ML20206U724
| ML20206U724 | |
| Person / Time | |
|---|---|
| Issue date: | 06/12/1994 |
| From: | Virgilio M Office of Nuclear Reactor Regulation |
| To: | Thadani A Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20206U672 | List: |
| References | |
| FOIA-99-82 TAC-M85648, NUDOCS 9902170314 | |
| Download: ML20206U724 (100) | |
Text
,
g s
((kjf(I
,E
,i UNIfl0 RTATil
}
:, '.7 /
NUCLEAR REQULATORY COMMISSION
,/
W A nillN Q I o H, 0,0 #0666 9001' June 12,1994 -
1 j
MfMORANDUM TOR:
Ashok C. Thadant, Associate Director for Inspection and Technical Assessment i
i IROM:
H:,rtin J. Virgilio,{ Acting Olrector.
i Division of. Systems; Safety and Analysis i
)
SUBJftl:
RESUt15 Of Tilt SUlWEY OF EQ EXPERTS ([Q TAP ACTION 11CH 3.c) 3 (TAC. H05648) m
~
~
1 At 016euund in the staff's Environmental Que.lification Task Action Plan (10-1 AP) of Juno 16, 1993, we are performing a programmatic review of l
cnvironmental qualification (EQ) for electrical equipment.
Our efforts in j
l this regard are specifically defined under Action item 2 of.the EQ TAP, which m
r,2 j
includos tho following claments:
3.a Review ticonso Ronowal Background information j
3.b Review fire protoction Reassessment Report i
j 3.c f1telt Opinions from Others (Regions, EQ Exports) 3.d Review Cxisting E0 program Requirements 3.m Roview NRC Audit / Inspection Practicos 3.f Roview Licensoe Implomentation Prar.tices l
l 3.g Iinallt.o Rovlow Results,,
I cur objoctivo in completing items 3.a-through 3,f (above is to identify potuntial f0 lasuoi and concarns that may deserve further) staff consideration.
l i
it is importAnt LO.'acognize that this part of our programmatic review la not intended to resolvo or to otherwise address any of the EQ issues that aro i
bient t riod.
After items 3.a through 3.f of the EQ TAP have been completed, All of the f0 lisues will be consolidated and :)ecifically addressed in tho i
st af f's final report undar item 3.g,"" Finalize leview Rosults," which will Includo rocommandations as appropriate.
Our final report is scheduled to bo eompletud by August 30, 1994, i
(I fl 31l ? <1 t 'l,
49 7 ' O * ','%
il i:
o,
!i
)f A'h 9902170314 940612 I !' b e
l KOHN99-82 pop g
,p.3.y w a ',
qG&0 0
l
1 c
Ashok lhadant
$2. ;
June 12, 1994 j
totth this report wo have completed the review associated with item 3.c of the LQ-TAP," Elicit 0intonsfromOthers.(Reions.EQExkentiniissuesthatwere arts)'andour j
ovaluation is enc osed for your informat on.
The po i
i Idontiflod during this review will be assembled and.. addressed in.our final i
report alnno with any other potential issues that are identified as we I
complots the other items in the EQ-TAP, Please contact me if you should havo j
any questions regarding the onclosed evaluation, l
Ollf\\nal Slgnea By Hartin J. Virgilio, Acting Director i
Olvision of Systems Safety and Analysis I
Inclosuro:
l 10 SURVly RISULTS OF NRC AND INDUSTRY EQ EXPERTS Appundix A - Environmontal Qualification Survey i
Appondix 0 - Outstanding issues, Problems, and Recommendations identiflod i
by Survey Participants
^
Appendix C Environmontal Qualification Survey full-Text Responses from L
NRC and Industry Experts; DISlR10U110N:
Contral filo SPIB EQ file i
tsRus sell CMcCratkon i
Gilobbard i
J1atum Coralton 1
A00nuno r Cllprlingor, ILLD MSlosson, lilCH
owhorry, Pul,R PShomanski, POLR Al l llais toni SPSU JCrato, RES SAguarwal, RIS AScrkis. Rf5 CRourk, Rf5 f),
S PL H : 05.jAL/
SPLD:C D:DSSA CGrat t od.4 )
Jiatur MVirgilio 4 / l-
/94 4/ 3 /94 A/Is/94
/,
/, l ' l ' " 'I ' l I
i e t:
l
i Enclosure !
EQ SURVEY RESULTS OF NRC AND INDUSTRY EQ EXPERTS (TAC NO. M85648) 1.0 INTRODUC110N As discussed in the Environmental Qualification ~ Task Action Plan (EQ-TAP) of Juno 16, 1993, the staff is performing a reassessment of the NRC environmental qualification requirements for electrical equipment. Action item 3 of the E0-TAP lists those actions that pertain to the programmatic review of EQ, which includo:
3.a Review License Ronewal Background Information 3.b Rovinw fire Protection Reassessment Report 3.c Elicit Opinions from Othars (Regions. EQ Exports) 3.d Review Existing EQ Program Roquirements 3.o Review NRf Audit /Inspoction Practices 3.f Revlos. Licensoc Implementation Practicos 39 finalize Review Results thi, pa.ticular evaluatic.n is intended to address E0-TAP Action item 3.c.
filcit Optitions from others-(Rtgions. EQ Ex>erts)." The specific objective of this ovaluation is to ld0ntify potential lQ issues and Concerns by survoying LO exports within the NRC and the industry.
The E0 survey was not
'4 Intended to be a comprehensive review of all' issues related to EQ, but rather a format for a targot group of experts to express their opinions and concerns rogarding the development and implementation of the EQ rule, and other issues related to E0.
Ultimately, all of tho issues and concerns that are identified during the EQ programmatic review will be consolidated and discussed in the final report (l0-1AP Action ltom 3 g).
Therefore, this evaluation does not include specific recommendations for further staff actions.
2.0 PURP05f AND SCOPE Of THE SURVEY lho !01Al' includos, amung other things, a programmatic review of the NRC E0
)
program.
As part of tho E0 programmatic review, the staff developed a survey I
to gather information from NRC and industry E0 experts so that potential prnidoms that may still crist with the current EQ regulations could bo idontified, documented, and addressed as part of the task action plan.
Iho staff selected 20 NRC and 14 industry experts to participate in the survoy.
The survey participants represented a wide ran from inspectors involved in E0 compliance inspections, ge of EQ exportiso, to NRC managers rosponsiblo for [0 program development, to industry consultants responsibio
_ _ _ _. _ _ - - - ~ - - - -
2 --
for implemontlng licensee [Q programs, to test engineert at private and sublic laboratories responsible for CcfKucting the [Q qualification testing, l lo survey was voluntary and the participants wors instructed to answer only those questions in their areas of expti'ise, iho staff advised the NRC regional of f ho exports that their individual expertise was desired and that a "reginnal* responso was not required, Of the 34 [Q experts solicited for this survey, ?? providad written responses for a response rata of apprortmately 651 ihe survey gatnered information about fiur areas relat1d 10 the development and impiomantatton of Ihe CQ rule (10 CfA E0,il9) namely"Programmatle Miscellaneous,"
Roquiromants," Operating [xperience,' 'EQ Inspections.* and the topics were selected to give the participant'r the opportunity to express their views on a wide variety of topics related to [Q, A copy of the survey i.o n t to the industry (Q experts is attached 48 Appendix A.
While it was thn l
Intent of the staf f to l(sep the NRC and industry surveys identical, one quottton (NRC Quottion 05), requesting the names of additional CQ experts to be contacted in this effort, was not included in the Industry's version of the I
w rver and was replac0d With another [Q-related question, the 'taff contaClod eaCh participant by phone prior to distributing the survny.
lhe participants were briefed on the purpose of the turvey and given the opportunity to ask questions.
The participants were given 30 days to (omplete the survey, I
.). O
SUMMARY
Of Itif 'O SURVEY RfSPONSf1 A surnm4ry of the connent s, opinions, and recommendations made by the experts pai t iiipatIng in the survey is provided below, the outstanding issues, pinbiems, and rornmmondations contained within the responses to the survey ha-,
tioon extrar.ted and are included as Appendix B, h.s t,iil orports answerod all questinns, therefore, the synopses glvon below representative of only those experts that did provide a response to a are part icular quest ion, A complete, unedited text of all survey responsos is provided in Appendix t.,
3.1 l'rogrammatic Requiromantsi--Section A Quotion M:
Are the rMulations 10 CFR 50,49) adequate to ensure plant safoty during and follow'ng design (basis events?
Do you believe there are unwarranted requirements? Are the differences in EQ recuirements for older vs, nower plants justifled? What specific changes woult you recomend?
Ivon though most respondents replied that the current requirements (e.g.,
10 flR 50.49) were adoquato to ensure plant safety during and af ter a deslan b4 sis acrident, and that the regulatinns were neither too strict nor unwareanted, many respondents provided recommendations for improvomonts in tho reonieemonts.
Somo respondants stated that some issues, for examplo, allowinq j
nidor plantt to be quallflod to requiremonts othnr than NURIG 0588, Category 1, and using of " sound roanons to the contrary' when replacing [Q aquipmont, l
3 I
These issues have been idontiflod may bo inadoquate to ensure plant safety.lnCluded in Appendix 8.
by tho siafI as Outsianding prnbloms and Approximately half of the respondants stated that the differences in tho requirnmonts betwoon oldor plants and newer plants were justified; the respondents wore aqually split on Whether the marginal safety benefit gained from upgrading older plants to the noWar standard Was justified by the significant rosts associated with upgrading the plants.
Oue s t.l on., A2 :
Does the existing qualification methodo11gy provide sufficient basis to conclude that electrical equipment will be able to mitiqate the offects of all postulated accidents over ths entire range of qua'ified life?
Are the current standards, procedures, and techniques used to conduct i
component typo-testing satisfactory for establishing the bases for onvironmental qualification? What specific coments and recomendations would you make rotativo to the qualification methodology fort a) oldor plants? b) nower plants?
Mohl ecspondonts felt the existing qualification methodology provided a sufItclent basis to qualify equipment. However, some respondants expressed r.nncerns over the standards, procedures and techniques used to type-tost individual inmpononts.
Ihn respondents also provided some specific rec onmondat ions on improving qualification mothodology, which havo been intloded in Appnndix 0.
Y Que s.t ion, A3 :
Is the burden of qualification appropriate to the importance of tho equipmont being qualified? Does the safety significance of the equipment justify the [Q requiromants that are being imposedt What role, if any, should risk significance (e.g. probabilistic risk assessment) play in formulating E0 requiromonts? What changos would you recomend?
While most of thn respondents agrood that the burden of qualification was appropriate to the importancn of the equipment being quellflod, some comented that the u npo of tho equipmont (e.g., equipment in radiation harsh only areas) includod in the rule should be reevaluated.
While there was no c onwnsus about the benefits of incorporating risk analysis into the [0 proc ess, the rospondents provided a broad rango of opinions and reconnnndat ions ahnut the safety significance of EQ equipment and the rolo that probabilit ist ic risk astossment (PRA) should play in [Q regulation, somo of whtrh have been included in tho outstanding issues and rocomondations in Appendix H.
Question A4-.What are the strengths and weaknesses of the existing EQ requirements as they portain to:
a) older plants?. b) newer plants?
lho major strengths attributed to the 30 requirements as soon by the respondents Included:
10 regulat ions are based on (1) the collectivo technical know-how and Judgment of ut ilit los (including EPRI), standards comittoos (including NRL pai t.ir ipant s), nat tunal laboratory staf fs, and the NRC staf f who iuued the requiromonts, and (2) twonly years of qualification to) ting
t 4
oxperience, about fifteen years of research testing at Sandia, hundreds of years of plant operating experience, and accident experience at THl.
for newer plant equipment, the LOCA/MSLB testing of aged samples rasulted in somewhat more realistic predictions of equipment performance under accident conditions as opposed to the older plant equipment which could be qualified using separate effects / analytical consideration of 49Ing.
% # % d;>. win s _..
Somo viewed the significant: environmental.and operating time con-servatism inherent in the qualification process as a strength, lho major weaknesses identified by the respondents are presented here and have been consolidated and included in.the list of issues and problems in Appendix 0.
Some of the terms and acceptance criteria for qualification used in the rulo nood to clarlfted so that the implementation of the rule will be more consistent, A componont noods to successfully pass only a single test to be cuallflod, the requirements to qualify equipment at older plant were not as stringont as those at newer plants; for example, requirements covering aging, margin, and synergistic effects.
No provisions are made in the qualification testing of [Q equipment to cable damage from trampling.' g the normal, service life; for example, account for abuses seen durin Oldor plants were allowed to qualify components using analysis for some of fects, such as the effectxof: spraying..
Ihn E0 process does not adequately account for the " weak links" in I
safety systems such as cable installation damage.
Hild onvironmont equipment is not covered by current regulations, j
Changes in manufacturing techniques and materials used when refurbishing oculpment may not ho adequately addressed by the original qualification documentation.
Ouqition3:
Are you aware of any specific problems or difficulties that currently exist or that existed in tae past with implementing EQ program 1
requiremonts?
Please distinguish between older and newer plants.
Although many respondants replied that they knew of no s)ecific problem that currently exists, some respondents did identify some pro)lems that they consider unresolved.
A summary of the problems is included below and have also been included in the problems listed in Appendix B.
ny
' yg ;+
o 7.
4 s
5-ticonsees need to walkdown and document the location of all safety equipment so that appropriate DBA parameters can'be used to. qualify the equipment.
lho NRC currnntly recognizes three different qualification standards.
Interproting the test data and standards by the NRC and the licenscos has caused problems.
when qualifying equipment, licensees use bulk containment temporatures instnad of more accurate localized temperatures.
Adoquato similarity betwoon the tested equipment and.the equipment.
Installed in the plant has been a problem.
Plants built before the current' qualification standards woro sub, ject to purchasing requiromants thatecontained little in the way of performanco i
standards for cable systems-(i.e., cables and their connectors) in harsh conditions.
~
l The cost of EQ testing is a barrier to entry for the introduction of now product and now technologies.
~
Sevoral problems exist with EQ testing, such as, inadequate simulation of containment spray of facts, improper use of accelerated aging techniques, a lack of consistency in what is required to be includod in j
the test program (especially in older plants), and test requirements that aro too conservative.
Documentation requirements are too burdensome with little safety
{
benefit.
Ques t.toa.M:
Are the current EQ recuirements for older and newer plants adoquate for plant operations beyonc the current 40-year operating license (i.e., for license renewal)? Which EQ issues need to be addressed for continued plant operations '.ieyond 40 years? What modifications would you make i
to the existing EQ requirements for license renewal?
j Most rospondents agracd that the current regulations were adequate for oporations beyond 40 years, although many respondents also included recommendations for modifications.and improvements that would enhance the regulations, NIK respondants cited aging, marg!ns, synergistic affects, differing testing i
and documentation requiremonts (betwoon the qualification standards). lack of continuing inspector training on [Q, lack of a continuing EQ inspection program, and the identification of maintenance activities necessary to extend j
thn life of components to the end of the renewal perloo as issues that nood to be addressed for continued plant operation beyond 40 years.
Industry respondents stated that thero wore no specific EQ issues to address, but noted that Ilronsons. shnuld be allowod mora flexibility to develop analytical solutions to spocific E0 related equipment issues.
i i
j l0utstandingissuesandrecommendationson,thissubject.havebeenincorporated into Appondix 0.
) 3.2 Operating Experience - Section 8
~
um.+-
~
l OsunoLM:
In general, how is EQ equipment actually aging (in service)
- compared with the equipment's predicted life?
1 l Most respondents stated that, generally speaking, qualified' equipment is aging i as orpocted or more slowly than expected.
Respondents acknowledged several jinstanceswhereequipmentfailedduetothermalaging,butexplainedthatin j tho majority of thoso casos the service environment was more severe than the i design environment.
Most of the NRC exports responded that they had i insufficiont information regarding the condition of equipment to make an l
l Informod responso.
j
! Question 82:
Describe problems you have encountered with EQ equipment. Do i some compononts routinely fail before the end of qualified life? Are i
i component qualiflod temperatures and radiation exposure levels consistent with ltheiractualin-serviceenvironment) i lMostrespondantsagreedthatequipmentisnotroutinelyfailingduetoaging
{
! cffects; however, quallflod components that have failed due to the effects of l aging have boon the subject of Information. Notices Bulletins, and other Industry experts:4 responded that the actual; operating (genericcommunications.onvironmontshaveconsistentlylowertemperaturesthan'the' pre
! temperatures used in the qualified life calculations, while NRC experts noted j that, even in service conditions milder than the design environment, cables and other components have failed due to higher than~ predicted locallied temperatures or radiation.
- QuuttoLB3:
Do you believe that maintenance performed on qualified equipment i is suf ficient to maintain the equipment's qualification?
Is there maintenance ibeing performed on equipment or components that may have an adverse affect on I E07 1
!Mostrespondentsbellovethatmaintenancepracticosatnuclearpowerplants
- are adoquate to maintain the qualification of EQ eculpment.
Maintenance l practicos identiflod as 3roblematic in the past anc documented in NRC linspectionreportshaveacenaddressed.
Most respondents also stated that jcurrent maintenance practicos are not having an adverse offect on [Q
! cau lpment.
! 0nultoLD3 :
Discuss your views and opinions of specific cases (current)
- whero replacomont equipment was not upgraded to 10 CFR 50.49 requirements lbecauseliconseesreference'soundreasons.tothecontrary(R.G.1.89)?'
lHostindustryrespondantsagreedthatolderequipmentisappropriatelybeing
! upgraded to the requirements of 10 CFR 50.49 and that the use of " sound
! reasons to the contrary
- is measured and justified.. Some NRC respondents lbolleve that the sound reasons to the contrary were intended to be used on a lonn-timo basis and not to bn used repeatedly as an excuso not to upgrade e
i m
---t--r
-- =--
.V
- S; f
i
. i equipmont as it is replaced.
Refer to the full text responses appearing in Appondix C for a complete discussion of this topic.
3.3 EQ Inspection Activities - Section C Quntion_Cl:
Were the NRC EQ inspections conducted with the appropriate scope and depth and in a consistent manner? Are you aware of any specific weaknesses associated with those inspections that need-to be addressed b
relative tot a) older plantsfg ))ne
,glantst g
- m.,
i Doth industry and NRC respondents' state'd that the inspections of EQ programs conducted in the mid-1980s were performed with the appropriate scope anc depth. No weaknesses in scope or depth were cited, tiough some inspectors stated that the four days (or less) of actual inspection time did not allow thom to perform as in-depth an assessment of the licensse's program as some had desired.
Some reasondents noted that the consistency between inspections may have varlod, but t1at it was to be expected, and the teams became more consistent after gaining experience.~
Outstanding issues ar.d problems associated with EQ inspection activities have been incorporated into Appendix 8.
QusMinn_C2: What safety-significant issues have bosn identified as a result of EQ inspections?
In your opinion, have these issues been adequately resolved?
The respondents, mostly NRC experts, described some of the safety issues identified during the CQ program inspections.
According to the respondents, nn safety significant issues remain outstanding, and the issues identified and the enforcement action taken are we11' documented in the inspection. reports.
One respondent stated that the benefittof the EQ inspections was to refocus the E0 activities on the actual installed conoition of the equipment rather than just on the adequacy of documentation.-
Qucu len C3: Is sufficient emphasis being placed on EQ in the current inspection programt Are inspectors suff'ciently trained? What changes would you recommend?
Many of t.ho respondents answered that little or no. routine inssection activity is conducted in the area of EQ.
NRC respondents stated that tits has resulted in a reduction in regional ~xpertise. To address this issue, the NRC e
respondents suggested developing training modules, including the discussion of E0 topics during inspector counterparts meetings, and development of'a periodic inspection program.
3.4 Miscellaneous - Section D qugn1pajll:
To what extent should maintenance and surveillance / condition monitoring be credited for demonstrating continued equipment qualification?
Are you aware of any surveillance or condition monitoring techniques that can be used to provide some assurance of remaining service l'fe? Do you have any specific recommendations in this regard?
_ _ _ _ _ _ = _ _
)
.8-Host respondents agreed that condition monitoring should be an important part of any continuing aging management program, although many suggested that more research is needed to develop useful condition monitoring teciniques.
The abiltty to determine whether EQ equipment maintains the margin to withstand thn effects of an accident is still a concern.
Some stated that routine survalliances and maintenance alone are not enough to verify continued qualification of most equipment,,$pecific:recomendations have been included in Appendix B.
Ouc51Jon_.D2 :
Should credit be given for other initiatives such as the maintenance rule for estabitshing and monitoring \\ maintaining equipment quallfication? Do you have any specific recommendations in this regard?
Most respondents stated that the information obtained through activities related to the maintenance rule should help in determining the qualification st4tus of IQ equipment, but that the information should not be the sole basis in detormtning the equipment's qualification.
The performance-based approach j
to the maintenance rule would identify Cases Where harsh environment equipment i
experiences unacceptable failure rates during operation, it would not identify aging degradation that is not advanced enough to im)act the equipment failure rate in the benign environments of plant operation, sut may be l
AdVanred Rnough to Compromlse thG safety function of the equipment under harsh environment common cause stressors in a design basis accident.
This reduction J
in margin can only be addressed by environmental. qualification testing.
Quc3 LlonJ).1:
What other options or approaches to establishing and maintaining
[Q requirements would you recosseend?
1 1he respnndents provided several llternative approaches for establishing and maintaining [0 reautrements, some of which have been incorporated into the j
nutstanding issues found in Appendix B.
For a complete susmary of options and approaches provided by the respondents, refer to Appendix C.
Question D4:
(Industry Only) Describe any specific EQ issues or topics that j
you believe deserve further research.
Ihree 10 issues were identified by respondents as needing further research, i
the first postulated that mild environment Components subjected to severe 4
operating conditions, such as self-heating, should be required to determine a qualifted life.
The second involved whether the actual installed stresses seen by cable systems are represented in the EQ programs for qualification.
1he final respondent offers a recomendation to develop a joint NRC/ industry research program to test equipment from plants that have operated for longer than 20 years.
Quettlon,03: (HRC Question D4) Do you have any additional coassents or observations relative to the adequacy of EQ or EQ program requirements at comercial nuclear power plants?
1 i
5everal c oments related to the adequacy of tQ program requirements wore made by the respondents.
Soc Appendix C for th0 full text responses to this questinn.
l i
~
7
.t.
non in related ILnya wan Identified anc int.luded n Appendix 11.
The little
.in.nues the ineffective policy provtring pretor ptive regulation for enmpl n 1,o hnii.I lunan turh an aging.
9.n, t inna in on you know of any literature that play be helpful in addreL Aing thu innus, 6uch an publinhed reportt i studtel i artteleni ste.1 A.c
...li. ( o,nialn* the complete lint of literature recomended by the i o n.uii.inantn.
s, 4n suMMAla
% M ip.* A sEO lin-coiputa ni thin nurypy wal in identify potential illunt that may 6t11I iih ihn s uirent 10 regislat tonn il lhat they could be addronipd al part v..o
.. i 1 6.- 14 LAP the Llaff han reviewed 'he nurvey responson and aniembled a iiu of inuni, problem 6, and raonmandallonn that were identified by the NRf, u.i uniost ry espurt t and han included them in ADiendiX lli lhp itaff Influded inoa6 and problem 6 adequately dolcri)ed by the turve
..onolid4tud thoto intuet and probleml With a comon theme, y renpondent6 unno
.o.i,
.u n i u,,....sultsorthesurveyhaveA16nbeensharedwiththeOfflesofRansarch lio it.nir une in developing the ransarch afforts called for in the IQ IAp.
In
.niinii.in mumher6 of plant lynttimt hranch uned the nurvey retillis to develop i
in.. i i.nl dnt uuinn tupir n inr the Information gathering tite visit t j
..uoi... ii.I under I en k 3, f ni t ha 10- I Ap, l
'K
.u l.]
l l
i l
l l
APPENDIX A
~
ENVIRONMENTAL Ql/ALiflCplOQURVEYl y
1 i
Background Information l
Name (Optional):
l EQ Experience:
i i
1 i
I.
j NOTh 1 or the purpose of this survey, older plants" will be those plants qualified under 1
the Dnision of Operaung Reactors' (DOR) Guidelines, or NUREG 0588, l
Category li requirements: " newer plants" will be those plants qualified under the l
Nt!Hl.G-0588. Category I requirements. If you need copies of these documents i
to respond to this sursey, please contact Christopher Gratton.
l A.
Procram.matic Reauirements
,,#s]')s,.
f 1.
Are the regulations (10 CFR 50.49) adequate to ensure plant safety during and l
followmp design basis events? Do you believe there are unwarranted requirements?
Are the ihfferences in EO requirements for older vs. newer plants justi0ed? What 4
speciin t hanges would you recommend' i
i f
l 2.
Does the existing qualification methodology provide sufGcient basis to conclude that i
electraal equipment will he able to mitigate the effects of all postulated accidents over l
the entire range of quali6ed life? Are the current standards, procedurcs, and l
techmques used to conduct component type-testing satisfactory for establishing the
{
bases for environmental 4tiali6 cation? What specific comments and recommendations j
would wu make relative to the quali0 cation methodology for: a) older plants? b) j newer jumts?
L 1s the bmden of quahtication appropnate to the importance of the equipment being quahfied Does the safety signi..cance of the equipment justify the EQ requirements that are hemp imposed? What role. if any,'should risk significance (e.g. probabilistic rnL aucument) play in formulating EQ requirements? What changes would you recommend?
4 What are the strengths and weaknesses of the existing EQ requirements as they pertam to a) ohler plants? b) newer piants?
A-2 1
5.
Arc you aware of any specinc problems or dif0culties that currently exist or that existed in the past with implementing EQ program requirements? Picase distinguish between older and newer plants.
I 6.
Arc the current EQ requirements for older and newer plants adequate for plant operations beyond the current 40 year operating license (i.e.. for licer.se renewal)?
Which EQ issues need to be addressed for continued plant operations beyond 40 l
years? What modi 0 cations would you make to the existing EQ requirements for license renewal?
l 1
IL Dmrptine Exocriener j
i l.
In general how is liq equipment actually aging (in service) compared with the eqmpment's predicted life?
l i
j
?.
Descnhe problems you have encountered with EQ equipment. I)o some components l
routinely f ail before the end of qualified lifc? Are component qualified temperatures
)
and r.nhatmn exposure lescis consistent with their actual in service environment?
1
!)o you beheve that mainter.ance performed on qualified equipment is suf0cient 'to-mamtam the cquipment's qualification? Is there maintenance being performed on l
equipment or components that may have an adverse affect on EQ?
1 i
4 I)iscou your views and opinions of specinc cases (current) where replacement equipment was not upgraded to 10 CFR 50.49 requirements because licensecs reference " sound reasons to the contrary (R.G.' t.89)?"
C EfQJnspeelion Activities I.
Were the NRr EQ inspections conducted with the appropriate scope and depth and in a conmient manner? Arc you aware of any specific weaknesses associated with those inspet tions that need to be addressed relative to; a) older plants? b) newer plants?
t 2.
What safety sigmficant issues have been identined as a result of EQ inspections? In your opinmn. have these issues been adequately resolved?
.t is surtiejent empham bemp placed on EQ in the current inspection program? Are inspeuon sulfiaently trained? What changes would you recommend?
i itA ~
,.g::p D.
Miscellaneous M.
,3,;. ;
g,, ;gf j
l.
To what cxtent should maintenance and surveillance / condition monitoring be credited -
for demonstrating continued equipment qualiR ation? ' Are you aware of any surveillance or condition monitoring techr.iques that can be used to provide some l
assurance of remaining service life? Do you have any specinc recommendations in this regard?
2.
Should credit be given for other initiatives such as the maintenance rule for-l establishing and monitoring maintaining equ*pment qualification? Do you have any specific recommendations in this regard?
3.
Whal other options or approaches to establishing and maintaining EQ requirements
]
would you recommend?
4, Describe any specinc EQ issues or topic that'you b~elieve deserve further research.
k 5.
Do you have any additional comments'or observations' relative to the adequacy of EQ or li() program requirements at commercial nuclear power plants?
6.
Do you know of any literature that may be helpful in addressing this issue, such as pubbshed reports, studies, articles, etc.?
,~
- 3
+
~
~
APPENDIX 3 Outstanding issues, Problens, and Recommendations identified by Survey Participants i
The staff reviewed all of the responses submitted'as-part of'this ;urvey and extracted those responses containing issues or problems related to EQ.
Responses with similar issues or problems were consolidated into single statements.
The full text of all responses is provided in Appendix C.
The following issues. problems, and recomendations are presented for further consideration by the staff:
(a) The following issues and problems relate to the current qualification testing methodology:
The test conditions require er.vironments that could never occur simultaneously or sequentia')y to any device.
The requirements for outside containment HLLB equipment and radiation harsh only equipment are also not credible.
There are currently three different qualification methods accepted by the NRC.
This creates confusion for the licensee and inconsistency for the regulator.
The secuence of exposing samples to full radiation dosages before LOCA exposure is unrealistic and too conservative.
A single qualifying methodology for all 'E0 equipment is r.ot cost-or safety-effective.
Allowing qualification of E0 equipment based on the results of a single laboratory test without requiring periodic retests does not provide enough information on which to base qualification.
Worst case voltage during the accident is not a qualification test requirement.
lime and dose rate requirements for testing suggest that equipment qualifnation extends beyond the DBA and into severe accident space.
Testing does not account for steam leaks and inadvertent spray actuations during normal operation (00R plants), and the self-heating effects of cables in the worst case cable bundle.
The effects of containment spray have not been adequately simulated during qualiftcation testing.
Cable cualification does not account adequately for deformation of Jacket and insulation at high stress points, the effects of high humidity and high temperature, and local vibration.
Older plant equipment qualification is not as rigorous as NUREG-0588 hecause the components have been qualified without aging, margins, or considering synergistic effects.
s:, -
]d.e
>^
, ((4
~
~
[
[e 1
' y,',
,, }.@h '
f B-2 Some significant aging mechanisms cannot be accelerated.
Some aging mechanisms required to be simulated for all samples may not be significant.
Mandrel bend testing EQ cable. samples does not simulate the conditions seen in the plant.
There is no regulatory guidance on the accuracy.of ~instiumentatico during EQ testing.
e.
m.
Thermally based aging calculations are not' precise and maintenance and inspection activities have not been developed to assist in the aging management of EQ components..,
Excessive reliance is placed on analytical aging calculations that may not be as reliable as testing, especially,in older. plants.
.m.,..
.g. e
.v Synergistic effects policy should be reevaluated because-in most plants, the rad;ation dose rates are low enough that the threshold for synergistic effect, i never eached.
S a it - nf commercial or generically named components is a Th< m pai
=
hea ih m ta is cover various manufacturers, vintages, or designs v. " ~ e vi tntar icing components.
The fol't w,ni t* 6t ng concerns remain unaddressed:
(i) interface effects between components tested separately for qualification; (ii) the effects of mechanical installation stresses on cables; (iii) continuous submergence prior to harsh exposure; and (iv) the momentary electrical effects from the postulated initial peak' temperature and radiation stresses.,
~
Changes in manufat.turing techniques.and materials used when refurbishing i
equipment may not be adequately addressed by the original qualification
--C N I
documentation for the equipment. 6 (b) The following issues and problems relate to the use of risk assessments in EQ programs:
l
. PRAS should be used as th'e basis for developing graded EQ standards.
A post-accident needs analysis should also factor into which EQ systems i
and components are required, j
. PRAs do not model passive components, such as cable systems.
. The accuracy and availability of the data upon which PRAs are based does i
not match the mathematical sophistication of the models.
Consensus of expert opinion-would be preferable.
..c yp
& * !)f
^.'
- e,;., Q, 4 m.~. g. u r
s e.pW$hM'8xL u.
- B-3, s
+ Several levels of oualification would be expensive to implement, licensees would ultimately qualify equipment to the most severe application at the plant..-
s The following recomendations and discussions taken from the responses in Appendix C generally support the role of PRA in the development and implementation of E0 regulation:
- Both the Sandia EQ Scoping Study and an EPRI report on the risk significance of equipment show that a risk-based cpproach to safety classification would lead to a substantially different list of Class lE equipment than given by traditional deterministic methods.
Criteria for the safety classification of equipment based on safety analysis should be developed 'o allow utilities to reclassify.their equipment. Allow the approach so be voluntary.
. PRA should play a ma.ior role in; formulating EQ requirements such as:
time windows of operability, instrument accuracy requirements, circuit design requirements for.~continuoussindication of circuit condition, and testability in normal conditions of?the' potential circuit integrity under harsh environments ~.'.^.PRAs must be realistic if they are to be used and they should be used to. focus the licensee's limited resources on the vital circuits.
By contra,t, these recommendations also taken from Appendix C generally do not support the use of PRA in the development and implementation of E0 regulation:
- Risk analyses should only play a secondary role in deciding safety issues and influencing EQ.
. Risk significance should play a minimal rolo in developing requirements.
1 9tsk significance depends on component failure rates during DBE and this I
Jata is simply not available.
- Do aq.1 try to classify
- levels of risk" or "importance to safety" unless a better method is developed, and only then if there is a very great potential far improvements in safety and performance at reasonable cost.
l 1
(c) The following problems relatetto the NRC's EQ ' inspection activitios of the I
mid-1980s end include problems with the current inspection program:
- The mid-1980s inspections emphasized documentation, without a corresponding emphasis on the equipment's safety significance.
The mid-1980s inspections found that many programs lacked the documentation necessary to support the EQ inspection activities, especially at the older plants.
The EQ in.pections were conducted on a one-time basis.
A periodic inspection program is not being conducted, i
a m..-
c:
.,Ae's a
w J
a.e4 A.
.~s.s 4..
%JA
.'sJa--
_.-m.
_.a 4.-
41 s a
am4 8--
- d-A 4'L m
B-4
~.
- Current inspectors are not sensitive:to!EQ issu s.they do not receive anddid w..,
training on EQ issues andpstand4rds,p'g;.gy,.not par.ticipa 1980s EQ inspections. Jg g.l, g g y.jpg, m9 j
There were inconsistencies (in'.the' interpretation of requirements and a
test results..
(d) The following methods are presented as recommendations for addressing EQ at license renewal:
To continue the qualified life of equipment beyond 40 years, submit a revised qualification package.. Any aging mechanism that might become significant only during the last 20 years of operation would have to be addressed, which is the standard practice in any qualification program.
The original aging and type-testing program could be reevaluated to justify a 60-year life because the utility would present measured operating environments that are far less severe tian the conservative design values used in the original qualification program.
This reevaluation would need to demonstrate the same degree of margins required by standards and regulations.
If the reevaluation does not support a 60 year life, new qualific1 tion test' data may have to be generated. or the environmeat of the equipment may have to be mitigated, or condition monitoring.could be used to-show that'the actual aging degradation of' the item}tswould beetcaraplace,the item.-> Extension of-s il i
program.
The last resor qualification would be governed by.the EQ requirements. appropriate to the current Itcensing basis of the plant.,
. Mainten nce programs for equipment in harsh environments should' include monitoring of certain critical characteristics.. The actions required for license renewal should. consist of the'following steps:
1.
Review all maintenance and replacement activity procedures and z
records to assure that the equipment has been maintained in a manner which retains the qualification status.
ii.
Determine whether the equipment has been operated in the environments for which it was qualified, i
l til.
Determine whether the equipment has been operated in the manner (modes) for which it was qualified.
iv.
Determine whether age is a significant failure mechanism in properly maintained;. equipment."
y-e' (c) Holsture transmission through cracks'in cable insulation or into the cable core through diffusion may. compromise an adjacent connector or terminal equipment not designed to: Withstand cable transmitted water.
(f) [0 requirements should not include equipment located outside the containment and exposed to short-term steam conditions or radiation harsh only equipment due to their low probability of causing core damage and i
i i
.=
y
' y y
,.e wm i
w;, y
.m;
-?
~
,_ r. m
- s. f,.'?
f
~'
- ,w.3 g.sw%2
- ,m a
4 i
i theirbriefexposuretotheharlhenvironmentNomparedwithcomponents 1
j inside containment).
(g) Maintenance programs should be required to monitor characteristics critical for maintaining EQ.
I (h) The maintenance required to maintain E0 equipment in a qualified i
configuration and the maintenance frequency interval should be specified.
(1) Documentation requirements for [Q components are too burdensome with i
little safety benefit, especially for equipment stored for future use.
(.j) The NRC and the industry should endorse a list'of EQ components to I
4 l
eliminate the differences in interpreting EQ test data..The testing of all listed EQ components should be approved by'the.NRC.
t
~
- s.3 (k) There has been insufficient testing on condition monitoring techniques or on the parameters to be trended to allow a technique to be used to t
determine remaining service life M
- ';A (1) Research on naturally aged cables should be conducted so that the current aging formulas can be validated.
The NRC and the industry should also cooperate to develop a program for the testing of other equipment removed i
from plants after 20 or more years of service.
(m) The cost of qualification testing is a barrier to the introduction or
[
adaptation of new products to the nuclear industry, t
i i
(n) tIcensees do not adequately evaluate the ambient temperatures around EQ equipment, relying on the average bulk temperature instead of the local temperature, i
(n) fquipmon; that is qualified to the DOR Guidelines and is well suited for its appilcation must be replaced with NUREG-0588 Category I equipment k
regardless of whether the upgraded equipment can perform the desired i
i function as well as the older equipment..These important performance parameters are not included in the regulations under " sound reasons to the 3
contrary."
j Oc yg i
(p) " Sound reasons.to the contrary" were originally intended to ease the i
transition into the EQ equipment replacement requirements of 10CFR50.49.
Those provisions have become outdated and should be removed from the regul a t.l on.
(q) The 00R Cuidelines state that ongoing programs should exist to review I
surveillance and maintenance records to assure that equipment that exhibits degradation (e.g, cables) will be identified and addressed as necessary.
Programs such as these are not generally in place at these plants.
n.
i
B-6 (r) Under the current requirements, active and passive EQ equipment are lumped together in the development of perforsknce requirements, design require-ments, maintenance programs, and safety priorities..,.
~
. ?S,p CFRA,n%;'.
-(
(s)10CFR50.49doesnotdefinetheterms'simi)ar'-((and'significant." Guidance on how to us+ these terms should be made" avail 4ble.
(t) The regulations do not clearly' state the ~ acceptance criteria for qualifying a component based on operating experience.
(u) Seismic Qualification Utility Group guidelines do not recognize the performance requirements of equipment during a design basis event (DBI),
only the damage to equipment that results from a DRt. More specifte seismic qualification requirements may be.needed..-
(v) for plants qualified under the~ DOR Cuidel 'is and up to the mid-1970s, the licensee's vendor specifications for EQ equipment contained few performance requirements describing the. acceptable performance of cable systems under harsh conditions.
(w) Safety-related equipment located in a mild environment and that crperiences severe environmental conditions due to its operating conditlon, such as self-heating from being continually energited, should be ovaluated for a quallfled life.
(x) Prescriptive regulatory approaches for complex issues, such as aging, are typically ineffective.
it is recommended that utilities be allowed to
)
analyze solutions to aging and.other.EQ-related: issues.
_ cy _
.n.q,..
(y) Maintenance data for older plants should'br compared with newer plants to see whether the newe., more stringent EQ requirements resulted in better equipment being installed.
(/) Licensees need to walkdown and document'the. location of all; safety equipment so that appropriate;DBA parameters' can be used to qualify the equipment.
1 j
i i
~
g- ;.;.e9yw++e.
.?y '
=
y 4
- .. - n.
p..
- P.. A
. S-., c,
.. r.y,.
.\\.
+.,
APPENDIX C
[NVIRONMENTAL QUAllflCATION SURVEY FULL-TEXT RESPONSES IROM HRC AND INCUSTRY CXPERTS NOT[: This survey was conducted anonymously.
Each respondent has been given a unique identifier (e.g.
(a), (b)) for cross-referencing of responses, and the responses have been alphabetically arranged for each question.
A.
Erntatatal IL.RcIltd Ecment.s 1.
Are the regulations (10 CFR 50.49) adequate to ensure plant safety during or following design basis events?
(A) SevPre ACfident\\ degraded core should be Contidored in sobrCe term.
Guidance on f0 of f.Gl's.
Rule for seismic and mechanical CQ, Seismic
$0VG doesn't recognize performance requirements durttig DBE. only the
- +
hG'~W damage after.
~
j (d) Region v believes that the regulations are adequate for design basis event'.
There have been occasions where our. inspectors have seen whero
}
not all the proper requirements were met.
i l
(e) (.at egory I are adequate.
Category 11 and 00R may not be, especially due j
to aging effects.
i to)
Synergism and dose rate effects need better treatment. *[ngineering 1
judgement' should be addressed.
I (i) fos l
(j) lhe qualifu ation requirements are adequate to a%sure qualification if I
reasonably appded, they can be easily over applied causing more severe I
qual i f ic al ion requirement s t han necessary.
One such method of making l
requirements too stringent is the desire for one all encompassing accident i
environment tast profile, lhe resulting profile contains conditions that l
inuld never occur simultaneously or sequentially.to any device.
00R and i
NURIG 0508. Category 11 qualifications are satisfactory as~long at normal (onditions are not severe and reasonable replacement frequencies are in
- place, for inside containment, the accident condition is generally much more severe than the aging condition except for components on hot proccos lines or can'tinuously energized solenoids where care must be taken to l
c ont rol aqing.
I don't believe changes are necessary to the rules.
When replac ement s are required, the components should kither be replaced with 50.49 qualified devices or retested to 50.49 requirements.
(L) 1hn current regulations are adequate but could be improved.
l t
l l
,y"
'f; (*.,3 ff, 8
??p ent w.,
_(
i C-2 (1) The regulatlons are deficient in the area of equipment upgrades.
They only require that the licensee document reasons to the contrary for not upgrading to 50.49.
Thus equipment may never be upgraded to 50.49.
The regulation also does not tie together the requirements of HUREG-0588 Category 1. and RG 1.89 as being an: acceptable methods;for,. meeting.the 1
regulation. What is happening I'. that: newer standards have been' issued which do not have the same regul'faments'for'tXample:as!!EEE'323-1974 which is endorsed by RG '.89 and NUREG-0588. +The NRC guidance has not kept up 4
with the industry.
Things have become stagnant.in this area.
I (m; ' lost licensees put tr. place an EQ program in the late eighties.
I don't think the program is well implemented." When NRC' backed off from the E0 inspections. licer: sees dissolved 'the"EQ group and treated -it as i
overybody's responsibility, i
The regulation as such is good enough but I doubt if there is adequate adherence to the requirements.
(n) the regulations are, in my opinion. generally adequate.
However, our experient e shows arnas where improvements can bn made.
These are covered I
in t he r ummon t s t ha t f o l l o.v.
One of the most significant contributions made by 50. 4 9 w A!, the requirement in.[ paragraph) (d) to generate the now f amous 0 t ist. I believe this has done almost as much for plant safety as I
has t he ont ire {
programl This is due to the fact that the Q-tist gave a incus to the sa cty (6 alures of the plant equipment that'had not been 4
t available previously, lhe very method needed to generate such a' list forced the plant operators to analy2e.their systems' and components' f unctions and gave tham a tool (the List) for, establishing, tracking, and.
rontrolling maintenance, replacement, and other activities affecting this,.
important equipment I trust the Commission isl maintaining vigilance in this area and not allowing this valuable tool'to rust'away.
)
(o) Regulations and t b existing qualification methodology are adequate to ensure plant safety during and following design basis events.
I believe t hat t he requirr'nents are warranted and appropriate. in view of the
< rit It al f untilons provided by Qualified equipment.
rp) the prinr ical di t inct ion between these two requirements is that DOR and t a t rgory 11 permit equipment aging to be evaluated by analysis whereas 4
tategory 11 requires that aging be addressed by testing. The introduction of age condit ioning in IEEE Std 323-73 and its endorsement by RG 1.89 1
appeared to be sensible when it took place. Hownver, the deficiencies of aging acc hnology became immediately so evident that IEEE quickly publisned an addendum in 1(([ Std 323 stating it was not its intent that industry
'hnuld undortake research to advance the state of the art to address the agtnq requirement in qualification.
t he gap bet ween t he requirement to demonstrate a qualified life by atteleratert aging (other options rarely being viable) and the status of agtnq technningy created a dilemma for those engaged in E0. Manufacturers, S
y
, yw, m-
' 'r ',:h
'f C
'b
_ j.,
' ;y%n ;. Q N M, y,, Q I; Q N M f
'D
^
g C-3 utilities, and test laboratories did the best they could: gradually the NRC accepted certain approaches; and these were adopted as benchmarks by 3
t he indust ry.
In the af termath of TMl, the NRC began to question whether the E0 programs l
that the utilities had been accepting were indeed in compliance with regulatory requirements. A broad'E0 review-undertaken at that time revealed extensive deficiencias, most of which were ganarously labelled as deficientles in documentation. The deficiencies were ultimately resolved:
mostly by analysts, but in some cases by additional testing.
l My own approach to the dilemma was to conduct EQ programs as well ar, pou nble within the limits of existing teshnology, to urge strongly that a
the technological limitations be taken into account in interpreting the rrquits, and to recommend that other approaches assuring plant safety (e.g., (ondition monitoring, surveillance, service condition monitoring, failure analysis and feedback) be given more emphasis'than they were 4
receiving at that time, for example. I took%the position /that a qualified Iife (0L) established by an accelerated aging. program was at best an est imate with a large uncertainty :1 felt that OL estimatos should bc subject.to revision as more information became'available about service conditions and equipment performance and as aging technology advanced, I repeatedly poinped out that lifetime predictions of hundreds, and even thousands, of years were practically meaningless, because the uncertaintles in such predictions cotid be comparable to the numbers themselves, $1milarly, the claimed;or implied accuracy of Ols established by age conditioning were exaggerated..e.g..
it. made'no sense to report a 4
i Ot of al years and claiming a margin of 1 year, if one neglected to j
evaluate t he uncert ainty in the 01, In careful laborat ory experiment a, on slabs of insulating materials (with an arruracy f ar exceeding that M hievable with oquipment such as Cables) typical uncertainties wera of the order of 10 and +100 years.
Sand i a N,et inna l l abora t or ief, (SNL) has achieved nigh accuracy in the a, c elerat ed aqinq of rahl insulat ion as demonstrated by the self-innsistency nf Ihe expor, mental results and agrooment of the data with a theory of the superposition of thermal and radiation aging effects, However, the precision of SNt's work is not at all typical of the ac celerated aging conducted in E0 programs. Furthermore, the evidence that t he rondit tons produced by accelerated aging are the same as those produced by aging in service is still far weaker than the evidence of self-consistency.
The age-c onditioning of even relatively simple devices has sometimes proved surprisingly difficult. As an example, one can cite the NRC research on solenold-operated valves (SOVs)
Although the program allotted far more resources than would be devoted to equipment aging by industry, onexpected dif ficulties caused the quality of the aging achieved to f ali
,ignifit.antly below the goals that had been set.[1]
i lt is generally recognized that it is easier to accommodate the aginq tomponent of f0 by ar,alysis t han to do so by t ett ing. That is, it is I
Q;U.)ZY pl} }, ~
.. w v.~
3
- !p.% glm
)
i C-4 eas!er to satisfy regulatory requirements by analysist but it is doubtful' that analysis produces the same level of technical rigor as that which can be accomplished by testing. At a recent technical meeting, a utility representative reported that analysis predicted a life in excess of 1000 years (given to four significant digits!) while testing failed to produce j
even a 40-year life for the same equipment. As discussed above, both l
mothods aro subject to consionrable gncertaintylebut expertence.shows that testing 15 the more realistic of the.two methodsh' f
j i
)
==
Conclusion:==
The bottom line of the foregoing-discussion is that neither' 3
J Category I nor Category !! are iden)lapproachesito3 EQiHowever,especially-forequipmentsubjecttoharshenvironments.thiCategoryIapproachis i
i suportor: if equipment can pass an accident simulation after it has been degraded by age-conditioning, one has greater confidence in its capability than one could have if the accident'. simulation werefnot preceded by j
ago conditioning.
(q) " Adequate to ensure plant safety" is used as if it Were a definite i
condition or criteria, it is not, of~ course, as is implicit in the NRC's defining this adPquacy differonlly for two identical plants diff0 ring in 4
j nothing but the date of construction'and/or licensing.
Therefore the i
question only has rationality if one is prepared to speak in relative terms su(h as. '4s plant A as safe now as it would be with some changed i
requirement?"
or "--as it would be with il million spent for certain equip changos?" or "as safe as plant B built to some different criteria?"
]
nte.
I personally believe there are some shortcomings in the basic grandfathering premises involved in 50.49 as wil.1 he clear from further l
comments below.
l (r) Comment 1: Requirements are adequate and warranted.
./ v v.,s o.
w I
Comment ?:
No.
The current accepta.nctiof;4eparate. effects testing as a.
i bases precludes consideration of potenti'al' synergies' for older equipment.
A test-based determination of qualified life is necessary.
Further, many l
older plants do not regard 10 CfR 50.49.(J) as meaningful, lacking a real NRC effort to enforce it.
4 Comment 3; Older plant:.' requirements are lacking in the seismic area.
SQUC is a stop in the right direction, but not good enough.
Need to l
rnnsider more spectftc seismic qualification of equipment. Similarity i
arguments o.f 5006 i ton shalluw..
2 4
Comment 4:
Justifiable only on a practical basis that replacement or l
upgrading wuuld be more detrimental to safety.
(s) the regulatlons in 10 CfR 50.49 have been adequate to ensure plant safety in the ten years since they were published (no one has challenced their adecuacy) and in the past few years the utility industry has given strong i
arguments for the appropriatennss and adequacy of the same Rule for i
est ablishing qualif ted lives as long as 60 years.
i b)
( g, $ :....
.,n.
(t) The regulations (50.49) provide lA basis for.An effective program to establish tho qualification of electrical equipment important to safety.
The regulations do not, however,' ensure' safety. ilmplementation of the EQ program onsures plant safety..
(u) This question is somewhat misleading since compliance with 50.49 alone cannot onsuro plant safety.
The more appropriate. question would be does t he re.julation ensure adoquate' electrical equipment perforniance during' events.
I holleve tho answor to this question is yes, since 50.49 Is broadly written to address; equipment scopo, aging, event stressors, and methods of qualification.
The broad tenants of the regulation provide appropriate consideration of the critical clomonts of environmental qualifIcatton.
ifowever, portions of NRC E0 guidance documents. (e.g..
Rorjulatory fiulde 1.89. 00R Guidelinos), or NRC staff interpretations aro unnoccuar t ly res t ric t ive.
1 hose are described elsewhere in this '.urvey r e s po n '. e.
(v) 10 LIR 50.49 providos adoquate guidance to ensure plant safety during and following a design basis event when used in conjunction with other requirements such as calibration and maintenance and the resulting data is evaluated (or impact on qualifind 1ife and porformance during and after an accident.
Tracking and trending of a component performanco is critical to maintain confidence that the component will perform its functlon.
g Are any of the requirements too strict or unwarranted 7 i
(b) 1 do eat believe that the f0 requirements a f too'strictl The' question is really quito simple: a system must be in service for an extendvd period of time, what is necessary to demonstrate that it will work? Testing new equipment clearly won't do it unless it is clear that the equipment will not be adversely affected during its service life.
Since we know that radiat inn and temperature both degrade cable insulation and other material, the test-it-new approach doesn't wo k Without a lot of additinnal informatlon.
(e) None of the requirements are too strict or unwarranted because E0 can be
.1ust ified by a singlo sample and this justifies the extra margin.
(i) No. They are nnt strict ennugh.
(h) I believe that the regulations are adequate and not too strict or unwarranted.
(1) No (q) At this time of writing. I do not have the 50.49 text available for detailed review. but one outstanding aspect of unwarranted requirement tomes to mind.
The almost simultaneous peak radiation and temperature c ondit ions po'.t ulated f or 10f.A is incredible in the minds of all t ho physic ist s and engineer' of the national labs, and at Mll with whom I have oti sc uued t hn ma t t er.
My concern has been that my evaluation of tho
h,5liffN5% WP.
.c s
.JQ }
,. \\ v C6 oporability of many cable systems throuch such a condition indicated cicarly that failure.would occur.
This.was. reported in some detail by the writor to the ACR5 many years ago.
The not result has been that those who understand the problem are politically forced to ignore the unrealistic requiromont and supply equipment without a complete _E0 demonstration, flawod requirements giving rise to flawed responses does not soom to be a smart or ethical approach to public safety.
What other response can the industry make?
(s) I know of no unwarrantnd requirements.
(t) Yns, in the aron of qualifying.acui?mente. located outside. containment and.
thosn components subjected.to rac ta tion only.,.0utside containment equipment exposed to HELD's'are typically not contributors to an increase in core molt frequency.
Qualification of equipment to Post-LOCA radiation is based on very conservative source terms..
(u) If ono limits requirements to the provisions of.50.49 then two requirements may be considered as unwarranted.
First, the requirement to age precondition for all significant: aging mechanisms prior to accident testing appears overly rostrictive-in practice.
The effects of some aging mechanisms simpiv cannot be accelerated..,s.In other. cases, traditional l
stressors (o.q., temperature)maysimplyno!besinificant yet, Staff interpretationsgenerallyconsiderthermalandradAtionagkngtobe absoluto requiromonts, if the term "significant" was properly clarified then the regulatory objective may be more appropriately achieved.
The second unwarranted requirement, contained in 10CFR50.49(j). requires qualification documentation to be maintainod for equipment " stored for futuro use",
if this were deleted the existing language would still i
ensure that documentation was maintained for the equipment's entiro l
installod duration.
f urthor, this requirement is confusing.
l l
As noted in below, the applic61on of a uniform methodology for all equipment within the scope o' 50.49 is unwarranted.
It is simply not cost or safoty-offective to apply the same methodologies to an in-containment device requiring active performance during..severo LOCA steam and radiation I
conditions and to some outside' containment device experiencing, in l
comparison, relatively benign radiation-only conditions.
i 1
(v) unw rranted requiremontt imposed on or assumed by licensees stem from 1
different interpretatio1s made by the NRC with respect to the documentation and data seeded to prove qualification to 10 CFR 50.49.
Consistent application of the rules ar.d regulations by the NRC would eliminate this.
Arc differences in E0 requirements for older vs. newer plants justified?
(a) No.
p-) Ihr differenre'. betworn novrJ cnd older plant fQ requirements have never
)
been torbn tr ally just if tte ' n t tpth.
j
- 1 C7 (f) No they are not justified.
Safo shut down of the plant must be a requiremont (primary concern),
(Improper) Maintenance and (doviations in) in,orvice condition has not been factored into the qualification of the equipment, (l) No (k) Olf ferences in the treatment of Old Vs.unoW plants' are justiflod.
(n) I cannot stato unequivocally that the differences in requirements for older vs. nower plants aro justiflod but, as a practical matter, I still believe it was the only practical thing to do at the time and I further belleve time has shown the wisdom of the decision, in fact, as I will di'.cu n further in [quustion) 2.,.it may be that the decision to make the requirements tougher for newor plants was the questionable onal (o) lho issue of the adoquacy of the.[0 for older plants vornus newer plants is difficult, at best - particularly when the issue of continued operation hoyond 40 years is considered.
Newer plants who properly performed EQ have at t ual aging related data for lifotimer, up to 40 years and should bo in a annd pnsit ton to prnjort lifo beyond 40 years using tools such as the Arrehntus model.
However, this is not true for older plants.
Many older plants do not hevo aging data for the equipment of concern, and at best-only have test da,4 for some of the materials-In the equipment, lo show qualific atinn to 4L vaars required considerable analysis, The Arrehnius model assumes a basi,,f known performance at known conditions frora which to pro.tect performance at differspt..but. loss sovere conditions.
Because I
alder plants would probably beginif romca: basi ~ 3.that:made ~ heavy use of '
s analysis rather than actual data,Jtiey may:not;be able to justifiably uso i
the Arrehnlus model and may not have a good basis from which to show life beyond 40 years.
They would probably have to repeat their original analyst'. and providn a snur.d technical basis to show that a lifetime greater than 40 years, such as 60 years istoxpected.
This issue is most difficult for electrical equipment whose performance or condition has not been periodically verified to be as expected, such as electrical cables.
While most items of equipment can be testod, inspected, or monitorod.
(. ables have not been the subject of a regular maintenance program of insportinn and monitoring, if such a program was in place the licenson could t ake advantage of the provisions of Regulatoie Guide 1.89, paragraph C.S.d.. which Indicate that the results of periodic surveillance and test ing programs would be acceptable as ongoing qualification to mndify qualified life.
However, I am not aware that this type of informat6nn is available for electrical cabins, Iho present LQ programs do not provide for data that could be of value when at tempt ing t o use in-situ losting to show that thnre is remaining useful life.
Specifically, critical.proporties are not evaluated at tho end of Ihe preaging period t.nd prior to-testing 8t simulated Accident iondittons, As a result it is nearly impossible to datormine how much margin is required at the and of qualified life for a particular item of equipment to survive design basis accidents, It is this margin that must be maint ained throughout equipment life, including oxtended life, lhe
.._.i___.__-.
v.
y
,.I Nd
)
5bb kNNI
- b
^
,g(gy% 1 C 8; emphasis of oxisting EQ is to show that equipment can age for the l
spocified porlod of timo (40 years) and'thon survive design basis accidents.
Rogardlag the issuo of the adoquacy of current EQ requirements for oldor and newer plants for operations boyond 40 years, I believe that the curront practicos are adoquato for nower plants that have prongod-electrical equipment and that the current requirements may not be adequata i
for older plants whero proaging was not porformed.
1 believe that the i
proforrod approach would be if environmorttal qualification (EQ) was demonstrated to include the licenso renewal porlod (typically qualified for 60 years) and the CQ was based on actual test data in accordance with Regulatory Guido 1.09 and ICCf Std 323-1974.
This would include preaging l
heed on the speciflod service conditions for a ported of timo that includod the Itconso ronewal porlod or demonstrating that previous i
preaqing (typically based on 40 years of servico life) contained sufficient conservatism to ho valid for the liconto ronewal period.
Whon conservatism is utilized the amount of conservatism must have a documented technical basis that thoroughly justifies its use. When EQ for the liconso renewal period cannot bo establishod for the identical: type of equipmont, E0 may be established by demonstrating similarity to equipment that was quallflod to the specifications of IEEE Std 323-1974.
for older plants whoru oquipment whose original EQ did not require preaging. the minimum acceptable approach:should:be to establish EQ through the license renewal period by a combination of reanalysis and informatinn obtainod from the roquired on-going 3rograms that review surveillance and maintenance records to assure tTat equipment which is exhibitInq age rolatud dogradation will be identified and replaced as necessary.
(Soc discussion in paragraph C. below.)
The reanalysis mus.t-he based on technically justified information concerning the aging of the material'. and the assumod service conditions.
Howevor, this is not the preforred approach.
The proforrod approach would be if EQ was nstablithod i
t o inc lude the license renewal porlod based on actual tost data in accordance with Regulatory Guido 1.89 and IEEE Std 323-1974, as discussed ahnve.
4 (q) f rom a pure engineering perspect ivo definitely not.
The heavy i
responsiblitty for proventing the torribio consequoncos of an uncontrolled
.u r i d e n t, both for innocent victims and for tho olectrical industry tIself. would strongly point to'the grandfathering of less rigorous practicos only if supplomontary design features, opkrating conditions. or orportonce f actors Indicate the risk of serious failure is insignificant (I'RM) or incrediblo, for passivo aquipment, normal experience as an indicator of adequacy is, however, almost useless when considoring equipment operability under LotA or HELB conditions.
(s) The differences in (0 requirements for older vs. newer plants aro
, justified by the woll-established grandfathering principlo, 10 justify the t rat of backfitt ing requiromonts to plants already constructed and operating it must be demonstrated that the original requirements and standards were inadequate in terms of safety, ihn introduction of the
l lll
/,
3 i
C-9 Guidelines states "The objective of the evaluations using these' guidelines.
is to identify Class lE equipment whose documentation does not provide-i reasonable assurance of environma tal qualification." The 00R-Guidelines dg require that aging effects oe accounted for in establishing the installed life of a component (note that for DOR Guideline equipment the-installed life is tantamount to the quallfled life). The document states _
"This positInn (of not having to demonstrate a spncific qualified life)
.[
does not, hownver, exclude equipment using materials that have been i
identifled as being susceptible to significant_ degradation due to thermal.
and radiation aging."
a-.
r, (t) Yes.
The regulations require upgrading 'of' equipment ' for older plants-unless there is sound reason to the contrary.: Essentially, there is no difference in the requirements except for' pre-aging prior to LOCA.
'[
Industry operating experience indicates that:the pre-aging is very conservative in that equipment is not aging as-fast as predicted by j
accelerated aging methodology, imposing this requirement on oldor plant is not justiflod.
1 (u) The two most significant differences between " grandfathered" and new i
requirements are the method of addressing preaging and margin considerations.
The exclusion of margin is justified based on the con-servatisms inherently contained in most environmental definitions and performance requirements.
The lack of preaging prior to accident testing difference is also justified since aging, per se, must still be addressed in the qualification evaluation (for grandfathered equipment greater reliance on analysis, coupled with some maiorial aging informatinn, is permitled).
Conceptually, both aging approaches are appropriate.
In practtee, adequacy depends on the quality, completeness, and conservatism of each analysis.
As noted above it is impossible to. adequately acceleraic all aging mechanisms.
Consequently, t)me form of aging management may be appropriate for the, success;of either aging approach.
What spect fic changes would you recommend? -
(a) Phased upgrade or rotesting [of the' old'er plant to meet the new plant-i requirements). Develop method (eg., EPRI indenter) for in-situ condition monitoring.
(b) Clearly a single standard must be developed.
The Commission was very clear in its direction to the staff to develop a single standard, and recognized that potential differences would exist, and the importance of l
providing a technical justification for accepting differences from the standard.
The older standard - no pre-aging prior to LOCA testing-is in my opinion, clearly not adequate to provide reasonable assurance that o!cctrical equipment will function after an extended service life.
The acceptability of the older stand ed has not been justified.
(e) All plants should be mado to meet the Catngory 1(323 71) guidelines.
(h) I believe that the differences in requirements for the replacement / upgrade uf components should be modified to require the older plants to upgrade to
{
?
C-10 I
i Category I when components are replaced, oither'at the end of life or for l
corrective maintenance.
(1) I would add a paragraph to require plants licensed -after November 30.
1985, to be in full compliance prior to licensee issuance, (The present rule does not address these plants,).
. c. w..
i j
i (j) More uniform guidance for interpretation of requirements might be useful j
especially for organitations with high turnover.
l i
(k) Would consider adding the requirements for surveillance and/or condition monitoring to both plant categories.
(n) Rather than " specific changes", I recommend some specific courses of actlon below which may or may not load to changes.
1 I
(p)
Conclusion:
lho bottom ilne of the foregoing discussion is that neither i
i Category I nor Category !! are ideal approaches to E0. However, especially t
i for equipment subject to harsh environments, the Category I appr;ach is superior: if equipment can pass an accident simulation af ter it has been i
j degraded by ago. conditioning, one has greater confidence in its capability l
than one could have if the accident simulation Were not p Mceded by l
age-conditioning.
l (s) I would not recommend any changes.tHowever,rwhatal.dsg]4 recommend is tbst
]
the NRC either accept (as it has during the.past tenryears)'the-qualified i
ststus of equipment in licensed older'oporating"plantt'or~1nitiate a reviow of qualification packages forithat equipment, in either case, the 00R Guidelines requirements should remain the criteria, I fool that the i
requirements are adequate, but it is incumbent upon the NRC to render j
sound engineering Judgmants on whether the basis for qualification j
documented by the licensee is adequato, (t ) Re-evaluation of sourte terms and the need to qualify equipment located outsido containment, and equipment subjected to.a radiation only harsh environment.
(u) I will limit this input to the following three areas:
methodology, 1
dnrumentation, and arcident scope, Currently, a single methodology, lj (preaging tombined with accident simulation testing) is viewed as the gn,Ly acceptable method of complying with 50.49.
This methodology (and all the document at ion and related licensee activities associate with it) is j
applled to All equipment with the scope of 50.49 regardless of its safety-J significance or the severity of its accident onvironmont.
Consequently, qualification of inside containment PORVs (very safety significant and l
exposed to severe LOCA conditions) and outsjde containment. position I
indication limit switches on a cooling water. containment ~ isolation valve (little safety significance and exposed to moderate LOCA radiation and I
possibly a moderate short-time steam exposure from a HELB) must comply i
j with the same [0 methodologies.
1 k
i a
4
~
.m.
y:
qe. ;,
i v
4 tv d$ff %.~
" -. c.
v 1
]
C Il i
I Secondly, individual utilities exi.end significant resources refining and updating EQ documentation. Much of the current documentation framework is derived from the NRC proforences during priors NRC~. audits, In france, EDF j
approved equipment manufacturers (coupled with EDF review and acceptance) i i
are responsible for maintaining the adequacy of qualification l
{
documentation and its applicability for supplied equipment.
Each french j
plant need not maintain E0 files. Absent standardization,-we are i
apparently unable to apply the french approach to current plants, i
Howevnr, some measures (not suggestod here) should be implementod to j
minimize much of the unnocessary paperwork burden associated with mair.taining [0 files.
Regarding the advanced reactors, I would strongly urge that standardized CQ programs, including single sets of normal i
operattun and accident conditions, be impiamented.
Those efforts should j
ho focused on minimizing individual utility CQ paperwork cost burdens.
l tastly, I have several olo p ations rogare rg accident' scope and qualificatlon.
Currently, W containmont steam, temperature, pressure, and submergence qualification is based on DBA (double-ended break) LOCA and MStB conditions.
It appropriately assumed that adequate performance i
for the DBA environmental conditions provides reasonable assurance of h
equipment performance for other LOCA and MSLB events with potentially (but slightly) different conditions and porformance needs. for outside containment pipe breaks, DBA break conditions: typically produce the highest pressures and temperatures but can bO of.sigriificantly shorter i
duration than smallor size breaks.
j l
However, two in-containment requirements DBA radiation doses per 110-14844 and the "one hour minimum" operating condition, suggest that the qualification provisions extend (rather informally) beyond DBAs intu the l
area of severe accidents.
The need for these two provisions, assuming l
qualification is limited to DBAs, has never boni. :laarly articulated, The l
rattonale for these provisions heromes somowhat clearer 11 one assumos, j
within the context of defense in-depth, they exists to provide some l
assurance that the equipment will function'for some severe accidents with i
delayed [CCS (one hour minimum operating time) and some significance core l
damage (110 14844).
linfortunately, since these provisions exist within i
the context of DBA qualification they can create qualification problems l
that are percolved as DBA arj not sevoro accident related, for examplo, most materials, particularl/ cable insulation / jackets, are significantly degraded by the in-containment T10-14844 doses. -Yet -0BA mitigated LOCAs i
exhibit only a fraction of the radiation assumed by T10-14844.
further, much 'f tho LOCA mitigating equipment, particularly the sense and o
i command features (e.g., transmitters) and associated equipment (e.g.,
cables, splices, penetrations) have little significance for most severe j
accidents.
The NRC should either delete these apparently beyond DBA provisions or clarify the basis for their nood, i
1 l
for older plants 10 CIR 50.49 should be implemented in conjunction with a l
(v) condition monit oring program.
The condition evaluation will eliminato the assumptions made by older plants in their [Q evaluation.
l j
i
C-12 2.
Does the existing qualification methodology provide sufficient basis to conclude that electrical equipment will be able to mitigate the effects of all postulated accidents over the entire range of qualified life?
I (a) [ Methodology) does not recognize that some things are more vulnerable at r
80t..
(b) The 323 1974 Std is a >retty goodistarti.gHoweverconly' testing:a single component such as a ca)le gE t and.not":' requiring periodic testing is not i
as meanlugful as may be required.
In the long run 323-1974 ght additional testing by the vendor or by licensees may be found to be appropriate.
I (d) Region V believes that the current qualification methodologies are adequate to conclude that the electrical equipment will be qualified.
Howevor, the Arrhenius equation may be too conservative (and underestimate useful life).
l (e) Adequate for the newer plants.
(h) If we assume that quallfled life means 40 years and that the Arrhent'us calculitions are acceptable, then the proper maintenance of equipment 15 required to ensure that age sensitive components are replaced periodically to crtend the life to 40 voars.
I am not certain that the present test data could demonstrate a,ualified life beyond 40 years for most of the t
equipment.
(') IfweacceptArrheniuspredictions,present;mo@odologyand.testingare-marginally adequate.
~
(J) Yes.
(1) The methodologies used to qualify equipment wnre based on the most current NRC information and [0 testing conducted in the industry.
This provided reasonable baseline data on the capabilities of f0 equipment assuming that required maintenance was performed.
Mechanical aging which should also include seismic aging is necessary to predict satisfactory performance over the range of the cualified life, HoweVer, operatinnal occurrences such as steam leaks and inadvertent containment spray actuations can cause degradation of equipment beyond that initially evaluated in [Q test
- results, lhe regulations do not provide for considering these occurrences in the test program.
Also using Arrehnius methods to extend the qualified 11fe beyond the actual tested profile does not appear to be the most prudent actton to take when a device is needed for long periods of time such as 30 days or 1 year post DBA.
The industry has over-used Arrehnius in this royard to extend test data to snvelope a longer quallfled life.
(n) I believe it does, in fact it may be excessive (see following_ comments).
In fact, it may be flawed (See A. 5.).
(q) Regulations are definitely not adequate for cable m l s t (cable, cunnecttons, seals).
l x
w.
C-13
"~
I (r) Comment 1:
llave less confident.c in equipment where qualified life is extrapolated to 40 years based on shurt aging time, i
l Comment 2:
Bases for determining a mandatory replacement interval for i
Class lE equipment in older plants is weak or" lacking. Older plants should have a requirement to upgrade.
Comment 3:
Environmental record keeping for equipment location, DBE testing of in-plant equipment.
(s) It is my judgment that ne existing qualification methodology ;s adequate, so long as du. attention is paid to areas of uncertainty that are well known thanks to the extensive attention that has been and is being paid to them over the last two decades. This attention has been mainly in the form of research tonducted by Sandla National Laboratories under NRC sponsorship and by several investigators under EPRI sponsorship.
These include such areas as limitations of aging models (Arrhenius for thermal aging and the equal-deso/ equal-damage model for radiation aging),
dif ferences betwoon simultaneous and sequential-imposition of aging and accident strest. ors, dose-rato effects, and differences between natural and artificial aging, Sandia has extensively researched all thess areas but the last, which is now underway in the University of connecticut inplant Natural vs. Artificial Aging Program sponsored by EPRI (" Natural Versus Artificial Aging of Nucinar Power Plant Components r EPRI Interim Report IR-100245, January 1992). All of these research efforts are described and referencedintheEPRIEquipmentQualificationReferepcoManual(Technical Report 100844, November 1992), especially Section 13. This section also I
gives an overview of the NRC Nuclear Plant Aging Research (NPAR) program studies that are pertinent to equipment aging and qualification, It is important to note that many of the lessons learned from all this research has had an effect on standards and requirements (e.g. known synergistir, dnst-rate, and sequential effects on aging simulation must be accounted for), but none of the research results has invalidated the basic qualifIcatton methods established by lEEE and endorsed by NRC requirement s. Any such invalidation could have been reflected in the 1983 version of ifEE 323, but the methods in that version are essentiallv those in the 1974 version, lhe bottom line is that, in the judgment of st andards writers, qualification engineers should be aware of the areas of uncer-tainties identified by all this research and should account for them in specifying qualification tests or in performing qualification evaluations, but none of the research findings invalidate the approaches to l
qualificatton referenced by standards and regulations.
'The NRC staff involved in evaluating the results of this survey could benefit from the in-depth treatment of EQ uncertainties in this section of the E0 Reference Manual. In the spirit of EQ research cooperation that has existed between TPRI and the NRC since the mid-1970's. EPRI in keeping with its long established pelicy, sont copies of the manual to more than 20 members of Iho NRC staf f at nn cost. Unfortunately, I have been informed that these manuals'are being returned to EPRI because an NRC lawyer views these important technical report s as inappropriate personal gif ts to NRC staf f members!
C-14 The following quotes from a section titled " Evaluating the Qualification information" on pagos 7-14 and 7-15 of thehreference manual are just small examples of the guidance to utilities regarding the establishment of a quallfled life for equipment in newer and older plants: "The equipment's installed life may he limited not by thermal aging effects but by other aging mechanisms (e.g., operational cycles).
All significant operational and environmental aging mechanisms should be addressed and the life-limiting mechanism determined.
Every effort should be made, either j
quantitatively, through the use of accelerated aging data, or qualitatively relying on oxperience, inspections, and maintenanco, to define a qualified life."...."Early environmental tests were often i
l conducted without preaging of the specimens.cWhen:.using.these test data, the plant-specific ovaluation must attempt to determine the 3ermissible levol of inservice deterioration that would not' invalidate tie lost conclusions."...." finally, inservice inspections,' tests, and maintenance may be used to ensure the equipment remains.in a condition' bounded by j
results of the aging evaluation."
(t) 1he existing qualification methodology provides an excellent basis to 1
conclude that electrical equipment Will be able to mitigate the effects of all postulated accidents over the entire range of qualified life.
Additional to the testing, ongoing evaluation,gtrending and monitoring of u
equipment for the installed locations provide reinforcement to that basis.
Part ?l's, if Not ices and industry groups also provide valuable information for evaluating testing assumptions and results.
tu) Ihis question addresses equipment adequacy for all " postulated accidents".
l 1 assumed thl; is intended to mean design basis accidents (DBAs).
Sne the prior response regarding E0 requirements that appear to be related to beyond DBA conditions.
In my opinion the simpic answer to this questic,n is yes.
the fundamental nurpose of [Q is to provide a basis for concluding that equipment is designed, installed..and maintained such that limiting harsh accident conditions in combination with operational. aging will not result in equipment common-mode failures.
Qualification of representative equipment, including tho limiting case of single tost samples, coupled with proper manufacturing, installation, maintenance, and operation provides reasonable assurance that environmental or aging induced common-mode failures will not occur.
This adequacy is based in i
large part on the conservatisms associated with the harsh environmental conditions used for E0 when compared to the conditions potentially of.rurring during the most probable accidents.
Virtually all [Q equipment potentially exposed to pipe-break steam conditions are qualified by simulation tests.
Since the assumod environmental condit ions reflect very conservative analyses of mass, energy, and heat sink characteristics for guillotine double-ended breaks and severe core degradation for radiation conditions, they are conservative representations of the conditions possible during the most likely accidents.
Further, much of the equipment is tested to generic profiles that rnntain additional conservatism when compared to plant-
.per i f ic condi t ions.
If reprtuentative equipment can survive these steam test conditions and subsequently installed equipmont is properly manu-
l C-15 factured, installod, and maintained, then the installed equipment should j
be capable of similar functionality under-plant accidents.
3 4
l further conservatism is dortvod from the assumed accident radiation
]
conditions which do GAL represent those resulting from a DBA mitigated LOCA and are, in fact, significantly more severe; 'for'many materials these extremely conservative radiation dose-assumptions' produce dc0radation that far exceeds the degradation experienced by prolonged
{
exposure to normal operating temperatures. ~ Recognizing that the conservative accident radiation exposure can account somewhat for the effects of thermal aging, even the test performance of some " unaged" i
equipment is reflective of a signi'icant degree of material degradation.
When thermal aging is addressed through either preaging or thermal aging analysis, additional assurance of equipment performance in the aged state
]
is obtained.
For " grandfathered" equipment, I believe aging analysis when 4
1 properly performed with an adoquato level of analytical conservatism can i
arldress the significance of operational, thormal, and radiation aging j
l mnchanisms.
f (v) The standards are sufficient, however, what is most important is that the l
actual test itself duplicate the components plant configuration and the environment that the equipment will be subjected to.
Certain techniques are overly conservative, such as the requiromont to uso a mandrel for
)
cable losting (this does not reflect in plant conditions), while tho standards lack guidance on instrument accuracy with respect to harsh environments.
Most tests did not simulate worst case' voltage 'io.,m:.ndit' ion's that can exist mu during all postulated accidents.
The worst case voltage will influence the operability of the safety-related equipment.
This applies to both i
j older and newer plants.
I Are the current standards, procedures, and techniques used to conduct I
component type-testing (by both research and development labs and qualification testing labs) satisfactory for establishing the bases for environmental qualification?
l (a) Synergistic effects are not well documented.
l l
(c) (urrent standards, procedures and techniques are at. (323-71),
i j
(g) Irstrument loop accuracy was often poorly addressed as it related to j
functional roquirements for terminal blocks and seals.
(j) Y rn.
Reasonable application of current guidance is adequate for performing qualification.
(n) My experience leads to the belief that the requirements to perform all tests on a single sample (see f.! of 50.49) to try to stimulate all of the 3
conditions expected to be experienced by the equipment during its lifetime may not he necessary.
Test results seem to indicate that failure to perform to specifications was always (1 cannot remember a single i
1
,g..w3.p.g%,
u w a
,p
?
~
c.16 exception) due to a design weakness that showed up when that weakness was
- tested, By this,1 mean the equipmentiftiled when exposed to' radiation because there were components or materials,with unacceptable sensitivity to radiation, or pressure, or power source variation, or seismic / dynamic vibration / acceleration, or moisture, or temperature.
I cannot remember any that were due to age (except radiation degradation which is a special case of material and componont selection) or combination testing.
Operational testing to worst case extremes exposed all of the design flaws I can remember.
(q) EQ requirements should bo hn stringent for some cable systems and mafJt stringent for others, depending upon safety significance.re PRA type analysis or time of and duration of functional need, (t) Yes, if anything the existing qualification methodology is too conservative.
For example, standards require that a place of equipment be exposed to the full LOCA radiation dose prior to LOCA testing even though-the equipment may only be required to operate for 5 minutes into the accidert and would not see an increase in. radiation during the time it must perform its safety function.
Additionally, emphasis should be placed on appropriate application and function of equipment within the design of:
that equipment instead of relying on deterministic methodologies that are i
at best uncertain.
(u) in general, current standards, procedures, and' qualification techniques are adequate for addressing harsh environment conditions and the offects of those aging mechanisms that can be reasonably accelerated.
There has historically hoen excessive reliance by the NRC. IEEE, and the industry on accslerated aging to precisely define a " qualified life".
There has been a growing industry recognition that these thermally-based life calculations are not precise and that other aging effects which cannot be accelerated must be addressed by maintenance and inspections, I note that tho (Q requirements from several nations (e.g., France and Japan) with advanced nucicar programs do not include provisions for defining equipment qualified life values.
Recognizing the uncertainty in such life definitions, these countries have focused on maintenance and inspections as thn aging management methods and have included some form of aging' simulation-(without a qualified life definition) in their qualification testing programs.
In large measure, I agree with their. approach.
What specific comments and recommendations would you make relative to the qualification methodology for:
a) older plants? b) newer plants?
(b) Whethe'r or not adequate periodic testing methods can be developed to determine the state of qualification" is a matter of conjecture.
I believe that any such tests are years in the future, Specific comments / recommendations may be made but unless each is justified technically, they are more wishful thinking than anything else.
As we move forward to resolve the various EQ questions it is important not to-repeat the mistakes made in the past.
The recommendations and positions should have a solid technical base and any exceptions clearly identified.
~
i s
C-17 Perhaps the biggest contribution to resolution of this issue will be the NRC's resolvo to clearly state the issues and then answer them.
(e) Older plants should be brought up to the newer standards.
This is very important for operations beyond 40 years..
(h) I believe that the Industry.and the NRC should work together to develop a list of qualified components 1that both' groups can agree to.
This would make compliance easier for the'11censees..I think that the qualification testing should be approved and accepted.by the NRC prior to a component being placed on the ilst of quallfled components.
(J) I do believe that there is needless fear of synergistic effects which do
{
not seem to bo important for the bulk of the power plants. Most power plants have low aging doses such that the threshold for synergism is not reached.
(k) The existing qualification methodology /it' generally' adequate, however results should be indexed against actual in plant samples exposed to normal aging conditions and appropriate adjustments made.
This would be most appropriate when quallfled life extensions are considered.
(n) If my experience is representative. I would recommend that the wealth of data resulting from the LQ Program be analyzed and, if justified by the data, ma, lor steps he taken to bring the requirements more in line with experience.
Perhaps EPRI could be given the task to collect (they have most of tha data now) and evaluate EQ failure data.
They should look at older and newer plants'-[Q results, especially where the same equipment was qualified both ways.
They should also evaluate maintenance and failuro (replacement) data for older and newer plants.
If the increased requirements for E0 for newer plants resulted in better equipment being
- nstalled (it should have if the newer requirements are meaningful) there should be fewer failures.
You have no idea how much it troubles me to say this, because I was one of the developers of the concept of combination testing, aging, and the rest.
It seemed the right thing to do at the time because no one was able to prove it wasn't needed and the arguments pro-and~ con were' unresolved.
Since the work is essentially finished and the money spent, you may ask why we should hother.
Talking to several old colleagues who are still in the business, I find that the added cost of our present [Q program requirement 5 coupled with the small market presented by the waning nucicar i
power option dampens the developmental ardor of both the suppliers and the plant operators.
Evan though operating experience is excellent in the U.S.. I believe the performance,' safety, and reliability could be j
significantly improved with the development of more modern control and safety system equipment (distributed Intelligence to name just one).
All avenues including reduced E0 costs should be examined and steps taken where possible to encourage development activity.
(q) I need 50.49 text to adequately respond.
Sorry.
h 9
4 6g C-18
.O 1
(t) fronding and condition monitoring for both older and newer plants is established or is being established in most' Utilities.
Additional evaluation of equipmeM will be required under the maintenance rule.
Ther4 fore, there is no eeed for additional ' recommendations relative to' l
the [Q methodology.
(u) I believe current practice for the qualificat'lon of most outside containment nquipment exposed to short-time steam conditions is unnecessary, for much of this equipment, the harsh conditions.-low tem-perature wet steam (e.g., 100% humidity at 175'f and extremely low pressures) for fractions of an hour, are not sufficiently severe to threaten operability of properly designed equipment.
This coupled With the lower core-damage threat suggests that the use of equipment designed for these higher temperatures coupled with thermal proof-tests, protective enclosures, and maintenance based aging management is sufficient to establish operability, i t should be noted that for EQ equi) ment, including equipment qualified for outside containment pipe-)reaks, the single failure criterion must be preserved.
Yet, regulations related to other plant events. such as fire, require only one train to be free of damage.
i Secondly, it appears the current TID 14844 radiation dose requirements coupled with the common practice of'tubjecting?the equipment to a i
radiation simulation prior to the LOCA' steam exposure is excessively conservative.
for many materials and electronic equipment this radiation and not thermal aging or steam testing is the most significant stressor, for many insulating materials (e.g. XPLC) this radiation and not thermal aging results in significant loss of elongation and brittleness, further.
4 sub.jecting equipment to radiation degradation prior to the peak accident steam conditions is mechanistically incorrect and overly severe.
For most mitigated DBAs these radiation levels are unrealistic.
Even for events-with delay ECC5 actuattun, significant releases would not occur until after the peak temperature conditions.
Lastly, current practice results in utilities establishing post-accident long-term operability for durations of 30 days to beyond 1 year.
The often significant efforts associated with establishing long term operability appear unwarranted.
I believe operability only for accident mit igat ton is necessary.
Several days post-reactor trip, decay heat i
levels are sufficiently low that natural convective containment cooling j
may be sufficient to maintain temperatures and pressurns within acceptable j
- limits, Under these conditions no subsequent in-containment equipment operab.ility may be necessary to provide adequate plant safety.
3.
Is the burden of qualification appropriate to the importance of the equipment being qualified?
(b) This is an interesting question.
EQ discussions frequently focus on this que'.tton and the result is a misconception that perhaps the scope of the 10 rule is ton broad or that qualification should not be required.
The i
scenarins in which EQ may be important arp not limited to a major pipe rupture.
A steam environment in a small area due to a minor gaskot leak t
e-
.. m a k,
v
- p-M b qi r
~
k z& ' '
+
C-19" or valve packing leak could provide enough; humidity to cause electrical equipment in the area to begin to short-out, and to cause instruments and/or their cables in have higher current to ground values.
The'results (nuld he inoperable equipment and inoperable or inaccurate instruments.
l Innperable equipment in and of itself is significant.
The issue.ls a potential less of redundancy and diversity compounded by additional
- failures, i
in addition to inoperable equipment thinkcwhat~this could mean at 4:00 am during a normal plant evolution or transient; -Which instruments and annunciators will the operators rely uponf At THI-2, the transient turned into an accident when operators did not understand that a single non safety-related valve indication was not*a position indication but a demand signal.
If they had shut a single non safety-related valve the transient would have been stopped.
Since March 28,n1979 we have learned a lot about operators and control room instrumentation.
One of the things that we see in daily reports is that operators, when faced with different indications / alarms can and do make mistakes, The burden of qualification is not too great; The safety significance of I
potential common mode failures of multiple systems and instruments is j
great.
(d) Our inspectnrs have an impression that the documentation to prove i
qualificalion seems excessive.
(c) The burden is justified for SSE (f) Yes, definitely.
(h) I believe so.
(k) The E0 burden is appropriate and justified in terms of s a ty significance.
(j) Many of the outsido containment components'in rad only areas seem to have excessive expenditures to cover the qualification given that many of these components will not be exposed to'anyiraalichallenge.
In-containment and HELB area qualifications are 1ecestary.
(1) Yes, I bellove the burden of qualification is appro)riate for the importance and safety significance of the equipment aeing qualified.
(n) This is an excellent question and one I have thought about, and argued about for a long long time (actually, ever since Jacobs, Gallagher, etc.
first proposed the concept).
I presently hold the position that all plant equipment should be designed, manufactured, installed, and operated in accordance with the determined requirements regardless of "importance to i
safety" because the continued operation of the plant is as important to the peace and well-being of the public as is the safety nf the plant.
Obviously for legal and regulatory reasons, safety is the major concern, but if all equipment were treated the same, in my opininn, many beneflis I
1 C-20 would accrue.
Furthermore, there has not yet been:a fool-proof.
methodology developed for quantification of safety. importance that I can-1 support.
We tried an experiment 3on?thetlEEE<Hucleart PowerJEngineering.
Committoo (NPEC) a few years ago,..We7gave.bgroup of industry exports (an HPEC Subcommittae) the guidelines in effect'at the~ time-and had them evaluate a list of equipment from various systems. We had each of the experts rate all of the equipment on the list and compared the results for consistency.
The results were not' completely random but so scattered as to make it quite evident that the method was' severely flawed.
Thn mothod was later adopted in Europe but I don't'see how it could possibly be of any practical use to them. 'Perhaps the NRC should check with their-European colleagues on this.
Especially to find out if it has been of any measurable benefit.
(r) Comment 1:
Yes.
The cost of qualification, at present, is generally minimal.
Programs are generally in a maintenance mode. Significant r
investment is only realized in mandatory replacement of equipment.
Cnmment 2: Yes.
The safety significance does justify the cost.
Conwnen t 3 : pRA methods and tools, at present, lack sufficient refinement and rigor to contribute significantly in the formulation of specific EQ requirements.
i However, it should be used to determine wh'ich equipment should receive higher levels of scrutiny would be in order, particularly in the case of older plants.
The risk-based significance of equipment located in' mild-environments should be as a. basis for determining an expansion of tho' existing 10CFR 50.49 scope.
(t) for the equipment truly important to safety, the burden of qualification is appropriate, But as implied above. the scope of equipment in the program as required by regulation'is'too big and therefore burdensome.
Does the safety significance of the equipment justify the EQ requirements that are being imposed?
(e) The safety significance of the equipment justifies the requirements.
(n) I think I covered this in the above response.
We should definitely not try to classify " levels of risk" or "importance to safety" unless a bottor method is developed and only then if there is very great potential for improvements in safety and performance at reasonable cost.
(q) 1-believe that, yes, the importance of the equipment should dict ~ ate different levels of design criteria and thus different qualification 1
- criteria, (q) In my mind the NRC was correct in placing clearly different burdens on i
equipment in harsh vs normal or mild environments (better than the IEEE).
-However, now that tools are better (PRA and other analytical experience) to differentiate in the functions of and safety significances of equipment, that too should be reflected in different EQ burdens.
l i
E
~
y,;g
- ,w
.m.
p y v..C-ty s..-
What role, if any, should risk significance play regarding E0 requirements? What changes would you' recommend?-
(a) PRA should only be used to prioritize E0 actions.
(p) Risk significance should play a minimal, if any, role In the requirements.
Risk sign depends on componer,t failure rates during DBE and this data is simply not available.
Recommend that all plants be brought to the_ current standards.
(f) PRA should not be a factor in determining EQ requirements.
Requirements should be hasod on reality (facts, occurrentes) rather than PRA charts.
Rocommendations: the licensees should be required to conduct in-service testing to prove continued qualification.
(g) Somo sort of graduated requirement system would be a good idea, it should be based on the safety importance or risk-of the equipment - not on what distinction can be easily implemented C and;l believe that would be difficult to cccomplish... g
, n.,.:r g m W. 7 (h) I would not recommend that risk significance be brought into the issue, it would nnly make things moro difficult than they already are.
(i) I would make no changes based'en someone's. determination of importance.
The additional complication would be an unnecessary burden.
(k) Risk significance should not play a major role in the assessment of EQ requirements sinco common mode failure is the major concern and the probability of initiating events (LOCA/MSLB) are not insignificant..
(j) It is not clear that changing levels of qualification to agree with risk significance would save much funds.
The analysis to justify.the levnis vuld bn expon'.tvo.
Since similar corr.ponnnts are used in multiple applications, the most severe application would govern the qualification requirements anyhow.
(j) I would allow the option to grade qualification, but I'm not sure that I would use the option.
(1) Risk significance was considered'w'han the~ master list ~of E0 equipnent was developed.
The components on the list are those required to mitigate the design basis accidents.
I.would not recommend any changes in regards to risk significance.
(n) How would you graduate qualification? Smaller margins? Half the SSE?
Two thirds of 4.1.0CA? Nonsensel Test to the performance requirements.
1 Everything' in the plant should be designed and proven to meet specifications.
We need nuclear power and we need positive public opinion.
Be firm in your requirements and )ublicize them'- you (NRC) do not do enough of this, by the way.
None [c.innges), in this aron.
.}
(+
, y;3 2 ^
+ JuhM OMhAdNb~v--E \\ i-
'4 M +
M w.g l
C-tt
{p) I believe risk analyses ~should play a secondary role in deciding. safety-issues and-influencing E0. It is my impression that the accuracy of the:
l data used does not match the sophistication of_the mathematical mothods<
i in partIcular, analyses which have attempted.to take aging into account do-not appear.very convincing. Efforts-to use--plant experience to estimate the offect of aging on equipment failure.ratetshave not been successful.
In Iarge part the reason is.that maintenance and refurbishing tend. to diminish the offect of aging: instead of regretting.that this prevents us from obtaining the type of failure: data. desired'for the risk analyses. We should be happy that maintenance Is doing its job, Horeover,att is not failure in normal service that should interest ust obviously, if equipment has a history of f ailure in normal service, it'mtght not be able tof perform its safety function under conditions of an appilcable accident -
l even if it were in satisfactory condition prior to the accident. More to-the point than failure rates ~1n normalsservice>should benths failure rate
~
under accident conditions, but1such data are clearly not available.
I would tend to place more confidence on the consensus of expert opinion on the relative importance of safety-related equipment than on the outcome of risk analysos.
(q) It should play a maior role in formulating EQ requirements such as:
time windows of operability, or accuracy circuit design requirements for continuous indication of circuit condition, and testability in normal j
j conditions of the potential circuit integrity under wet. conditions.
I The question invites a long treatise. One immediate thought onlyi PRAs now seem to deal with reliability numbers for' active components only while completely ignoring the cable systems upon which almost all other electrical equipment depends.
This is understandable and perhaps justif uble for equipment in mild environments where cable. system i
reliabliity is excellent, but'seems unreasonable and inexcusable when considering either young or aged cable -connectors, and seals when under harsh accident conditions.
PRAs must be realistic and honest if they are.
to be used and they should be used to focus.onsthe vital circuits in order to spend limited funds to harden-up only the most critical circuits by better designs and E0 practices.
(s) Both the Sandia [Q Scoping Study (NUREG/CR-5313. " Equipment Qualification
([Q) Risk Scoping Study " Sandla National Laboratories and SAIC, January 1989) and an [pRI report on the rlsk significance of equipment ("Use of Safety.importance Rankings in Equipment Qualification: A Study of Big Rock point." [PRI final Report NSAC-036, January 1984) 1how as expected that a risk based approach to safety classification would lead to a substantially different list of Class l[ equipment than given by traditional deterministic a.ethods.
I recommend th.it 'he NRC develop' criteria for the safety classification of
.cquipment based.on risk /pRA and allow utvlltlis to use the approach and criteria.f or reclassifying their. equipmeM. l fee! the approach should bc l voluntary (based on economic considerati&ns-by g ilities) and not.
compulsory, because it is highly ~unlikely thatscht new approach would find-1 O
x -
.-,~n-.s-u- -. -
-a..n-
-a
~---+ ' --'
-- ~ - - * "
~
I 1
i C-23 i
that any risk-significant equipmenf. Was not Al're dy~ classified as lE. The i
first several plant applications could be used to confirm this, (t) PRA's should be considered in identifying those function which truly contribute tc an increase of core meltdown frequency and enable the utilities to dedicate resources in more significant safety areas.
b j
Consider PRA's, new source terms, and the safety significance of outside enntainment liflB's in identifying those safety function which should be i
i within the scope of the [Q program, (u) To date PRA has played an insignificant role in EQ.
In order for EQ to be i
more cost and safety-effective, PRA insights should he integrated into the l
IQ st andarth.
Since PRA is nno of the few tools we have to address
{
relative safety signi(Icance, we should use it to identify the safety significant C0 equipment and issues.
Reg:rding PRAs and the TAP, I am 4
toncerned that the current IAP efforts are focused on simply modeling EQ j
equipment and accident failure rates, Although helpful, these' difficult activities will not provide short-term assistance prioritizing EQ issues.
i without performing any additional analyses, PRA practitioners can provide t
insights regarding the need for long-term post-accident: operability, the relative safety significance of certain systemscand. functions, the j
relative importance of different types of in-containment primary l
equioment, the apparent frequency of requiring (0, and the importanco of n, differing types of initiating events 4
on-[Q equipment'In' accident l
mit1 41lon.
9 1
l (v) Probability Risk Assessment based on Core Meltdown frequency may play a significant role in establishing the acceptance criteria that a component must ment for use in [0 appilcations.
Current methodology may ter,t a j
component whose function in accident mitigationLis minimal to the 7.afe I
triteria as a component that has a significant role in accident l
mitigation 4
What are the strengths and weaknesses of the existing EQ requirements as l
they pertain to:
a) older plants? b) newer plants?
I (a) [Weaknou Neeri to have a) better definition of " similarity" and how tu vorify it and the performance requirements.
(e.g.,
"like for-like" definttton in Cl-91-05) l 4
1 (b) One of the weak areas in the current EQ rule is the concept of " harsh" environments.
The basis for this conclusion is that normal service l
conditions of high temperature, for example greater than 150 degrees F, may not be classified as harsh sinco no DBA h involved, But day in and day out electrical components may see these high temperatures and not be in an EQ program.
The issun is really one of design and selection of equipment.
What we have found over the years is that temperatures are signiftrantly higher than nriginally expected during the design process.
Without pre-aging, and recognizing the recent failures ~during testing of l
cables, the questinn which needs to be answered is:
What i
(
i
- m
~
~. - '
~ -
m ye.&gu Lh..
- z. - u ~
~ ~.
m 4
C-24 test / qualification process is technically sufficient to provide an adequate or reasonable level of confidence that electrical equipment will perform its intendnd function in the 40th year of plant operation? The
)
answer to this que tion will dominate the answers to the majority of the 1.0 questlons.
The requirements aro woak in that a single test success is valid forever.
1his aspect of the qualification process should be i
changed.
(e) Older plant are not required to considor aging, margin, or synergistic effects of radiation and temperature.
Weakness.in older plants include relaxation in testing and documentation requirements.
Strengths in newer plants are more stringent requirements that simulate actual plant conditions.
(f) Weakness of both requirements is: that the qualification comes at the 4
beginning.
Over time, components are abused in situ.
Thest in servicq problems are not accounted for in the service" life.'[Xample: Cabics are often trampled or stepped on during outages possibly damaging them, shortening their usnabic lifa.
However, no in service testing is done to prove the appropriate IR.
(h) A major weakness with the EQ rule is that licensees requesting construction permits / operating licenses after November 30, 1985, are not covered by the rule.
(i) I find no strengths in this ivgulation.
Newer plants are excluded.
(j) The strengths and the weaknesses are within the individual applications of the requirements not necessarily the requirements.
Some DDR qualif b at ions are very strong.
Some NUREG 0588 Category I qualifttations are weak.
It depends on the nature of the qualification test and the subsequent activities to assure that it is adequate.
The documentation can be strong with a weak installation means the device is not adequate.
Perhaps where the standards become weak is that they are not strong on as-installed meeting the as qualified condition.
}
(A) for newer plant equipment the LOCA/MSLB testing yas. required to be performed on aged (accelerated) specimens resulting in somewhat more realistic prediction of equipment behavior as opposed to the older plant equipment which cou H be qualified using separate effects / analytical consideration of aging.
(1) Obviously the DDR guidelines did not require pre-aging or testing for
)
spray effects.
It allowed the licensee to perform an analysis for these i
effects.
Unless the licensee replaces DDR items and does not document i
reasons to the contrary DDR qualified items could remain in the plant forever.
I believe this is a problem and should be addressed in any future regulations for oldor plants.
(n) My major ronrern (nr all plants regarding IQ. both older and newer, is whether or not the operating conditions assumed during the qualification (many of which were generic in order to cover a multitude of plants using I
l C-25 tho same equipmnnt) accurately reflect-the:. actual / plant (operating conditions during the intervening years.and.at'present, is new or replacement equipment being qualif ted to' hypothesized conditions that are too stringent or not stringent enough?l Do we have data on present plant operating conditions to justify the qualification envelopes? One area I was never comfortable with is'the actual? temperature rise in the centar of the worst bundle of safety-related cables in the Cable Spreading Room.
How many are and have ben at temperatures far in excess of those used in l
qcalification? What about the hot spot temperatures in the densely packed control room panels. Admittedly, I am less concerned about equipment.in mild (accessible) locaticas but still concerned about unexacted multiple failures during a potentially. common cause event (earthquace, HVAC failure).
Maybe you have covered this area with regulatory action sir.cc l've been busy with other things, if'so, fine, My concern may seem frivolous in view of my stated experience regarding aging, but the concern is more directed to possible unknown cases in which the operating environment may be (and have been) in excess of the design capabilltles (i.e.. material state changes) as opposed to operation within acceptable limits over long periods of time.
(p) A major deficiency of the EQ process is that > lt does not' ace.ount for many 4
of the weak links in safety systems. In the case of cables. for example.
[0 does not account adequately for installation' damage deformation of t
Jacket and insulation at high stress points the ef fects of high humidity t
and high temporature (i.e., in excess.of: levels. assumed as-the service conditions), and local vibration, While such conditions are evaluated when uncovered in a plant, the analyses usually performed do not provide firm l
evidence that the weak links would survive accident conditions, (r) Comment 1:
Most significant.to all plants, no attention has been paid to mild-environment equipmont. Many newer plants. Initiated mild-environment E0 programs, but have since deleted them.in the absence of any NRC ef fort
]
for follow through in this area, Comment 2:
Problems aro with implomenting requirements.
Test sequences cause conservatism, but is the only practical method.
rnnment 3:
for older plants, weakness lies in specific seismic qualification tests and too 100s0 an intnrpretation of similarity.
Comment 4:
Strengths--more inspectlons to observe degradation on older
+
plants, qualification demonstrateo by test on newer plant 3.
Weaknesses--
non-conservative testing and more analysis used on older plants, reliance on qualifted life and less inspections on newer plants, (s) The strengths for both older and newer plants are that they are based on (1) the collertIve technical know-how and judgment of utilities (including EPRI), standards committoes (including NRC participants). national laboratory staffs, and the NRC staff who issued the requirements and (2) twenty years of qualification testing experience, about fifteen years of research testing; at Sandia, hundreds of years of plant operating experience, and an accident at THl.
,g je g
,(
- A., Q f.,.4 s. ;,,14 m.v.mL ' A; C-26 The weaknesses are rooted in the same problerrs that partain to any requirement: they are weakened by less-than-perfect clarity of wording, uneven interpretation and completeness in Imp?ementation, and uneven enforcement. None of these problems justify significant. efforts to revise requirements.
(t) As stated above, the requirements,for older vs. newer plants are essentially the same.
The only exceptionsis the requirement for pre-aging prior to LOCA testing in older. DDR guideline, plants, it should be noted that industry experience indicatet the.present methodology for pre-aging t
is extremoly conservative. in that equipment'is:not aging at the rates predicted by accelerated aging.
Otherwise-the requirements provide the basis for insuring equipment will perforra its safety function under harsh conditions.
(u) These are generally discussed elsewhere'in the survey response.
The most important strength is the significant environmar.tal and operating time consarvatism inherent in the qualification process.
Two possible weaknesses may be excessive reliance on analytical aging calculations and adequate consideration of the enmbined effects of manuf acturing/ material changes when old qualification testing is-applied to newly manufactured equipment.
(v) With respect to both older and newer plants the NRC shoulo clearly state the acceptance criteria for qualifying a component based oc operating '
crperience (while the industry standards allow the use of ope,iting crperience. It is not clear where the NRC stands.and what is t.id is not acceptable).
5.
Are you aware of any specific problems or' difficulties that :urrently exist or that existed in the past with implementation.of EQ program requirements? Please distinguish between older and newer plants.
(a) Many Just read some of my inspection reports. -for example: Use of generic *se called" worst case DBA'parameterst but old plants dnn't know where equipment is.
A major effort'it needed to walk down and document for older plants just what equipment they have where and what are the actual ambient / service conditions. (Long term).
(b) lhe primary problem was the NRC staff's changing position on IQ and what was required.
The result was confusion in the staff as well as in the industry.
To a large extent it silll exists today and is evidenced by a i
number of the nuestions in this survey. i.e. What is required? Why is it required?.
(e) The biggest problem te the 3 levels of [Q requirements, it existed in the past exist s now and w'!I ront inue to crists unless the NRC fires it I
don't believe the older plants are as safe as tha newer plants when it comes to f0 protection against DB[s.
(f) Many documented [Q stories Butyl cable. Incompatible lubricants in timitorque valves, others. All those in Region til have been addrev.ed.
fn
u S.~
ys cW. -
s C-27 (h) The largest issue has been the differences'in interpreting the test data.
By developing a standard list of components. this could be eliminated.
Another problem has been licensees failing to adequately evaluate the ambient temperatures that equipment is subjected to.
Many want to rely on average bulk temperatures, rather than localized temperatures.
Another issue is what maintenance is requited to maintain the equipment in a qualified configuration and how of ten should it be done.
(i) Interpretation disagreements between.llcensees and NRC.
(j) Resolution of such problems is my work.
There is no one specific set of problems.
(k) One of the more difficult areas in establishing qualification bases or in reviewing the adequacy.of.. qualification has.been in determining whether' adequate similarity exists between the item quallfled and the item installed in the plant (differences in material formulat lon.
configuration, arrangement).
l (1) I am not aware of any specific problems or difficulties that current 1.v exist with implementation of E0 program requirements.
i l
(n) I am only aware of those problems mentioned in the preceding paragrapns.
In additton. I share with nthers some concern that there may be excessive ;
optimism that accelerated aging truly represents real time aging.
I heard 4 some programs being discussed some time back to d9 some additional in.est ina t inns,
Hope they are being implemented. My inmfort :one would be graatly improved if we relied less on accelerated agtny and more on
{
,urveillance and degradation monitoring in real time.
l
,Q) the writer's esperience has been that for earlier plants the purchaser's
'pers for cable contained little in the way of performance requirements i
aimed at adequate performance of Cable iYstems under harsh Conditions.
Ref erence was made to standard spect plus limited radiat ion ggias resistance and ability to operate through a high temperature cycle.
Not until the middle to~ late '70s did more focused requirements come in cabic I
and rcnnector manufacturers.
[ven at present. inadequate performante requirements ate used -- as indicated by otner survey respunses of 1he writer.
A sarnnd serious weakness f requent ly observed a the at t est ant a by u cr5 and the NRC of [Q tests of commercial or generically named component s as
)
eovering various manufacturers or vintages or construction designs of i
(ables and interfacing components. It is well known and has been often demonstrated that great variation exists in the aging effects and harsh environment performance of components subject in these variatinns in mild environment s most of the'* variat ions are inconsequenttal. but in l
potentially harsh environments they may be trittially important
{
I
5 M
a
-C-28 Perhaps an underlying problem with our EQ practices has been our lumping.
4 together active and passive equipment in our EQ. thinking for. performance requirements, design requirements -(single falltire Criterion ' for example), '
~
maintenance programs - (testability), and importance-to-safety priority differentiation.
Unless we focus sharply on the functionally most critical cable systems and spend our limited staff and dollars on these. I believe it is very unlikely any significant improvement will be made in public safety through accident mitigation related to cable systems.
The l
9.yerall class 1[ cable plant in a nuclear station is just too massive and rnstly to upgrade.
the NRC has done better than the balance of the industry. I believe. In focusing the E0 program on harsh environment areas. As a IIG participant on IEEE 323 and 383 I have repeatedly advocated that we should focus on comon cause f ailure prevention during and af ter DBEs and that 111111 EQ emphasis should be put on normal performanc.e where good industrial practice and utility self interest in keeping failures low should suffice.
Even today, a cursory survey of E0-ralated technical pspers shows (with a few notable exception.) a preponderant emp' asis on normal service f ailure n
rates, wear out, drift, replacement parts, shelf life, leakage rates, j
etc., and a relatively cursory. 'oh yes..and be ture :it will be functional if it's in an accident or other DBE environment P Common ~cause' failures are given little. If any, attention-in many.(mostt) IQ papers..'
' :. 7,
,<3 The only way the writer can see id break todays' impasse for upgrading public safety by improving cable systems in either old or new stations is to start with only the few most critically important systems as revealed by PRA or other acceptable methodologies.
An objective and careful removal. analysis, and test of those few romponents removed in any such program should quit.kly reveal the justificat ion or lack thereof for continuance of or extansion of the program to lower priority systems.
(r) Coment 1:
Losts of testing.
This is a barrier to the introduction or adapt at ion of new product s to the nuclear industry.
This is a parti ular barrier in the l&C world, the rapid advance of technology has created a dilema f or maint ainahtlity nf cert ain qualified systems.
For older i
plants, this coup en with uncertainty in the licensing prorou of st at e of the art syst ems, prevent s moderni24t ion which may ult imately drive plant s of older vinta.,e into early retirement
( nmen t ?
NRL iesit anr y def ine t he herenary rigor f or Analy*,ip ha sed
- methods, in general, muc h of the Industry assumes that qualification based solely on analyses are taboo, even where sufficient rigor-can adequately be achieved.
C omen t 3:
Ior older plants. the use of " engineering judgement
- Anal the
" spirit of SQL.'G" i s being abused, Coment 4:
Adrquate simulat ion of ef fec t s of containment spray, abu'.tng ac celerated aging, lack of consistenry in qualification programs (primarily on older plants), sequential testing 100 conservative.
i I
i f
i C-29 (s) 1 know of no specific problems or' difficulties'thitt s.urrently exist or Ihat enisted in Ihe past with implementIngJQ. program: requirements.,.
4 4
i (t) In the [Q "world", the burden of maintaining [0 " PAPER" has always been costly and cumbersome with no significant safety benefit.
This'is true 4
for both c.lner and newer plants.-
3 (u) This is much too broad a topic to be adequal ly discussed in this survey.
However, one philosophical observation may. help the NRC staff in its IAP efforts.
As noted long ago by $andla's Lloyd Bonzon, much of the difficult in the IQ arena stems from the NRC and industry's difficulty defending the state-of.the art.
[Q unlike many other technical Areas is not precise.
the objective 11 achieving reasonable assurance of operability.
Yet, we often engage in evaluations and " discussions" on specific technical issues that appear to pre. suppose an overall analytical cert ainty that generally does not exist. 'Ihis of ten unjustified focus on -
precision t an espend significant resources arguing and fine-tuning posittant that may have little importance when overall IQ uncertainties 4nd safety 5iqntiiranre are rnnsidered.
(v) No.
6.
Are the current EQ requirements'for older and newer plants adequate for plant operations beyond the current 40-year.. operating license.(i.e., for license renewal)?
'e (Thisisperhapsmyfavoritequeillonlon'the. Survey.)
y.
- - %l_% -
(b)
Given that 1((I
)?3 1974 is the requirement for newer plants and it requires pre-AginQ to determInc qualtfled life, the quallfled ll(p for many long lived elec t ric al components such at cables is typically 40 years.
What is the technte al just 1firation to conclude that these cables are qualified for
- 50. 60. 65 year e Maintainino the CLB for older plants leads to the enrolusinn that nlder (ables are qualIfled forever or unt 11 they f4 1 which ever comes Iirst.
I believe that given the current [Q requirement.,
the Bip for l0 and Renewal is the best approach.
(P) IQ recutrements for newer plants are adequate for renewal.
Older plant j
requirement. are nnt adequate for renewal.
(q) two (oncerns; (a) Arrhenius.is not good for 60 years as A minimum, neott
.omo sort nf benchmark along the way. (b) Provision should be made in yiven design periodically, maybe every ?O years or so.
cequa ' ? f 4 (h)
A, I stated above. ! believe the replacement clause should be modified to remove the statement "unless there are sound reasont'to the i
contrary."
(t) The present rule is barely adequate for the original term of the license.
Iitensinn of the term of the 1itensee should include a verifIcatinn that all age and environment conditiont that have existed and that can i
.potenttally ncrur ar* 4 onsidered In reevaluating < nnt inued qual 1f a at inn.
I
4: <r..st Mip
~
C-30 (j) The current licensing basis can be used successfully if adequate steps are l
taken to assure that aging is not causing significant deterioration.
The nature of the beast is that by 40 years very few DOR components will l
remain.
Cables, penetrations and motors are what would be expected.
Any other components would have to be in very benign locations, and if they were, aging would not be a concern.
I have a belief that for those components that will be used for more than 40 years. condition monitoring should be used to assure continued safe use in quallfled application.
This would cover any uncertainties'in aging models f.1 would not change '
4-existing qualification rules.:.for;licensefrenewalM f aging ef fects _ (t'hermal and l
(k) No'.
Need to have a better assessment radiation).
(1) The current E0 requirements for older and newer plants may not be adequate l
for plant operations beyond the.currenti40 year caerating license.
The.
effects of aging /quallfled life would be one of tie areas that comes to mind that should be considered as part of plant life extension.
The older plants did not have to pre age components'at:part of the test program and operational occurrences su;h as.steamiltakt'and inadvertent spray.
actuations could cause degradation of. equipment.
These events were not considered by the licenson's when they estabitsbed qualification for their
[0 equipment.
Also test programs utilized Arrhenius calculations to damonstrate qualification for periods of time far beyond the tested duration.
These tests may rot have been adequate to extend the life of equipment beyond the 40 year plant life.
(n) f test of all. the maintenance program for the equipment in harsh environment *. should include monitoring of certain critical characteristics. Our IMI-2 Technical EvaluationiGroup discoveredethat-equipment survived the accident and worked proped y only if. properly applied (i.e.. application requirements' consistent with design specifications) and maintained.
A significant~ number of fatiures could be traced to seal failure.
Had the maintenance program included seal evaluation and surveillance.the overall outcome might have been different.
Based on my aforestated experiente, I believe that action ~that snould be required fne [0 for 1trense renewal should consist of the following stops:
i.
Review all maintenance and replacement activity procedures and records to assure that'the equipment has been maintained in a manner which is consistent with the need to retain the qualificat ion statu'.
By this I mean, crit ! cal characteritt ics (mat erials. part s, et.c. )
were not degraded by either maintenance action or design change, it.
Determine whether or not the equipment has been operated in the environt ent s f or which it was quallfled, iii. Determino whether or not the equipment has been ope ated in the i
manner (modes) for which it was qualified.
l l
.e w
.+
skh&AkAhbN. SWhc
~<
$!v p!
.:.y iv.
Based on my suggestions in 2., determine.whether or not age is a significant failure mechanism (other than (Iniform random) in properly maintained equipment.
This should be all that is necessary.
We should' strive toward this kind of quality, cost-effective approach.
Make. history work for us.
(p)foreculpmentthathasbeenqua'llfied'ina2cordancewithCategory1.It would be only slightly facetious to~ state that 20 years of additional service falls within the noise level of the accuracy of the qualified Iivc5 e5tablished. Ior equipment qualifled in accordance with Category 11 and also subject to harsh environments, additional testing to comply with (ategory I wnuld increase confidence in the equipment *s functional capa-tulity. [ven for Category 11 equipment in mild environments, a blanket esemptinn from additional testing is not advisable. Continuously energized equipmant and equipment subject to high load rates might have failure r4tn near the end of their qualified lives that are inconsistent with the dependability desired after an accident.ceven thotigh the service conditions do not change.
(') in principle i see no engineering justification for judging a station's 10 q
adequate at 39 and inadoquate at 41.. ilt is. unfortunate that the funy concept " adequate
- has already been stretched to incredulity by the grandfathering of less stringent EQ practices for older plants.
How can it be a Justifiable reference in considering license renewal?
l believa U) ' adequacy
- should be considered at all Limes ghen i sign,f,Qatt t_1.1Qb1 and that the 1Iconso renewal ritual should be to double check that, in fact, was being done all duringithe initialilicense period.
I can think of npthing relative to cable systems which one could judge adequate at 39 and then questlonable at 41, 11 questionable at 41, then we botter atte u the 511uatton and be prepared for CorrertiVo Attinn at 25 or 30 (r; tnmment 1.
T h.> louer rigor in [0 requirements f or older plant makes thom r e s i a. t a n t in upgrades of equipment lhe impact of this will be a stoop investment no enary to estend plant life beyond 40 years, Some degree of (nmprnmise may ha warranted, but requirements as currently written, interpreted and enforced will preclude older plants from upgrading on an ongoinq hasis to an extent that the plant life extension is economically viable f.nmment 2:
Most programs don't really justify 40 years and repl'acement of equipment 'hould occur periodically. If 40 years has been accepted then it probably can be extrapolated on the same basis, t omont 3:
Imphasira more testing on naturally aged equipment.
(s) lhe [0 reautrements for determining oualification up to 40 years are perfectly valid and adequate for determining qualiftration up to 60 year..
in eontinue the qualtfled 1ife of equipmont beyond 40 years, the inense renewal appll( an! Wnuld have to tuhmlt a revised Qualification package to
A C-32 1
the NRC. Of course any aging mechanism that might became significant nnly during the last 20 years of operation would have to be addressed, but this is a standard part of any qualification mothod.:In many cases the original aging and type-testing program could be : reevaluated to. justify a 60-year.
life because the utility would present measured operating environments that are far less than the conservative design values that were used in the original qualification program,'This reevaluation would need to demonstrate the same degree of margins raquired by standards and regulations.
If the reevaluation does not demonstrate a'60 year life, new qualification test data may have to be generated. Or the environment of the equipment may have to be mitigated. Or condition monitoring.could be used to show-that the actual aging degradation of the itemils less than that in the original cualification program' The last resort would he to replace the itam not al1 this would be governed by the EQ requirements appropriate to the current licent,tng basis of the plant, tvidently some are questioning tne adequacy of the EQ requirements in the 00R Guidelines for justifying a 60-year qualified life, those persons may not be giving enough credit to tb key requirements contained therein.. The 00R Culdelines require specifical in of service condittons including margins and conservatisms on accid 'nt-conditions. and type testing under.
accident conditions with aging conditions evaluated by analysis "a m igd br i n uitt_a" and a surveillance and maintenance program "to assure that equipment whl(h is exhibiting age related degradation will be identIflod and replat.ed as necessary." lhe key message is that the DOR Guldallnes expItcitly address the same " age related degradation" that is of concern for Itcense renewal. There is no basis for disallowing the use of DOR Guidelines to qualify equipment for 60 years'.'It is up to the regulators to exercise their duty And professional judgment as to the adequacy of the DDR qualification package submitted.in conjunction with a' license renewal' appliratlon.
(t) Unequivocally, yes.
l (u) Ihe current regulatory and technical (Q framework is adequate for plant i
operat ion beyonti 40 years.
I do not believe exittinq (0 regulations or quidance documents need modification specifically to address [0 tarhnic al luues in the license renewal period.
flowe ve r, the safety-signiftrant technical t uuas must cont inue to be adequately addrened by ongoing [0 programs.
The mnst technically justifind license renewal [0 conrern relates to the condition of long Ilved equipment when compared to the equipment's ralculated life.
[ven in this contart current guidance
.ippears adequate.
For example, Regulatory Guide 1.89 Rev. I currently states that state o'-the art preconditioning IechntQues are not capable of simulating all sogntiscent t," pes of degradatson....fsperience nggests that consIdcrat Ion should be ge ven to a combtnet bon of (1) preconditiontng.. and (f) survtlliance, t est ing, and analnt enanct. '
V i m m
y a
.u.
~
g.g C~33..
% ma
~
(v)Licenserenewalshouldbebaseduponcondlilonassessmentandobjective evidence that the installed equipment will. survive beyond its qualified life plus accident exposure.
Which EQ issues need to be addressed for continued plant operations beyond 40 years?
(P) Ihe main issues for older plants are aging, margin, synergistic effeCis, relaxed testing requirements and documentation requirements.
(f) inspectors shoulo be trained and check EO on a continuing basis, not when problems arise.
(h) The issues that need to be addressed are wnat components are quallfled, what their qualified life is, and what maintenance arttvitiek are roquared te extend the life to the end of the renewal period.
(t) 1here are no specific "EQ lssues".to be. addressed for license renewal.
There are, however, specific equipment'Issilht khich need to be addressed in comply with the license renewal rule. "Under the rule'. E0 li covered untler the current licensing hatis and the utilities should be allowed to
" engineer" into the license renewal period.
Thit' engineering may include sound analyses and/or testing, or wholesale replacement.
The important aspect is that the rule and regulations are adequate.
What modifications would you make to the existing EQ requirements for license renewal?
(e) for license renewal, modify 10CFR 50.49 to upgrade older plants.
(fj All plants.hnold be requireo to meet the requirements of the rule for litense runewal.
In additton, all equipment should be requaltfled for life beyond 40 years.
Due to the environment the equipment has seen for 1
11s serviri I tfe. *here 14, no quarantee that its qualification 15 st111 good.
Some mathod of in situ test ing or the taking n( samples would d
'ufftre for this requirement (k) (.onsideration should be given to LOCA testing selected lamples of plant equipment (primarily cable) before and after additic9al aging; (t) Nona, errept for those pointi made in other parts of this section.
l I
l a
C 34 B.
Op.trAuns.CepsIJ.sncs l.
In general. 'ow is EQ equi pent actually: aging,.(in service) compared with the equipmen.'s predicted.liftf e -
y y J.
9 (a) Anecdotal data-only, need more testing.I' ""
(h) (l'm not sure that you can get a good answer'to this question.) Tre reporting data bases such as NPR0s and LCRs will only give us part of the answer.
We do know that a lot of equipment, including cables, is being replaced as a result of aging / failures.. To'the extent that the equipment was quallfled for 40 years, the actual service life is much less than the quallfled life.
(d) Generally the f0 equipmentlit aging slower that preditted.
6 (e) Don't know, NRC / Industry hasn't looked.
(f) Unknown. must watt and see, fspecially for components that are effected tw Inw levels of radialton, ett.
(h) in c ases of ASCO solenoid operated valves, we have seen what appeared to be premature failures until the actual ~ operating' conditions (ambient temperature and normally energiaod/deenergited) were. properly evaluated.
When that was done, the calculated I.ife Closely approximated the actual life prior.to failure.
(l} lhis area is not being inspected at this. time;'therefore, I have no rorrents.
(J) Host applications are aging more slowly. Thermal conditions that dominate most component aging are less severe for the' bulk of equipment.
Localtred hotspots are Causing some components to age relatively fait but not at a rate faster than models sugges.t when actual. conditions are considered.
(n) I am not qua l i f io.1 t o answer t hes* Ques t ions ~ al though l 10 have npinions oh some of them' 1
(c) Some of the previous discussion is applicable to this section.
4 It has not been uncommon f or electrical c ables to f ail because local operatinq cnndittons were more severe than estimated.
Some have argued that this is not the f ault of the equipment but is evidence of m i *. a p p l i c a l t o n.
While there is some truth to this argument, it appears thAt f0 didn't out the mustard, even if the problem was improperly
<perified conditinns Snot ifical ton of conditions is a part of [0.
Briause of these events. It <. c o m s reasonable to ask. how many other similar condittons exist but are not known because tho (alile didn't fall.
]
yet the cable mtyht not be adequate in survive an accident?
This relates directly to the dt'.cussion above about the fact that cables arr not 1
subjected to a program of periodit. survaillance and tetting.
I
- u
.w o, v 9 Ae,;a.
m.H C-35 4
(q) In aenerd, aging ef fects are being, found favorable compared to what would have been expected under the ooJtulated conditions. Whore ambient conditions were wet, or as high a temperature as specified, or as high radiation, premature aging or failures have been found on a number of cable system components, Of course, ambients g a severe than specified have also caused failures. Mpu ambients are well below the specified values (and equipment ratings) so that failures in normal service are rare and aging rates low.
Service experience with cable systems gives only.
minimal rnlevant Information as to its operability under accident conditions.
l (r) Comment l' I
fxcept for rare cases of gross error, where equipment has failed in-service due to much higher than estimated service conditions (e.g.,
normally energlied solenoid valves), little work is being done on which to
{
base a general assessment.
In general, plants are taking a hands-off approach fearing less than desirable results.
- s. v Comment 2:
(xperience indicates that components that have problems in service do so because assumptions in determining life prove inadequate, e.g. service temperature higher than anticipated, self-heating not properly accounted
)
for, materials changing (poor 0A).
1 (s) I have no dirert knowledge of how (Q equipment is actually aging in i
service compared with the equipment's predicted life. However. the EPRI in plant study has sper imens of cables and small elect rical component s i
installed in nine operating reactors since 1985, but they have not been aging long enough to come to conclusions regarding in-plant versus art ificial aqing. Also, a $andia study for life extension of Cables (Du(harme, A R.
and t. D. Bustard, " Technical tvaluation of In Containment Cables for U.S Nucloar plant Lif e Extension," Proceedings nf the International Conference on Operability of Nuclear Systems in Normal and Adverse Invironments. OpfRA 89. Vol. 1, Lyon, france, September 18 22. 1989, p. 605.) has concluded that th0re have been only a handful of
((Rs that indicate any aging degradation of cables.
(t) Generally, equipment is aging at a rate significantly slower then predicted by aging methodology typically used in E0 tetting.
(u) This que'. tion is best directed to utilities but I have several observations.
With the possible exception of cable, uttitties generally do not compare the equipment condition at the end of its qualif ted life wtth some " prod u t ed" condit ton.
Since most qualiftcation te'.t prnurams do not measure equipment properties after accelerated aging, no baseline end of I n f e (andliion data it available to support such a comparison.
However, for in containment LOCA qualified er4ulpment, the condition at the
{
cnd of qualifted life is generally excellent.
Uttilty maintenance J
personnel often question why equipment, in apparently like-new conditlon, is being reptared when it has reached its end of qualif ted lif e.
This cacellent c ondit inn generally stems from minimum thermal and other aging I
degradatinn.
As an illustration, please ref erom e the attached turves for
\\
l 1
l
.t
.C-36 the Raychem WCSF material.
Based on Arrhenius the 40 year at 90'C thermalagingline(figure 2CurveB?.results)napproximately60%
~
~,
retention of the original" unaged fult mate ~elong'ation (513%).
(See figure I using two Curve B data points - 800 hours0.00926 days <br />0.222 hours <br />0.00132 weeks <br />3.044e-4 months <br /> at 150'C and 2500 hours0.0289 days <br />0.694 hours <br />0.00413 weeks <br />9.5125e-4 months <br /> at 136'C).
In other words, after material thermal aging to the equivalent of 40 years at 90*C this material has an ultimate elongation in excess of 300%.
Since actual operating temperatures are generally below those assumed for qualification, even less thermal degradation should occur.
Qualification and research tests have shown that cables typically fall only after significant degradation results in material cracking (i.e., the material has essentially retained no elongation capability).
One could conclude that temperature is not a significant aging mechanism for this
- material, for many equipment materials, the accident LOCA dose (e.g., 200 Mrads) and not normal aging causes the most significant degradation.
(v) Most of E0 equipment has not aged as fast as predicted by equipment qualified life.
However. certain equipment is degrading at a faster rate (i.e., Kerite cable).
2.
Describe problems you have encountered with EQ equipment. Do some components routinely fail before the end of qualified life?
(e)
Many of the problems are described in>the' numerous lNs, Bulletins, and GLs issued on E0.
(f)
Many maintenance personnel are unfamiliar with EQ requirements and conduct repairs or maintenance that "unqualifies" the equipment.
[xample - transmitter 56als, if they are not replaced during the component repair, qualificat ion. can not be usured, Other phy%I al problems, like water (condonsation) in clo:4d cable conduits, and splice practices that are not LAW the EQ instructions (that are stricter than normal electrical spice practices).
(h) As I stated above, many licensees do not know what the local temperature of components are.
A'. a result, the calculated life is often much.any than the actual life.
()) very few components are falling.
Some are found deteriorated more than one would 1ike but failures of qualifled equipment are not frequent.
Most radiation conditions are at or below design conditions.
Current 40 year doses for many plants are 2 Mrad or less.
(Some have 10 Mrad doses but these seem to be in the minority).
None have the 5'.) Mrad deses suggested by IEEE 323-1974.
(1) (.nmpononts may fail for various reasons which may or may not be related to (0 attributes.
However, components have failed from steam' leaks.
As far as component quallfled temperatures and radiation exposures I believe that these environmental parameters have been adequately addressed by the I
licensee's fQ programs through detail analysis and study of the areas for such problems as hot spots.
However steam leaks and spray downs have not been adequately addressed.
^;
x,
~,. +
f
.p.
. ~. ',g+s s'
C-37 (q) Sec 1 above.
tr) Comment:
for older plant. revised accident analysis. create a substantial increase in accident environment conditions, lhls is due to many factors. most significant are fuel upgrades and pump and fan cooler. degradation.
At issue is the fact that qualification per the DOR guidelines is marginal.~
at best. to the original EQ requirements.x Upward revision of the E0' requirements can be costly.
ilD 14844 is viewed as excessively conservative.
Current requirements.
include a 10% margin in addition toithe TID'14844' based values.
A more
+
realistic basis for determination of radiation accideit environments is needed.
I (s) I have encountered no problems with E0 equipment. Certainly. I have no' knowl.dge of components that"" routinely fail b9 fore.the end of;qualiflod lifo " Any such knowledge would appear to be. reportable under part 21 and
[
not through a survey, j
l The EPRI plant agt.1q study is showing that component qualified temperatures and radiation exposure levels are not consistent with their i
actual in-service environment -- almost all measured temperatures are less l
than qualification temperatures and all measured radiation doses are much less than qualification doses.
(t) We have experienced no problems with E0 equipment and we_have not-experienced routino f ailures of E0 equipment before the end of _ qualified life.
It should be noted that E0 equipment <is no~different from other equipment except for the testing which has been done to it and the i
requirements to maintain its quallfled configuration (i.e.. sealing.
mounting, qualified life etc.).
Therefore..the data base.for equipmen't trending and failure rates is much larger then just the equipment in the E0 program.
i (u) lhese operational and equipment failure questions are best directed to utilitles.
Regarding operational vs. assumed environments. l'have-several-comments.
For virtually all equipment, the actual operational radiation level is significantly below the assumed normal radiation dose (often assumed as 50 Mrads.) and material damage thresholds.
Similarly, i
operating temperatures are generally below those assumed fer qualification purposes.
Further. plant temperatures fluctuate due ta seasonal and operattonal (hanges while for qualificalton purposes utilit N typtrally auump a single cnnt inuous value (e.g., peak normal or maximt.m datiqn t om-perature),
lhe one possible exception. plant hnt spot areas, arn generally addressed by monitoring and other methods.
(v) [quipment rellability can be enhanced by using sound reasons to the contrary." A more stringent requirement for replacement part 5 and components are required to enhance equipment reliability.
i
..< w
~
._ ps ' ; @gh;.pgg3 3..
u.
.x e.~., p.
w C Are component qualified temperatures an'd fa'dlation exposure consistent with their actual in-service environment?
(a) in many cases, no.
Example:.the ANO fiasco.
(d) The problems we've seen are where the full expected temperature range was not considered, or considered inappropriately.
Specifically, some outdoor transmitters not fully quallfled for cold temperatures or the high temperature next to a steam line were not adequately calculated or measured.
(e) Some components like cables have failed in normal environs prior to the end of their qualified life.
While industry argues the actual temp / radiation levels are lower in-service than that assumed for E0, there have been many instances of equipment damage due to temperature / radiation hotspots.
~
(h) As for radiation, the level of. radiation Is almost. always' an ' assumption based on calculationt I do not know of any li.censee who has measured localized radiation levels during plant operation.
(j) Most temperatures are averaging'20*F or.more below design normal maximums.
A few locations near M51Vs pressuri2crs, and primary 100ps are higher.
Some plants have summer containment temperatures.that are higher than originally expected, but even for these the aging rate is nnt excessive for the bulk of the equipment.
(t) lypically, the actual in-service temperatures and radiation levels are lower than qualified parameters.
3.
Are you aware of any weaknesses associated with'the maintenance practices being performed on E0 equipment?
(a) Surveillanca and maintenance in qeneral rould be much improved.
(d) Region V is not aware of any poor malatenance practices associated with (Q euuipment, (c) Maintenance in general af. pears to be good and has not resultes,.n
<ignificant C0 problems.
As components age, this issue needs to be looked at in more detail.
4 (q) The E0 inspections of ten showed that [Q ongineers and plant personnel old not always interface well.
Check the escalated enforcement files.
Lubricants and tape splires were two problem areas.
Some plants addressed maintenance requiroments much better than others in the [Q documentatlon.
(h) Many licensees take a very narrow approach to required E0 maintenance, they consider reconnendations of maintenance by the vendor as just that, recommendations.
In many instances, the vendors make tlatements that they a,'.ume a perindir preventive maintenance program, as identified in the technical manual, is being implemented in order to ensure the equipment
C-39 remains in a qualified condition.
lhat is because some age sensitive components (e.g. lubricants) were not subjected to thermal or radiation aging.
(j) The bulk of the maintenance is adequate..Some plants have better control l
over actual Installed conditions.' Host? current maintenance is related to replacement except for NOVs3and.motorsglj(donf tiseenmaintenance as being adverse for E0 equipment.
- p. e.
(1) I am not aware of any weaknesses associated with the maintenance pratlices being performed on.E0 equipment.
I believe.for those E0 components where maintenance is required for qualification that it is adequate.
- However, all to components do not have required routine EQ maintenance.
(q) for passive cable systems, no. Maintenance activities on passive, mostly inaccessible components are minimal, yield'very limited useful dat A And, if it invnives the movement of cable components, may adversely affect operability of the equipment if subjected to a harsh environment.
(r) Comment 1:
for many fla u lf components this is not an option.
Warranty and llatillities tuues, not specific to the nuclear industry, have forced some manufacturers to curtail the sale of refurbishment kits.
This is a parttrularly problem for A0V appurtenance'..
In some cases, maintenance activities of any sort is strongly discouraged.
vandors of Class It equipmnnt are forced to forbid maintenance, and encourage full replacement.
Current 3EQ programs'do not qualify maintenance activities, and:little ~benefitWould be realized for undertaking substantial investmentito.do?so.; '
Comment 2: Maintenance that is required to maintain qualification ihnuld be specIfind by the qualifying party and become mandatnry at the plant, Routine maintenance programs must be reviewed to determine potential effect on the qualified product. Maintenance' indicated by more recent programs may not be backfit.
(s) I have seen no evidence that maintenance performed on quallfled equipment in keeping with the qualification program is not sufficient to maintain the equipment's qualification. I have seen no evidente that properly performed maintenance has ever had an adverse affert nn f0 (t) Rnuttne maintenance programs are structured to include specifir *f0 maintenance" and therefore are adequat e to maintain quali f icat ion.
(u) I have several perspec t ive<. regarding " sound reasons t o the cont rary"
~
Upgrading, simply to have quellfication based on testing that includes p r ef a g i ng, is unnecenary for equipment not experiencing significant degradation due to normal or accident conditions.
As a broad generalitation. most outside containment equipment should fall into this category of no significant aging degradation.
For example, with few exceptions, equipment materials are highly tolerant of the several Mrds of post LOCA radiation occurring outside containment'and do not experienre i:-
y:m i
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - ----- d
t Y
I C 40 E s
.e
.s significant aging during~ normal operation.
Yet, on replacement this equipment must be replaced with other equipment which has been qualified using tests with preaging and bounding accident conditions. An analogous situation exists for outside containment equipment experiencing relatively low-temperature (e.g., 175'f) short-time (e.g., 30 minutes) pipe-break steam conditions.
All else being equal, the upgraded qualification is somewhat beneficial by reducing qualification uncertaintles.
The upgraded equipment could but not always exhibit other performance or operational improvements over the existing equipment, However. In many cases there can be negative upgrading effects.. importantly.and based on experience, the current equipment is generally well'tuited to its. applications from reliability, performance. operation.;and maintenance perspectives.
The new equipment, although E0 upgraded, may be less suited to the application from these perspectives.
Yet.nthese important operational considerations are not currently identified as example sound 1 reasons in' Regulatory Guide 1.09.2/
I believe overall suitability and not solely [Q should form the basis for utility declitons regarding the use of upgraded equipment.
Finally, the use of upgraded equipment inevitably involves some design and physical plant changes to acconmodate differing characteristics.
Always present is the potential for problems arising.from such changes or lack of experience with the new. equipment...
Is the maintenance performed on E0 equipment sufficient to maintain equipment qualification?
[Therewerenoresponsestothisquestion]
)
15 there maintenance being performed on equipment or components that may have an adverse affect on EQ?
4 (h) I do not know of any maintenance that is being performed that is 4
detrtmental to the equipment, however, maintenance that should,be performed and is not may be detrimental.
(1) I am not aware of any maintenance.being performed on equipment or components that may have an adverse affect on [Q.
l (t) Not to our knowledge.
This 1% why utilities have elahnrato review l
processes for procedures and training to' avoid just this type of problem.
I 2/
laterestingly. performance and reliabillty considerations were addressed in the original sound reasons discussion rontained in an NRC gen-crit letter but were specifically excluded by the Staff when the regulatory qut<te was iv.ued, i
C
~ - -..
1 C-41 4
Discuss your views and opinions of specific cases (current) where l
replacement equipment was not upgraded to 10 CFA 50.49 requirements because licensees reference " sound reasons to the contrary (R.G. 1.89)?"
4 i
(a) This often leads to trying to qualify what'are'now' parts and subcomponents j
that are ??? only [ rarely] available - if at all. As CGl's - this is not easy.
e 3
(e) The sound reasons to the contrary were intended to be used on a onetime i
basis and not intended to be used repeatedly as an excuse not to upgrade 4
equip as it is replaced.
RG 1.89 should be modified to eliminate this w.ndfall for the industry.
There should be no need to invoke this any j
longer. We need to find out from the industry to what extent this is still j
being used.
j (9) (No) i (h) I do not know of any cases at this 1tme where ! disagreed with a i
l u.mee on the determination of sound reasons to the centrary.
1 do l
feel that by the time the older plants request a license extension.
i much of its equipment will be obsolete and will.recuire replacement wIth equipment qualI(led to the higher standard f 10 f.FR 50.49.
o
(!) I know that equipment is not being'upgradad Ifue to'" sound' reasons to the contrary" however i do not have Azcurrent spe&iflCJCase because^ l'have not i
performed many (0 inspections over the past 3 yeart.
I do know that most 1
of the licensee's L0 programs have provisions for documenting " sound reasons to the contrary" (or not upgrading to 10 CFR 50 A9.
fj) I don't 500 (,ound reasons to the contrary as a big probiem.
Most plants j
are reptartnq components to modern standards when end of life occurs for 4 l
DuR or fateqory Il device.
(r; Performanic of equipment during and following the accident has not boen i
adequ.itely addressed by regulations.
The credibility.of values used in det ormininn ut point s is quest ionable. Some are hold io a 95/95 i
determination and other' are not. 11 may be uf.eieu in qualify equipment if porinrmante is i q n o r t'd.
l
}
- ! - omo n t RG 1.89 is t oo loose in t his area. Equ'pment should be j
opuraded (or at least reviewod for adequa(y) unleis by some justifi5ation the upr;rade woult; have a negative impact on safety, i
j (t) the seven " Sound Reasons to the Contrary" (SRC't) appropriately maintain emphasis 0.n '.af ety while not adding to lhe complexity-of the already l
4omplex maintenance task.
We have no knowledge of misuse of the SRL's and typically the emphasis is biased toward upgrade rather that invoking the
% R(. ' s.
C 42 C.
LQ._1.rupcc.Llun_ac t i v i t i es 1.
Were the NRC EQ inspections conducted with the appropriate scope and depth' and in a consistent manner?
(a) More scope and depth would have been better.; ~
(d) Region V is not aware of any m I,fic+ gk\\eCM;g n on c 0
- t
'M spec Wea heiset of the ? scope' of-previous [0 inspections.
The EQ inspectionlprogram conducted some years ago was of sufficient scope and dnpth. '
(e) They were conducted consistently with appropriate-scope and depth (ronducted 15).
(f) Yes.
There did appear to be regional differences implementing the DDR guidelines.
(h)Iamnotsurethatwewereveryconsbtent(aianagenry)inlooking at the maintenance activities, or the selettion of ambient conditions, for example.
I think we Inarned as we went and 1 believe we tried to ma.ntain consistency.
I do believe. however, that we did lack consistency in the area of the two examples, above.
(k) Inspe(tions were appropriate in scope and were consittently.porformed.
I am not aware of any specific weaknesses associated with these insportione.
(1) The (Q inspections were in my opinion conduc.ted:With the. appropriate scope ~
s.
and depth and were done in a consistentimannet f Teamhleaders and members received hours of training on EQ.' The regiohnlbleamslwould always have a represent at ive f rom the Vendor inspection iltanch on the team as an Assistant Team leader or the,itam Leader. ' Region ~11 performed the inspections using inspectors out of the' Plant,5ystem Section-in ORS.
The inspectors were either ficctrical engineers or had'several years of esperience in [0.
in addition, the (Q files reviewed by the EQ iaam were i
principally reviewed by NRC contractors hired for their f0 expertise.
i aware of any weaknesses associated with those inspections.
am not (n) I am not qualified to answer these questions althetigh I do have opinions on some of them!
~
(s) 1 have lit t le t o no knowledge of the scope and depth of the NRC LQ inspertlons. I do know however, that a number of utilitles were fined ior inadequaries in their programs as a result of the inspertions. That tolls me there was *ome robust nen t o their unpo and depth.
(t; Iho Nkt (.Q inspec t infis were consistent with a broad scope and to-considerable depth.
Please note that we were inspected under hoth DDR and f,ategory ll requirements.
(u) Ihese questtons are best addressed by NRC personnel.
However6 except for the NRL Staff activities in support of original plant. operation, I believe v
h;
. qWp be-4 e e.
g:
C-43 that the NRC has ever performed.
j the NRC [0 inspections were the most comprehensive compliance evaluations i
(v) The NRC [0 inspection process lacked adequate depth and consistency from Region to Region.
The process typically allowed I week for preparation. I week on site and 1-2 weeks for inspection report preparation and approval and the inspection team generally consisted of 5-6 people.
This equates to about 3 1/2 to a days of actual inspection, the scope of the Inspection in terms of the components selected for review was adequate but not enough time was allotted to conduct a thorough in-depth. investigation..
Are you aware of any specific weaknesses associated with those inspections that need to be addressed relative to:
a) older plants? b) newer plants?
(e)
The documentation in the older plant supporting.the EQ inspections were a problem.
They were not as good as'the newer plants.
(f) Not from the Region !!! plants.
(g) Instrument loop accuracy was sometimes overlooked, depending on inspection 1 team leader and members; (1) 1he area of [0 should be reinspected on a periodic schedule.
At present only random issues identified during other activities are evaluated.
(m) [0 knowledge does not exist in the working level herause of reduced emphasis.
This is the case with NRC inspectors as'well as licensees.
The populatinn that gained a good working knowledge of [0 has moved on to higher positions or different positionfi.
In my experience of'second round audits to lOok Al preVicus open items, maintenance Was the weak area.
(o) 00R Guidelins Section 7.0, Aging. states (underlining has been added for emphas1s);
implicit in the staff position in Regulatory Guide 1.89 with rogard to backfitting 1(([ Std. 323 1974 it the staff's Conclusion that the incremental improvement in safety from arbitrarily requiring that a specified qualified life be demonstrated for all Class 1[ equipment is not suf ficient to justify the expense for plants already construc ted and operating.
This position does not; however, exclude equipment using materials that have been identified as being susceptible to significant degradation due to thermal and radiation aging.
Component maintenance or replacement schedules should include considerations of the specific aging characteristics of the component materials.
Qnggjng_pf.Qgam should exist at the olan Lto__ review sur_yelllance ard maintenance rfD I.dk LQ_nturt 1hJLtguioment which is exhibit inn ace relatsLdearadat io3 will_ tee Hign,L(f_LqLand reglaced as necesum I am not aware that the above stated ongoing programs generally exist for (ables, t.hich certainly contain material known to be susceptible to both thermal and radiation aging effects.
It appears that NRC's inspect ion act ivitim may be latking in t his area.
I
l
'Y Awf$.
t-
- n 94, 4
4 C-44
)
(t) The emphasis was biased toward " paper" without..&. corresponding emphasis toward the safety significanco of the equipment s 2.
What safety-significant issues have been identified.as'a result of EQ inspections?
7 l
(d) At Trojan, there has boon a significant safety-related issue associated with the qualification of containment electrical penetrations assemblies (IPAs).
The enforcement actions have been deferred due to potential wrongdoing by the licensen.
These issues are still being inspected.
The IPA seals were replaced prior to restart from the refueling outage.
i 1
(e) Cable problems Limitorque MOVs, solenoid valves, terminal blocks, penetrations, grease, and transmitters were identified, 1
I (h) I cannot i n all all the safety significant issues, however, they woro
)
j weli documented.
(ij ine findings and resultant actions were documented in inspectton reporis.
(J) The benefit of the [0 inspections was to refocus the E0 activities towards I
t he ac tually installed condit ion rat.hcr.than just towards hav,ing adequate J
documentation.
&;:?: Mima
+
bi (k) A number of safety issues incl (id[ng Linadeqilate%%D l
ilualification of cable, i
rable splices, containment penetrations, solenoid valves, valve onerators, terminal blocks, etc. were identified.
l (il Numerous F0 issues were tdentified based on specific E0 problems at l
r ert ain sites.
A good portion of the problems dealt with the qualifitatinn nf electrical interfaces (i.e., splices and terminatic.ns),
several escalated enforcement cases were handled dealing with these t s '. u e s.
Limitorque MOV quallflCations were also'a problem and the tubject i
of numerous NRC Notices.
This was basically resulting from the' fact that the installed configuratinns did not match the tested configuration.
Most of t he f 0 t unes were r.".olvod at the end of the first round of f0
]
i n '.por t t on s.
l (t) The inspertinns Idontifind generic 1stues such as MOV testinq and
.ipplicat on of heat 'hrink splices.
The inspec t ions also highlighted the n.ed for training and tonfiguration control.
a (s) I take it this question is directed mainly to NRC staff who are in the best positson to say what safety-ligniftCanttiltues have been. identified as a result of NRf 10 inspections.
(u) Iht s quest inn is best addressed by NRL and utility personnel.
Without comment ing on their safety-significance, I believe that most f 0 issues dostrihed by NRC Bulletins, Informat ton Not ices, and f.eneric Let ter', were neiginally self identtfled by utilitles and manufacturers or resulted from 4
researth or f0 qualification testing.
~
~
l
..,,t l
~
Cd5 (v) failure of AMP Nylon and Xynar splices to perform their safety functinn when subjected to postulated accident conditions.
What enforcement actions have been taken?
(a) See the Record.
i (e) About 20 utilities received. fines /for.,EQ1 nfractions from.these n'
inspections.
(g) Review of the escalated enforcement. records would be useful, there were about 30.
(h) Many enforcement actions were taken, including ASLB hearings for the farley case.
(k) inforcement actions, including civil penalties, were taken against several utilities relative to these issues.,The.,sigolficant inspection issues h.1ve been satisfactorily resolved.
Have all significant inspection issues been adequately resolved?
(a) No (b) You should take a look at the proposed enforcement actions to get a feel for the answer to this one.
(c) All signifitant issues have-been resolved.,
(h) 10 the best of my knowledge. the issues have.been resolved.
3.
Is sufficient emphasis being placed on EQ in the current inspection program?
(4) Nn.
(d) Sufficient emphatl5 was placed upon IQ during the last round of
)
inspections.
As a result, licensee's established an asce proceduralized administrative controls over'the IQ area. ptable set of Assuming that 1trensees are following those controls, one would conclude that licon ves are plating suffitlent emphasis on [0, (e) The current inspection program l'. really not looking at IO.
The '84 '86 (0 insper t inns are cons triered complet e and no f urIher inspec t tons are t
warranted (many NRf./ industry people believe this).
I believe periorilt I
inspettions should be ronducted.
Since no 10 inspections are being e onduc t ed, the regional expertise is fading away.
l (q) I don't know, lhe regions sttil have a few inspartnes with some 1980s esper t enc e plu'. maybe ot hers wit h some knnwlerige.
I believe that mn'.t people don't understand [Q well.
)
- _.-. - - ~ -
g wijf C-46
- O " ~ '"
~
~
(h) There is no real emphasis placed on EQ,in the: inspection program.
I
~
attempt to look at the EQ issues when possible", but.that is not really j
gond enough.
(i) lhere is no emphasis being placed on EQ in the-inspection program and no 1
i training is taking place nor are EQ issues being'. highlighted.
.a;
.,a n / :. -
j (j) CQ focus varies greatly from reglo& :. w...;,> m, Regions 1 and 2 seem to n to region.
j have a stronger focus.
Region 3 doesn't seem tc have as much interest.
There do not seem to be a large number of personnel that understand EQ intimately.
(1) As far as I know. E0 is not being inspected as part of the routine 1
inspection program in Region !!.
It is typically examined as part of the reactive inspection program.
There has been no formalized EQ training in j
NRC since 1987.
I would say that most inspectors do not thoroughly j
understand the past positions NRC has taken in regard to environmental qualification.
(s) 1 didn't even know that there are " current" EQ inspection programs.
l (t) Yes.
The original insportions verified that programs within the utilities were established to meet the requirements of 10CFR50.49 and the present inspection program ensures the. implementation;is.being maintained.
(v) It appears that EQ is not a focus at the. Regional-level.
j Are inspectors sufficiently trained?
3 (a) No.
4 l
(d) If the NRC wishes to re-examine the EQ area. an inspection cadre would have to be trained and the inspection plans redeIloed.
)
(q) I don't know what, if any. EQ tral'ing is provided.
n i
j (h) Inspectors. In general, are not very well trained in EQ.
lhere are very i
few i nspectors that I bcIlove are adequately trained and quallfled to j
inspect in this area.
There are probably more inspectors who could evaluate test re port s that could adequately determine if its installation j
and care and handling are appropriate.
In order 10 inspect any new plants j
that are hu111. additional training will bG noussary.
Ihis is to ensure 4
adequacy of the inspection efforts as well as obtain some degree of ronsistency.
(t) Most inspectors are not familiar with some of the unique knowledge part scular to EQ (i.e.. Arrhenius, LOCA profile requirements..etc).
(v) It appears that "new" Region based inspectnr$ that have [Q responsibilitles have not rerPiVed training, i
i I
-e-,
w
)
i C-47 What changes would you recommend?
Is the inspection plan ' focus, scope,
]
and depth adequate?
i (a) It appears to be haphazard now.
I l
(1) ! believe the regulations should be updated to reflect our current positions on E0 If E0 is going to be part of.the routine inspection J
program there should be some type of formal NRC training and a module developed for the inspection effort.
19 addition, counterpart meetings should be held to keep inspectors' abreast of EQ. issues in the industry..
I i
(t) All inspectors should be consistent in their: technical interpretations of 1
EQ adequacy.
For example, if a utility can demonstrate that a technical concern was adequately addressed in a prior inspection, the " rehashing,"
of the issue because of a technical difference in opinion between l
l Inspectors is not warranted in a rew' inspection but should be dealt with among the inspectors, i
(v) Additional training and including EQ as part of other inspection activities, i
l x;e. m :
I l
l
\\
f a
l l
l i
l i
i
-g--
w-
-ny, w
g,..y..,
w
?:
,? fc:l$f '
^
D.
Miscellaneous 1.
To what extent should maintenance and surveillance / condition monitoring be credited for demonstrating continued equipment qualification?.
j (a) To the extent that the aging mechanisms'and symptoms are well known, i
backed up by continued sampling and testing.
Need more [ heat rise?)
monitoring.
Effects of chemical contamination.(including self-sources like off gassing and ????.
The parameter most likely to be found inaccurate is the activation energy - then is the [ actual vs assumed) service condition.
(d) Region V does not recommend any change to E0 requirements for equipment in harsh (e.g. containment) environments, for equipment in mild EQ cnvironments, we think that a PM program to replace certain critical parts (e.g 0 rings) along with surveillance testing should be sufficient.
(e) Maintenance and surveillance / ondition. monitoring can only be used to detect and mitigate aging.
Iero credit should be given to demonstrating continued E0 E0 should only be demonstrated using LOCA testing on preaged samples.
(f) Condition monitoring shotild get a. lot of ' redit toward qualification.
c
]
(9) Condition monitoring is more of a dream than a reality. For much equipment, the normal environment degradation is small compared to accident effects, and the.small changes,that may occur in normal service are difficult to measure and not verysimportant.
The EPRI cabic condition monitor should be useful ~ to determine the specific JEAliqa of maximum cable degradation, so that we can ensure that the environment iLt_tho w locations is taken into account (h) I do not believe that routine maintenance and surveillance activities alone are adequate to determine or maintain the qualificalion of equipment.
(1) Routtne maintenance and surveillance can not be used to verify continued qualification of most equipment.
(k) I am not aware of any reliable and proven. condition monitoring or surveillance techniques that could be reliably used in a quantitative way to determine the remaining service life of qualified equipment.
(j) For components that will actually be used for long durations, condition monitoring with acceptance criteria related to accident capability will assure that continued use is conservative.
Of course such CM techniques may not be developable or implementable at a reasonable cost.
In that case, other alternatives should be considered.
Indenter and OIT senm promising.
Techniques for penetrations are desirable.
(1) The problem with condition monitoring as a means for demonstrating continued equipment qualification is that not enough testing has been
i 1
C-49 performed to establish which parameters should be trended to determine remaining service life.
The bulk of the qualification testing done to date only looked at insulation resistance values for electrical performance during the actual test, if the values exceeded a typical acceptance criteria of 1 meg ohm the specimen passed the test.
- However, this parameter alone may not ce enough to demonstrate continued equipment qualification.
I believe more testing'is'necessary to correlate measured l
test parameters with direct attributes of degradation.
(n) for equipment in mild environments, coutine calibration, maintenance, and surveillance testing should suffice.
The question becomes much more dif ficu t to answer for hersh environments. As.l. stated at.sve in A.6..
historial data and inclusion of the periodic, measure'nent of those characteristics critical to the proper operation of the equipment and the maintenar.ce activities require <, ',e mitigation should suffice for the rest (including surveillance, etc.).
(o) ! believe that maintenance and surveillance / condition monitoring should be credited to demonstrating continued EQ only if it effectively assures that the margin required for successful performance during accident conditions is maintained at all times during the life of the equipment.
As previously discussed, this maybe nearly impossible for some items of equipment where the required margin is not known.
(p) At this stage in the li'e of operating plants, maintenance is relatively more important than qualification. A key element,of this collection of activitios should be engineering evaluation of maintenance problems and f ailures coupled with feedback to increase tho' assurance of safety by warranted changes in equipment, service conditions. and operating conditions. We must avoid the type of incident'in which post evaluation reveals that prior events warning of a problem.had been ignored.
(q) Refer to [my) [O response.
(r) Yes, the monitoring of onvironments can be signiflCant in demonstrating excess conservatism in estimates of service conditions.
In many cases the basis assumption that the maximum normal ambient temperature MUST be the basis for calculating qualified life is unduly punitive.
110 14844 based radiation lives are equally excessive in conservatism.
This is particularly the cases where single nodo dosages are determined to envelop the entire containment building.
Equipment, such as the Westinghouse Lifetime monitor (this is not a commercial), has been designed specifically to monitor local temperature and radiation environments.
(s) Part of proper qialification is the specification of the maintenance and surveillance actions that are r:ecessary to support the established oualified life (see Section 9 of the EQ Reference Manual). Condition
[
I
e.
gy g -
-,s%
C-50 monitoring (which measures condition indicators that can be trended)l/
is normally not performed because (1) condition monitoring is easier said than done -- these is a scarcity of proven methods for in-situ measurements that quantify the remaining life of equipment (see for example " Condition Monitoring of Nuclear Plant Electrical Equipment," EPRI final Report NP-3357, february 1984 and many NPAR reports that address condition monitoring) and (2) traditional EQ methods rely almost exclusively on testing, analysis, and maintenance / surveillance, which are viewed as adequate without condition monitoring. Condition monitoring h used fairly extensively in plants when life-cycle decisions on expensive l
equipment like generators and large motors are being made or when a problem with any type of equipment is observed.
Because cables are so expensive to replace a. lot of attention has been given to their condition monitoring. EPRI-has held two industry workshops on cable condition monitoring' monitor 4(Toman.70.f' Cable Life Evaluatio.
and.has. developed,a commercialized y
technique, the indenter aging n
Services, Cable Indenter Aging Monitor " Ogden' Environmental and Energy Services Company, 1991). Although this technique can be used to troubleshoot cables that are suspected of being prematurely aged (e.g.
cable subjected inadvertently to periods with greater-than-usual temperatures), it may prove to be most beneficial as a last resort method (short of replacement) for justifying 20 years of life extension for some cables during license renewal (see response to A6). I also feel that more j
use should be made of the indenter in technical support of research, l
efforts, for example, it is unfortunate that the indenter was not used to characterize the aged condition of the Okonite cables tested in a recent aging and LOCA test program at Sandia.
This data would have gone a long way 9 comparing the aged condition of these cables to the condition of sir ar cables aged and tested in a previous Sandia cable program (which, by the way, were tested with the indenter - back in the days of William f armer, NRC Research, when we had the conenon sense to coordinate industry and NRC research programs).
(t) Maintenance and surveillance / condition monitoring are tools of a complete program to demonstrate equl> ment qualification._ Operating experience and trending are also part of tiese programti
- l (u) I believe that equipment inspection, maintenance, surveillance, and condition monitoring (if available) should be important elements of aging management.
Since accelerated aging simulations cannot address the of fccts of all significant aging mechanisms, there should be reliance on such on-going activities, for many EQ equipment applications, particularly outside containment equipment not experiencing significant aging or accident stresses, I believe these on-going aging management 3/
There is a distinction between the terms " surveillance" (go-no-go tests) and " condition monitoring" (trending). I recommend that NRC staff persons involved in the f0 Task Action Plan use the terminology in Nuclear Power Plant Common Aging Terminology, fPRI final Report TR 100844, Doromber
.99?. The NRC parttclpated in developing this.
- ~~~
C-Si techniques are preferred to reliance on preagjpg Almul&Llons.. for the-sake of brevity, I will reserve my' input;regarding such techniques until the November workshop.
(v) Non-intrusive types of monitoring needed to-be explored as a means to monitor equipment qualification.
There were many methods discussed a'. the NRC Aging Conference which may be useful for EQ.
Motor current signature and cable indenter methodologies are examples of non intrusive measures that could be used to demonstrate remaining life.
Are you aware of any surveillances or condition monitoring techniques that can be used to provide some assurance'of remaining service life? Do you have any specific recommendations in this regard?
(a) The EpR! Indenter looks promising.
(c) for cables, break point elongation and indenter may be useful in determining whether or not a cable is aging and thus becoming more brittle. It will not tell you whether a cable will pass an LOCA test. LOCA testing is the preferred method of demonstrating qualification.
(h) I am not aware of any monitoring methods that/could determine remaining service life, however, that does not mean that-some do not exist or could be developed in the future.
(j) Some of the indentation test methods for cabl[ monitoring appear to have a potential for producing useful results if indexed against samples of similar materials of known amount of degradation.
Such methods could be applied on a case by case basis.
(n) Techniques and equipment are available to measure the electrical characteristics of insulated wires and"cablesEsuch~that methods and requirements could be developed to allow periodic evaluation of the l
" health" of such components.
Some newly developed l&C equipment (primarily for Control Room use) have microprocesser-based self test ing systems that can be quite sophisticated.
These could be developed further to test certain critical characteristics that would be measures of
" health" and periodic.nlly (or on demand) report them to the operator.
Since microprocessors do not work too well in harsh environments. I visualize a dedicated " health monitor" remotely connected to critical iomponents of various transmitters motors. etc, constantly evaluating their condition.
Ihis could supplement operational surveillance testing should you determine it is necessary.
The technology is available.
(q) ! have worked extensively in this area and have imparted my ideas in the context of EpRl-sponsored Cable Condition Monitoring Workthops.
for lack of time, may i note only one item here: - I believe the current lack of concern for moisture transmission'through Jackets by their cracking or by transmitting of moisture into the cable core by diffusion has been wrongly ignored in that it may quickly compromise an adjacent connector or terminal equipment that was not designed to withstand cable-transmitted water.
Ibarefore Jacket condition monitoring can be important. 1 believe
w,. x,gpomgMWoi : Em l
l C 52 j
the only method of cable Jacket condition monitoring used today is visual or " feel" observations at points (usual terminal or junction boxes) accessed in the course of other equipment maintenance / surveillance activities.
Other methods of ascertaining the presence of mechanical integrity of Jackets are the use of moderate voltage electrical testing I
from cable shields to ground or pneumatic ~ testing. These would be applicable to many cable designs but by no means all. Use of the cable indenter test can reveal the Jacket (and by inference, the insulation) condition somewhat qualitztively to allow some life predictions for many cables. No sophisticated research is needed to evaluate and implement such tests: simply an impetus to overcome utilities' inertia and their dread of any new test that miqhi yield bad-news.
~
~
(t) [PRI nas several ongoing programs, such at the cable indentor, which show promise.
Predictin remaininglifeofqmponentsthroughdeterministic methodology is stil in its infancy A logical approach would be to (1) reproduce the accelerated aging which was part of an E0 test and (2) compare the installed equipment condition to the condition predicted by the accelerated aging.
As long as the installed equipment is not degraded to the tested degradation there is relative assurance of remaining service life.
However, it should be noted that any condition monitoring technique which requires removing or extensive handling of aged equipment may result in inducing further degradation not representative of the installed condition.
2.
Should credit be given for other initiatives such as the maintenance rule for establishing and monitoring maintaining equipment qualification? Do you have any specific recommendations in this regard?
(a) Yes. if E0 concerns are covered.
(d) Region V does not think the maintenance rule should be given crcdit in assessing E0 requirements for ifAr.ib environments..We think that credit can be given from the maintenance rule to relax EQ requirements for mild environments (e.g. auxiliary building).
(e) 1ho main rule does nothing for demonstrating qualtitcat ion and no (rodit should be given for it.
(f) The maintenance rule does not get into E0 at all.
(q) I have not worked with the maintenance rule. It *hould definitely address
[0, (h) Only if the maintenance rule requires the performance of maintenance activities to ensure the equipment remains in a qualt fled condition.
The maintenance requirements should be spelled out very clearly and.
if not strictly adhered in, enforcement action should be taken.
(i) No L______
l
.m
- 0. _, y&,:&:ffj.
gj@M"kf@MWhM V
(j) EQ already requires maintenance at'the end of qtiallfled life, If intervice f ailures occur on qualified equipment, action is needed to determine what was the cause.
I don't.believe..that'the Maintenance Rule-r will cause much change to the3[Q arena.' *.
(1) I am not familiar with the maintenance rule and how it will be implemented.
But as I stated earlier several EQ ltems may not have i
required maintenance.
So the maintenance rule Will not impact these compononts.
(n) Based on my previous statements, my answer is obvious
" Absolutely!"
(q) Cable and most connectlons being totally passive components in safety-related systems are related only very tangentially to the usual maintenance programs.
(r) Degradation of components could be detected by evaluation of surveillance data. Given sufficient attention to this purpose in maintenance program planning, significant benefit _s could be realized.
...,g g, ' a.
a (s) The maintenance rule can be glVon a lot ofscredit(for establishing and maintaining equipment qualification for equipment in a mild environment (in fact, the effectIvoness of such maintenance for ensuring that aging does not produce common mode failures is one of'the main reasons the NRC did not include mild environment equipment in the scope of 10 CfR 50.49),
for harsh environment equipment, the maintenance rule can be given credit for promoting the implementation of surveillanca and servicing that may be DEfnAr_Y for maintaining the quallfled status of equipment (although.such measures are necessary even without the maintenanC0 rule). The performance-based approach in the rule would' identify cases where harsh environment equipment experiences unacceptable failure rates durinq operatton. But it would not identify cases where aging degradatlon is not
- advanced ennuuh to impact the failure rate of the equipment during tha benign environments of plant operation, but may be advanced enough to com-promise the safety funct ion 6f the equipment under harsh-environment enmmon cause stressors in a design basis accident. It is this even'uality that is addressed by environmental qualification. Thus, the maintenance rule is not suff M eat to ensure the quellfled status of equipment in a harsh environment - a sound E0 program, including EQ maintenance, is sufficient.
(t) fes. this would be part of the overall program referr&d to above.
The industr.y programs for maintaining equipment, both periodic and preventative, are designed and tailored to mainta!n equipment in optimal cnndition for safety and operational reasons. Mair,taining equipment in optimal conditnon contributes to relative assurance that the equipment will perform its safety function if required in'a harsh environment.
Maintenance of EQ equipment is only part of an integrated maintenance program which enhances not onlrEQ btit safe and efficient operation of plants.
Utilities should be allowed to demonstrate the ability of thase programs to maintain (Q but [0 should nnt be driving maintenance.
&Yhh8[
?
^
d ',
,y c.$4 er 's, e ' t
,a d'
(u) In a word - Yes.
See response to question D.l.
(v) Yes. I think that to the extent that a maintenance activity is a detailed evaluation of a component (detailed piece part inspection, test etc.) the maintenance rule should allow a licensee to take credit for this data in evaluating equipment qualification.
1 3.
What other options or approaches to establishing and maintaining EQ requirements would you recommendt (e) Periodic DBE testing of components is the only way to assure that they will ibnction during and following a DBE.
A single test good for 40 years is a scary thing.
(f) Periodic testing is the only to know absolutely the status of quallfled equipment.
(h) I would recommend that a group be formed in HQ'that would develop an 1
inspectionprocedureand.periodicallydnspectfall<the-plantsinthe-countrytoensure~licensetarecomplfYngwiththeruleandtohave t
consistency.
(1) I recomend that the licensees continue.to' implement an IQ program that is periodically inspected.
(k) Should attempt to get more realistic assessment of equipment aging (cable insulation, splices. etc.) by obtaining and' testing specimens of in-plant I
equipment and comparing the results to accelerated aging predictions.
i This information should be used to reassess the existing EQ requirements.
J (1) As stated earlier. the EQ rule should be updated and maintained currant.
The E0 standards are being revised by the industry: howevar. NRC ha<. not adopted any of the later standards in Regulatory Guides.
(q) This subject 15 covered in othcr responses given.
(s) Since it is my judgment that current approaches for establishing and maintaining [0 in older and newer plants are sufficient. I dn not recommend any other optionh or approaches,.
(t) put less emphasis on " paper" and more emphasis' actual operating experience and condition of equipment.
(u) See responses to A.! and A.2.
As noted above I maintain that ge current approach of applying a consistent methodology to all equipment pithin tne scope of 50.49 is i ippropriate.
Rigorous qualification combined with inspections and agli.g management should be applied to the most 'tafety i
significant equipment located inside containment.
Little. if any. 50.49 i
type qualification need be. applied to radiation only harsh and outside containcent equipment exputed to short-time low temperature steam effects.
For other equipment, varying approaches could be applied.
I suspect an overall safety gain would result, if ex hting industry EQ resources were i
C-55 i
redirected and focused on a narrower equipment scope with more rigorously maintained and scrutinized equipment performance and qualification.
4.
Describe any spect fic EQ issues or topics 3.that you belleve deserve further research.
~ *
(f)
[mphasis should be placed on training workers and contractors to recognize f0 requirements, b en if they arentt working on [Q sysicms they can damage corponents and affect their qualification.
(p) I recommend that the NRC and industry cooperate to develop a program for the testing of equipment removed from plants after 20 or more years of service. Whenever a quastion arises about the adequacy of installed equipment, it appears that industry has invariably insisted that the problem could be addressed by analysit and hat objected to testing. A prime concern is that a test failure could have major consequences, even if there were evidence that the failure was caused by equipment handling and other factors not related to equipment capability. Consequently, it is essential that the proposed program be designed to eliminate the risk of the feared consequences. The objective should be to obtain ininemation nn the effects of real aging on installed equipment and thereby overcome some of the deficiencies of the traditional methods of evaluat ing aging effects. T he NR(. '. Nuclear Plant Aging Research Program has done some of thi5 but Iittle of it has involved equipment in current use in operatipg plants. 10 reduce the overall cost of such-an undertaking, the program could be limited to equipment considered to have the greatest impact on safety, and each utility (or group of utilities) would be. assigned pae item to test.
This suggestion is undoubtedly controversial'. and great ef fort wo.sid be required for its successful planning ' however, ~l have long felt t hat there is no substitute for testing equipment that has seen long service in an operating plant. The adversarial atmosphere that of ten exists t.ctween the NRC and industry would have to be overcome in the common interest of public safety.
THf FRENCH APPROACH in the extent that I am familiar with it. I think the basic element
- nf the f rench approach to [Q. are a relatively standardi;*d age-condit toning coupled with an ef fort to develop condition monitoring methods capable of revealln] when equipment is approaching the point where it will no longer im able to perform its safety functinn. While standardized ageJconditioning eliminates the variableness in approaches by different manufacturers and utilities, it does not account for the different aging rates of different materials'and parts. However, the weakest point in the french approach is the expectation that predictive condition monitoring techniques can be developed in time to be used effectively. At present. I am not aware of any viable predictive condition monitoring methods; and the rate at which their development is taking place is noi very promising.
a I
C 56 (a) Some of the concernt of the writer have been those expressed in two publications known to NRC staff (Ref, below). They deal dominantly with i
harsh wet environment effects that have rarely, if ever, been part of cable system (Q programs, [ramples are: interfare effert s between component s tested separately; use of a single test for widely different designs, vintages, and manulArturerst marhanical Installation stresses on i
cable (sharp bonds, vertical supports, sea) compression deformation);
continuous submers ton prf or io harsh exposure; and combined momentary I
eiertrical ef fer t s from postulateri initial peak tempe dure and radiatlon j
stresses these are concerns that are applic.',ble'to all plants J
old, lhe degree of Concern for such: ltemizadnoted(would very young or m
substant tally,from station to station depending upon.the specific cable 1
designs, component designs, installation practices and EQ program of the i
statton. Although these generic issues have been raisert in the past, I I
unde r s t anti that there are no outttanding rahle generic iss si on the NRC 1,
agenda. What.t} the proper venue to address these concern J 4
(s) littllties must 4
i ensure that plant-specific environments or maintenance prac t tres do not neoatively affect the quallfled status of their I
in this regard I take this opportunity to reinforce preVinusly equipment l
S uurd utdante regarding the quallfled st&Lul of.polyamide insulated wire (virtua ly all of it in nuclear plants (1 Kapton, trade mark of DuPont)
]
used for electrical teads for containment penetrations and other i
safety-related electrical coutpment. The qualified status of such leads is i
maintained only if they are (l) compiian(e witn manufarturers' installed and maintained with met iculous inttruct tons to avoid sharp bends ant!
l inadvertent damage, (?) kept free of wetting for long periods during plant i
operation to m ;imi/e thr agtng mechanism of hydrolysis in tight bands.
and (3) prot ec ted dor tnq an acrident from direct spray impingement more
. igg re s s i ve than the spray conditions for which it was qualified by test.
Also, handling of "Kapton" insulated wire after exposureito a radiation' i
dose of more than 1 Mrad should be minimlied due cto.the. potential failurn of degraded "inflon" adhesive, which holds the spiral-wrapped polyamide i
tape insulation in place problems due to handling damage of "Kapton* wire I
led the Nuclear Regulatory Commission to issue an Information Notice (No i
88 89). EPRI published the report " Review of Polyamide Insulated Wire in Huclear Power Plants," final Report NP 7189, february 1991, to provide utilities with guidance on practices that will maximize its reliability 1
i under operating and accident cnnditions. The report states that "If plant
( ondit ions approach levels identified in this report. it would be prudent i
to review the qualification file and as bnilt configurations related to i
i the t echn ic al issues rovered here."
j fv) There are sa fet y relat ed component s locat ed in a " mild environment'* whirb experience a " harsh" environment due to their operattnq condition.
1hese componerts (i.e., continunusly energlied solenoid valve) should be j
ovaluated for a cualified life i
NRL ininrmat ton Not tres should contain more spet ific information regarding Ihe identifind prnblem.
the addrened by each licensee thatidentifled problem or concerns should be l
is affectori.
{
\\
y,.pge.w. o m
,,72 9
4
.my 3
.C.$7' (w) P.onded Jacket cables - Based on the fact that 11C cables with thin jackets
(<a5 mils) were not specifically qualification tested, are these Okonite cables quallfled today? Should they be qualification tested? What do we do to resolve overall bonded Jacket. cable,it' sue?
5 Do you have any additional comments or observations relative to the adequacy of E0 or EQ program requirements at commercial nuclear power plants?
(e) lhe industry hai Adopted A italus qu0 attitude on fQ and it's !Ime for !he NRC to shake the tree.
[Q is not a one time issue that gets resolved for 40 to 60 years.
Periodic verification of component qualified life is warranted (h) i sometimos get the 'eeling that llcensees, and sometimes NRC personnel, think in - the IQ lisue is dead.
(k) Have no direct knowledge but suspect that overall emphasis on IQ has slynificantly diminished since tie completion of the NRC [Q inspections, (m) I don't thtnk the approach of e1iablithing qualifled Iife Wiih replacement equipment is properly handled. 'The design changes that go into a later model of a product may make the qualification much worse and sometimes
- beller, The licensee is going to refer to the spec., and the salesman will rely on his catalogue for supplying the.aqulyalent product, I think it is time we revisited this issue to look at the effectiveness of the licensee's [Q program.
(t ) Prescript ive regulatory approachet, for romples issues, such as aging, are typically ineffecttve.
The result is to " meet the rule" and not to proactively " engineer" resolutions and enhancements, Utilities should be allowerl t o " engineer" t hrough the aging, and other f0 is sues, in concert with the NR(. to insure compliance with regulatory requirements are met and sa f e, ef f icient operat ions of commercial nuclear power plants, (w)
With regard', to the Okontte aging methodology - 15 it acceptable?
If we determine that condit ion monitoring, in situ testing, or some other test method (IST?) is acceptable, does it matter?.How conservative is arcelerated aging versus natural aging?
If we determine that condition monitoring, in situ testing, or some other test method (!$i?) is required to assure that qualified equipment is still i
good, is'that sufficient for license renewal or do utilities still have to prove qualification for the additional renewal period, in other words do they have to do additional qualification testing to prove equipment is qualified for 60 years.
a
' < y :)
y,
~4
, p.;
.m.
t.gg, o
6.
Do you know of any literature that may.be helpful in addressing this issue, such as published reports,.studiosi. articles, etc.?
(d) IPRI has published a mandaI (Nuclear Power. Plant Equipment Qualification Peference Manual) and the survey may benefit' ffom-inquiring whf ther Itcensees consider the manual to provide proper guidance, whether 1teensees are using the manual and whether it would be useful for the NRC to recognize the manual vla-a Reg. Guide, it looks Itko there is an industry standard on IQ which the NRC should address in some way.
t (e) 5 Aggarwal has a compendium of E0 information.
(j; IFRl Cable (.ondit ion Mnnitoring Workshop, f ebruary 1993. (Papers)
(n) The l&C i[G report on IMl 2 might'be of' interest since it addresseA Instrument failures in accident conditions.
If you can find and contact 5.[. lonpar in DOI Headquarters or R.D. Melninger at fG&G Idaho. Inc.. you may be able to get a copy.
I have a dral.t but loaned my final version to l
someone and did not get it back.
l've f.een away froih l0 for about th'ee years so was not able to answer all r
your questions, but I hope this hel It' appears from the questions you are headed in the right directlon. ps. Good. fortune.
(p) I have enclosed a copy of a' paper l' prepared'for an NRC workshop on plant
~
aging in 1982. which I believe still has pertinent information.
(q) Refetcocc5 expressing the writer's concerns are; NUREG/CR4DI "Rosidual Ltfe Assessment of Major ligh'. Water Reactor Components Overvi v Vol.
- 2. Chap. 13 Cables & Connett'ons in Containment (Nov. 1989). and Proc:
1993 [PRI Wnrkshop nn Power Plant Cable Condition Monitoring pp. 2 1 thru 2 10 " Cable Condition Monitoring - Ihe Challen proper f0 perspective it, given in IEEE Trans. ge Before Us." An example of on Nuclear Science. Aug.
'93. p. 1263 " Role of Iraining in Maintaining Equipment Qualification" by
$lii.or i Kasturi (first two paragraphs).
(s) yns. I know lit erat ure that may be helpful in addrost.ing this iuue, lhore are over 500 reference *. cited in [pRI's [quipment-Qualif ttation Referent.e Manual, lhe stated ob.lectives of the manual are:
Consolidate and preserve the substantial, but scattered, body of-equipment qualification technology developed in the past decados Sustain untform good practice of equipment qualification during long-term plant operation facilitate the training of future equipment qualificatior, engineers and other plant personnel f or Whom equipment quallfication is important, but 1
may not theli ortmary job responsibility Support utills irograms for maintaining the qualified status of existing equipn, u and qualifying replacement equipment. In turn minimittnq forced ni, mes
. [nhance and promote ti.
- "mic benef its of equipmrint life exten:.lon and plant lic.cnse renewal
.r
._x
.~
e,
.A 2
i C-59 Pntnt not areas in which equipment qualification technology is still deveinping and may be further Improved ihe entire manual is pert inent to the questions asked by.this survey.
Section !? " Qualification fxperience' is especially pertinent. It is intendad to
- bring to the user's attention factuel Information regarding the performance of certain equipment types in past tests or plant service,"
~
~
j the manual is an example of the:many Wayt'.thatVth4 Utility' Industry j
rnnstantly promotes good practicos and learnt from experiences in (0, As i
problems are identified, they are addressed generically or within individual utility E0 programs. I am confident ~that any valid concerns l
identified by NRC's reassessment of EQ program requirements will be
{
addressed by the industry. 1 am 'eqtlally confident that the reassessment will not identify any valhl faute for modifying e:Isting time tested [Q j
methads and reoutrements.
i j
(u)
The literature associated with qualification testing and aging 4
simulations for nuclear pnwer plant and other related information li vast and ion extensive to be provided here.
Suggested reading are l
NUR[b/CR-a301. Statu% Report on Iquipment Qualification issues Research and Resolution, and the fPRI E0 Reference Manual.
During the par.t few i
years it has bnen my expertence that many in the NRC Staff involved wi'h l
(0, some viewed 4s 10 knowledgeabini are not conversant with the 1
las iqht s prov tried b*,
't of the mnre arressible literature snurre'.
(nulside the appltrable standards and regulations).
I have wondered if i
the (0 TAP and the significant Staff and industry resources that will be consumed by the program have been catalyted byfthis lack of insight and i
historical perspective.
l i
i i
L i
j l
t q
yJ s
.I i
.-