12. q'ndance is needed on how to use these terms. . l  ! Peer Review Commaa(J: J l

a. I am not aware that this i> still an tssue. l j

b. c ee the enforcement (iles

c. We can't have it both ways. On the one hand'we criticize the NRC for being too prescriptive, and on the other we want to prescribe l i

even ordinary and commonly.used terms- Which.~in their usage in EQ context, have none other,than'their7 dictionary meaning. l

                                                 .A-ll!

Q d

                                               - +- ms.?y..r.w
                                        ' x; ^
                                               ,g                                         - ,

em ,

d. An NRC/EPRI dictionary:was-p"Mred to provide definitions and agree on terminology.-. ~ .

c .

e.  % rits analytical resolution.,

f. Todate,tneNRCstaffhas'inthpreted-theterm"similar"inthe context of EQ to mean "the samen;cthe term "significant" in the contextofE('hasbeenprimari)y;1efttothe.discretionof licensees anc, applicants..

                                                    . M. , ,p
           $1&ff Assessment:

fonfusion of this nature is best addressed through industry initiatives. The NRC staff should encourage and be supportive of such efforts, pr equ i lla ."d on the st af f's review of NRC oversight issues, the folluwing recommendations were made:

a. The NRC staff should encourage industry initiatives to improve the state of equipment qualification, and-changes.in the regulation (10 CfR 50,49) should be made to facil.jtate this approach. Speciflc methodologies, approaches, techniques,ind det&lls that;are' acceptable to the NRC staff for establishing and maintaining EQ should be provid::d through the issuance of Regulatory Guides,.the-Standard Review Plan,.NUPEGs. and ~

other documents where changes can easily be made as more Informat,on' b.comes available and advances are made.in the state of E0 technoin9y. b The NRC staf f thould be' supportive of dndustry initiatives to: -(a). improve and streamline (Q requirements, methods, and practIres based on the knowledge that has been devcloped over the last two decades; and (b) use PRA and other techniques to better focus EQ requirements and to help place EC issues in proper perspective, i The NRC stafI shouiJ establish a more focused program of 'Q oversight by: (a) establishing and maintaining a high level of expr tise in the c.rea of f0. (b) maintaining E0 guidance documents and the SRP up to date based on advances that are nade through research and industry initiatives; (c) developing and implementing ongoing E0 audit and inspect ion programs; (d) setting a low U,reshold f or report ing of E0 deficiencies 50 that the staff will be better informed of E0 problems that are being identified and better able to recognize and resolve emerging E0 issues; and (e) better managing an< directing 10 research activities. d The one issue that remains to be addressed relative to JCOs is'that the process does not recognize the need for an exery* 4cn from the LQ rule per 10 CFR 50.12. E0 is a very specialized area and requires a certain level of understanding and expertise. While it is not necessary (nor desired) e Inr all NRC field inspectors to be experts in this are'. some level of A-ll2

                                                .             1 o, o                            :, p     , ,

training and qualification is necessary *a.d should be prov'ded. Resolution of specific- E0 problems'that are controversial or beyond the knowledqe level of field inspectors, should be discussed with:the HQ staff. f Research 't good to a point, but there are limitations to what can be accomplis,co. The results of past research efforts as well as otner insightfu) information relative-to EQ'should be catalogued and the information should be well underbtood and related to specific equipment applications, as appropriate. Any.further rasearch should: a) be based on a well defined need for. additional'information, b) be pursued only if there is a good likelihood that.theldesired'information will be. obtained, and c). be pursued:onlyilfiithes' c gst of research is justified 'in.

                                                                                                                        'Y terms of the expected 1 benefit 1to'.publ.iClfheilth and safety. '

c. While it is important'to alert the industry of potential problems with equipment qualification, the NRC' staff should take care not to act prematurely before enough is known and understood about the specific problem that has been.. identified.

S . . ,.. 1 i l

                                                                .n     .
                                                    '_ 1 -

A-lli

M=a

                   ^'

e s , e e j C. bj i.,. ell aneous Peer Review Comment's

a. The- list appears to have~ at 'leastTtwoTproblemsibcFirst,'it presents both ~

sides of most concernsildentifieds(un4the(one:. hand, the old criteria,; 1

requirements,methodologywerc1tooJrJg6rousppreteniptive6andonerous; '

andontheotherhand,'theywere'notistlffi,enodgh)I/;'Second,*4thelist doesn*t reflect much study 6f shat was'done~inllicensing and operating reactor EQ actions in.the 1984-89:. time.? frame.t*For. example, ground rules weredocumentedconcerningcoldshbidown','ify~escalatedenforcement actionsfor30-someplantsitepir'Aectidnsil bodsced'c'onsidibable consistency with regard to sionif hintIVioli ~ont',' and'the EQ inspection l procedures addressed PRA for the -sample : selection process. - i i Here's an example of what has been'done. .The' Frankil~n TERs of the early to mid 1960, documented EQ reviews,for all of the master list equipment ' at all of tl.e operating reactors. . During1the EQ.inspecticrs, we found good and bad aspects, including the failure of many licensees to identify all equipment requiring qualification, and we built on the 2 ' faundation of the TE'As. One of the strong. points of the TERs was that iney addressed the plant functidnal TeQuirements .for 11$h'. Component; the component had to be shown to'be capable of performing specified funct lons for a 'specified envirnnment (i.e.,ltdon't askover-simplified wasn't me if it's qualified, ask me what it's~ qualified:for).The/ Franklin TERs weighed several ' generic hurJreo pounds. analysis, it was real'World..:Was the information In them 4V&lutted 48; i f ormulating the present list? 950me cof..ther. Criticisms .m.ghtiread ' ' l' differently if they were. 7

b. It's too bad that EQ was consigned toilimbo2footsoimany years price in ,

the present resurrection.'fContinuity.wastpretty thoroughly. disrupted, and those of us who plowed thousandsrof hours into EQ in the 1980s have i discarded and forgotten considerable information that might be useful now. l l I 1 6 . [y, E. 4 g l i A-114 l I _}}