ML20235T331
| ML20235T331 | |
| Person / Time | |
|---|---|
| Site: | Humboldt Bay |
| Issue date: | 02/23/1989 |
| From: | Shiffer J PACIFIC GAS & ELECTRIC CO. |
| To: | Chilk S NRC OFFICE OF THE SECRETARY (SECY) |
| References | |
| FRN-53FR47822, RULE-PR-50 53FR47822-00022, 53FR47822-22, NUDOCS 8903080243 | |
| Download: ML20235T331 (113) | |
Text
_ .. .
Pacific Gas and Electile Company 77 Ees e S"ce: JamesD Siar San Pa~ ::: CA 9406 %ce Pres:::em "59727:3 fbclear Pew Gre'a:.on
... e, g' I{- DOCKET UUMBERi ,
PRO. POSED RULE ;:
4 - W' L -
(Q M l/Nh '89 FEB 27 P2 :08 February-23, 1989- h.- ,
Mr. Samuel J. Chilk Secretary U.S Nuclear Regulatory Commission Washington, DC 20555
Dear Mr. Chilk:
ATTENTION: Docketing and Service Branch RE: Proposed Rule 10 CFR Part 50 Ensuring the Effectiveness of Maintenance Programs for Nuclear Power Plants 53 FR 47822 (November 28, 1988)
REQUEST FOR COMMENTS This document is provided in response to the Nuclear Regulatory Commission's (NRC) request for comments relative to the referenced proposed Maintenance Rule. In addition to the following general comments, Attachment A provides more detailed observations.
PG&E is opposed to the Maintenance Rule. Our opposition to the rule is based primarily on the following points:
The NRC-has failed to demonstrate that there is an industry maintenance problem and, therefore, lacks justification for the rule.
The rule addresses maintenance processes rather than the results.
The extension of rule based maintenance requirements to Balance of Plant (BOP) will inevitably lead to a dilution of attention
~
paid to high safety priority work.
The cost of the rule will significantly outweigh any marginal benefits.
In particular:
- 1) JUSTIFICATION FOR RULE The NRC contends that there is a lack of effective maintenance at nuclear plants. This conclusion is based primarily on NUREG 1212; a now obsolete 1985 study of the state of nuclear industry maintenance.
8903000243 890223 PDR
$[53h47022
y - - - _ _ _
Mr. Samuel J. Chilk y February 23, 1989, Page 2 In part due to NUREG 1212, the industry as wel_1 as the NRC became sensitized to maintenance issues in the 1984/1985--
timeframe. :Since that time there have been numerous industry initiatives to help improve maintenance including:-
INPO Maintenance Guidelines (85-038) issued in 1985 provided basic program standards Self evaluations against the INP0 standards have.been performed by all plants INPO Maintenance Assistance Review Teams (MART) have been targeted at maintenance outliers.
INP0 Maintenance Indicators have been developed to. track performance .
EPRI Nuclear Maintenance Assistance Center was established in 1988 to help address significant generic industry maintenance issues During the'same period basic measures of plant safety performance such as scram rate, ESF actuations and safety system availability have improved significantly. More directly related to maintenance, SALP scores have also improved. In 1984, there were 27 plants (~26%) with Maintenance /
l Surveillance SALP 3's. In 1987'there were only 7 (~6%) such plants and we anticipate the number will.be lower for 1988.
Since a SALP 1-or 2 by the NRC's'own admission indicates an effective program, the problem. is clearly not a widespread industry problem but one of a small number of outliers. It-would be counterproductive to develop a minimum standard which could potentially lower the level of performance for the entire industry when only a handful of plants are experiencing
. problems.
Because of the intense industry attention to maintenance'and
~
. the continuing favorable trend in safety performance since 1985, the need and justification for this rule is clearly questionable ~ .
- 2) PROGRAMMATIC VS RESULTS BASED RULEMAKING The rule implies that by requiring all utilities to implement a prescriptive standard for plant maintenance processes, the NRC's basic objectives for plant performance will improve to an acceptable level.
The NRC has been unable to develop any quantifiable correlation between the comprehensiveness and formality of maintenance programs (which the rule would create) and the safety results it ultimately wishes to accomplish: reduction in challenges to safety systems and increased safety system reliability.
1 Mr. Samuel J. Chilk February 20, 1989 Page 3 l
In addition, the NRC has not defined what constitutes sufficient performance. The industry will~ continue to be ,
ratcheted on programs until such time as chere are objective l plant performance standards relative to maintenance, e.g.
quantification of an acceptable level of risk.
- 3) BALANCE OF PLANT (BOP)
The rule will require the maintenance program to be applied to all balance of plant components within the protected area.
It is our opinion that extension of NRC regulation to BOP will dilute resources which would more profitably be spent on critical safety systems and components.
Further, once NRC regulation of non-safety related equipment is established, it will likely set the precedent for a continuous upgrading of requirements for BOP and additional dilution of limited resources.
Scope of BOP - While the NRC currently states that the scope will be limited to equipment within the protected area (PA),
reliability analysis shows that there are considerable challenges to plant safety features originating from outside the PA (offsite power is a primary example.) There may well be pressure to address maintenance activities outside the PA even in areas that may have already been analyzed such as Station Blackout.
Quality Activities Relative to B0P - At this point the rule's primary emphasis is on a " systematic evaluation" of BOP systems to assure understanding of their relationship to safety function and that appropriate maintenance is applied.
However, once we set the precedent that plants are not designed to accommodate infrequent BOP challenges, the force of logic may ultimately require an extension of much of the QA program to BOP including nonconformance, procurement controls, QC, configuration management, records management, ,
etc. l
- 4) BACKFIT RULE The NRC states that the requirement under 10CFR50.109 (Backfit i Rule) to analyze the cost / benefit a.ssociated witn the maintenance rule is not required based on their perception that the current state of maintenance at nuclear power plants does l not provide for adequate public protection. This is clearly not the case. As evidenced by SALP trends, maintenance is <
better now than it was 4 years ago when the NRC maintenance study NUREG 1212 was completed. The results of that study did not lead the NRC to shut down any plants based on inadequate l maintenance.
f !
p -- - - . _ _ _ _ _ _ _ _ . _ - _ - _ _ _ _ _ - _ _ _
i U
Mr. Samuel'J. Chilk February'23, 1989 Page 4.
While_ recognizing ~the importance.of effective maintenance in protection ofithe public health and: safety, PG&E feels that the ..
proposed Maintenance Rule is not in the best' interest'of the public.or industry and, at' worst, may result in a degradation' of.already good maintenance performance.
?
. JDS:whc Attachment-I I
t
}-
m _. _ . _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _
- y. ,,
I~ i.- .
[*, .:e .
}
i r
I 1
/
. ATT A C HM E NT ' 'A L-
~\',.
This attachment contains-2 parts. The first is. a:-
.. reproduction'of.the proposed rule and.. regulatory analysis with-indices for each paragraph. -The-second part contains detailed PG&E coments-referenced to the indices of part~1. .
I e
'l 5 3
1 i
f- .
[7590-01). i M NUCLEAR REGL'LAICRY CCMM15510N 10 CTR Part 50 Ensuring the Effectiveness of Maintenance Programs for Nuclear Power Plants A~INCY: Nuclear Re;d aicry C:r-issi:n. -
AC~"N: "rc;;ses Rule.
' h SUMMARn ~ne C:mmissien is pre;: sir; :: ame c its regulations to require c:m ercial N: lear ;: er piant li:ensees to strengthen their maintenance -
a:tivities in er:er to recu:e the likelitece of failures and events caused-by t e is:( cf effe:ti,e mainte.ance. The C:mmission telieves safety can:
an: rust te ennantec ty cefining an a:e:;uate maintenance program to i ensure tre effectiveness of sucn ;r: grams througneut the nuclear industry.
The ;r:::se: e.le requires clant -aintt%nce programs to include specific activities, including t .e eenitoring of the ef festiveness of plant maintena pregrams.
h CATE: F::- e,t :eries exotrei aanuary 27, tseil Comments received after t
this cate will te censi:erec if it is practical to do so, but assura9ce of consideration cannot be given except as to comments . received on or before this date,
~
Mail written comments to: Secretary, U.S. Nuclear Regulatory --
@ ADDRESSES:
Commission, Washington, D. I. 20555 Attention: Docketing and service Branch.
Deliver comments to: 11155 Rockvile Pike, Rockville, MD.
between 7:30 am and 4:15 pm weekdays.
i ...
l l .'
l p
-E__-.____
-_a _ ._.I-._ -- _ - - _-_ - . _ _ _ _ ___--__ _ - - - . - _ _ - - - _ _ - - - - - . - - _ _ _ _ . _ _ - - _ _ _ _ _ _ - - _ . - - _ . _ - - - - - - _ _ . - _
- y. -
} -
l n'
< l t
it
! i o
- c. '
-h. Cocies cf the caper on rulemasing options, transcript and proceedings -
of.the Public WorksheD, draft NUREG repp t,-draft regulatory analysis, i
!4 environmental assessment and finding of no significant impact, the support statement sLbmittes to OM8, and comments received say be examined at: -the-hRC Public Document Room, 2120 L Street, Lower Level, N.Wi,' Washington, D.C.
- e FOR FURTHER INFC M f!CN CONTACT: Moni Dey,' Of fice cf Fuelear. Regukatory Research, U.S. Nuclear Regulatory Commission, Washie; ton,'O.C. 120555,,
(301) 492 3730, 5;P7tEMI C Rf.:NFC W II",N:
SACK 3ROUND Cn Maren 23, 1955, the Ccemission puolished a final Policy Statement en Mainte a .ce cf Nsclear Po.er Plants. -In tne Policy Statement, the -
Comission stated t*at it ex:ected to publish a Acti:e of. Proposed Rule aking in tne rear future ame ;roviced the general framework for. the proposeo rulei g The Cermissien has a program to continually evaluate the operational ~
performance of nuclear power plants. Analysis of operational events has ,
shown that, in some cases, nuclear power plant equi; ment is not being maintained at a level to ensure that the equipment will perform,' with a high degree'of reliability, its intended function when required. A' limited NRC examination of nuclear power plant maintenance programs has found a wide variation in the ef festiveness of these programs. At some plants, maintenance has been a significant contributor to plant reliability, problems and hence, is of safety concern. The Commission believes safety can and must be enhanced by strengthening the ef festiveness of maintenance .
programs throughout the nuclear industry and this is the objective of this proposed rule.
q 34 -
y vn
,g N i i
- 3 I ql ' '
~ . ,
+]
- P
- 1
.DESCRIPT!CN
~
- h. ItisthecbjectiveoftheCommissionthatallcomconents. systems
[nd structures of nuclear. power plants'be effectively maintained so.that-plant equipment will perform its intended function when required.c The -
scope of the proposed rufe is b'med to cover all <$ms, structures -
ang,,1ce:cnentc;inc1ccine th1se in__the Ra1ance of plaet (BOP). To acccmplish this cojective, the preposed rule would require each commercial nuclear power plant to develop and implement a well-cefined program to-assure that maintenance activities are cenducted to preserve or restore.
-ith prer;t repair, tre availability,' performance and reliability of plant'
. structures, s>ste 5 anc cce: nents. The program shculd clearly define'-
the cc :: ents and act.ivities included..as well as the management systems used te c:.tr:1 th:se activities. Further, the program should include -
IL feectack cf specific results to ensure corrective actions,-provisions for overall pre; ram evalwation, and the identification of possiDie component' cr system cesign prc:lems. Ccmpliance with the rule would be verified by-htC 4., cit and H.spection.
.The proposed rule d:es not require that licensees report Maintenance .
FerfereanceIndicators(MPIs). However, each licensee would be required to have his own system for tenitoring maintenance effectiveness which would te subject to Nh. review.] The Cen. mission solicits cor..ents .on the. ' -
acclimation'and usefulness of MPIs as part of the rule..and whether a set 66PONSE of RP!s exists which'could ind'este the effectiveness of plant DND maintenance programs. In addition, t$e Commission solicits feedback on whether to require reporting a specific set of MP!s to the NRC as part of-the rule.
PUBLIC WORK $ HOP
@ The Commission held a Public Workshop on July 11-13, 1988 in-Washington, D.C. to solicit early input for the formulation of the rule from the public and regulated industry. Prior to the Workshop, a paper on rulemaking options was distributed to interested parties 'to facilitate '
g
,1 A
r i
1 J
Verkshop discussions._ The paper on rulemakbg' optionsI , and the transcript and proces'aings (NUREG/CP - 0099)2,f gn, Workshop are 4
available for inspection in the hRC Public Document Room ' 2120 L Street.;
Lower Level, N W. , Washington, D.C.
i As a result of Work' shop discussions, the Commission has'ccme to the y following cenc165' ions:
Q 1. Rulemaking should enceurage industry initiativ'es cirected toware irproving maintenance,,since such initiatives promote industry responsit'ility for preolem identification and resolution;
.g 2. . Prescriptive rule aking eptions cay' impece industry initiatives ar.c respppsibility t: improve maintenance; and
- 3. Rwle-aking snould te directed teward specifying tte htC's?
I expectations-.in maintenance =and reoviring licensee menitoring of:
tre ef fecti.emess of maintenance prograts.
@ Therefore, the C: missica propeses a maintenance rule which gives incenthe fer industry to develep a standare for a maintenance program,' wh
, NRC may enc:rse in a Regulatcry Guide.
' ~
J.
- . h Memorandum from victor-Stello,. Jr. , Executive Director for Operations, to the Commissioners, Preposed Rulemaking for the Maintenance of Nuclear Power Plants," dated June 27,1988.
@ 2 '
Copies of NUREG series reports ma be purchased through the U.S.
- CovernmentPrintingOfficebycalfing(202)275-2060 or by writing to the U.S. Government Printing Of fice, P.O. Box 37082. Washington, D.C.
20013 7082. Copies may'also be purchased from the National Technical:
Information Service, U.S. Capartment of Commerce, 5285 Port Royal-i' . .
Road, Springfield, VA 22161. A ccpy is available for inspection .
or copying for a fee in the NRC Public Document Room, 2120 L Street, Lower Level, N.W., Washington, D.C.
a.
~
)
(:
EXPECTATIONS FOR MAlhiENANCE STANDIRD The Ccmmission encourages industry to develop a Maintenance
- Standard-.
which will provide guidance for complying with requirements of the propos rule. The Commission believes that the development of a standard will gu current industry initiNives towarcs developing and implement {ng'acceptabL l maintenance prograts, and that utility participation in preparation ~of a Maintenance Stancarc will provide aeditional ir.centive and responsibility i :r:virg plant aittenance pregrams.
h The C: sissica :lans to cevel p a Regulatcry Guide to provide guidant fer c:- 6 ia; with tte rule if incastry c:es ret develop an a:e uate stant P..ever, tte (cm-issien prefers to enc:rse an'ind.stry devele;ed standard, To eeet tr.e Ccm:nissien's plans fer imple entati:n of the rule the industr3 c: mittent to develes a Maintena9ce Starcard should te made r.:w and a fini .
star.ca*: st:uld te ;res: sea no later inan Se:tet:er 1,1989. The Commiss' expects 10 0 0115n a Regulatory Guide endersing & Maintenance Standard or pr:.idi ; VC guidarce in N:ve-ter 1999. The c:merekensive program-re: wire erts of tre preposec rule would te re:uired to be fully implemente -
within t. years fell: wing publication of the final Rule.
h The ;r:::ses rule cefines th:se attributes the C:mmission c:nsicers necess for.an acce; table Maintenance Program. To te acce; table, any standard develeped to imple".ent the rule should have the foll: wing characteristics:
h
- Should define the plant ,ystems, structures and components included in the maintenance program (the scope of the rule covers all systems, structures and components including those in the 60P); *
- Should require a systematic evaluation (" systems approach") of ;
g g thefunctionsandobjectivesofplantsystems,componentsand
'g Ilt structures to determine maintenance activities and requirements;
[ m.bouer.g pp t-r g '
should provide citar and specific programmatic requirements that QD l
_ _ _ _ _ - _ _ _ - - - - - _ _ - _ _ _ _ _ - - - - - - - - _ - - - - - _ - - - - _ - - _ - - - - - - - - - - - - - - - - - _ - - - - - - - - - - - - - - - - - - - - - -- - - -- - - - -d
-y----_----.--_-------._-----
L
[2 ,
v l(_'
can te practically implemented to achieve high' reliability; h) '
Shculd be'ccr.orehensive.in addressing the activities and-functions included in the proposed rule plus provisions for self-assessment; Should reference stancards or guidelines such as those developed' by 'ANS, ASME. IEEE, ASTM, INPO, er EFRI where practical' toi
, provide (a) s;ecific gregra matic requirements cr (b) guidance for maintenance of s;ecific types.cf equip ent; .
h
5'c !d alic flexibility for ade: den cf rew innovative technolog as trey are ,alidated; anc' h
- She.!d previce fer sufficient deccentaticn 50 that program-ef festiveness and cc pliance with requirements of. the standard can te eve 16ated.
~
' @- ' The (c--issica nas conducted studies to review effective maintenance,
- a;;rcaches and practices in other countries and industries and has documented the findirgs in a draf t NLREG series report.3 The Commission seeks com ents on the craf t NURIC report which may be submitted to the hRC as indicated uncer the ADORE!!ES teacing.
3 A free single copy of draft NUREG 1333, " Maintenance Approaches and Practices in Selected Foreign Nuclear Power Programs and Other U.S.
Industries: Review and Lessons Learned," to the extent of supply, may b $
obtained by writing to the Distribution Services Section Document - y control Branch, Olvision of Information and Support Services, U.S. Nuc1 l Regulatory Commission, Washington, D.C. 20555. A copy is also availabl
. for inspection or copying for a fee in the NRC Public Document Room, 211 L Street, Lower Level, W.W., Washingten, D.C.
y i
_- - - _ - -_ _- -- - - - = - - - -
- y. . . - g a -
i t
l_ _
i it, 1 ,
g :. i n Y .
l
'j jp "
1 The craf t N'JREG report cencludes that the fc11c.ing~are practices in '
foreign c:untries and other U.5,' industrial maintenance programs which:-
have been found to c:ntribute significantly to effective maintenance: .
'h
- 1. Focuson'longtermmaintenanceobjectives;establishaproactive
' maintenance program as c; posed to reactive maintenance; h' 2. Use of a reliability centered' approach to maintenanc'e,lincluding censiderati:n of tBe can -achine interface; --
, h 3. C:llecticn and e g neering evah.atien of f ailure cata (root ~
c asse- oraly s i s);
-h. 4 Use f 47 i te;ratec infermati:n cystem f:r collecting data and-e: nit: ring the e' festiveness cf a maintenance prograi; h 5. Use of maintenance tecnnician training / certification pr arams;- <
1 h 6. Cerive:la.ningancscred.;lingf*cmeverallprogramobjectives;=
h 7. Enhance envireneentie:tivatien of maintenance technicians (e.g.
thru cross training, " Crew Chief". c ncept); and
~$
~
E. Clearly define interf aces between maintenance and other activities (engineering supp:rt, cperations QA, QC,' corporate offices, safety review).
h The Commission encourages in the development of.any Maintenance Standard consideration of the findings in the above referenced draft NUREG report and any other pertinent studies on effective maintenance practices, h The Ccmmission also encourages the industry to consider incorporating appropriate maintenance' derived from plant aging studies in the Maintenance Standard.
Iis e
_ _ _ _ _ _ . _ _-..__m__ _-- - - _ -- - - - - - - - -
p
[
ff n.
l l:
( .
i.
m D The C: emission views maintenance rulemaking as an opportunity to L encourage (a) good maintenance practices, (b) the aception of common l
maintenance standarcs, and (c)'the development of valid quantitative measures of ef festiveness which could become the basis for regulation by
- - outecees rather than pro, cesses. The herein proposed rule addresses (a) and(b)butnot(c). follow-on rulemaking is envisioned which would build.
en the f.ases established in this rule and would defir.e those validated tasurable cuantities or indicaters that could credibly beccme the basis-fer regulatery attenti:n or acticn in the future instead of the systems f:r maintenance uncer (a) and (t) of this pre;: sed rule.
As note: at:.e, an inte;N1 ; art of a g::d raintenaece program is the monit: ring anc feed:a:k of results. Programs sheuld utilize quantitative reasures to monit:r and adjust the maintenance program activities. Peasures that are based u;cn actual c: ::nent reliability and f ail.re hist:ry previde useful indication of esir.tenance effective ess. Such reasures need a well structured and c:e;cnent c'ientes syste : (e.g., the Oclear Plant Reliability Cata System (NPR05))
to ca;ture a c track e:wi; ent history data. The C:mmission notes and ere:gra;es the use of the industry wide hPR05 for this purpose in view of tre culti;!e uses f:r the data. '
CERTIFICATION OF PLANT PJLINTENANCE FR00 RAMS
@ The Commission wishes to encourage industry initiatives and responsibility for problem identification and resolution. Therefore, as a further way of encouraging industry participation and responsibility, the Commission is willing to consider a process of utilizing a designated third party, similar to the ASME code system, to review and certify licensee maintenance programs for conformarge with the Maintenance Standard. This process, if successful, could alleviate the need for detailed NRC inspection of all licensee maintenance programs. The Commission solicits proposals for conducting such a certification process.
CCFFENTS REQUESTED In aeditfor. to comments on the content of this Notice of Pi. <osed Rulemak'ng..the C:mmission specifically solicits input on the following:
Q 1. !s it appropriate for the nuclear power industry to develop a Maintenance StaMard and, if so, would the industry develop such a Maintenance Stancare?
J h 2. What lesel cf detail should te included in the Maintenance S tar.dar:?
{, 3. Is two years a reascrable tire to develep, and im;1ement a stan: arc?
Q 's it a::r:;riate for a desig.atea thira p4rty to certify plant
~
4 maicterae:e programs to c:rply with the Maintenance Standard; and, if so, would an crganizati:n te willing to perform such certification?
[O '
- 5. The Ccemission plans to issue by Novem:er 1989, a Regulatory :
Guide estaD1ishing stancards a*c criteria for determining what c:nstitates an effective maintenance program. This Regulatory Guide'is being develeped in parallel with the final rulemaking. j The Commission encourages the industry to develop standards and acceptance criteria. If an acceptable industry standard is available in.this time frame, the Comission will consider endorsing the industry standard in the Regulatory ;
Guide. An industry comitment to develop a maintenance standard, consistent with the Comission's schedule to issue a final Regulatory Guide by Never.ber 1989, would be necessary during this public comment period.
(D
w.
1 h 6. The C:tmission believes that the pre;osed maintenance rule shoulc te censidered unoer 10 CFR 50.109 (a) (4) cf the backfit rule wnich would exempt the maintenance rule from backfit recuiretents based on the precepts that effective maintenance is.
~
necessary to assure adecuate public protection and that the preposed rule codifies and standardizes previously existing Cem.eission requirements, t:th explicit and implicit, in plant technical specifications, licensee safety analysis reports, and 10 CFR 50 Appencix B. The C:mmission recuests public comment c:n:erning the reed f:r a tackfit analysis for this rulemaking.
Q 7. **e C: missi:n :elieves that the inclusien of balance of plant (50P) e;uiprent in the pre;esed maintenance rule is necessary aa: :r:;er. Pc ever, the Cc=ission also reccgnizes that some licensee maintenance pecgrams, as presently configured, apply to
.struct.res, systems, and compenents that are, without question, irrete, ant 1: prctection of public health'and safety from ra:' hgical hazards ass:ciated with the c:eration of the r u: lear ;:.er plant. The Ccmission requests public comment-c:ncerning what limitaticn, if any, shculd be placed on the
final maintenance rule te previde some licensee flexibility in this regard.
,3 8. The C:? mission telieves that individual worker accountability.
k plays an important role in an effective maintenance program.
The Comnission is, therefore, soliciting ccmments on the means to '
incorporating this consideration into a licensee's maintenance program. ,
@ 9. The Commission desires to establish criteria within the UfA maintenance rule which would form the basis for determining when N a maintenance program is fully effective and additional improvement i_s_not warranted frem a safety standpoint. -Such ;
criteria might be either quantitative or qualitative and could i 5
o Le baseo en specific measurable attributes, on overall plant performance, on program results, or on otter attributes. The a
- Commission requests public comment concerring the need' for suct criteria, the form of such criteria, and the criteria themselves, d 10. Are performance indicators that are being used by industry, ma) be used in the future, or have been used in the past, appropriate candidates as quantitative eetsares of maintenance 1 e f fectis er.e ss? The Ccemission is particularly ir.terested in
~
escerience er analysis concerning ir.cicaters or the use of i .cicatc*s cf cce;cnea.t re1f acility as mair.tenance performance indicatcrs.
h 11. Should an incustry wice component f ailure reporting system, e.g. , h:US, te useo ty all plants in orcer to support the sharing of generic maintenance ex;erience ano facilitate moniterir; cf maintenance effectiveness?
- @ 12. Conissic*er Peterts hao the felicwing vievs: Icannotjoin the majcrity in supporting the preposed ru:emaking on maintenance. In order to have the tenefit of the public's comments, it has teen my custcm to agree te publication of preposec rulemakings. I cannet do so in t'.is instance. I have asked one fu,camental questien. What are we trying to
' accomplish with this rule that cannot more effectively and
- innovatively be accomplished without a regulation? I have not received a satisfactory answer. I do not telieve the case has been made that licensees do not have established maintenance
- programs. Mest importantly to me, there has been no demonstration that this rule would improve implementation of
" existing pregrams. Neither have ! been provided with compelling documentation on what the problem is and how, specifically, this rule will fix it. On the contrary, the trends staff has provided shew continued improvement in the maintenance area.
w._ -
si i
The prep sed rule the Commission is now publishing fails to
. provide a basis for determining when a maintenance program is :
effective er,when improvements.are " appropriate." 'We are even; delaying publication of the ' accompanying Regulatory Guide until'
.the final rule. Without being' afforded the opportunity to; review this imoletentation document, the Commission'is left in;-
the position of approving a specious rule. It is no. wonder that this rulemaking would elicit such widespread opposition.' The-public is teing asied to c:mment' on a rule of fers but no--
tutstance. I believe it would be more productive to delay isstance cf this preposed rule until the draft regulatory guide
~is availatle'fer c:e ent. Only then can'we receive meaningful cerents en the rulemaiting package Q !'am concernec that this rule goes beyene our authcrity. I' cannot agree with a rule that would have the.NRC regulating '
maintenance en all systems, str';ctures and components regardless ofw$ethertheyhave'anexustoradic1cgicalisafetyornot. I as tre.: lee ty the attituce dem:nstrated when we request public:
.com*ents on what limitations, if any, should be placed on.the?
I final rule to address structures, systems and components' that.
are " wit 5 cut cuestion irrelevant (my. emphasis) to'the protection of public health and safety." ;This clearly abdicates our responsibility to show that a regulation is needed.- We must ask ourselves: are we proceeoing with this rulemaking for the sake
, of the rule itself? As attested to by the cases where the Commission' cited licensees,-the NRC already'has the authority to enforce compliance in the maintenance area.
h The arguments advanced by both the staff and the Commission in trying to comply with the requirements of the backfit rule have played a significant role in my decision not to support this proposed rulemaking. The staff argument for the rule's compliance with 50.109 has been made on the basis of cost. The
~
l
. __ - - - _ - - - a
.o l
staff states that the backfit analysis snews that " .. the rule hCT.
, (qviQ .ill provide a substantial increase in the protection of the public tealth and safety without any additional cost." I am l- .
skeptical of the assumptions made in the tackfit and regulatory analysis and request comments on both these documents. I also request com ents on the views of the ACRS. They state that "...
there are characteristics of regulations, and especially the way in which they are typically enforced, that lead us to believe that, uncer a rule, a move tcsard uniformity would occur, and this is likely to decrease the effecti'.eness of some of the tetter existir; programs." I share t'eir concern that the esiste ce cf t'.is rule could maxe things worse a.d diminish estter than enhance tre protection of the puolic.
g Ee;a aing "ade:: ate protection," the Cc mission a;; ears to be sajing that since effecthe maintenance is necessary to maintain a:ecuate ;rotection, this rule should be excepted under 50.109(a:
Inis ext ; tion would ;rchibit staff from taking im: lamentation et into censiceration. However, it would recuire that a documented evaluation te prepared for public com ent. Therefore, my opposit to the exce; tion is not to the exception itself but to the prececential nature of the use of the adecuate protection argumer Let me state that I too strongly believe that effective maintenar is necessary to assure that nuclear power plants are safe and to-I provide adequate protection to the public. Ialsobelieve,just strongly, that this rule is not necessary to provide that crotect and that as the ACRS noted it may well have the opposite effect.
believe that we Cannot afford to be careless about the use of the
" adequate protection" argument for exception to the backfit rule.
The Commission is in litigation about this very issue. The Commission addressed this point in detail under the heading
" Adequate Protection" in the Response to Ccmments on the final 10 CFR Part 50 Revision of Backfit Process for Power Reactors. L us remember that there had been concerns that in dealing with the
_ g. .
A Y:l
)
j 1
1 backfit rule, the (cr. mission would use the phrase " adequate
&b
~
protection" arbitrarily. The Commission could' unwittingly be giving credence to that view /
3
-1 h Additionally,~it seems to me that the Commission position on adeounte protaction is: internally inconsistent. TheCommission-I f
needs to recegnize that when it states that-this rule is needed'-]
to raintain acequate protection, it is saying that the current - f operating plants new p:se undue risk to the public which we are i presently toleratir.g. IfIbelievedthat.Iwouldsuggest(as -l
,. I'm st.re would the rest of the Co=ission) that this rule become
, i- a:iately effective. This is clearly not tne case. As the Cc--issi:n in the very same cceent shows, "' .. the proposed rulec:cifiesandstancarcizespreviuslyexistir.c(my' emphasis)'.
C: = ssion requirements, both explicit and implicit, in plant' tec n ical s;ecificatices, licensee safety analysis reports,' and' 1 10 CFR 50, A;;eacix B." It seems to me that the Cem ission can'.t-have it t3th ways.
I rec;est c:=ents en ey views.
IMPACT Although the C:=ission believes that this proposed rul'e is by virtue-of10CFR50.1C9(a)(4)notsubjecttotherequirementforabackfit finding and analysis, it has nevertheless performed an analysis of cost and other backfit factors as an alternative ground for proceedirg with the proposed rule and to facilitate public comment on the backfit issue.
Q The economic impact of the proposed requirement on licensees should be negligible. The accompanying draft regulatory analysis shows that, although initial financial investments will be required by some licensees to establish a systematic and comprehensive maintenance program, the- .
savings due to decreased corrective maintenance costs and increased plant availabilityshouldoutweightheinveste.entcosts.[TheCommission_
l- p l
l l-l l
7, I.
- n; _ _
i l
. q 4
ggpoust. RE.Qu TAO ,
Q , gW 'srecuest c^c_c'peration of informed sources in order to develop .
details and verification of this analysis. -
J m
FINDING OF NO SIGNIFICANT ENVIRONMENTAL IMPACT: AVAILABILITY'
@ The Commission har" determined under the National Environmental Policy Act of 1969, as amenced, and the C:mmission's regulations in Subpart A of 10 CFR Part 51, this rule, if ade:ted, would not be a major Federal action:
significantly af fectir.g the quality of the human environment and; therefor an envireneental impa:t state est is not required.
g The p ::: sed acti:n is directed to preserve and preestly restore c;erability, reliability, and a,ailability of, or to prevent the failure.
ef, plant structures, systems, and ccepenents whose f ailure could threate,7 the health a'.: safety cf the public. Since the prep: sed action is
- directed to ie;r:ving the m3intenance of plant syste s, comp:nents and '-
-structures an: d:es r:t require any r:cification of tne plant,'it will not affect the quality of the human envire ment.
g The environmental assessment and finoing of no significant impact on'.
which this ceterminati n is based are available for inspection at the NRC Pu:lic Document Recm 2120 L Street, Lcwer Level, N.W., Washington, D.C.
Single copies of the environmental assess ent and finding of no
'significant inpact are available frem Moni Cey, Office of Nuclear Regulatory Research, Telephone: (301)492-3730, U.S. Nuclear Regulatory Commission, Washington,D.C. 20555.
PAPERWORK RIDUCTION ACT STATEMENT Q This proposed rule amends reccrdkeeping requirements that are subject to the Paperwork Reduction Act of 1980 (24 U.S.C. 3501 et seq.) This rule.
has been submitted to the Of' ice of Management and Budget for review and approval of the paperwork requirements.
t g.
l
W fa~~T e
g r.y3_W.Y ,7 i !
r j b 1 t
e l
h Public recorting turden for this recordkeeping of information is
'~
' estimated to average 1000 hours0.0116 days <br />0.278 hours <br />0.00165 weeks <br />3.805e-4 months <br /> per plant per year including the time for reviewing instructions, searching existing data sources, gathering and 4 .
maintaining the data needed, and completing and reviewing the collection 1
s of information. Send cepents regarding this burden estimate or any other aspect of this collection of information, including suggestions for k
reducing this burden, to the Records and Reports Management tranch.
Civision of Information Support Services /IRM, Office of Administration and c
es:urces Paaa;e ent, U.S. Nwclear Regulatory Ccmaission, Vashington, D.C.
2:555; anc to tre Office of Information and Regulatory Affairs, Office of Famage ent ano Eve;et, ' washington, D.C. 2C503.
REGULATCRY ANALYSIS M The C:mmissien has pre;ared a draft regulatory analysis on this
- re;
- sec regulatien. The analysis examines the costs and benefits of the alterratises considered by the Commission. The draf t analysis is a.ailatle f:r irs;ecti:n in the hRC P blic Cocument Rcom, 2120 L St., N.W.
washirgten, D.C. Sieg!e copies of the draf t analysis may be
- taired f t:m M:ni Cey, U.S. Nuclear Regulatory Commis sion ' Washington, D.C. 20555 (301) 492 3730.
The C:Pmission recuests public comment on the draft regulatory analysis. Comments on the draf t analysis may be submitted to the hRC as indicated uncer the ADDRESSES heading.
BACKFIT ANALYSIS h As noted above, the Commission has completed a tackfit analysis for the proposed rule. The Commission has determined, based on this analysis, that backfitting to comply with the requirements of this proposed rule will provice a substantial increase in protection to public health and safety without any additional costs. The backfit analysis on which this determination is based is as follows:
E
-__m - _ _ - -
'o' q;
)qAm,g.m.:;M.
y yr ,.A.1.E,.
y ,
e f
Y g 't Analysis are Ceturentatien for the Frecesitien that the Precesed Rule i
$atisfies the "Adtcuate Protectien" Esrectton in 10 CFR 50.1C9f aMal:
Analysis ene cetereiration That the Pelemakiro to Amend 10 CFR 50-l Concernire Maintenance Complies with tackfit Rule 10 CrR 50.109
.- In Any Event g While the Cemission's existing regulattens contain several previsicts wnich, either implicitly or explicitly, accress the need for effective Fairterance,' they co net explicitly address the requirements of an effective mainterarce ;regram in any easily treersteed and cc ;reaersive 'astien. The Cemissien telieves that safety can and must te eneanced ty imprcving the effectiveness of rainterance programs thrcugreut the nucles* ir.custry. The prc;cted rule recuires nuclear poner plart rainterarce prograns to have a set of functices and activities which the Cemissice telieves are essential for a cor;rehensive and effective program.
Q s ,_,,. -The fircirgs and cenclusiens of the Cemission's assesse:ent of the*
w
'/ effectiveness of plant maintenance programs can be fourd in NURIC-1212. _
Q* Status of f*aintenance in the U.S. fluelear Pcwer Industry.* [The study shcwed weakresses at plants in a nutter of areas of maintenance which the Commissicn telieves is critical for an everall effective er" " -
tter, analysis of c;erational events has shown trat, in some cases,
' nuclear power plant equipment is not beirg maintained with a high degree
. of reliability that the ecuipment will perform its intended function when i ecuired.[The Comission believes that effective maintenance is necessary~
for aceguate protection. The objective of the proposed. rule is to require improvements for those plants that are poorly maintained, and also to . I prevent the declining performance of plants that are well-maintained. The proposed rule, by addressing in an easily understood way the requirement that there be an effective maintenance program, would advance the goal of having a comprehensive set of regulations which define what is needed for
.zi adequate protection and; therefore, satisfies the 'Adecuate Protection' Exceptionin10CFR50.109(a)(4).
N
~ ~ ~~ ~ ~
5@MfMQEif[6.
p ,
- g. < 3 L'
1 I (
i c ,;.
/
j
)
i h In addition,' the Cceriission has conducted an aralysis and determine.*
that the prepesed r.aintenance rule complies with section (3) of the backfit.
rele,10 CFR 50.109. The tenefit of improvements in maintenance has been estirated to be Lin the order of :50,000 person rem for the'11fetime of all nuclear power plants. Cddon sense also suggests that a well maintained' plant poses less risk than ene peerly maintained. The proposed rule'willf help ensure a satisfactory level of performance for maintenance programs by specifying these activities and practices which, through esperience,-
have teen shewn to te key eierents of effective rainterance_ programs' and - ,
.shculd resstt in a substantial enhancement cf public hecith and safety.
Irpreve ents in rainterarce serve tc rett dsal etjectives: '(1) erhance safety; and alse (2) increase plant availability and capacity factor, therefore, it tas teen estirated that entantement in safety as a result of ircrevec maintenance ray te accerplished withcut any addition!1 cests, dee te ccst savings free reduced plant ecwntire. Further, a preactive are well plar.ree eateterance prograh decreases costs.of2 ccrrective are repeat maintenance,
@ The folicateg discussion presents the surinary' of t'hs backilt -
analysis. Further details ray te fcund in the Pegulatory Anal is for the prepesed rule.
t.nalysis of i 50.109fe) Factors
@ 1. Statement of the specific objectives that the backfit is designed to !
achieve. ,
I Q The purpose of the maintenance rule is to improve maintenance effectiveness, and thereby enhance overall safety, by establishing-basic requirements for plant maintenance programs. In establishing these requirements, the Comission intends te consider the fr.dustry wide efforts that have g1 ready been initiated.
b 1
1
_ _ _ _ _ , - - - - _ - - , - - - - - - - - . - - - - , - - _a-.-n-- - a _ - . _ - - - - - - _ _ - - - - - - - - - _ __ _ _ . , - - - _ _ _ _ - _ - - - - - - - - _ - , - , - - - - _ . - - _ - - . - - _ - - - - - - - - - - - - , - - _ - _ _ - . - - - - _ _ _ - - - - - - - , - - - _ _ - - _ _ - Y
. !6 i i - m j h The objectives of the maintenance rule are as follows:
(1) To define NRC's expectations for plant maintenance; and '
(2) to improve licensee maintenance programs by requiring th; effective conduct of a set of functions and activities.
h With implementation of the rule, it is expected that the current w'
, variation in maintenance performance will be reduced so that the performance of plants that lag behind the industry as a whole will brought up to the level of performance of the majority of the industry. Second, the overall average level of industry's performance should also improve.[AssoM.s dAA4atiaas A5
,. \ SE5cR60CD ## #ostgg ittt -I ti.,.
An important part of the structure of,the rule is to achieve impros maintenance performance in a way that allows licensees the flexibility to determine the details of their individual maintenanc.
program so that plant-specific factors can be taken into"'ac' count $
This flexible approach will enhance both safety and cost-effectiveness, compared to a rigid and prescriptive rulemaking approach.
,/ . 4 li
- 2. General description of the activity required by the licensee er.
Ad ? "
m applicant in order to complete the backfit.
43
&> W ' .' i i
h The licensees will be required to have a documented and effective maintenance program which shall include the following activities:
3 ,
g 1 (1) Technology in the areas of:
(1) Corrective maintenance, (ii) Preventive maintenance, (111) Predictive maintenance, and (iv) Maintenance Surveillance; 4
y in ,
s I
(2) Engineering in support of' maintenance;
, (3) ' Quality assurance and quality contr:1 of maintenance'.
- activities; (4) Incorporation of plant modifications.into the maintenance.
program; i.-
(5) Equiptent histcry and trencing; (6) Maintenance rec:rd keeping;.
(7) Managetent of. carts, tools, and facilities; (8) Maintenance procedures; (9) Post Maintenance testing and return to service activities;;
(10) Peasures of overall maintenance program effectiveness; (11) Maintenance canager.ent and organization in the areas of;-
(i) Planning, (ii) Scheduling, (iii) Staffing, !
(iv)'Shiftcoverage,and (v) Resource allocation; (12) Control of contracted maintenance services;
]
(11) Radiological exposure control (including ALARA) during maintenance activities; i
7 y -- m
(:
1 (14)Maintenancepers:nnelQualificationandtraining; (15)Internalcommunicationsbetweenthemaintenance
' organization ano plant cperations and support groups; (16) C:mmunicatiens tetween plant and corporate management and the maintenance organization; and (17) C:nsideratien of maintenance recc=encatiens er requiremen.
< of incividual ver: ors.
h Criteria fer a::e::5:ility :f the c:ncuct of the above activities will ta ;roviceo in the prepesed Maintenance Standard er Regulatory Guice.
@ 3. Potential c*.an;e in the risk to the public frem the accidental ef fsite release of radioactive material.
Q b;le ettati:n e' tne pre;cseo maintenance rule will result in an esthated t:tal risk reduction to the public ranging from 50,000 to 500,000 persen-rem with a p: int estimate of asoct 250,000 person-rem M4 Fotential impact en radiological exposure of facility employees.
A large fraction (two thirds to three fcurths) of the occupational
- radiation exposure incurred at nuclear power plants is associated with maintenance, on the order of 300 person-rem per reactor year in 1987. Improvements in maintenance programs can affect collective occupational exposures both positively and negatively. Increases in maintenance activity due to expanded preventive maintenance or more aggressivecorrectivemaintenance(toreducebacklogs,forexample) will tend to increase exposure, and productivity and reductions in
. the amount of rework will tend to reduce exposures. The net effect of these positive and negktive trends is believed to be beneficial, l
7 y_'
f 4
but small ccmpared to the other cests and benefits of improved maintenance.-
.e 0 .
h 5. Installation and continuing costs associated with backfit, including the cost of facility cowntime or_ the cost of construction delay.
Q For 100 operating reactors, the total estimated cost associated with the proposed maintenance rule is -5198 million, The minus sign
~
denotes a ccst savings. This estimate breaks down as follows:
Ind;strv
. Cest Ele emt Millions of 1988 Cellars Maintenance Plan 4 Recordkeeping and Repcrtieg 12.3-Maintenance Stancara 2.8 Reg. Guide Workshops 0.4 Implement Improvements (at some plants)
Preventive Mainte.ance 1500-Maint. -Ops. Coord. 2.1 Maint. Info. System 150 Procedures 14 Increased Staffing 50 to Reduce Overtime
. Reduced Risks of Onsite Property Damage -33 Cost Savings Due to improved Availability and Reduced Corrective Maintenance -1900 Total Net !.ndustry Costs -198 -
NOT[5: 1. Negative signs denote cost sayings.
- 2. Values in table are rounded.
l @
l a.
m -- - -
3:37w ---'---------;--9 J
')
1 l
J l
- 6. The potential safety impact of changes in plant or operational complexity.-including the relationship to proposed and existing regulatory requirements.
The proposed rule would require certain elements'in a plant maintenance prograr and should not add to plant or operational complexity, !sproved maintenance should result in a decrease of.
challenges to safety systems and fcrced outages, and theritfere, should decrease the ccmplexity of cperations.
@ The procesec maintenance rule is related to the follcwing existing reg:.;1atory requirements:
(1) Surveillance requirements for safety systems is required in the technical specifications. These requirements are not duplicated f or the Dr < M5ed maintenance rule.
(2) In service inspection requirements are covered under 10 CFR 50,55a Codes and Standards. These requirements of a preventive e.aintenance program are not duplicated under the preposed maintenance tule. .
(3) Criterion 1 Appenclx A,10 CFR 50, concerning General Design Criteria (GDC), requires that a quality assurance (QA) program _ be established and implemented in order to provide adequate assurance that Nuclear Power Plant (NpP) structures, systems, and components will satisfactorily l perform their safety functions. Appendix B of 10 CFR 50 establishes QA requirements for the design, construction, and operation of those structures, systems, and components that prevent or mitigate the consequences of postulated accidents that could cause un.due risk to the health and safety of the public. Regulatory Guide 1.33 has been promulgated to describe acceptable metheds for complying i
i u
-' -~ -- -
- - - = - - - - - - - --
a---
77 7c7;w ---
f i'
, with the provisions of these appendices of 10 CFR 50 for Cswuo t# the operational phase of NPPs.
1 .
These requirements are directed towards maintaining a quality assurance program and do not explicitly address j maintenance as defined in the proposed rule. The proposed Maintenance Standard and Regulatory Guide will directly accress the maintenance activities in the proposed rule, 7 The estimated resource burcen on the NRC associated with the tackfit anc the availability of such resources.
The estimated total cost for hRC review of industry'submittals recuired by the proposed maintenance rule is sa00,000 based on the need for developing a Regulatory Gaide to provide the basis for review of an industry-developed Maintenance Standard or to provice guidance on implementation of the rule if a Maintenance Stancard is_not develepe d incivicual plant maintenance.
g ,j ,,,, programs will te subject to the ongoing inspection process but gE
- ft wouldnotterecuiredtotesubmittedtomaNRqfor,r,evie,wan(
{ap:reval.
8.f 9%Is.
- 8. The potential impact of difference in f acility type oesign, or age on the relevancy and practicality of. the backfit.
The proposed maintenance rule establishes generic requirements applicable to all types of facilities and designs, regardless of their age.
, , . 9. Whether the backfit is interim or final and, if interim, the s ... A
- justification for imposing the backfit on an interim basis. The proposed maintenance rule would be a final requirement.
- l l
1 I
v -
f; L
o REGULATORY FLEX 18ILITY CERTIFICATION In accordance with the Regu'1 story Fleittility Act'of 1980,' 5 U.S.C.
605(b), the Commissien hereby certifies that this proposed regulation'util not, if promulgated, have a significant economic impact on a substantial nurcer of small entities. This proposed regulation'affects licensees that-own and operate nuclear utilization f actitties licensed under Sections 102 and IC4 of the Atomic Energy Act of 1954. as amended. -These,1icensees do
. not fall wittin tre definition of s all bustress set forth in Section 3 of the Small Eusir.ess Act, 15 U.S.C. 632, or =1 thin the 5:411 Business Size Stancares set ferth in 13 CFR Part 121.
PART 50 DCEESTIC LICEh5!NG CF FRCCCCTICh AO LTIL!!ATICN FACILITIES
- 1. The Autterity citation for Part 50 continues to read as follows:
Autterity: S e c s . 102, 103. 104, 105, 161, 182, 183, 186, 189, - 68 Stat. 936, 937, 538, 948, 953, 954, 955, 556, as amended. Secs. 234 83 Stat.1244, as arended (42 U.S.C. 2132, 2133, 2134, 2135,' 2201, 2232, 2233, 2236, 2239, 2282); Sees. 201, as arenced, 2C2, 206, 88 :
Stat.1242, as amenced 1244,1246 (42 U.S.C. 5541, 5842,1846). -
Section 50.7 also issued under Pub. L.: 95 601, Sec. 10,-92 Stat.
2951 (42 U.S.C. 5851). Sections 50.10 also issued under Secs.101,
,185, 68, Stat. 936, 955, asamended(42U.S.C.4332). Sections 50.23, 50.35, 50.55 and 5056 also issued under Sec.185, 68, Statt.
955(42U.S.C.2235). Sections 50.33a, 50.55a, and Appendia Q also issued uncer Sec.102, Publ. L.91-190, 83 Stat. 853 (42 U.S.C.
4332). Sections 50.34 and 5054 also issued under Sec. 204, 88 Stat.
1245(42U.S.C.5884). Sections 50.58, 5091, and 50.92 also issued under Publ. L.97-415, 96 Stat. 2073 (4'e U.S.C. 2239). Section 50.78 else issued under Sec.122, 68 Stat. 939 (42 U.S.C. 2152). Sections 50.80 50.81 also issued under Sec.184, 68 Stat. 954, as amended (42
-i p
p -
U.S.C.2234). Section 50.103 also issued under Sec.- 108, 68 939, as arended (42 U.S.C. 2237).
i For the purposes'of Sec. 223, 68 Stat. 958, as amende'd (42 U.
2273),16 50.10(a), (b), and (c) 50.44, 50.46, 50.46 and 50.! '
50.54(a) are issued under Sec.161b, 6S Stat. 948, as amendec U.S.C. 2201(b)); il 50.10(b) and (c) and 50.54 are issued unc 1611, 68, stat. 949, as amended (42 U.S.C. 2201(i)); and il !
50.55(e), 50.59(b), 50.70, 50.71, 50.72,10.73 and 50.78 are under sec. IElu, EB Stat. 950, as amendec (42 U.S.C. 2201(c))
h 2. A'new' t 50.65 is acced to read as follows:
so i 50.65 Recuirereets to ensure the effectiveness of r.aintenante procrars for nuclear power plants.
g (a) Applicability. The requirernents of this section apply te-nuclear pc.er reacters licensed under.1 !0.21 b or 50.22 h (b) Definitions. For the purpose of this section, the Commi cetines r.aintenance as the aggregate of those planned an syster.atic actions required to preserve and promptly res operat,ility, reliability, and availability of, or to pre -
failure of, plant structures, systems,'and components.
Comrission 11 tends the score of the rule to cover all_ sy' s_tructures ar.[ components, jncluding those in the
$alance.cf-Flant. Maintenance includes not only activit traditionally associated with identifying and correcting or potential degraded conditions, i.e., repair, surveill.
diagnostic examinations, and preventive measures; but est all supporting functions for the conduct of these activi Maintenance includes the following activities:
4 m- _ - - - , , - - - - , . - - - - - - - - - - - - - - - - - - - - - . - - - - - - - - - _ - - - - - - - _ . _ . - - - - - -- . . . - - - - . - - - - - _ _ - - - - - - _ _ - , - - - - - _ _ - _ _ _ ---_u - - - _ - - - --------------------------------------u - - ----.- _ - - -a------
s .., . - - - .
-. o.
' y .:
],
- i. J '
U' ,
l i
h (1) Technology'in the areas of:
(1) Corrective maintenance, {
(11) Preventive maintenance,
{
'(iii) Predictive maintenance, and (iv) Maintenance Surveillance; (2) Engineering in support of maintenance;
?
(2) Cuality assurance and cuality control of main't activities; (a) Incorporation of plant modifications into'the mair.ter.ance program; (5) Equipment history and trending; (6) P.aintenance recorc keeping; 1
(7) Management of parts, tools, and f acilities;'
(B) P:ainter.ance procedures; (9) Post-Maintenance testing and return to service activities;
, (10) Measures of overall maintenance program effecti (11) Maintenance management and organization in the c,f:
(1) Planning.
(ii) Scheduling,
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . . _ _ _ . _ _ . _ _ _ _ _ _ _ __.______.___________.______________________.________.________._.________.m
l Lc/L.' .
Mi ,
(CNrh E .
_(ggg)sggggng,
,_ (iv) $hif t coverage, and
. (v) Resource allocation; (12) Control of contracted maintenance services;.
(13), Radiological exposure control (including ALARA maintenance activities; '
(14) Maintenance personnel qualification an'dtriini (15) Internal communications between the mainttnanc
+-
organ 128 tion and plant cperations and supportJ
.(16 ) Cor:rnunications between plant and corporate ran and the maintenance organization; and (17: Consideration of maintenance recocine'ndations' o requirements of individus1 vendors.
h (c) Feevirements. Each holder of an operating Ilcanse subjc this section shall (1) establish, implement, and maintale effective and documented maintenance program that addres:
elements and activities in paragraph (b) of this section.
(2) regularly assess the effectiveness of this maintenant program and, based upon this assessment, mke. improvement appropriate.
(d) Imple entation. By [ insert a date 2 years after the effe date of the amendment) each licensee shall certify, by le ,
the Director of the Office of Nuclear Reactor Regulation, '
comprehensive documented maintenance program is being mai and implemented, which addresses all elements and activit '
paragraph (b) of this section including ressures to ronit j
i
- _ - - . - - - _ _ - - - - - - - - - - - -- - - - - - - - - - - - - - - - -- - a
t
- 4
,g
..y h.
- effecttv' eness of the maintenance program and to improve -
program where appropriate. In addition,' each licensee t =
9 develop [ insert a date 3 scnths after the effective date i amendment)'a timely and expeditious plan and schedule (1.
Key Milestones) for meeting the requirements of this see Dated at Rockville, Maryland this M Y day of e/ede_v For the Nuclear seguistory Coer -
~
I u
4
- 0. .
s SamuelJ. Chi 1}, -
Secretary of the Cocrnission.
e
'1l
@ i
._____m mm_ _ _ _ ._.m. - _m.. . _ _ _ _ - . _ . _m_.. ._-_-.___..-___-____.___.._-_..-.___-._.m__m.__. -
___m._ _ ____.__.__ _.._.-_...-._.--._.____-___--__.--___-_-___m_.-_____u.__.-.--___._-_m._)
, \
DRAFT REGULATORY ANALYSIS FOR FF.0 POSED RULEMAKING ON MAlhlEhANCE OF h0 CLEAR
< POWER PLANTS November 1988 i
l
h5
}, ;
I EXECUTIVE
SUMMARY
The Nuclear Regulatory Comission (NRC) has initiated a rulemaking to -
establish baiic requirements for nuclear power plant maintenance programs.
The Comission's intent to proceed with the rulemaking was announced in a Policy Statement on the Maintenance of Nuclear Power Plants, which was published in the Federal Register on March 23, 1988. The Policy Statement also provided initial guidance to the industry on the objectives and scope of maintenance and on the activities which form the basis of e.n adequate maintenance program.
This document is a regulatory analysis of the proposed rule.
STATEMENT OF THE PROBLEM n
g operations has been widely recognized for a long time. -In recent years however, maintenance has become the focus of increasing attention Doth from the NRC and the nuclear industry. NRC has identified a number of maintenance.
related safety concerns that form tte basis for its view that improvements,in maintenance are necessary. The concerns were summarized in the Policy Statement: )
Q *The Comission has a program to continually evaluate the operational performance of nuclear power plants. Analysis of operational events has shown that, in some cases. nuclear power plant equipment is not being mainteined at a level which ensures, with a high degree of reliability, that the equipment will perform its intended function when required. A.
limited NRC examination of nuclear power plant maintenance programs has found a wide variation in the effectiveness of these programs. Inadequa maintenar.ce at some plants has been a significant contributor to plant reliability prcblems and, hence, is a safety concern. The Commission believes safety can be enhanced by improving the effectiveness of maintenance programs throughout.the nuclear industry."
h A more extensive discussion of maintenance-related safety concerns was published in NUREG-1212, Status of Maintenance in the U.S. Nuclear Power 1 Industry 19BE.
- @ in recent years.The nucleartoindustry In addition has safety, the also focused industry has strongincreasing economic atte incentives to provide effective maintenance. The industry has undertaken a wide variety of initiatives to improve maintenance and has comitted substant' resources to the effort. Major programs have been underway for several years under the auspices of NUMARC, INPO, EPR1, owner's groups, and individual j utilities. Although there remain areas for improvement, the ERC and the !
- industry agree that the initiatives have already resulted in substantial progress, with continued advances expected in the years ahead. The NRC intent -
to recognize and take account of these initiatives in structuring and implementing the proposed maintenance rule, iii G
l
y V y h
h OBJECTIVES OF THE PROPOSED MAINTENANCE REGULATION In its Final Policy Statement on Maintenance of Nuclear Power Plants .
.the Comission declared that it believes safety can be enhanced by improving the effectiveness of maintenance prcgrams throughout the industry. The purpt of the maintenance rule is to improve maintenance effectiveness, and thereby.
enhance overall safety, by establishing basic requirements for plant:
maintenanceprograms. In establishing these requirements the Comission intends to consider the industry. wide efforts that have already been initiett h The objectives of the maintenance rule are as follows:
- 1. To define NRC's expectations for plant maintenance.-
- 2. To ensure timely action by industry to upgrade maintenance'where necessary.
- 3. To build uponiindustr initiatives to promote responsibility and ownerst by industry .for resol ing maintenance problems
- 4. To improve licensee maintenance programs.
With implementation of the rule, it is expected that the current wide variation in maintenance perfomance will be reduced so that the perfomance
, of plants that lag behind the industry as;a whole will be brought up to the
' level of the majority of the industry. Second, the overall average level of industry's perfomance should also improve.
@ An important part of the structure of the rule is to achieve improved ~
maintenance perfomance in a way that allows licensees the flexibility to detemine the details of their individual maintenance program so that plant-specific factors can be taken into account. This flexible approach will enhance both safety and cost. effectiveness, compared to a rigid and prescriptive rulemaking approach.
p" Also an important part of the structure of the rule is to recognize and reinforce the. positive initiatives that the industry has already undertaken.
The rule is intended to encourage the continuation and the timely completion of these initiatives.
AL1ERNATIVES h Five principal rulemaking options have been considered. Additional options, including the status eco option (no rule) and various combinations of the five basic options, are also possible, but are not being considered presently. Taken as a whole the options reflect a range of regulatory approaches, with differing degrees of prescriptiveness, and varying strategie
. for defining and measuring maintenance effectiveness. Consistent with the Policy Statement on Maintenance, all of.the rulemaking options cover a broad scope of structures, systems, components and activities, including balance. ; ,
cf plant. '
iv 1
- $ _ y y- , y --- -
yy Y
W L
, 4,o ,
M .
b' i y
' 00 tier. I '-- General Performance Based Rule
- Each licensee would be required to (1) have an effective and'docum*ented[
s
' maintenance pro ram;-(2) establish quantitative measures of the effectiveness 1 L ' 7". of.the program i.e., maintenance performance indicators); (3) use the measure
' to periodically evaluate the program and, if appropriate, . improve it; end (4) , f
.reportthemaintenanceperformanceindicatorstotheNRCcnaquarterlybasisq This option is performance-based because it relias on maintenance
' performance indicators as the mechanism for regulatory oversight of the
! effectiveness of a licensee's program. Licensees would have considerable o
flexibility in designing a maintenance program consistent with their,own plant l' ' specific needs and conditions, provided they cocid demonstrate through the .
performance indicators that the program was effective. ' This option also allos for licensee flexibility ig selecting'and reporting on measures of effectiveness. A Regulatory Guide would be written by the NRC to provide a recomended approach, but the licensee would be able to offer alternatives if.
appropriate.
' b.h A major feature of Option 1 is that it leaves some important points; undefined, principal openallowing flexibility' issues are: (1) W hat constitutes an effective maintenance
- program 7 and (2) How will NRC judge the. acceptability of-licensee Performance Option 2 -- Prescriptive Performance-Based Rule h This option is similar to Option 1 in most respects, except that it -
prescribes the quantitative measures of maintenance effectiveness-(maintenanc performance indicators) to be used. The rule would spell out which indicator, must be reported. There would probably be a separate document that contained instructions on reporting formats,- precise definitions,'and oth;er details, similar to the guidance for LERs.
Ootion 3 -- General Rule. With industry Standard As with Options 1 and 2, each W ensee would be required to'have an effective and documented maintenance program. NRC's regulatory position on the elements of an effective maintenance program would be defined in a.
y, '
Regulatory Guide.that referenced an industry maintenance standard, after-appropriate NRC review and endorsement. There would be no requirement to.-
report maintenance performance indicators.
By building on an industry standard, Option 3 provides an efficient mechanism for recognizing and incorporating the maintenance initiatives undertaken by industry in recent years. -Many of the elements of such a standard are well-developed as a result of industry's efforts and could be factored into a standard. 'This option would also encourage standardization of industry maintenance practices. A disadvantage of Option 3 is that the completion of a standard is a tinee-consuming process; however, development -
and validation of maintenance erformance indicators and issuance of a RegulatoryGuide,asrequiredbyOptions1and2,canalsobealengthy f process.
. v
W qn . g -
e
~ ~
.o ,
e
[k ' , ~ '
Q,: ,
)' M QO t-
-)
'a
'[' n s
(
(( . ,
ji >
V<h{,- .y ,
i y' gg -
_q. ; l }i '
.( '
}
3 , _
! Option a -- Genefal Rule, With'Resulatory Guide; J
~
bb;h(
MW "
.. Option 4 resembles Option 3. except that NRC would not endorse an indus1 i f" ' '
- ) standard in its Regulatory Guide. Rather,- NRC would establish its regulator..I c
4 i $ . position independently, without reference to industry standards. A disadvan i Sc 6' i of. Option 4,- relative to Option 3,'is that it might not fully incorporate th R O _ , y benefits of industry's experience and detailed plant knowledge.- However, it j
, would relieve industry of the burden of developing a standard.-
Cot ion' 5 -- Prescriptive Rule
~
'g f qr '
h Option 5 is to romulgate a detailed rule that would spell out complete in the regulations t e requirements for nuclear power plant maintenance programs. It would provide a comprehensive definition of what is required o
,- licensees, but could impact ~the industry initiatives currently underway. -
n, "'
-Major disadvantages of Option 5 are the lack of flexibility to accommodate?
< varying conditions and changing technology, and the difficulty of defining '
Hl[ a effective maintenance with the necessary specificity.s Considerable time end.
g, 3 NRC resources would be needed to develop a prescriptive rule.-
re y 1
' CONSEQUENCES-e h The expected consequences of the proposed maintenance rule are grouped _
into five major categories: effects on public health and safety, effects on.
,.4 , occupational radiation exposure, industry costs and cost savings, NRC costs, and other regulatory considerations..
Public Health and Safety h Precise quantification of the public health and safety benefits of, improved maintenance is difficult. However,--it is possible to approximate
- the benefits by performing risk sensitivity calculations under a range of ;
assumptions.
Q e stimated Two approach'es were used to estimate risk reduction. First, it was that everage baseline risks due to internal events at typical plan with acceptable maintenance programs are on the order of several hundred person rem per reqctor year: 300 person rem per reactor-year was used as az nominal estimate.ta) .Now,- to satisfy the propesed rule, some plants will need to implement maintenance improvements.- If the average risk from these plants is currently two to five times as high as the risk from typical plenty
@ (a) Existingprobabilisticriskassessmerts(PRAs)typicallyyield' estimates in this ranges Draf t NUREG-1150, Vol. 2, Appendices B-F, contains estimates in this range for five reactors. Earlier PRAs also yield sim-results. For example, NUREG/CR-4330, Vol. 2 (Mullen et al. 1986. . Revh of Licht Water Reculatory Requirements: Assessment of Selected Reavireme That May Have Marcinal Importance to Risk) presents calculations based.
on the Reactor 5afety 5tucy and the R5WP PRAs documented in NUREG/CR-1j 1659. ;
vi i
E! ' '
f.
~' );,
u I
P (this assumption w111 be discussed further in a moment), it would be some 6C 4
.to 1500 person rem per reactor-year; the potential risk reduction, if these-
. plants could be improved to the level of the typical plants, is then 300 to '
1200 person rem per reactor-year. Since maintenance is likely to be only or of the reasons for the increased risk at these plants, not all of the potent risk reduction would be achievable through improved maintenance alone. Howe given that maintenance accour.ts-for a substantial fraction of all activity -
the plant (70% according to a presentation by the Chaitsan of MUMARC's Work 1 Group on Maintenance at the Public Workshop on the maintenance rule),
improvements in maintenance might reduce the extra risk at these plants by-two-thirds, i.e., b
- to 30 such plants (y this200 to 800isperson estimate consistent remwith perdata reactor-year.
presented ;inifNUREG-there are L.- 1212), and the average remaining plant lifetime is 30 years, the total poter.
risk reduction is on the order of 120,000 to 720,000 person-ren.-
To support the' factor of 2 to 5 increase in risk postulated 'above, risk sensitivity calculations dere performed. Based on a statistical analysis of the relationship between safety system failures (one of NRC's performance-indicators) and maintenance quality as measured by the Systematic Assessment of Licensee Performance ($ ALP),,it was estimated that plants with weak maintenance perfomance experience safety system failures at 2 to 3 times.tt
. rate of typical plants. This estimate was inserted into PRA models and the
- change in risk due to the increased rate of safety system failures was calculated. The risks were found to increase by a factor of 2 to Si depend' on the PRA used.
(Z i As a second point of reference, a prioritiration analysis of the Maintenance and Surveillance Program was reviewed. It was published in NURI.
0933. Key assumptions in the calculations were thats (1) plant aging would-increase the frequency of transients caused by balance of plant by 10%;-(2) full implementation of tne Maintenance and Surveillance Program would result in substantial reductions in the frequency of transients (by a factor of 2.*'
-(3) there would also be a small decrease in the unavailability due to'
. maintenance downtime. The total estimated risk reduction was on the order ~
of 300,000 person-rem.'
{ rem. The estimates are In considering theall roughlyofinthe benefits themaintenance range of that industry is already making improvements in maintenance based on inittat 100,000 to 1,000,000 pers, rule, it should be rect!
i currently underway. Thus, a substantial fraction of the risk reduction estimated above will be gained under the status quo, even without a maintene :
rule. It is difficult to estimate precisely how much of the benefits will t gained even in the absence of a maintenance rule. However, it is likely the-the existing industry initiatives will significantly reduce the number of plants needing maintenance improvements. Thus, the abcVe estimates of risk 1 reduction are probably too high by a factor of two or more.
' ^
kg attributed to the maintenance rule will be estimated as 50,000 to 500,00 1 person rem; 250,000 will be used as a point estimate.
vii i
i
)
w Occupational Radiation Erresure A large fraction (two thirds to three fourths) of the occupatierfal radiation exposure incurred at nuclear power plants is associated with maintenance, on the order of 300 person rem per reactor-year in 1987.
Improvements in maintenance programs can affect collective occupational exposures both positively and negatively. Increases in maintenance activity due to expanded preventive maintenance or more aggressive corrective maintenance (to reduce backlogs, for example) will tend to increase exposure, at least initially. On the other hahd, improvements in maintenance efficient and productivity and reductions in the amount of rework will tend to reduce exposures. The net effect of these positive and negative trends is believed to be beneficial, but small compared to the other costs and benefits of impre maintenance. To achieve large reductions in collective radiation exposures, it is necessary to look beyond maintenance per se and consider other measures e.g., design changes to reduce contamination throughout the plant, facilitatt easy access by workers, provide better shielding, tools, methods and equipmet and if possible eliminate the need for certain tasks like snubber testing that can cause high exposures.
Industry Costs and Cest Savines h Table 5.1 sumarires the industry costs and cost savings attributable t the various options; the table displays the base case estimates (in the colu labeled "B"), along with lower and upper estimates (labeled 'L' and 'U' respectively) to provide some indication of sensitivity to variations in cer major assumptions. The base case estimates are explained in detail in Chapt-5 of the regulatory analysis. The lower and upper estimates are covered in i Appendix A. All costs are expressed in 1988 dollars. Future costs are discounted to present value at a 10% real discount rate as suggested by the teoulatory Analysis Guidelines, NUREG/BR-0058, Revision 1. Two points shoul be noted. First, for the base case, each of the options is estimated to yie a net cost savings, on the order of $100 million to $200 million over the remaining lifetimes of current plants. The cost savings are mainly due to a estimated 5 percentage point improvement in plant capacity factors at these plants that are currently lagging behind the industry as a whole in terms of maintenance effectiveness. This improvement in capacity factor was estimate on the basis of a statistical analysis of the relationship between plant capacity factors in 1985 1987 and maintenance quality as measured by the Systematic Assessment of Licensee perforinance (SALP). Other cost savings er also expected, such as reductions in the amount of corrective maintenance at rework, but the irnprovement in capacity factor is the dominant savings.
Second, options 3 arii 4 have a net cost advantage compared to Options 1 and 2. This difference is mainly due to the increased recordkeeping and reporting costs that plants would incur to collect and report maintenance performance indicators under Options 1 ano 2. These recordkeeping and repor costs would apply to all plants and would centinue throughout the remaining lifetimes of the plants. Options 3 and 4 do not require reporting of performance indicators.
viii I
_ _ _ _ I
w.
?g _
e e W D
- e. .9 S; e O
S; e =
- e e
'ee e
- e. e.e e e
to e a em e S W ' ese 33
- e
- mam e. me.
& w 89 m e 9e e e e e e e e W ** e es en as we e e e8 ** O e to
- e es e 0
ed
- e.
50 S on '
S e os e e se e e g e S O en w og e eo e8 emt 99 e e e S S pe se S e
- '= b3 $
b =
em = 9 3 e. me. e. e.
hi - a e- - a =
e e
- E e e e e e e 2 e e
e O, P 4s b e e e. O. -
e e O se e
S 4F9 _ej, om= O e en- em S
9 e 0
. e e. = se.
4 e e. es e et ce e' e se e 's e es et *=
b G & se e em os em et e e e as ' se e one G G em se $
Da S e 9
go == g . e. se.
89 SD at es e e me en e es e a y e
- es e es er pt e em 4Y == 93 3 Se o ea**
o
- e. ee e .
e e e
.S . s G - ee e W ee A D 6
- es eS ee e= e e= e es e e e O e
- e e
ee b e b e e se e e e W 9e G WG es se te e e em 4 es ga er . es w e
- m. S e G 4 b es G er se == A
.e5 S m
- 49 eS e 9 +=as e ese e* e ce e e er e 59 e #e 9 9 e em se e em os em e e e S S 89 9 e se es se 5 e
- S S b ene em e'6 9 e8 99 b e S 9 es e S
- - O
- Of 9 4....' ' , , , , ,. j ,'/ ) , ,, , j . i , t , i . . i
.. , , A e s og en e e se en e se e se e ** 4 en o es e O w e o e e s 9 == FD e et e S *
>= w 9e se # D e 9 &9
- 9 9 es e+ e e #+ 3 e D O es a W9' e9 ** e- e e* e M e e e e #8 'S Et e ese e e e e de o e ==
.S 9
9 se Se to e e as e > no 9
- # es e e e es -
- O S e S
- ep 9 e 9 g g g e w. es. e 9 e=
- e en e em e ce e e e e sp 98 9 9 es ses e g
& em a e em en am e e e e se es es ese 4 ene tp 9 e 9 5 se me e ee ef e t 9 b o e.
Emo one ' em en e e4
- b e 9 9 6 80 e e e se o e es e e O *e e e op 9 e es w at e e GO Gb - =a O eb g F5 e 4 9 4 9e em G e
en 4 O
e #
S A- A 9se -
G E
es e4 eE ew G 9
b g 9
as
- em esaea G
es G S g e .
b W 9 & O O E G en O G S 9 5 48 9
& es S S e S 48
- e e & G se a w W G
p *= S se S 9 G S S 9 e S e 9 5 ** # -*9 9 e 9 $ & G S 4 e6 4 S S S S S De 48 9 ed 4 S e 9 e et b e == == m es E E b e S
- ep O N *W d S S S og Om en em 9 e8 e ** O O
== S S b G D ** O WD e8 % es ab= 9 N 88 8 9 W A eJ 9 S S 3D De e 9 tdJ e
==
e 4
- eBI h
O b
e 3
W
+
+=
en S
as S
9 9 e Eb S 9 S **e S 9
4D e
9 4.
9 e
S S
W e8 e
G &
S a=e S
p e
- wO S em
.3 e
9 S
WD 9
e 9
S
.e e
S G
e 9 De 3 o et S
se*
ce SG .
W **& e& ~ pe S W == m ** 88 W
8 9 WD N
9 e8 **e6 9=
b 5
e 9 3 & es
> ==
- D O E
9 Sai e 9 De += 5 S ==S N e 9 ein b S er b 9 9 - O es es* em 9 9 e 4 es eA ==et af G D et O e > S W b E 9 b eft e W *=8 +=
e es b e 9 O em .S. ==S 3 & W 9 9 9 9 3 S 9 9 . b a a e b g 9 3 & es .- es e
= U to e an.a a- a a a a. .e se = e e b.
e e e.* !
S . O S S . O b o e. O e-Se a M De & a=s S & W ES W B 1
i f*
M l l
- l i 1 1
1 i
_ _ _ _ _ . . . _ . . _ _ . _ _ . _ _ _ _. _ _ _ _ _ . _ _ _ _ _ _ _ _ , _ _ _ _ _ _ _ _ s
g_ _
I h
i h Table 5.2. Total NRC Costs over 30 Years for All Plants (1988 Present Worth in Hillions of 1988 Dollars) ~!
}
NRC Cost Element DevelopRegulatoryGuides(s) 0.1 0.1 0.4 0.4 Endorse Industry Standard. (4.1)
Analyre Licensee Reports' 3.5 3.5 Total NRC Costs , 3.6 -- 3.6 0.4 0.4 :
NRC Costs.
Table 5.2 Sumarizes the base case NRC costs associated with the g maintenance rulemaking options, sumed over 30 years. All costs are in 1988
+ dollars. Future costs are discounted to present value at a 10% real discoun rate as suggested by.the Reculatory Analysis Guidelines. The NRC costs are small relative to those iacurred by inowstry. , They are also small relative' to NRC's budget. A to Options 1 and 2. gain, Options 3 and 4 have a net cost advantage com indicator reports under Options 1 and 2 is the major reason for the differen Other Reculatory Considerations h As stated in the Handboek for value-impact assessment, the quantitative portions of a value impact assessment or regulatory analysis are not intende to serve as the sole or even the principal basis for regulatory decisions.
Other inputs are needed and other considerations can affect the decision:
process. In the case of the maintenance rulemaking, these other considerate.
include such factors as:
- The timeliness with which each option can be implemente' d and thus begin-to provide the desired maintenance improvements.
- The possibility that industry initiatives might be hampered.
- Potential benefits for NRC's enforcement program from having a maintena re ulation that clearly states NRC's expectations with regard to ma nienance
- The potential adverse impacts on the industry if a clearly defined and' predictable regulatory position on maintenance is not established.
X i
G l
l
'9 .
- Impacts of the maintenance regulation on other requirements, as well as interactions with'other reDulatcry initiatives (e.g., Safety System Functional Inspections, and certain generic safety issues related to equipment reliability) that address some of the same risks.
The distribution of the' costs and benefits..
h The last point refers to a key aspect of the pro osed maintenance rule. Plar that currently have weak maintenance programs wi 1 account for the bulk of th
, costs and benefits of the rule. Plants with effective maintenance programs will be relatively unaffected.
DECISION RATIONALE
@ Option 3 is recomended as the preferred option. This recommendation is supported by the quantitat.he cost-benefit analysis and by other consideratic, that are not quantitative.
s Q The base case costs and benefits of Option 3 are summarized in Table 5.3. Based on these quantitative cost benefit estimates, Option 3 has a positive net benefit. Risks to public health and safety will be reduced substantially, as the perfomance of plants with weak maintenance is improved Occupational exposure riskt, to workers will also be reduced. On the cost side, although industry and NRC will initially incur increased costs, these cost increases are counterbalanced by cost savings due to improvements in plant capacity f actors, reductions in the amount of corrective maintenance -
and rework, and reduced risks of onsite property damage staming from acciden Q Of the four options examined in the regulatory analysis, Options 3 and 4 place the least burden on industry and on the NRC, and are estimated to be as effective in reducing risk as Options 1 and 2. Option 3 is also supported by other regulatory considerations not quantified in the regulatory analysis.
It can be implemented in a timely manner, and produce the desired improvement.
in maintenance within a reasonable period of time. It has the least potentia to adversely impact industry initiatives, because it builds upon them through the endorsement of an industry standard. Via the industry standard and the Regulatory Guide that endorses the standard, Option 3 will provide a clear statement of NRC's expectations with regard to maintenance, and will thereby obviate potential difficulties in interpretation and regulatory unpredictabit that could arise with less clearly defined options. Finally, by relying on an industry standard, Option 3 is least likely to impact plants that already have effective maintenance, since these plants are likely to meet the standarr thus, the impact of the rule will be concentrated on the subset of plants that are lagging behind the industry as a whole.
6 xi
~
l I l f
.i l-
- . Table 5.3. Summary of Costs and Benefits-h,
)
Dollars Person-Rem (Millions)
Pubile Risk Reduction- 250,000 250(1) l l Occupational Radiation Exposure Reduction $se11(2) , Sma11(2).
l Net Industry Cost Savings 197(3)
NRC Costs .- ' .D.4 (4)
Net Benefit 447 (1) Person-rem converted to ' dollars on basis of $1000/ person-rem.
<. (2) Effect of maintenance rule on occupational exposures is estimated to be
.' beneficial, but relatively small cornpared to the reduction in public ri and the industry cost savings.
(3) Detailed breakdown contained in Table 5.1.
(4) Negative sign indicates a cost increase. Detailed breakdown contained Teble 5.2.
es an zit l .
8 1
f 1.0 STATEMENT OF THE PROBLEM "
1 b The Nuclear Regulatory Connission (NRC) has initiated a rulemaking on the maintenance of nuclear power plants. This report is a regulatory analg of the proposed rule. This section discusses the background of NRC's maintenance rulemaking, the need for a maintenance rule, and the contents G this regulatory analysis.
1.1 BACKGROUND
The importance of good maintenance in assuring safe and reliable plan 0
@ operations has been widely recognized for a long time ' In rec the rapid growth of the numb.er of operating plants in the U.S., the approa3 end of the nuclear construction era in the U.S.,-and the accumulation of significant reactor operating experience, maintenance has become the focus t increasing attention both from the NRC and from the industry.
1.1.1 NRC Maintenance Procrams Historically, NRC regulator attentiontomaintenancehasbeenbasedD
' a considerable extent on the qua ity assurance requirements of 10 CFR 50, Appendix B, the surveillance requirements in 10 CFR 50.36, and the inservia inspection requirements in 10 CFR 50.55a. While requirements such as thesa, along with related guidance, cover some of the major elements of nuclear pa plant maintenance, the regulatioris do not explicitly state NRC's espectatia on the full range of activities that fom the basis of a maintenance progrq In the early 1980s, a series of studies and programs undertaken by NRC, IN4 and others began to focus increased attention on maintenance, and led to 12 development of HRC's Maintenance and Surveillance Program Plan in 1985 (5E@t 85-129, April 12. 1985). The Plan was designed to address five principal' maintenance safety issues that were found to exist at some plants. The PICI sunrtarized these issues, and the corresponding Maintenance and Survellianco Program (M5P) objectives, as follows:
h *1. There is indication that needed maintenance is not being accomplished or is not performed effectively. The objective is to detemine the effectiveness of current nuclear power plant maintenance programs, to detect the causes and effects of equipment perfomance degradation, a3 to identify corrective action to minimize equipment failures and unavailability.
h 2. Many failures result from improper performance of maintenance. The objective is to reduce failures from improper maintenance by identifyic the fundamental causes of hutaan maintenance errors, by identifying practices which reduce the error rate and which increase the probabill(
of error detection prior to system demand, and to assess the effectiv0(
of Ilcensee strategies for improved maintenance performance.
@ 3. The interface between maintenance and operations is presently inadequat The objective is to detemine the causes of poor coordination among activities, to assure proper integration of maintenance, operations, a other organizational interfaces for maintenance activities.
1.1
l r
l l
l h 4. The number of maintenance-related challenges to safety systems i*s excessive. The objective is to determine the causes of the high rate of challenges and improve the effectiveness of nuclear power plant maintenance programs in assuring operability of safety systems.
~
b 5. A major portion of occupational radiation exposure (over 754) and manfie radiological hazards eccur to personnel perfoming maintenance activi (NUREG-0713). Improved planning and control of maintenance activities p may significantly reduce occupational exposure to as much as half of the present exposure. The objective is to assure that by performing prevent and corrective maintenance in e planned and optimited manner, the unnecessary and unanticipated radiological exposure of maintenance personnel will be reduced."
h The Maintenance and Surveillance Program Plan, Phase I, was implemented bepinning in 1g85. Results from Phase I were published in June 1g86 in a twc vt ame report, NUREG 1212. The basic objective of Phase I was to survey maintenance practices in the U.S. nuclear industry and evaluate their.
effectiveness; in so doing, Phase ! was also intended to address the princips maintenance safety issues identified in the Plan. Among the major elements of the survey and assessment were a review and evaluation of available data on maintenance status and performance; a review of the preventive maintenance proprem that was put in place at Salem following the 1g83 ATW5 event; development and administration of a questionnaire to obtain plant specific maintenance information from NRC resident inspectors; and eight site visits evaluate plant-specific maintenance programs using a systematic protocol to compile and structure the infomation. Phase I also integrated relevant information from a large number of related programs and activities in the NR:
and in the nuclear induttry. Examples of these related programs included a comparative analysis of U.5. and Japanese maintenance programs, NRC's Nucles' Plant Aging Research Program, and the maintenance activities of INPO, NUMARC and EFRI.
h NUREG-1212 reported the in-depth findings and conclusions of Phase !.
Several princirci conclusions were stressed. First overall industry as indicated b reliability and plant performance were found to be Improvinglso confir declining forced outage and scram rates. Howeser, Phase I a that the five maintenance safety problems identified in the Maintenance and Surveillance Pre; ram Plan were evident to varying degrees across the U.S.
nuclear industry. Wide variations were found in maintenance practices and effectiveness. Industry initiatives to improve maintenance were noted, but definitive assessments of the effectiveness of these programs were not yet
- possible. In summary, Phase ! confimed the existence of safety concerns rehted to maintenance and concluded that further NRC and industry attention wa'. needed to address these concerns, h in response to the findings of Phase I, the Commission directed the NRC staff to develop a Policy Statement to forsalize the Commission's position o maintenance. The Final Comission Policy Statement on Maintenance of Nucles Power Plants was published in the Federal Register on March 23,1g68. The Policy Statement noted that analyses of operational events have in some case 1.2
A indicated Reviews of individual plant maintenance (c%,
mt programs,asweaknesses dascribed in NUREG-1212 in maintenance.
for example, showed that the ef fectivent of the programs varied widely. In addition inadeguate maintenance was fount -
to be a significant contributor to reliabi1Ity problems and safety concerns.
The Policy Statement declared the Commission's intert to proceed with a rulemaking aimed at improving the effectiveness of maintenance programs throughout the nuclear industry. To provide interim guidance to industry -
while the rulemaking was underway, the Policy Statement also clarified the Commission's views on various maintenance issues, such as the definition and scope of maintenance and the elements of a maintenance program.
1.1.2 Industry initiatives b Since at least the early 1980s, industry has been actively engaged in a wide range of initiatives to improve nuclear power plant maintenance programt Many individual utilities have developed their own plant or utility specifir maintenance improvement programs. Industry groups such as IMPO, EPRI, and NUMARC have been particularly active.
The industry initiatives have been extensively reported at conferences
@ and at various hkC/ industry meetings, including the Public Worksho Maintenance Rulemaking held in Washington, D.C., July 11 13, 1988. The following examples provide an indication of the broad scope of the initiative
- The development of Guidelines for-the tenduct of Maintenance et Noelet,5 Po er Sta' ions, JhPO E5-038, a comprehensive ano cetailed guide to maintenance.
- Maintenance Self Assessment, an industry-wide effort by each utility to evaluate the adequacy of its maintenance program, based on the INPD guidelines.
- Maintenance Assistance and Review Teams.
- Maintenance Peer Evaluation Program.
- Outage Management Assistance.
-
- Maintenance Training Accreditation
- Maintenance Workshops
- Nuclear Maintenance Assistance.
h Collectively, these activities represent a substantial investment of resource and management commitment tomi improved mainter.ance. According to industry l representatives, full impleme..tation of the initiatives will require 3 to 5 years or longer. Although there remain areas for improvement, the NRC and the utilities agree that the initiatives have already resulted in substantia
- progress, with continued advances espected in the years ahead. NRC intends to recognize and take account of these industry initiatives in its regulator) l l
1.3 4
_ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ m _____.._ _ _ _ _ .___. __ _ _ _ _ _ _ _ . _ _ _ _ . . _ . . _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ .__._____.__._______-.___________________J
program, and in particular in structuring and implementing the proposed
- maintenance rule 1.2 NEED FOR IMPROVED MAINTENANCE hc In developing its regulatory position on nuclear power plant maintenance the NRC has identified a number of maintenance-related safety concerns that form the basis for its view that improvements in maintenance effectiveness are necessary.' These concerns have been discussed in the reports on Phase 1:
of the Maintenance and Surveillance Program (NUREG.1212, Volumes 1 and 2),
the Final Comission Policy Statement on Maintenance of Nuclear Power Plants (53 Federal Register 9430 March 23.1988) and the Public Workshop on NRC's MaintenanceRulemakingheld,inWashington,D.C., July 11-13, 1988.
1.2.1 Maintenance Practices and Effectiveness vary Widelv
-Variation in maintenance practices was extensively documented in NbREG-1212, based on a survey of 66 NRC resident inspectors at operating plants anc in-depth reviews of maintenance programs at eight sites. Among others,,
examples noted in NUREG-1212 included:
kN
- In the opinion of the resident inspectors, about one quarter of the plar:
wre judged to have an extensive preventive maintenance program,' about one half have an adequate preventive mair.tenance program, and about one quarter have a minimal preventive maintenance program.
A majority (63%) of the plants systematically monitor some type of
@ ~
- maintenance performance measures, while the restMost doplants, not.
however, have trouble correlating their perfomance measures with specti-aspects of their maintenance program, and analysi:
- k.
- While 70% ofonly of failures, thetwoplants reportedly of the sites use a asystem visited evidenced work for trend for trending and analysis of root causes.
@
- The common addition of 'inforsal labels' on plant equipment indicates that a large number (40%) of plants need to improve labeling programs.
One fourth of the plants were reported to have a high backlog of h
- maintenance work in the opinion of resident inspectors.
The data presented in NUREG.1212 were collected mainly in 1985 an'd the '
early part of 1986. In view of the significant efforts devoted by industry to maintenance initiatives in recent years, the findings in NUREG 1212 are-believed to be somewhat dated. Both industry and NRC have presented evidence I of improving overall maintenance performance in' recent years. Overall plant performance indicators monitored by NRC and INPO, for example, show clear signs of an improving trend in overall performance; stronger maintenance programs are recognized as one of the factors driving the improvement.
Despite the overall improvement, there remain wide variations in OGL performance that are of concern to the NRC. While the industry as a whole is 1.4 l @
l L
l l
h improving on average, a fraction of the plants are lagging behind the overal' Nb . trend. Figures 1.la and 1.lb show this point. Figure 1.la is a histogram tr shows the distribution of the quarterly average number of unplanned automatic scrams, one of the performance indicators monitored by NRC and INPO. Many plants average less than one scram per quarter. However, there is a certain; fraction of the plants that experience scrams at three to four times the ave' rate. Figure 1.lb shows the distribution of safety system failures, an indicator monitored by NRC. Again, most plants are concentrated at the left of the distribution with relatively few safety system failures, but a certait fraction of plants experience safety system failures at a noticeably higher rate. Examination of the distribution of perfomance for other perfomance indicators and other time periods gives similar results: a relatively small fraction of the plants lag behind the industry as a whole.
1.2.2 Maintenance Weaknesses Have Been implicated in Operational Events ht factorsMaintenance weaknesses have been reco nized as important contributing in a number of significant abnoma occurrences, such as the water.
hanner event at San Onofre Unit 1 on November 21, 1985, the overcooling evenn at Rancho Seco on December 26, 1985, and the loss of feedwater at Davis-Bess-on June 9,1985. These events and their contributing causes have been extensively documented. Maintenance has also played a role in other significant events, as well as in a large fraction of LERs.
h In addition to individual event analyses, there have also been studies by both industry and NRC that examined large numbers of events and identifie the role of maintenance on a broader basis. For example, in his presentation to the Public Workshcp on hRC's Maintenance Rulemaking, Mr. C. 0. Woody, Chaiman of the NUMARC Working Group on Maintenance, mentioned several studie by industry. One study examined the root causes of events occurring between 1980 and 1984 and determined that some 38% were maintenance related. NRC ha.
perfomed l imilar studies. A recent AE00 study of significant events involv-proceduresta) contaited the following findings with regard to maintenance procedures:
Q
- Based on a review of LERs from 1984 and 1985, 101 events were identifie that involved procedures. Of-these, 22 involved deficiencies in maintenance procedures.
Q
- Of 18 procedure related events involving automatic reactor trips,10 involved inadequate maintenance procedures.
Other NRC analyses have indicated that maintenance problems account for a substantial fraction of abnomal occurrences LERs. equipment failures, ESF actuations, safety system failures, and outages. Pertinent studies have considered wrong unit / wrong train events, events involving deficiencies in labeling and identification systems, and trends and patterns in trips and ES' actuations. ,
(a) E. A. Trager. 1988. Special Study Report: Significant Events involving Procedures. AEOD/5801.
1.5
_w.
4 it =
30 -
21 -
m.neer 23 "
ef Plants '
E-D-
9 1-;~ .
l c 1 2 2 d !
e.
Average b rte
- of S:rs's Pe* C.a'te*
Fice*e I la. = Histeg*8? Cf the 5.#e*3ge ku-Der,ef U41a* Dei A 1&fd1%C 1Crats f ur Cwarter for L'.$.
hTJ in aggi.
3* =
4
- t ='
t tunt,er cf Plants !
e -i l
i -
l C=
e 5- -
"""1 """)
c- a s u s to 15 - 20 2$ 30 35 C0 CS Average s.soer of Safety lysten Failures Per Quartar_
(ef, r e cu* e 1.16. . Histogree of the Average maner of Safety Systee f ailures per Guarter for U.S. Plant i
e.
I L
i
.., I I: 4 l _ _ _ _ _ _ . _ _ _ . . . _ . _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ . _ _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ . _ . _ _ _ _ _
m-- - : - g -- -
- . a-e 1.3 CONTENTS OF THIS REPORT This report conta' ins five major sections, which cover the re analysistopicsspecifiedintheReculatoryAnalysisGuidelinesokulatory the U.S.
Nuclear Reculatory Comission. Following this introductory section, which discusses the background and need for a maintenance rule, Section 2 states the objectives of the proposed maintenance rule. Section 3 describes the k1 -
principal alternatives that have been considered in the maintenance rulemakir Section 4 presents estimates of the anticipated effects of the rule on the health ar.d safety of the pubite and workers. Section 5 of the report estimat costs of the rule to industry and to the NRC; estimated cost savings to indus-are also included.
is e
1.7
u 2.0 OBJEC1!VES OF THE PROPOSED MAINTENANCE RULE ,
h In its Final Policy Statement on Maintenance of Nuclear Power Plants, the Commission declared that it believes safety can be enhanced by improving the effectiveness of maintenance programs throughout the industry. The purpe of the maintenance rule is to improve maintenance effectiveness, and thereby enhance overall safety, by establishing basic requirements for plant maintenance programs. 19 establishing these requirements the coarnission intends to consider the industry-wide efforts that have already been initiate The objectives of the maintenance rule are as follows:
- 1. To define NRC's expectations for plant maintenance.
- 2. To ensure timely actio'n by industry to upgrade maintenance where necessary.
- 3. To build upon industry initiatives to promote responsibility and ownerst by industry for resolving maintenance problems.
4.-To improve licensee maintenance programs.
Q, With implementation of the rule, it is expected that the current wide variation in maintenance performance will be reduced so that the performance of plants that lag behind the industry'as a whole will be brought up to the level of the majority of the industry. Second, the overall average level of industry's perfomance should also improve.
Q~/ maintenance performance in a way that allows licensees the flexibility cetermine the details of their individual maintenance program so that plant-specific factors can be taken into account. This flexible approach will enhance both safety and cost effectiveness, compared to a rigid and prescriptive rulemaking approach.
Also en irportant part of the structure of the rule is to recognize and (h';reinforcethepositiveinitiativesthattheindustryhasalreadyunderta The rule is intended to encourage the continuation and the timely completion of these initiatives.
G 2.1
3.0 ALTERNA11VES h Five principal rulemaking options have been considered. They are sumarized in this section of the regulatory analysis. Taken as a whole the options reflect a range of regulatory approaches, with differing degrees of prescriptiveness, and varying strategies for defining and measuring maintenan:
effectiveness.- Consistent with the Policy Statement on Maintenance, the optit all cover a broad scope.cf structures, systems, components, and activities, including balance-of-plant.
3.1 OPTION 1 -- GENERAL PERFORMANCE BASED RULE h Each licensee w%i' be required to (1) have an effective and documented
- maintenance program; (2) establish quantitative measures of the effectiveness of the program (i.e., maintenance performance indicators); (3) use the measur?
to periodically evaluate the program and, if appropriate, improve it; and (4)-.
report the maintenance pt formance indicators to the NRC on a quarterly basis This option is perf reance based because it' relies on maintenance performance indicators *s the mechanism for regulatory oversight of the .
' effectiveness of a licensee's program. Licensees would have considerable:
. flexibility in designing a maintenance program consistent with their own plan' specific needs and conditions, provided they could demonstrate through the performance indicators that the program was effective. This option also allo for licensee flexibility in selecting and reporting on measures of effectiveness. A Regulatory Guide would be written by the NRC to provide a recomended approach, but the licensee would be able to offer alternatives if appropriate.
A major feature of Option 1 is that it leaves some important points
@ undefined, allowing flexibility in the application of the Among rule. the:
principal open issues ares (1) What constitutes an effective malr.tenance -
program 7 and (2) How will NRC judge the acceptability of licensee performance 3.2 OPTION 2 -- PRESCRIPTIVE PERFORMANCE.8A!ED RULE d, % ) This option is similar to Option 1 in most respects, except that it prescribes the quantitative measures of mainter.ance effectiveness (maintenanc performanceindicators). The rule would spell out which indicators must be reported. There woyld probably be a separate document (e.g., a Regulatory Guide or a NUREG)(as that contained instructions on reporting formats, precis' definitions, and other details, similar to the guidance for LERs.
h The advantage of this option compared to Option 1 is that it would give the NRC a uniform information base on maintenance performance industry-wide.
This would ease the NRC's task of evaluating and interpreting the licensees' f (a) Such a document might be analogous to NUREG-1022, ' Licensee Event Report
~ System: Description of System and Guidelines for Reporting ' published by the Office for Analysis and Evaluation of Operational Data in 1g83, following enmpletion of the amended LER rule.
3.1 EL l
I a
I
_ - ._____ ___ -_-_-__ __-- - _ _ _ - _ _ ___-_ _--_-_-__ _ ____l _- - ______ _____ -__ -_____- _ _ _ _ _ _ _ _ _
__-__u_, _ _ _. _ . -.
reports. Licensees would still have the flexibility to track additional or alternative indicaters for their own purposes. A disadvantage of Option 2 1' that the existence of a uniform, industry-wide data base might lend itself t:
inappropriate comparisons or rankings of plants.
3.3 OPTION 3 .- GENERAL RUtE, WITH INDUSTRY $TANDARD As with Options 1 and 2, each licenser would be required to have an effective and documented maintenance program. NRC's regulatory position on the elements of an effective maintenance program would be defined in a Regulatory Guide that referenced an industry maintenance standard, after appropriate NRC review and endorsement. There would be no requirement to report maintenance performan,ce indicators.
NN By building on an industry standard, Option 3 provides an efficient mechanism for recognizing and incorporating the maintenance initiatives undertaken by industry in recent years. Many of the elements of such a standard are well-developed as a result of industry's efforts and could be factored into a standard. This option would also encourage standardization of industry maintenance practices. A disadvantage of Option 3 is that the completion of a standard is a time-consuming process; however,' development and validation of maintenance perfomance indicators and issuance of a Regulatory Guide, as required by Options 1 and 2 can also be a lengthy proce-3.4 OPTION 4 -- GENERAL RULE WITH REGUtATORY GU10E
'$ij
~
Option 4 resembles Option 3, except that NRC would not endorse en indus
standard in its Regulatory Guide. Rather, NRC would establish its regulator, position independently, without reference to indestry standards. A disadvan of Option 4 relative to Option 3, is that it might not fully incorporate th benefits of industry's experience ana detailed plant knowledge. However, it would relieve industry of the burden of developing a standard.
3.5 OPTION 5 -- PRESCRIPTIVE RULE Option 5 is to promulgate a detailed rule that would spell out complete M in the regulations the requirements for nuclear power plant ma programs. It would provide in the regulations a comprehensive definition of what is required of licensees, but could impact the industry initiatives currently underway. Major disadvantages of Option 5 are the lack of flexibi to accommodate varying conditiers and changing technology, and the difficult; of defining effective maintenance with the necessary specificity. Considerat time and NRC resources would be needed to develop a prescriptive rule.
3.2-
.c 4.0 HEALTH AND SAFETY The health and safety consequences of improved maintenance are discussed in this section. Two categories of consequences are described, and to the extent possible, quantitative estimates are developed. The two categories are:
- Reduced risks to the health and safety of the public
- Reduced occupational radiation exp6sures to workers.
4.1 PUBLIC HEALTH AND SAFETY Two approaches will bt vsed to assess the public health and safety benefits of improved maintenance:
- Statistical analysis of relationships between maintenance effectiveness and plant performance.
- 4.1.1 Statistical analysis The objective of the statistical analysis is to detemine - based on available measures of maintenance effectiveness and plant performance, whethe there is a relationship between the measures. Do plants with relativel maintenance programs tend to perfom better on an overall plant If abasi relationship is found, it can serve as a quantitative expression of the safet benefits of improved maintenance.
m (7 ej As a surrogate measure of maintenance effectiveness, the statistical analysis uses scores from SALP evaluations of maintenance and surveillance, the two functional areas most closely related to maintenance program effectiveness. (The SALP program is now combining these into a single functionalarea.) The measure of maintenance effectiveness is defined as the average score across the two functional areas. It is recognized that this SALP based measure has limitations as a measure of maintenance effectiveness.
However, SALP does reflect a comprehensive, integrated assessment that incorporates relevant information from many sources, including operating experience, inspection reports, enforcement history, and evaluations by regic based and resident inspectors. Thus, it should provide a measure of maintena that can, despite its approximate nature, distinguish between relatively high and low levels of performance.
~
p As measures of plant performance, the statistical analysis uses six of hRC's performance it.dicators:
- Automatic scrams while critical !
- Safety system actuations
- Significant events 4.1 l
l
g . _ -
l l
l
- Safety system failures
- Forced outage rate
- Equipment forced outages per 1000 critical hours.
h' measured In order to examine the relationship between maintenance quality, as by the SALP surrogate measure described above, and the six perfomanc indicators, a variety of modgig were fit to the data by regression analysis, using the 5 software system. tai The dependent variables were the performance indicators. The independent (* predictor") variables included the maintenance score, plant size, and a measure of plant age or vintage. The measure of. age was the number of years sity:e the issuance of the construction persit.
Looking at the results as a whole, the hypothesized relationship between
@d maintenance quality and overall plant performance is supported by th However, not all the indicators showed statistically sign (ficant relationship!
with the maintenance measure. Scrams, for example, were not related to the maintenance measure. However, this negative result would likely change if the scram measure for each plant were adjusted for the cmount of time the
' plant was critical; that is, the number of scrams in a given period should be divided by the number of hours the plant was critical, thereby normali. ring the scram n.easure. (It will be shown later that the maintenance measure is clearly related to plant capacity factor, and hence to hours critical.) Safe-system failures, forced outage rate, and safety system actuations, on the other hand, were found to be related to the maintenance measure.
In sumary, the statistical analysis supports the hypothesis that plants with better maintenance programs tend to perform better in terms of the overa plant performance indicators. This provides objective evidence of the safety benefits of improved maintenance.
4.1.2 Risk Estimates
- Maintenance does not readily lend itself to a comprehensive evaluation based on probabilistic risk assessment techniques, in part because of a lack of data, and in part because of the limitations of PRA models. If detailed data were available on the differences in equipment reliability / availability as a function of maintenance effectiveness, then PRA models could be exercise in a fairly straightforward manner to calculate the sensitivity of risk to differing levels of maintenance effectiveness. However, such data do not exist. There is strong evidence that maintenance quality affects equipment reliability / availability, but quantifying the size of the effect is difficult except in limited cases, g (a) R. A. Becker and J. H. Chambers. 1984. $r An Interactive Environw nt for Data Analysis and Graphics. Wadsworth Advanced Book Program, BeTinen California.
4.2
, ic ~l
.+ 4 1
- {
11
- i. k
$35 1
'l Q / ,, Two sources of evidence will be mentioned briefly here. .
One source iss ~
comparative studies of plant performance in countries with differing maintenance philosophies and practices. The example that is most often cite
- is the comparison of U.S. and Japanese reactors. As documented in NUREG/CR-
. 3883 and in the presentation by Mr. Omoto of Tokyo Electric Power Company at -
the Public Workshop on the Maintenance Rulemaking.,the average scram rate atO ,
R Japanese plants is an order of magnitude lower than in U.S. plante.: It isi generally agreed that this difference is due in part to better esintenance,
.although other factors also account for part of the difference. European; .
plants such as those in France and West Germany generally experience fewer t U than U.S. plants, although the comparisons must be interpreted cautiously because of technical differences between plants in the various countries;~
French plants, for example,. commonly operate in a load-following mode which.
' tends to raise the potenti,al for trips. Based on this type of comparative-information, one can reasonably argue that improved maintenance in U.S. plar would lead to fewer transients,- but the magnitude of the improvement cannot be quantified with any precision.
h Data from U.S. plants are another source of evidence. As an illustratit it is possible to correlate safety system failures, as measured by NRC's performance indicator program,, and maintenance quality, as measured by the SALP program. Maintenance quality can be approximately measured as the aver SALP score across two functional areas maintenance and surveillance. To explore possible relationships between these variables, regression analyses were performed,-using safety system failures as the dependent variable and maintenance quality as the independent (predictor) variable; the effects of plant age and size were also modeled by including them as additional.indeper.
variables in the regression analysis. Based on this analysis, it was found-that safety system failure rates for the best maintained plants tend to be roughly a factor of two to three lower than for plants that score poorly in-the SALP maintenarde and surveillance functional areas. .One can also examir.
directly (refer back to Figure 1.1) the distribution of safety system failur in U.S. plants and observe variations of roughly this same magnitude. (Note however, that the variation seen in Figure 1.1 is due to additional factors -
besides maintenance; thus, not all the spread in the data can be attributed differences between well-maintained and poorly-maintained plants.) Again,-
these data and analyses provide some quantitative evidence of the effect of-maintenance on reliability / availability, but they are not at the level of detail that one would like for a thorough PRA analysis.
1 'PRA models, as they currer,tly exist, do reflect some aspects of maintenance and surveillance,.e.g., unavailability contributions due to test and maintenance, but they do net capture the full range of maintenance impac -
- on the plant.
In an effort to roughly estimate the risk reduction that could be achie due to improved maintenance, two somewhat different calculations are discuss and compared.
As a first approximation, assume that the majority of plants' are alreac.
gg>perforwingreasonablywellintermsoftheirmaintenanceeffectivenessand that the maintenance rule will not affect them very much. Suppose, however, 4.3
-- - _ _ . _ _ . _ . _ _ _ = _ = ._ . - _ _ _ - - - _ _ _ _ _ . _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
- o Y
t a
that 20 to 30 plants have programs and performance.(substantial as Assume that room most of forthe improvement in derives risk reduction their main from improvements.in these 20 to 30 plants.
"E h Based on prior studies, baseline risks due to internal events'for typic
- plants are on the order of. several hundred person-rem per reactor-year. The1 following ranges appear in draft NUREG-1150, Volume 2,1b) Appendices B - F, are one of several risk measures reported there ~
- Surry(withdirectheating) 30 - 150 person res/ reactor-year
- Surry (without direct heating)' 30. s 90 person res/ reactor year
- Zion - 7' - 200 person-rom / reactor-year
- Sequoyah 100 - 900 person-res/ reactor-year
- Peach Bcttom 3 - 280 person rem / reactor-year a Grand Gulf 2 - 90 person-res/ reactor year e m t g These values are based on suming exposures out to a rad ys of 50 miles, followin the comon practice in value-impact analysis.t i E ier PRAs als-yield si ilar estimates. For example,.NUREG/CR-4330, Vol. 2, presents calculations based on the Reactor Safety Study for Surry and Peach Bottom,-
and based on the R$$MP PRAs of Oconee and Grand Gulf. The estimated risks range from 71 to 250 person rem per reactor year. Note that all of these estimates consider internal events only. Consideration of risks due to exte-h (a) This estimate, although uncertain, is consistent with data presented in-NUREG 1212. Section 1.2.1 of this re vlatory anal sis discusses severs of the areas for improvement identifi d in NUREG-1 12, and quotes the-fractions of plants that may need improvement. Chapter 5 of this, regulatory analysis discusses these data more extensively.' For example NUREG-1212 found that 22% of the plants had ' minimal preventive mainten programs.' The precise number of plants that will be required by the rule to make substantial improvements in maintenance will depend on the eventual contents of the industry standard, and the success of industry initiatives in improving maintenance performance. NRC Maintenance Team-l: Inspections are currently being performed. These will provide additions information for assessing maintenance effectiveness industry-wide.
(b) U. 5. Nuclear Regulatory Commission. 1987. Reactor Risk Reference
" Documentt Draft for Comment. NUREG 1150, Volume 2. Office of nuclear Regulatory Research, february 1987.
(c) See A Handbook for Value impact Assessment, NUREG/CR 3568, PNL-4646, Pacific horthwest Laboratory.
(d) Mullen et al. 1986. Review of Licht Water Reactor Reculatory Requirements: Assessment of 5 elected Requirements That May Have Marcina' Importance to Risk. NUREG/CR 4330, Vol. 2. Pacific horthwest_Laborator Richiand, Washington.
4.4 j
1-I' w o
' events, which are not quantified in most PRAs, would increase these e
'Jb Assume for purposes of this discussion' that 300 person-ree/ reactor. year
- is representative of the ' typical" or.
- average" risk level at U.S. plants.
l' If the 20 to 30 plants with substantial room for improvement in maintenance:
have risks two to five times as high (this assumption will be discussed furthe in a moment),' those risks would be some 600 to !$00 person res/ reactor-year..
- l. .The potential risk reduction, if these plants could be improved to the level-of the typical plants, is then 300 to 1200 person res per. reactor year. :Since maintenance is likely to be only one of the possible reasons for the increasec '
risk at these plants, not'all of the potential risk reduction would be.
. achievable through improved maintenance alone. .However, given that maintenanc accounts for.a substantial fraction of all activity in the plant (70% accordin.-
to a presentation at the Public Workshop by the chairman of NUMARC's Working -
Group on Maintenance), improvements in maintenance might reduce the 'entra" .
risk by two thirds, i.e., by 200 to 800 person. rem per reactor-year. -Assuming 20 to 30 reactors and an average remaining life of 30 years, the total potenti benefit is on the order of 120,000 to 720,000 person-rem.
.%W . maintenance There are at'least two reasons why. risks might be higher at plants whose.
performance needs improvement: (1)asmentionedearlier, plants; with weak maintenance tend to experience safety system failures at a higher -
rate than typical plants, and (2) poor maintenance may contribute to a higher-rate of. transients. Based on the statistical analyses discussed previously,1 which examined the relationship between safety system failures and maintenance"
- quality, it was estimated that plants with weak maintenance could have safety:
system failure rates 2 to 3 times'as high as the industry as a whole.
G i' With regard to transients, statistical analyses were performed of the relationship between maintenance quality (based on the SALP measure) and the.
number of scrams (as measured by NRC's scram indicator, which is-defined as the number of unplanned automatic scrams while the plant is critical). No statistically significant relationship was found. Plants with weak maintenanc.
programs, although they may experience scrams at a higher rate while they are critical than do the plants with' strong maintenance programs, tend to have low -
availability and capacity factors, i.e., not as many hours critical. .$ince they have fewer operating hours, the total number of scrams is not as great as it would otherwise be.
~
f To explore the risk impacts of safety system failures, risk sensitivity calculations were performed, with the help of computer programs containing the dominant accident sequences and cut sets from PRAs for Oconee and Grand Gulf. Parameter values in the PRAs corresponding to safety system failures were increased and risk levels were recalculated. To estimate population doses, the dose consequence factors described in the Handbook for Value. impact Assessment, NUREG/CR 3558, were used. . These factors were based on WASH 1400 release categories and CRAC2 calculations that assumed a unifors population distribution of 340 persons per square mile, meteorology of a typical sidwest site (Braidwood), an exclusion area of 0.5 miles, and other generic assumptions g y ear,The increase assuming in riskofranged a factor from about two increase in the300 ratetoof1400 person safety res system per react failures.
4.5 M
1 i
- )
1 l,
'h Note that these were older PRAs than those reflected in NUREG 1150. These estimated increases are roughly consistent with the factor of 2 to 5 mentiont i
1' earlier. 3 As a s'econd point of reference, a prioritirat .
OM MaintenanceandSurveillanceProgramwasreviewed.joganalysisofNRC's'Lai As part of the NRC's . .
. program to prioritize generic safety issues, calculations were performed to
-estimate potential rise reduction from implementation of NRC's Maintenance-and $uryeillance Program Plan. The methodology was that described in NUREG/C 2800.(b) Key assumptions in the calculations were that (1)plantagingwoui increase the frequency of transients caused by the balance of-plant by 10%; .
(2) full implementation of the Program Plan would result in substantial - i l-
' reductions (by a factor of 2.7) in the frequency of transients; and-(3) ther would also be a small reduction in unavailability due to maintenance downtim(- e Based on these assumptions, risk sensitivity calculations were performed usfr:
. representative PWR and BWR PRA models.: The total estimated risk reduction-over the remaining lifetimes of U.S. reactors was'on the order of 300,000 person rem..
.g' . . These estimates are all in the range of 100,000 to 1,000,000' person rem.
and considering the uncertainties in'such calculations are reasonably consistent.
@% In con'sidering the benefits of the maintenance rule, it should be recall that industry is already showing improvements in maintenance based on initiatives currently underway. Thus, a substantial share of the risk-reduction estimated above will be gained under the status quo, even without a-maintenance rule. It is difficult to estimate precisely how much of the c benefit will be gained even in the absence of a maintenance rule. This will depend on the timeliness and effectiveness of the ongoing industry initiative.
However, it seems reasonable to suppese that the industry initiatives will reduce the number of maintenance outliers significantly over the next few years. Thus, the above estimates of risk reduction are probably too high by a factor of two or more.
h .In suneary, for purposes of this regulatory analysis, the risk reduction due to the n;aintenance rule will be estimated as 50,000 to 500,000 person-rem; 250,000 will be used as a point estimate.
(a) R. Emrit et al. A Prioritiration of Generic Safety issues. NUREG-0933.
The prioritization of the Maintenance and Surveillance Program appears in Section 4 as issue HF-02.
(b) W. B. Andrews et al. 1983. Guidelines for Nuclear Power Plant Safety' Issue Prioritiration Information Development. NUREG/CR 2800, PhL-42 H .
Pacific horthwest Laboratory, Richland, Washington.
4.6 1
, y __- _ _ __ _ -
- _ - - _ _ _ _ ~ _
h h ^
-4.2 OCCUPATIONAL EXPOSURE (ROUTIhE) -
hl . A large fraction of the occupational radiation exposure incurred at nuclear power plants is associated with maintenance. : Table 4.1 from NUREG .
0713, Volume 0,.shows the percentages of total plant collective, dose received by various work functions between 1978 and 1984.- Combining the 2 functions-of Routine Maintenance and $pecial Maintenance, one can see that maintenance accounts for two thirds to three fourths of the collective dose. If contributions ~ from inservice inspection and surveillance are added, the fract is even higher. Collective doses per plant have been declining significantly -
in recent years,-as can be seen in Appendix A. In 1987, according to INPO l
F . figures, the average collective dose at PWRs was 368 person res per. unit, whi at SWRs the collective dose averaged 521 person res per unit.'
h . Improvements in maint'enance programs can affect occupational exposures both positively and negatively. Expanded preventive maintenance and more-aggressive corrective maintenance to reduce backlogs will tend to increase-exposures, at least initially. since there will be more activity in the plant
. 0ver time, however, this trend may be reversed as more of the maintenance -
activities in the plant are planned and optimited as part of the preventive
' maintenance program.- Improvements in maintenance efficiency and productivity and rstductions.in reworki will also tend to reduce exposures. The net effect of these positive and negative trends is believed to be positive, but small .
compared to the other costs and benefits of maintenance upgrades.
h To achieve large reductions in collective occupational exposures,'it is necessary to look beyond the maintenance program per se and and consiSr othet '
measures, e.g., reduce contamination throughout the plant, facilitate easy -
access by workers so that jobs can be accomplished more quickly, improve -
training, provide better shielding, tools, methods and equipment (including robotic technology where appropriate) and if possible institute design modificationsthateliminateorminimIretneneedfortaskslikepipingrepeii-and inspection or snubber testing that cause high exposures.
An excellent example of this comprehensive approach to exposure reducifor.
[ was presented by Mr. Ometo of Tokyo Electric Power C Workshop on EC's Maintenance Rulemaking. Exposures during the first annual outage at TEPC0 plants have been reduced by an order of magnitude through such actions as installing a condensate refiltering system, reducing cobalt content in selected components, using more corrosign. resistant materials, and providing additional work space in selected areas.ta)
To illustrate the point that occupational exposure benefits from improvec
, maintenance are likely to be small compared to the other costs and benefits, b (a) See also: D. W. Moeller, " Achieving Progress in Radiation Protection --
The Next Steps,' Nuclear Plant Journal, March. April 1987, p. 36; H. Ocksr and C.J. Wood, " Radiation fiele Reduction,' Nuclear Plant Journal March.
April 1988, p. 43; E. I. Hunt, ' Reducing Personnel Exposure by Reducino Cobalt Levels in primary Systems,' Nuclear Plant Journal, May. June 1987,
- p. 24.
4.7
p 4
i Table 4.1 PERCENTAGES OF ANNUAL' COLLECTIVE' DOSE AT LWR 5 BY WORK FUNCTIONI.
Percent of Collective Dose Each Year.
Work Function 1978 -E 8 1980. ))81 g .JJ13 1984 Reacto'r Operations and Surveillance 13.3% 12.2% 9.5% 8.9%- 9.4% 10.1% 11.45 Routine Maintenance 31.5%. 29.2% 35.5% 36.1% 27.9% 29.7% 26.95 Inservice Inspection 7*.i4 9.0% 5.5% 5.3% 6.5% 7.6% '6.3t Special Maintenance 35.9% 39.4% 40.6% 40.5% 46.8% 43.9% .45.45 Waste Processing 5.0% 3.6% 3.0%' 4.2% 5.0% 4.6% 3.6' Refueling 6.6% 6.6%. 6.1% 5.0% 4.4% 4.1% 6.4' 1From NUREG-0713, Occupational Radiation Exposure at Comercial Nuclear Power Rea:ters and Other Facilities 1984 9
4.8 1
l
_ _J
S.
a bounding calculation will be provided. Supposethatthereactorinques@
has an average yearly collective dose of 400 person rem, which is intermedi:
between the 1987 figures for PWRs and BWRs. As discussed above, large reductions in these exposures will require actions that go beyond maintenaa to include design changes, advanced equipment and methods (e.g., remote ND1 and similar measures. It does not seem reasonable to attribute these kinda of potential benefits to the maintenance rule. Suppose that improved .
maintenance alone (e.g., through.such measures as better training, optisfra maintenance intervals, and reduced amou6ts cf rework) could reduce occupati:
exposures by 10%. The risk reduction would then be on the order of 40 peso rem per reactor-year, a fraction of the public risk reduction values discua!
above (i.e., 200 to 800 person rem per reactor-year). To convert person-0G of occupational exposure to dollars, a range of values has been considered <
the literature, from $100 par person-rem to several thousand dellars. A nD value of $1000 per person rem will be used here. Assuming again that 20 - j plants are affected, with an average remaining plant life of 30 years, andt estimated exposure reduction of 40 person-rem per reactor year, the estima3 occupational exposure benefit is 24,000 to 35,000 person rem, or $24 millis to $36 million, which is small relative to the other cost-benefit componen@
@ be relative y small, they are not to be ignored.Althou h the Rather, it will occupat be assume here that the intent of the maintenance rule'is to ensure that due weight @
given to occupational exposures in making decisions affecting or involving maintenance. There may be individual situations in which occupational exposures are the dominant factor in planning a particular maintenance activity. All nuclear power plants have radiation protection and ALARA programs, based on 10 CFR Part 20 and Regulatory Guide 8.8. Changes in maintenance programs will create opportunities to achieve additional cost-effective reductions in occupational exposure. By integrating ALARA considerations into the changes in the maintenance programs, it will be possible to optimize the changes so that occupational exposure benefits arO gained.
4.3 OCCUPAT10Nat EXPOSURE (ACCIDENTAL)
In addition to the routine ~ exposures discussed in section 4.2, workera can also incur an occupational exposure due to acctdents. There are two components of occupational exposure related to accidents, imediate and loa tern. The first occurs at the time of the accident and during the immedia@
management Of the emergenry. The second is a long tern exposure, at lower dose rates, associated with the cleanup and refurbishment or decomissionig of the damaged facility. The Handbeek for value impact Assessment estimata a collective exposure on the order of 20,000 person rem for these two components, most of which is attributable to the post-accident cleanup and refurbishment or decommissioning. When this value is multiplied by the ch3 in accident probability attributable to the maintenance rule, the result is negligibly small.
h The change in accident probability attributable to the maintenance rul(
can be bounded in the following stoner. Draft NUREG 1150 as well as earlia PRAs present estimates of the probability of severe core damage due to inte{
4.9 G
cvents. Tha Gstimatss ar2 typically in the range of 1/10,000 to 1/200,000 ONb)\ per reacter-year.. As in section 4.1, it cay be estimated that the risks in plants with substantial room for improvement in maintenance are 2 to 5 times as high, and that the maintenance rule will reduce the " extra
- risk by two-thirds. :The reduction in risk from accident-related occupational exposure .
is then on the order of a few person-rem per reactor-year, which is negligible compared to the public risk reduction estimated in section 4.1.
L',
9
,.,.,.m... ,,.
6 e
4.10 l
l.
l-l 1
I
- L_______.._____.______-..__-..-____ _ _ i
- y. , p .
~
y m
t
, f Q1 i.
r 0i 5.0 [0$T}
Estimated costs of the proposed maintenance rule are discussed in this 1 1 section. In Section 5.1, the major cost elements associated with the rulemaking options are identified. The analysis of industry costs is present-in Section 5.2. NRC cost estimates'are presented in Section 5.3.: The-total costs of the options are sumarized in Section 5.4.
5.1 IDENTIFICATION 0F MAJOR COST ELEMENTS.
... .in this section, each rulemaking option is'briefly discussed and the major cost elements that'it would entail are defined.
. 5.1.I' Option 1 -- General Perfomance Based Rule
- h. maintenance Eachlicenseewould-bkrequiEedtohaveaneffectiveanddocumented program, to establish measures to evaluate and improve the effectiveness of the program, and to report periodically on its effectiveness A Regulatory Guide would set forth acceptable methods for measuring the-effectiveness of maintenance programs,'and each licensee would either follow the Guide er establish its own alternative measuring and reporting system.
Elements Licensees would develop and maint'ain documentation on.the maintenance _
program. . Initially, this would involve the development of a maintenance-i plan. The plan would subsequently be updated from time to time. The plan would provide for a maintenance recordkeeping system to document 1 the activities of the maintenance program on an ongoing basis.
i k *- Licensees would develop and implement a reporting system on the-effectiveness of the maintenance program.
M'e N- - NRC wou?d develop the Regulatory Guide. Industry would have some form of input and participation.in this effort. .There might also be a need for supplem NUREG 1022.gn}ary ta; instructions, analogous to those provided for LERs in' ha NRC would analyze the licensee reports, and prepare a quarterly sumary report.
~I#* NRC would perform maintenance inspections, based in part on the licensees-reports of maintenance effectiveness.
l
@* Industry would implement maintenance improvements as needed to ensure maintenance effectiveness.
(a) " Licensee Event Report System: Description of System and Guidelines for Reporting,' NUREG-1022, Office for Analysis and Evaluation of Operational Data, U. 5. Nuclear Regulatory Comission,1983.
5.1 1
L
1 i
I
-]
5.1.2 ' Option 2 -- Prescriptive Perfomance Based Rule. j Option 2 is similar to Option 1, except that the rule would prescribe '
the maintenance reporting requirements in some detail. Option 2 would establish uniform reporting requirements, in contrast to Option 1.which wou - 4 allow more flexibility and plant to. plant variation. Thisoptionwouldrequj essentially the same actions by industry and NRC as Option 1. A Regulatory,-1 Guide or NUREG would probably s(till be needed to spell out certain reporting d precise definitions of indicators, data formats, etc. '
5.1.3 Option 3 -- General Rulec With Industry Standard h Option 3 would require each licensee to establish, document, and impleme a maintenance program. The. rule would not prescribe in detail the scope of required maintenance activities and the manner in which they should be conducted. Instead, NRC would prepare a Regulatory Guide endorsing an indust standard.
Elements h-
- Industry would develop a maintenance standar' d consistent 'with the framework in the Maintenance Policy Statement and rule. The standard-would be developed under the auspices of NUMARC, INPO, EPRI, or some -
other entity.
P Q
- NRC would review the standard and, assuming the standard is found to be acceptable, would endorse it. While industry develops its standard, NRC would work in parallel to develop the technical basis for its review and evaluation of the standard.
~.
kK
+
Licensees would develop and maintain' documentation on the maintenance program. Initially, this would involve the development of a maintenance.
plan. The plan would subsequently be updated from time to time. The-plan would provide for a maintenance recordkeeping system to document the activities of the maintenance program on an ongoing basis.
NRC would perfom maintenance inspections.
Industry would implement maintenance improvements as needed to ensure maintenance effectiveness.
5.1.4 Ootion 4 -- General Rule. With Reeulaterv Guide h Option 4 is similar to Option 3, except that no indust standard would be developed. Insteed, NRC wculd develop a relatively deta ed Regulatory Guide, or perhaps a series of guides at varying levels of detail, that descrit the scope of the maintenance activities that should be conducted and provide guidance on acceptable maintenance practices.
5.2
4 o
I l
1 I .
5.1.5 Option 5 -- Prescriptive Rsle NRC would develop a' detailed, prescriptive rule spelling out specific requirements that licensees would have to meet. Because the requirements would be embodied in the regulations, flexibility to adopt alternative maintenance approaches would be limited, although exemptions would still be available via the usual mechanisms (10 CFR 50.12). This option will not be considered in detail in this analysis.
5.2 INDUSTRY COSTS h Costs to the industry both to implement specific elements of the five
, tions, op such as developing recordkeeping and reporting systems, and to upgr.
maintenance programs to ensure that they are effective, are assessed in this section; ranges for the estimates are discussed in Appendix A. All costs.
are given in constant 1988 dollars, discounted at an annual rate of 10% as appropriate. Unless otherwise indicated, industry labor costs are assessed I at 554 per fully-loaded staff-hour, or $108,000 per staff year; for addition details en labor costs, see Generic Cost Estimates, NUREG/CR-4627.
5.2.1 Costs to implement specific Option Elements h
.Kk Industry will incur costs to develop and maintain maintenance plans, to develop and operatt recordkeeping and reporting systems, and to develop a maintenance standart'. Costs will also be incurred in providing input.to the NRC during its development of Regulatory Guides.
Maintenance Plan Development Costs kJ A
For all options, licensees will need to develop and document their maintenance program in a plan that will be available to NRC inspectors. Bec:
of the variability among licensees in the formality of current maintenance programs, and in the comprehensiveness of various maintenance program element; it is assumed that two thirds of the plants will require a modest effort to develop the plan, with the remaining third requiring a more substantial effo-The modest effort is assumed to require 3 staff months per plant, or $27,000 and the more substantial effort is assumed to require 6 staff-months per plar or 554,000, for a total industry cost of about 54 million for 110 plants.
Recordkeepino and Reportino Costs
- N
~
kT4 Under Options 1 and 2, licensees will need to implement recordkeeping ,
and reporting systems for the maintenance indicators. The costs to develop and implement these systems include the following:
- Costs to revise and develop recordkeeping procedures
- Costs to modify current maintenance databases
- Training costs for maintenance recordkeeping systems.
- Costs to collect and record data
- Coststopurchasenewequipment(computers, terminals,etc.)
- Equipment maintenance and repair costs 5.3
l
)
i l
It is assumed that no new recordkeepin procedures are needed for those i
mainten6x e indicators currently used by IN 0. Each plant is assumed to ]
l require four new maintenance recordkeeping procedures that describe the data l input needed to calculate new maintenance indicators. In addition, it is I assumed that revisions will be required to one procedure that describes the '1 i overall operation of the recordkeeping system. . Abstract 2.2.2 of the Generic 1 CostEstimatesprovidesestimatesofthecoststowriteorrewritemaintenancti procedures. Routine changes are assumed to cost $900, with a range of $750 to $1,000, and complex changes are assumed to cost $3,600, with a range of
$3,100 to $3,600.(in 1986 dollars). It is assumed that the new procedure
. costs are intermediate between the simple and complex change costs, and are . ;
estimatedtobe$2,300($2,400in1988 dollars),andthattheprocedurerevis
is a simple change ($900 in 1988 dollars). .Thus, the total per-plant cost to develop and revise maintenance recordkeeping procedures is estimated to be.
$10,100, and the total for,110 plants is about $1.1 million.
h Additional costs will be incurred in modifyi g current plant databases to accept the additional maintenance indicator in ornation and to produce the ;'
quarterly reports. It is estimated that 25 staff-weeks will be required to complete this task, for a per-plant cost of $54,000; for 110 plants, the. tota is 55,940,000.
(Q Training on the new system is assumed to require 2 staff-weeks for each v of 3 data entry technicians. At an assumed rate of $31 per fully-loaded staf.
hour, as suggested by Abstract 6.3 of the Generic Cost Estimates, the total-cost per plant is estimated to be $7,400. The total estimateo training cost.(
for 110 plants is thus $814,000.
Q ndicator i data and to subit the quarterly reports to the NRC. Operational
. The annual cost to collect the additional performance indicator date is estimated to be 1/2 staff-year, at $54 per fully-loaded staff hour. Data entry costs are assumed to be 1/4 staff-year for a data entry technician ($31 per hour).-
Report generation costs are assumed to be an additional 2 staff-weeks per quarter, or 8 staff-weeks per year, at $54 per hour. Therefore, the total per plant data collection and report generation cost is estimated to be $86,8 per year. The total discounted cost over 30 years for 110 plants is thus roughly $90 million.
~
f~; It is assumed that the maintenan:e indicator infomation can be entered via the terminals that are currently used to enter data for INPO maintenance indicators, and that the current computer system is adequate to store the information, so that no additional equipment purchase or maintenance costs are incurred. Thus, the total implementation and operating costs for the
. maintenance indicator information would be about $98 million.
t The above recordkeeping and reporting costs are applicable to Options 1 and 2. There is an additional recordkeeping cost element that applies to all four rulemaking options. For purposes of estimating information collection burden, as required under the Paperwork Reduction Act of 1980, the burden for recordkeeping required by the proposed maintenance rule is estimated to avera; 5.4
, , 4 i
3:
h 1,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> per plant per year, including the time for reviewing. instructions, gg searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Since many plants, E . already collect the needed information, the above estimate is assumed to appi) to a fraction of.the plants'(i.e., 25). Based on a labor rate of $54 per fully-loaded staff-hour, an average remaining plant . lifetime of 30 years,-and'
. a real discount rate of 10%, the total estimated cost is $12.7 million (l'000 ,
x 25 x $54 x 9.43).
- h. optionWith this addition of approximately $13 million to all four rulemaking
% the total cost estimates for recordkeeping and reporting are roughly
$111 million for Options 1 and 2, and $13 elliton for Options 3 and 4.
Industry Standard Development Costs
~ hL . Under Option 3. it is assumed that an industry standard would be develope that would address the areas specified by the NRC in the Policy Statement and the rule. The standard could be developed under the aegis of an industry group such as NUMARC, INPO, or EpRI. It is assumed for purposes of the regulatory analysis that the standard would be developed over a two year peric
.m h It is assumed that the costs to develop the standard include personne'l costs for a 20 person steering group and for 10-15 working groups with 6 members each. It is assumed that the steering group will meet 5-8 days per.
L. year, and the working groups will meet 10-12 days per year. Additional time will probably be spent outside the group meetings to prepare' documents; this is assumed to be I week per year for each person. In additica, it is assumed that the steering group will meet once per year with the NRC for two days each, and that a two day workshop will be held each year to solicit input from all licensees. Two persons from each of approximately'50 utilities are ass 9med to attend the workshops.
Thus, the total industry effort, exclusive of travel, is assumed to be 1500 3390 industry staff-days per year, or an annual cost of about $648,000 -
'51.46 million. Assuming a travel and per diem cost of $800 per trip, the
. annual travel costs are estimated to be $272,000 to $498.000. The total' discounted cost to industry to develop the standard is thus estimated to be
- $1.76 million to $3.73 million, with a nominal cost of about $2.75 million.
Reculatory Guide Development Costs h it is assumed that 2 workshops will be held to obtain industry input to the development of Regulatory Guides. Each workshop will last two days, and two staff members from each plant will attend. Thus the total industry cost k to attend the workshops is estimated to be aboct $380,000.
5.2.2 Costs to implement Needed Maintenance Improvements h Regardless of which regulatory option is selected, licensees with maintenance programs that do not satisfy the requirement for a 'well-defined
[
i 5.5 L_-_._-____-_-___-_----. _ . - - _ _ . . - _ - - - . . _ _ _ _ _ . .._ - - _ _ . - - - . _ _ _ _ . _ _ _ _ - . . _ _ - - - _
and effective
- maintenance program (a) will have to make certain improvements.
These improvements are the ultimate goal of the maintenance rule and, without question, will account for a dominant share of both the costs and benefits.
A r@
It is difficult to estimate with any precision the specific actions that each plant will have to take to improve its maintenance program based on the rule. There are two principal reasons for this. First, available data on 4 plant-specific maintenance programs and practices are limited. Second, at. '
this early stage in the maintenance rulemaking process,.there is not yet a /
clear, detailed definition of what constitutes an effective maintenance program. Nevertheless, in order to develop cost estimates for the regulatory analysis, it is necessary to make some assumptions about the maintenance improvements that will be needed.
As a guide for formulating these assumptions, we rely primarily on information in NUREG-1212, *5tatus of Maintenance in the U.S. Nuclear Power Industry 1985." Although the information in NUREG 1212 was collected in 1985 and early 1986 and is therefore somewhat dated, it remains the most complete and comprehensive publicly available compilation of data on U.S. maintenance practices. NRC is currently conducting maintenance team inspections of a sample of plants, and information from these is expected to become available as the rulemaking proceeds. Thus, the assumptions used in the regulatory
- analysis can be reviewed and refined based on more up-to-date information as the rulemaking progresses.
h in the sections that' follow, the relevant coments from NUREG 1212 are first quoted. Then the assumptions and data used in developing the cost estimates are discussed.
Preventive Maintenance Cg, g .
- 'Almost all (93%) plants have a formal preventive maintenance program, although at only two-thirds of the plants do these programs have written goals and objectives. Most (755,) plants also have a formal predictive maintenance program, with approximately half (48%) having written documents specifying geals and objectives. One quarter of the plants have no formal program in predictive maintenance.... Note that the majority of plants spend most staff time on corrective maintenance and no plants reported spending more than 50% of staff time on preventive maintenance.... The maintenance staff at site survey plants currently spend 10% - 50% of their time on preventive maintenance. However, maintenance managers et almost all site survey plants reported that they have future goals which range from !O4 - 90%. Several maintenance managers noted that an excessive amount of backlogged work was hindering h (a) 'Esch comercial nuclear power plant should develop and implement a well defined and effective program to assure that maintenance activities are conducted to preserve or restore the availability, performance and reliability of plant structures, systems, and components." Final Comission Policy Statement on Maintenance of Nuclear Power Plants, $3 Federal Register 9430.
5.6 4
I
T' ,
i their ability to increase the amount of time spe'nt on preventive maintenance.... The resident inspectors indicated that 22% of the plant.
have extensive preventive maintenance pro programs, and 224 have minimal programs." grams, 56% have adequate Assumption: The rule will increase the level of preventive maintenance (PM) that is considered appropriate, and all of the plants with minimal preventive maintenance programs according to .
NUREG 1212 will now need to upgrade them significantly. Based on the 1985 data, this translates to 22% of the plants. Improvements ..
in maintenance since 1985 will tend to lower this percentage, but rising expectations for preventive maintenance will tend to raise it. Therefore, 25% will be used as the base case assumption.f h The cost of upgrading a preventive maintenance program will vary.
considerably depending on plant-specific circumstances and choices. Many of the benefits will be economic (e.
maintenance and reduced downtime)g., cost savings due considerations
, and cost-effectiveness to reduced corrective ma hcve a controlling influence on the detailed implementation of the program.y These considerations are highly plant-specific and system-specific. EPRI has published *A Guide for Developing Preventive Maintenance Programs in Electric Power Plants," EPRI hP-3146, which explains how to develop a cost-effective
- program at an individual plant. Reliability-centered maintenance programs.
which originated in the aerospace industry and are now being applied in the nuclear industry, also provide similar features.
Q} One approach implemented at a large, two-unit PWR (Salem) i of the substantial costs that can be incurred in some cases.(a) First, somes ind 65 plant systems were categorized into 3 groups, based on safety significance.
A component data base was then developed on more than 30,000 components requiring preventive maintenance. Procedures and orders were developed for maintaining the components. The whole system was computerized. The initial effort to set up the system, including compiling and verifying the data base, and writing procedures and orders was on the order of 130,000 staff-hours, and it took one and a hclf years. At $54 per staff-hour, this effort amounts to aboutto$7 estimated million.
require Ongoing 5 people implementation to maintain of the upgraded and operate the computer system, PM
@ (a) J. D. Driscoll. ' Salem Generating Station Managed Maintenance Program."
In Transactions of the American Nuclear Society 12th Biennial Conference on Eeactor Opera *.ino Deerience. August 4-7, 1955. Williamsburg, Virginia.
OLO (b) These estimates can be compared with estimates for a similar system i implemented at Fort Calhoun. That system, known as CHAMPS (computerfred I history and maintenance planning system), requires 2 full-time clerks and two part-time maintenance engineers at the plant to process maintenanc orders, issue and collect preventive maintenance and surveillance tests, prepare reports, maintain daily files, support users, and define needed system improvements. In addition to the onsite staff, corporate support staff are also required: one project engineer, two programer-analysts,.
and one clerk. See: R. L. Jaworski, J. K, Gasper, and M. A. Klanderud.
5.7 dD l
_ J
y l
and 40 technicians / mechanics fulltime to do the maintenance during routine operations (during an outage the number of technicians / mechanics increases to 100). This translates to roughly $4.9 million per year. Assuming 30 years of remaining plant life, the total cost of the program (present value) is on the order of $$$ million. It should be noted that cost savings due to reduced.
corrective maintenance will compensate for part of this cost; the magnitude of this savings was not reported in the reference. An [PRI report on reliability-centered maintenance (RCM) estimated, based on experience in other industries, 30% to.40%.(that a) In RCH couldthe addition, reduce EPRIcorrective maintenance report pointed out thecosts potential by as savings much as associated with reduced downtime; this savings is estimated later in section 5.2.2.
L An example at a more detailed level may help to clarify the reason why a full-scale preventive maintenance program can be very ecstly. One licensee spent more than 4 staff-months developing a computerized system to track plant instrument calibration status and writing supporting procedures that described the use of the system and provided instructions on the calibration of pressure, temperature. and other measuring instrumentation. The system produced periodit-reports that indicated the instruments that needed to be calibrated, their previous calibration results, and the procedures to be used for each instrument The results would be monitored and used to re-evaluate the calibration frequent for instruments that consistently registered as in-calibration or out-of-calibration when checked. Assuming a salary of $54 per staff-hour, this progrt cost the licensee a minimum of $36,000. This example would have to be multiplied many times to arrive at a comprehensive program that covered an entire plant.
LQ Thus, assuming that 25% of the plants (i.e., 27 plants) will have to implement a detailed preventive maintenance program similar to the one described above for Salem, the total industry cost would be about $1.5 billion.
Maintenance-Operatiens Coordination
- " .. 13% of the plants have no position with explicit responsibility for Lh maintenance-operations coordination."
- Assumption: A few plants (10) will need to give more attention to the maintenance-cperations interface and will have to femally assigr.
the responsibility for maintenance operations coordination to a designated individual. This will impose an additional workload on the designated person, who will then teve to delegate some existing work to others. The overall net impact will be modest: assume one additional staff-day per week to formalize and systematize the interface through periodic meetings and written communications, h
- CHAMPS Implementation at Omaha Public Fower District.' Nuclear Plant safety, May-June 1986, p.16.
(a) R. Vasudeman, A. M. Smith, T. D. Matteson, and B. R. Hao. 1985.
Application ef Reliability Centered Maintenance to Component Coolino-Water System at Turkey Point Units 3 ano a. [FRI hP 4271.
5.8 4
,. m)
b.
I l
This works out to 400 hours0.00463 days <br />0.111 hours <br />6.613757e-4 weeks <br />1.522e-4 months <br /> per year or roughly $22,000 per plant, for a total cost for 10 plants of $220,000. It is conceivable that a plant with very poor communications might need to devote -
substantially more effort to improved communications, but the .
assumption here is that only an incremental change is needed to femalize the process. Sumed over 30 years and discounted to pres value at a 10% real discount rate, the total estimated cost is L approximately $2.1 million.
Maintenance Information Systems .
- Plants use a wide variety of systems for tracking information on maintenance activities to assist in planning for maintenance needs and programs. Most (63%) plants systematically monitor maintenance performance using some formal measurement system. However, most of the plants visited had p*oblems relating the measured performance parameters to actual maintenance program practices or needed improvements.....
Resident inspectors reported that most plants (70%) reportedly have a system ir, place for trending and analysis of repeated failures, and alec all plants (g3%) conduct post trip and pre-start up analyses when a
. maintenance related trip has occurred. However, only one of the sites surveyed was found to conduct formal equipment trending. The other plar.
. are either currently developing a system to do trending or rely primaril on the memories of experienced personnel."
{g -
Assumption: Given the increasing emphasis on performance indicator and the expanding availability of computer systems suitab*e for data collection and trending, it is assumed that the rule will requ significant upgrading of plant data collection and analysis. Even though many plants already have som of these elements, it is assum d, particularly in the that area their scope reliability of equipment will be expected trending. totp$61 Nee for this pu is becoming readily available in the nuclear serv, ices . market. It is assumed that a plant could acquire the necessary ha b ere and software and train staff for an' initial investment of $250',W, This includes computer, peripherals, terminals, software, trainifig, etc. Operation of the system would require 5 people full time, or
$540,000 per year. Much more extensive and sophisticated systems are possible and might be useful for other menagement reasons, but the above estimates are believed adequate for maintenance, assuming a healthy amount of competition in the services market to provide these systems to utilities. An example of a comprehensive (and considerably more costly) system particularly is the PIMS (Plant Canyon.jn{ortation tai In additionManagement to maintenance, System) it serves system suchimplemented functions as- at personnel, purchasing, radiation protection, electronic mail, word i processing and others. It consists of an IBM 3090 Sierr M eries 1 I
h (a) L. Perrelli, T. Black, and J. Brady. 'The Maintenance System for the 804 Moves Toward the Next Decade.' Nuclear Plant Jcurnal, Volume 6 No.
- 1. January February 1988. p. 34 j
5.9 l l
._7.
m machine and over 1,000 workstations throughout the company. PIMS '
and other similar systems are widely advertised for sale to utilities.
~
l 3 The total cost to industry to upgrade data collection and analysis -
activities at 110 plants would be roughly $5f>8 million. However, some plants will already have such systems and others will accompitsh these upgrades as-part of their PM upgrades, as discussed previously. This cost element will therefore be reduced by a factor of 4; i.e., it is assumed that only one quart of the plants will be affected. This reduces the cost to $147 million.
Maintainability r"%
(,Q) * "
... Maintainability programs (for designing and installing equipment so that it is easily maintained) ... are not well recognized or supportet I in the industry. Only 33% of the plants have some aspects of a.
maintainability impranament program.*
y The maintenance rule is not intended to require maintainability improvements. . Therefore, no cost will be attributed to this in the regulator; analysis. To provide some perspective on the maintainability issue, the follo ing paragraphs discuss several'of the basic concepts and illustrate potential cost impacts if mahtainability . improvements were to be implemented h Although the maintenance policy statement did not exolicitly mention maintainability, it did call for
- identification of possible component or system design problems.* In the military, the aerospace industry, and certair foreign maintenance programs such as that of Japan, maintainability is emphasized. Full scale maintainability improvement programs in the U.S. nucle industry could be very expensive, in that they could result in significant hard are/ plant design modifications.
that enter into such programs aretta) Examples of the kinds of consideratio e provision of platfom:s, stands, ladders, scaffolding, guardrails, handrails, ramps, lifting aids, and labels.
- control and optimization of environmental conditions such as Ifghting, temperature, humidity, noise, vibration and radiation.
+' *gn of work areas, tool cribs, warehouses and storage areas, laydown as, pull space.
+ design of tools, test equipment, controls and displays, mockups, hoists, cranes, and communications equipment.
- use of design features to minimize the possibility of human error in maintenance.
R. V. Badalamente et al. 1985. Recommendations to the NRC on Human OffA (a) Engineering Guidelines for Nuclear Power Plant Maintainability.
hukEG/CR 3517, PNL-4865. Pacific horthwest Laboratory.
5.10
a For purposes of an illustrative analysis, it will be assumed that 50% of' the plants might need to upgrade their efforts in this area.- A staff enginee will be assigned full time to this function. Improvements in labeling, mater-handlirD equipment (cranes, lifting aids, etc.), tools, and work areas will b implemented under the guidance of the maintainability engineer at an addition cost in the first two years of $500,000. In subsequent years, the maintainability engineer will identify additional improvements based on problems recognized during maintenance. However, it is assumed that these improvements will only be made if they are shown to result in cost savings that compensate for their increased costs. Thus, the cost of the engineer is the only net cor,t after the initial implementation period. The total industry cost to implement maintainability improvements is thus estimated to be about $82.2 million.
Since the maintenance rule is not intended to require maintainability.
@ Improvements, the actual cost for this element will be set at zero in regulatory analysis.
Spare Parts Q Lack of adequate spare parts is a problem at some plants, which indi, cat
- that a potential problem exists than halfin(57%)
spare of t parts ord atams.
According to resident inspectors, more sometimes operated in a limiting condition of operation over the past year due to the unavailability of spare parts. In 7% of the plants, lack of spare parts often holds up maintenance work. The site visits found a possible relationship between spare parts unavailability and lack of a computerized method of inventory control... According to the questionnaire. 84% of the plants have computerized spare parts inventory control systems."
Cyg,
. Assumption: Allowing for improvements since 1985, 90% of the plant:
now have computerized systems, and this provides adequate managemen.
of spare parts. The remaining 10% will need to acquire the same capability. However, it is assumed that these plants are included in the group that would need to significantly upgrade their PM programs, and that the computerized spare parts system would be included in the upgrade. Therefore, no incremental costs are inclur for this activity.
Procedures h * *A majority (55%) of the plants review their stintenance procedures on a regular interval. All of the visited plants have requirements to review at least some of their procedures on a regular basis as well as methods to implement temporary and persanent changes in procedures.... About half of the plants require the use of procedures for all maintenance work; at about a quarter of the plants, procedure use is only required for safety related work.... Two of the site visit plants had problems with procedural compliance. That is, craft workers were not using procedures in a verbatim fashion, as required by administrative policy. .
5.11 GD
__m.______ .___ _ _ _ _ _ _ _ _____..____ _
t j This was seen as part'of the reason for maintenance problems at these plants."
Assumption: Industry initiatives to improve procedures and their use have been successful in minimizing procedures-related maintenen in all but a handful of plants, i.e., 5. These plants will need to undertake major efforts to overhaul maintenance procedures. The effort will include writing, rewriting, editing, reviewing, verify'r and validating the iprocedures. NURIG 1212 reports that plants e'ga.
in major upgrades of their maintenance procedures typically ha? 15-40 people perfo sing these activities, and that these progrars took from 1 to 2.5 years to complete. The cost of this effort is thereft estimated at 25 staff-years, or about $2.7 million, per plant.
The total industry cost to improve procedures and their use would' be roughly $13.5 million. (This amount is also assumed to cover any costs for the remaining 105 plants to conduct more modest effort as needed.) The estimate of $2.i million per plant could be significantly low in some instances, if the existing procedures are particularly poor.
Staffine to Reduce Overtime
- k) *
'The site survey plants were asked what their overtime rate had been over the past two or three years. The overall amount of overtime ranged from 13% - 40%, which includes both nomal operations and outages. The maintenance manager at six of the.eight site visit plants indicated that his staff's overtime rate is too high and that he would like to reduce it."
A -
Assumption: The average maintenance overtime rate for the industry i) Implementation of the maintenance-rulea(whole is assumed to be 25%.s) wi)1 reduce this to 15 as needs significant improvement. If the oorkload is held constant, the reduction in overtime will cause a need for approximately 10%
more maintenance staff (125/115 = 1.09). At a one unit plant the average number of maintenance personnel was 107 in 1985; for two.
unit plants the number was 141. These numbers have increased since 1985; Bowers et al. (1987)(b) indicates that the number of maintenan-personnel is probably more on the order of ISO for one-unit plants, and 27D for two unit plants. Assuming that 40% of the plants are single units, this results in an average maintenance staff of 153 h (a) In a separate reDulatory action, NRC is considering setting limits on overtime. If that action is implemented before the maintenance rule, ,
I this cost element will be attributable to the overtime requirement, and not to the maintenance rule.
,g (b) H.1. Bowers et al. Cost Estimatino Relationships for Nuclear Power Plant Operation and Maintenance. Oak Ridge hational Laboratory, ORht/1M-10563, hovemoer 1957.
5.12 GD
m people. It is assumed that the impact is an average of 15 people per affected plant.
h Typical costs to increase sta fin levels and decrease overtime were assessedinVa11arioetal.(1985)a)forreactoroperators. These costs can be adapted to provide an indication of the costs to increase maintenance staffing levels.
Implementation Costs. Typical hiring costs for a control room operator, .
escalated to 1958 dollars, were estimated to be $12,000, with a range of $5,5 to $16,000. These costs include costs for interviewing, relocation, and new -
employee orientation training, and administrative costs associated with intak k it is assumed that these costs would be approximately the same for the hiring ofnewmaintenancepersonnel(althoughrelocationcosts,estimatedat$6.500, may be lower).
fiw itverage training costs for licensed and unlicensed reactor operators were estimated to be about $580 per day (including overheads), with much of the cost attributable to simulator costs. Operator training was estimated te take 55 days for nonlicensed operators, and 181 days for licensed operators and candidates. Maintenance personnel training should be less extensive and costly than that for reactor operators, so it was assumed that training would take 28 days at a cost of $290 per day. Thus, the total incremental training cost per hire is assumed to be $8,100, with a range of $6,000 to $10,000.
Total implementation Costs for Overtime Reduction Hiring Costs for 15 Employees $180,000 per plant Training Costs for 15 Employees 121,800 per plant Total 3301,600 per plant Total for 20 - 30 plants: $6,036,000 a $g,054,000 y Operatino Costs. It is assumed that the annual fully-loaded salary cost for the additional maintenance employees include an allowance for periodic training. Bowers et al. (1987) provides estimates of annual salaries for nuclear utility personnel. The average annual maintenance salary, in 1988 dollars, was estimated to be about $39,300, which corresponds to an annual cost of about $71.000, assuming an 80% adder for benefits and overheads as suggested by the Generic Cest Estimates Abstract 6.3. Thus, the annual sala-cost to a licensee for 15 new employees is estimated to be $1.07 million.
The licensee will also accrue a cost savings because of the reduction in overtime pay. Since all of the reduction in overtime is accounted for by the new hires, it is assumed (in accordance with Vallario et al.,1985) that the '
savings amounts to 1.5 times the base salary cost for the new hires. Thus, the annual sayings due to a reduction in overtime is estimated to be 1.5 time! '
(a) R.W. Va11ario et al. Reculatory Analysis of Recommendations for NRC Policy on Shift Schedulino and Overtime. Facific horthwest Laboratory, Pht-5593, September 1955.
5.13 3
l
$39,300, or $59,000 per hire. The total cost savings resulting from the hire-of 15 new employees is thus $B85,000 per year.
Total Doerational Costs for Overtime Reduction Annual Salary Cost $1,065,000- per plant Annual Overtime Savings 885.000- per plant Net Annual Cost 3180,000 per plant Annual Cost for 20.- 30 plants: $3,600,000 - $5,400,000
,E When the annual operational cost of $3.6 million to $5.4 million is sunrne
' over 30 years and discounted to present value at a 10% real discount rate,.
the total is about $34 million to $51 million. After adding the initial implementation costs of $6 million to $9 million, the total cost over thirty years for all af fected plants is estimated'at $40 million to $60 million. As a point estimate, a value of $50 million will be used in this regulatory analysis.
5.2.3 Industry Cost savinos O7 Although it is not the NRC's mission to pursue economic benefits for the nuclear industry, improvements in maintenance will yield very large economic
- benefits in the form of cost savings due to reduced downtime and reduced expenses for corrective maintenance; because the cost of downtime is so high, the cost savings due to reduced downtime is the dominant factor. In this section, the magnitude of this cost savings is estimated. First, a rough approximation is given to bound the benefits. Then results of a statistical analysis are described that lend additional support to the estimates. Finally some additional potential benefits related to reactor aging and life extension and averted onsite property damage are discussed briefly, e For a first approximation assume again that 20 to 30 plants have substantial capacity factorsroom becauseforofimprovemen,t in maintenance maintenance weaknesses. a) Suppose and(are exper that improved maintenance could increase the average capacity factors of these plants by 5 percentage points, or 18 days per year; this assumption will be discussed further below. The total num 540. At 5443,000 per day,(b)ber the of days of avoided increased output replacement is then power cost 360 to is $159 millit M (a) Between 1985 and 1987, 19 reactors (out of 81 that operated throughout the three-year period) had net everage capacity factors below 50%.
A W (b) See J. C. Van Kuiken et al. 1987. Replacement Eneroy Costs for Nuclear Electricity-Generatino Units in the Uniteo States: 1987-1991. huREG/CR-4D12, ANL- AA-30, Vol . 2. Argonne hational Laboratory Argonne, Illinois.
Based on the estimates in NUREG/CR 4012, the average replacement energy cost (in 1985 dollars) for 1988 is $410,000. When this is converted to 1988 dollars by applying the Gross National Product Price Oeflator, the estimate becomes $443,000. Based on Abstract 6.4 of the Generic Cost Estimates, the GNP Price Defletor for 1988 is projected as 121.4, while that for 1985 is 112.3; the ratio of the two is 1.08. This ratio is 5.14 G
r l
to $239 million per year. Over 30 years, assuming constant real prices and a 10% discount rate, this adds up to 9.43 times that amount, or $1.5 billion to
$2.3 billion. As a point estimate, the mid value, $1.9 b(11 ton, will be,used.
Q To provide some perspective on the postulated percentage point improvemen in capacity factor, it may be noted that the median capacity factor at U.S.
l plants for 1985 1987 was 63.6%. The top quarter of U.S. plants had capacity factors that were 10 to 25 percentage points higher.
OM any, A between statisticalplant analysis wasfactor capacity performed to examinequality.
and maintenance the relationship, if The hypothesis to be explored is that plants with better maintenance enjoy higher capacity factors and hence more favorable economic performance.
i l
Q To measure maintenance quality, a measure is constructed from the scores received in SALP evaluations. Two of the SALP functional areas are used:
maintenance and surveillance'. The measure of maintenance quality for a plant is the average of its scores in these two areas. The usual SALP scale applies except that there can be fractional values; a score of 1 is the best and a score of 3 is the worst. Since SALP evaluation periods are different for .
different plants, an algorithm was used to calculate a time weighted average I score for each calendar year. . For example, if a plant received SALP evaluations on March 31, 1985, September 30, 1985, and March 31, 1987,'its maintenance quality ir. 1985 would be a weighted average of its scores in the two evaluation periods covering 1986. Three fourths of the weight would be placed on the first period (which captured the first iine months of 1986),
and one fourth of the weight would be placed on the second period (which captured the last three months). The SALP data were obtained from NUREG-1214, Historical Data Sumary of the Systematic Assessment of Licensee Perfomance, which is upcated semi-annually.
g C Blake,gpgeity as who factor infomation sumarized wasthe data from obtained from aBlake Graybook.(b) recent article average reported by capacity factors for 1985-1987. Use of the three year average tends to smooth out variations due to refuelling outages ano other short-term factors.
bQ in order to explore the relationship between maintenance quality and capacity factor, a variety of models werg fit to the data by regression analysis, using the 5 software system.(c) Capacity factor was the dependent variable. Independent variables included the maintenance quality score.(scores y ( multiplied times $410,000 to obtain $443,000. For further discussion of i M ,D 1' time-related Estimates. cost adjustments, see Abstract 6.4 of the Generic Cost i
b (a) E. Michael Blake. ' Domestic Capacity Factors: More Improvement in the Mid Range.' Nuclear News. Volume 31, No. 7. May 1988. pp. 43 48.
T )(b) Licensed Operatino Reacterst Status Sumai v Report. NUREG-0020. Issued.
monthly.
{p(c) R. A. Becker and J. M. Chambers. 1984 St An Interactive Environment for Data Analysis and Graphics. Wadsworth Advanced Book Program, Belmont,.
California. !
5.15
-m
4 s -- - ,
, ;2 If
- l b \ , / (.
.i n ,
r 1 ;
for several different ralender- years were tested),1 plant type {PWR of BWR),c plant size,: and a measure ~of plant age or vintage. The measure of, age was-i , the number, of years since the issuance of the construction permit.. Pjodels 1
' including all possible subsets of the variables were tested to detemine the
.werecalculatedandplotted,'and0-0(quantile-quantilebest fit.?
plots were Scatterplot used to: of
" , examine the distribution of the residuals.. Influential observations and:
- 4. outliers were highlighted and the sensitivity.of the fitted models to these -
, , { data points was tested by deleting certain points and rFitting the models..
h Results'showed that capacity factor and maintenance quality are clearly (
related.1 A number of models fit reasonably well. One of the simpler models-was'the followings
- Capacity factor = 94 - 16.7 s maintenance score .
S @ This~ equation says that cepacity factor declines as maintenance score 1 increases, i.e., as maintenance quality decreases. (recall that the best -
maintained plants will have a score of I and the worst a secre of 3).' the L numerical value of the coefficient.16.7, should be recognized as uncertain.
It has an estimated standard error of 3.3 in this model; an approximate confidence interval would be 16.7 plus or minus 6.6. Thus a one point change, in the maintenance score is worth an estimated 10 to 20 percentage points in terms of capacity factor. Other models gave results in roughly the same rang'
> lt should be recognized that maintenance improvements may require increased l outages-in some cases, which would tend to limit the capacity factor : .
improvements estimated above, and would not be realized immediately; several.
years might be required to achieve significant improvements.. Also, some of, the improvement in capacity factor will be achieved even in the absence of a n . maintenance rule, because of the industry initiatives already undemay.-
-Therefore, the value'used in the regulatory analysis will be 5 percentage points, which is based on taking the low end of the 10 to 20 point range and 1 dividing by 2, to adjust for the improvement that may be achieved even in the.
absence of a maintenance rule.:
These caleviations yield results similar ta the values assumed at theu
@ beginning of this section and lend additional support'to the conclus -
-the potential economic benefits of improved maintenance are very large. '
. Industry has a substantial economic motivation to improve maintenance, and this provides ample incentive and justification for the major industry initiatives that have been undertaken in recent years.
@. damage. A benefit can also be estimated due to reduced risks of onsite property .
If a severe accident occurs at a reactor, onsite costs will be incur-in several categories . including: (1) the cost of interdicting and/or
'decontaminatingonsiteproperty;(2)thecostofreplacementenergy;and(3) ,
repair and refurbishment costs'. The Handbook for Value-impact Assessment, NUREG/CR 3568, gives guidance on estimating these costs and on calculating-the reduced e:onomic risks to onsite property when a proposed regulatory acti.
.would reduce the probability of a severe accident.
5.16 7 j
- - - . _ - _ _ - - - - - - - - - - _ - - - - - - - , - - - - - - -----------,-----,.__----_-__a m- --____ _
in section 4.3, the change in accident probability attributable to the ONK maintenance rule was discussed. For purposes of this calculation, it is asst based on NUREG 1150 estimates that the average baseline probability of-severe core damage due to internal events is 5.0E-05 at typical plants, and that it is five times as high in plants that have substantial room for maintenance improvement. If the maintenance rule reduces the extra risk by two thirds (cf.'the discussion in section 4.1.2), then the change in the probability of severe core damage is 1.3E-04. Based on the estimates in section 3.6 of the Handbook for Value Impact Assessment, the averted onsite property risk over 30 years is 1.3E-04 times 51.0E10, or $1.3 million. Summed over 20 to 30 plants the total is $26 million to $39 million: $33 million will be used as e point estimate. This estimate should be regarded as a first approximation.
More refined estimates could be calculated, but since this economic benefit is small relative to the cost savings identified earlier, the approximate value is adequate for purposes of this regulatory analysis.
T O An additional economit benefit of improved maintenance should be mentfor.
for completeness. Utilities are currently investigating the possibility of seeking license renewals for existing plants to extend their operating Ilves beyond 40 years. The potential industry-wide economic benefits of license reneral have been estimated to be on the order of hundreds of billions of dollars. It is clear that good maintenance will be a key to license renewal.
and improvements ressiting from the maintenance rule will help to establish
~
the required technical basis. A separate rulemaking is underway to establish the regulatory framework for license renewal.
5.3 NRC COSTS The NRC will incur costs to review and endorse the industry standard, d((# - develop and issue regulatory guides, analyze licensee reports, and perform plant inspections. Unless otherwise indicated, NRC labor costs are assessed at $41.30 per fully-loaded staff hour, or $74,400 per staff year, as suggeste by Abstract 5.2 of the Generic Cost Estimates.
Cost to Develoo and Issue Reculatery Guides Q u nder e 1 but Option 5.Re NRCulatory costs toGuines solicit and other industry supporting comments document as input to the development process are assumed to be negligible Q Under Options 1 and 2, a Regulatory Guide would be developed that provid.
guidance on acceptable methods for measuring the effectiveness of licensee maintenance programs and for reporting on the programs to the NRC. (However, each licensee could establish a measuring and reporting program that he considers acceptable.) Specifically, it is assumed that the Regulatory Guide and reporting system and define a >
, would set of maintenance describe an acceptable performance recordkeeping(MPI) that could be use indicators licensee performance. Specifications on reporting formats and frequencies would also be included.
[T below:
It is assumed that the Regulatory Guide will describe the ten HPIs listet 5.17 l
_ _ _ _ _ _ _ __ _ - _ _ - - - ----------------------------------------------------U
y --
1 ,
i ;
j i
11 Ratio of preventive to total maintenance (h#, @ 2h Corrective maintenance backlog greater than three months cid.
3hPreventivemaintenanceitemsoverdue
-4? liaintenance staff turnover rate Si Maintenance rework .
6h Ratio of deficiencies discovered during surveillance to those discovered during any demand Tl Number and duration of 80P equipment out of service 8 1 Safety system performance indicator 9
10c) Mean Backlog time to return to service of en performance gineering change notices related to equipment h Four of the indicators listed above (nos.1, 2, 3, and 8) are currently used by INPO to track licensee erforsance. All are defined in'a fairly straight.-
forward f ashion, so devel 'ing the definitions for use in the Regulatory Guid-should not be a major task in practice, however, it is recognized that difficulties in interpretation may still arise at the plants as they attempt to apply the definitions.
+
Previous studies that have assessed the impacts of proposed regulatory
~ changes have included estimates of the costs tg pevelop and issue Regulatory
' Guides that can be used for this study. Thesetas indicate that a 1 staff.
year effort is usually sufficient for the development and publication of a-Regulatory Guide. Regulatory Guides that merely endorse an industry standard, may require less effort to develop, while those that require extensive analys3 and methodology development may require more effort. Thus, a base-case estiet.
for the Regulatory Guide development may be 1 staff. year of effort, with a range of 0.5 to 1.5 staff years. At $74,400 per fully-loaded NRC staff year (Generic Cost Estimates, Abstract 5.2), the cost is estimated at $74,400, with a range of 337,200 to $111,600.
h tlnder Options 3 and 4, a detailed Regulatory Guide would be developed that would provide the scope and methods for the conduct of maintenance .
activities by licensees. . For Option 3, the intent is te endorse an industry standard; however, NRC would move ahead in parallel with industry to develop a technical basis for reviewing and approving the industry standard. If the.-
industry standard were found to be inadequate, NRC would then be prepared to issue its own guidance. For Option 4, NRC would develop the Regulatory Guide independently, without relying on an industry standard. It is assumed that the Guide would provide guidance in each of the 15 areas desi NRC Policy Statement on Maintenance of Nuclear Power Plants (gnated in the 24,1988) as " Activities Which Fors the Basis of a Maintenance Program." It .
is expected that such a Guide would represent a significant development effort OuT (a) Information provided by NRC personnel for inclusion in a regulatory analysis of proposed revisions to 10 CFR Part 20; also, TM1 Action Plan (NURIG 0660), and Guidelines for Nuclear Power Plant Safety issue Prioritiration Information Development, Supplement 2 (hWREG/CR 2800).
5.18
1 requiring 4 6 staff years. _Thus, the base case cost is estimated to' be 5 staff-years, or $372,000, with a range of $2g8,000 to $446,000.
NRC Costs to implement a Perfomance Indicator Tracking Program
[UU} Under Options'] and 2, the NRC will incur costs to receive and process the perfomance indicator infomation submitted by the licensees. These cost.
include costs to enter the data into a database, produce summary reports,.
monitor trends, and identify those licensees that require remedial action to
. address poor perfomance issues.
h NRC's Interoffice Task Group on Performance Indicators estimated (in 1986 that four to six full-time employees would be required to implement the initi.-
indicator program. It is assumed that, because Option 1 allows each licensee to report data on the set of indicators chosen by the licensee, a significant effort will be required by t,he NRC to process and analyze the data on a consistent basis. Thus, it is estimated that the same number of employees will be needed to implement the proposed indicator program using eight additional indicators, with five full time employees as the base case. The annual NRC staff cost is assumed to be $74.400; thus the annual NRC cost to implement the program is estimated to be $372,000, with a range of $2g8,000 to $446,000, Over 30 years at a 10% discount rate, this adds up to roughly
$3.5 million.
NRC Costs to Review and Endorse the Industry Standard
> It is estimated that six staff-months of effort will be required over two years to review the industry standard, to pose questions to the industry .
steering croup, and to endorse the standard or parts of it. Thus the total NRC cost Is estimated to be $35.500 -
NRC Inspection Costs Whichever rulemaking option is selected, NRC will perform maintenance inspections on plants that are judged to warrant increased regulatory attentis with respect to maintenance. The plants will be identified based on the full range of infomation available to NRC management, including observations by resident inspectors and regional staff, SALP evaluatinas, operating expertene.
and perfomance indicators. As a base case, it is asssmed that there will be no incremental inspection costs due to the maintenance rule. The rationale it that NRC already perfoms maintenance inspections and maintenance evaluations, the rule would serve to strengthen the basis for the inspections and evaluations, but would not require additional inspection effort.
To investigate the sensitivity of costs to the assumption of no increment b'inspectioneffort,calculationsofpossibleinspectioncostsunderan alternative set of assumptions were perfomed. Based on estimates of the effort involved in the maintenance team inspections initiated in July 1988, i the inspections to be perfomed under the rule would require approximately 35 j staff-weeks of labor. A team of 6 NRC staff, consisting of 4 from the region and 2 from headquarters, would spend 2 weeks on-site and one week at the regional office. There would be one week of preparation by the team and two 5.19 l
i 1
n- ,
1 p/ - -
weeks after the inspection to write up the results in an inspection report.
It is assumed that additional time would also be required for planning and organizing the inspection and for completing the inspection report after management reviews. This could add an estimated four staff-weeks of effort.
These estimates would vary somewhat depending on the circumstances. Forty staff-weeks will be used as the nominal estimate. Assuming $41.30 per hour,
, the NRC cost will be approximately $66,100 per inspection. 1 Only a fraction of the plants will warrant this level of increased
@ regulatory attention to their maintenance programs. Also some plants will already be inspected before the rule is in place. Itwillbeassumedhere' that 10 additional plants will be inspected initially, and two plants per year for the next five years, and one plant per year in subsequent years.
The total discounted inspection cost is thus $1.53 million. Licensees would spend a comparable amount,since they would need to provide information and assistance to the inspection team.
Q In sumary, the base case inspection cost is assumed to be zero in this regulatory analysis. The sensitivity calculations described above indicate a relatively small cost impact if some level of incremental inspection is performed.
5.4 SU M tY g Table 5.1 sumarires the industry costs and cost savings attributable to the various options; the table displays the base case estimates (in the colum-labeled 'B"), along with lower and upper estimates (labeled 'L' and 'U" respectively) to provide some indication of sensitivity to variations in cert major assumptions. The lower and upper estimates are developed in Appendix A. Two points should be noted. First, for the base case, each of the option is estimated to yield a net cost savings, on the order of $100 million to
$200 million over the remaining lifetimes of current plants. The cost saving:
are due to an estimated 5 percentage point improvement in plant capacity factors at those plants that are currently lagging behind the industry as a whole in terms of maintenance effectiveness. Other cost savings are also expected. such as reductions in the amount of corrective maintenance and rewo-but the improvement in capacity factor is the dominant savings.
OC- Second. options 3 and 4 have a net cost advantage compared to Options 1 and 2. This difference is mainly due to the increased recordkeeping and reporting costs that plants would incur to collect and report maintenance perfomance indicators under Options 1 and 2. These recordkeeping and report *.
costs would apply to all plants and would continue throughout the remaining lifetimes of the plants. Options 3 and 4 do not require reporting of performance indicators.
Table 5.2 summarizes the NRC costs, sumed over 30 years. The costs are small relative to those incurred by industry. They are also small relative to NRC's budget. Again, Options 3 and 4 have a net cost advantage compared to Options 1 and 2. The need to process and analyze licensees' perfomance indicator reports under Options 1 and 2 is the major reason for the difference 5.20 i
l J
___ _ _ J
h
- w e e W D
e W- e o e o- e e e e . .
O es en e e e e .e e I e 4 es e W so ' pt 0
0 W
- eD e a em.
& W erb e e gis e e g gg g g e s's e se ,ee as et e e as se a e to se se e
& m
- e9 WB s=8 e e e4 e e es e e o e e 9 es w et e en es een SD 4 Ob o S e ge sa 9 4 6 9 aus b -e W p e. DD. e.
em .-9 as se se e o e. e ce e e e e e-
== CD eg et se e se o e e e de e e*
em 43 e em e e er e2_ b e S 99 ee . e.s O es, se -e e e. .
e e9 W e @e e e St e W e 5% e $h.
O b h so e $$ - een se e b S e em se pg - O e e se 4 e en 9 e se e
>= S e *e t ee
- O +=O e*
et em SD st se see o e
es e e as et e e
e e e og et e se e... 0 e e 9 3 es se D , 9 0 e **S e. e9 8
b 3 er.
.e a es s e ==
e m e e seo es e e o e se e- &
9 a b Se 9e e
em o e e e e
to e O # h w 9 w e es e ** $
e S e*
-W e mm w 4 e 9 b es .O. e + ee. A et S et W sue e e go e e ee e -et en e o h es e as se em et se e
e
- 8 G
e 3 e e se se 9 e 4 g em aus 8 e se
- e 5 h 9 9 e9 e=e & 9 m S a ese* e E O M O O O " O M 9 en.cu.e.l.'I%?#hM$NM.
- - / f.8'@8'NN N. Pf "e" e .O. O O O es e e em O. e e .
e me 99 8 95 e S e t- esse Go en e D e 6 A D e e 9 9 w e= as S
- - D * == a e es e es as e- e e- e es e o e o es e e se e e e e se e e .S e e se se as e O
C w
e em em
. e o >.
e ** eas O
- e G W 9 9 9 3 e eO e 3
e ** e . se. en g 3 .
M W # *e e een e 9e e e e e WD e ce e 9 W es
& ** e e es en es os e e e a e e aus es as se g e se g 3 m p i S se *= 0 9 tEB 0 5 en e b o se a
4
. i t. .J se e o-e.eme e em e o ta e e e se as e e =e I S b 9 .
%# *D 9 e es w me e em 9 e===
9 pt t #b e 9 9 O On em O A 9 eft 9 e e e s= A e _
9 S A -# et S es e ee e8 eE ew 9 e es e S 9 b 4 as ==G S S *'
e- W G & S e 5 9 es e 9 e E es e W 9 9 e 9 e e8 S
h S e as S
e S
3 9 e 9 e
Pe D w w 9 m += == S es 9 8 *= e 9 O e 9 e & e S 9 as a e S G e e De em 9 w4 e W G S e# b d == O eD e S b 9 3 9 9 9 4 e S 9 se tio em **op S es e ma G S +=eb
- 5 W e em 5
S 9
b e
S as
> ==
e b 3 a w G grt
+
es
+=
es en-e ep em e
D ena e 9 9 9 es 9 s S e Cb e ** 9 S
e 9 e a ed a a e De e 9 as b e S e &
- e == .yp e e e e a e S e o e @ S e. A e .ar 9 9 De B se
- 4 9 me ame > & G e G 9 9 e o es se* fe fit *
& W ==& eft 9 ** e so S e W **e e4
>9 b CD 8%
S e
9 9 De +=9 es 3
as S e* * +=e 9 e e 9 9 e $
- = +=
a= e 9 9t D em B b
e*
S ass b 9 9 9 eD em as D e O 9 et eA -eD 9 9 et C3 9 > S S w b e 9 b et e =* ++
S e8 4 e e g .S= 3 ea & 9 == sa3 e a D E G B = b eee e **e b g 9 9 & e> es 9 9
S e
ef 9 e ==& e. 3 3 3 b es e4 *D G S em e bb S em 9 S e e 9 b e D G H S 8's 3 e se g me g & y eg y 3 e
ei 1
1 i
l 1
I
(;; ,
U l
s t
L l-Q h. Table 5.2. Total NRC Costs over 30 Years' for All' Plants .
(1988PresentWorthinMillionsof1988 Dollars).,
^
!- NRC Cost Element Develop Regulatory Guides (s)- 0.1 0.1 0.4 0.4 '.
Endorse Industry Standard (<0.1)
' Analyze Licensee Reports 3.5 3.5
? ' Total NRC Costs 3.6 3.6 0.4 0.4' s
6 5.22 I
_ _ - _ _ _ _ - - _ _ _ _ 1
,1 APPENDIX A RANSES FOR 1NDU$Tay COST ESTIMATES e
i
APPENDIX A RANGES FOR 1NDU$iRY COST ESTIMATES
.This appendix discusses ranges for the major industry cost elements in Chapter 5 and presents upper and lower estimates for each. The ranges, which are sumarized, along with the base case cost estimates, in Tables $.1 and 5.1, are intended to give some indication of the sensitivity of the cost-estimates to variations in some of the major assumptions. It should be stressed that the ranges are not statistical confidence intervals and should
~
not be interpreted as such.
h Before the ranges are discussed, several coments on uncertainties are i order. There are two basic uncertainties in the regulatory analysis that dominate all the others. Both are explicitly discussed in the regulatory analysis.
First, the number of plants that will require substantial maintenance-
[ improvements is not' precisely known. Both the costs and benefits can be strongly related to this factor. Industry asserts that the maintenance initiatives undertaken in recent years have resulted in major improvements and that new there are very few plants with weak maintenance. NRC asserts
.that a significant number of plants still need substantial improvement. Neiti NRC nor industry has a fim, up to-date, quantitative basis for estimating this number with precision. The best documented basis is still NUREG-1212, which was published more than 2 years ago. The regulatory analysis relies heavily on NUREG 1212.
@ 5econd, the improvements that plants will be required to make are not precisely defined. The standard that plants will have to meet has not yet-been written. And even if the standard already existed, the costs would be sensitive to how it was interpreted and implemented. Therefore, any estimate of costs to comply with the standard must be highly uncertain at this time.
Maintenance Plan Development Costs M As an upper estimate, it is assumed that the more substantial effort discussed in Chapter 5 (6 staff months per plant, or $54,000) applies to all 110 plants, for a total cost of $5.9 million. As a lower estimate, the smallt level of effort (3 staff-months per plant, or $27,000) is assumed to apply at all 110 plants, for a total cost of $3 elliton. The base case assumed a 1:2 mixture of these two levels of effort, for a total cost of $4 million.
Recordkeepino and Reportino Costs The base case estimate was $111 million for Options 1 and 2, and $13
@ million for Options 3 and 4. For Options 1 and 2, the dominant factor was the assumption that the average plant would need to devote an additional 1820 staff-hours per year to (This and report generation. perfomance consistedindicator data collection, of 1/2 staff-year, or 1000 data entry,f hours, o professional labor for data collection at $54 per hour,1/4 staff year, or 8
- _ - _ - -- _~ -
k.
h 500 hours0.00579 days <br />0.139 hours <br />8.267196e-4 weeks <br />1.9025e-4 months <br />,'for a data. entry technician at $31 320 hours0.0037 days <br />0.0889 hours <br />5.291005e-4 weeks <br />1.2176e-4 months <br /> of professional' labor for reporting)per .
hour, This addedandup8tostaff-weeks
$86.800 per or plant per year; when sumed over all plants and over 30 years, this element accounted for $90 million of the total recordkeeping and reporting cost of
$111 elllion for Options 1 and 2. In developing ranges for recordkeeping and 3 ' reporting costs, we will focus on this dominant $90 million estimate.
This estimate
@l data collection could very depending on the extent of existin activities. in-plant-If it is assumed that the average p ant already j' collects.the necessary data and stores it in a readily retrievable fom, the I
only incremental cost would be the cost of producing the reports, or 320 hour:
per year times $54 per hour. This amounts to $17,300 per plant per year..
Scrned over 110 plants and 30 years of remaining plant. life, the total is $17.
million, as compared to the base case of $90 million. Thus the total recordkceping and reportirts cost estimate would be roughly $39 elliton instest-of $111 milhon. This is an optimistic lower bound; many plants do not currently collect all of the infomation needed for the postulated maintenance.
perfomance indicators.
As an upper estimate, it is assumed that the average per-plant labor for
@ collecting perfomance indicator data, entering it into a data b generating reports is twice as high as the base case i c than $90 million. This assumption results in a total re.e., $180 million cordkeeping and rcthe -
reporting cost of $201 million.
Industry Standard Development Costs The regulatory analysis provided a range of $1.76 million to $3.73 millic for this cost element, with a nominal (base case) value of $2.75 million.
Reculatory Guide Development Costs h In the base case, it was assumed that 2 workshops would be held to obtain-industry input on the development of the Regulatory Guide, and that two staff members would attend from each plant.. As a lower estimate, a single workshop attended by only one person per plant may be assumed, which would reduce the estimated cost by a factor of 4, to $95,000. As an upper estimate, it may be assumed that a third workshop might be needed, which would increase the base case cost by 50% to $570,000.
Costs to implement Needed Maintenance improvements kh Before presenting ran maintenance improvements. It isges worthfor stressing the costs of each that theofprincipal the postulated source of uncertainty is not in the individual cost elements identified under this heading. Rather, as stated in the regulatory analysis, the more t,asic issue is to detemine what improvements will be required:
@ "It is difficult to estimate with any precision the specific actions that each plant will have to take to improve its maintenance program based on the rule. There are two principal reasons for this.. First, available data on plant specific maintenance programs and practices are limited. Second, at this early stage in the maintenance rulemaking
process, there is not yet a clear, detailed definition of what constitut an effective maintenance program.*
Preventive Maintenance h The b'ase case estimate was founded on essentially two factors:
- 1. Roughly 25% of the 110 plants would have to implement extensive upgrades of their preventive maintenance programs. This estimate was derived from information in NUREG-1212.
- 2. The average cost of the upgrade would be $56 elllion. This was based mainly on the Salem experience, h 'As a lower estimate, it say.be argued that industry initiatives in the last two to three years have arready addressed the issues discussed in NUREG-1212, so that only a handful (perhaps 5%) of plants would now need substantial improvement. Therefore, a factor of five reduction from the base case estimat will be used as the lower estimate.
g As an upper estimate, it may be argued that the maintenance rule will lead to ratcheting and that a larger number of plants will need to make .
substantial improvements, even those that were judged in NUREG-1212 to have adequate or extensive preventive maintenance programs. The only plants that might not be affected would be those that already have programs comparable to the one at Salem. If it is assumed that 75% of the plants would be significantly affected, the base case costs would be tripled.
Maintenance-Doerations Coordination Q For the base case, it was assumed that 10 plants would need to strengther the maintenance-operations interface by forna11y assigning the responsibility for maintenance operations coordination to a designated individual. NUREG.
1212 had found that 13% of the plants had no position with this responsibillt)
The additional effort was assumed to be 400 staff-hours per year, which is roughly one day a week. As a lower estimate, it may be argued that industry initiatives since the publication of NUREG-1212 have addressed this issue, and that only a handful of plants (i.e., 5) will still need to devote additier.
effort in this area. Thus, the base case estimate of $2.1 million would bc reduced by a factor of 2 to roughly $1 million.
@ hoursAsaanyear upper estimate, it may be argued that one day a week or 400 staf f-is not sufficient to achieve the necessary imp 10 plants. It might be necessary to assign a person full-time to this responsibility, i.e., 2000 hours0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> a
. cost estimate by a factor of 5 to $ year, 10.4which would increase the base case million.
Maintenance Information Systems h For the base case it was assumed that 25% of the plants would need to upgrade portions of their maintenance information systems to provide enhanced capabilities for trending of equipment reliability. The initial cost per
! affected plant to acquire additional hardnare and train staff was estimated as $250,000 +nd the annual operating cost per affected plant was estimated as l
l
_7 .
p 5540,000,' which assumed 5 people full time. After sunning the operating costs over 30 years and discounting to present value at a 10% real discount rate,
'the operating costs per affected plant were about $5.1 million and the total cost including the additional hardware and training was $5.3 million per affected plant. For all affected plants, the total estimated cost was $147 million.
As a lower estimate, it will be assumed that only 10% rather than 25% of '
@ the plants might incur such costs, which would reduce the est of 2.5. The resulting estimate is about $59 million.
As an upper estimate, it may be argued that all plants should have
@ sophisticated equipment reliability trending capabilities. -If 25%
plants already have such capabilities and another 25% will acquire them throup their preventive maintenance upgrades (as discussed separately), then the remaining 50s of the plants *would still need upgrades and should be included.
in this cost estimate. This would increase the cost by a factor of two. If, in addition, the average cost of the upgrade were doubled to allow for more extensive computer equipment and more staff, another factor of two increase in the costs would result. Thus, the upper estimete.will be taken as $600-million, or four times the base case.
Maintainability h For the base case, it was assumed that the rule would not require any' maintainability improvements. To provide some perspective on the maintainability issue, the regulatory analysis discussed a hypothetical' maintainability improvement effort and estimated its cost as about $82 million.
This can be used as an upper estimate, although it si'.:,uld be recognized that maintainability improvements costs could be significantly higher, depending '
on how extensive the requirements were.
Soare Parts PD The regulatory analysis discussed the pessible need for computerized spare parts management, and estimated that the few plants that needed improvement in this area would combine this upgrade with the preventive maintenance upgrade discussed separately. Therefore, the incremental cost under this heading would be zero. As an upper estimate, one could argue that a few plants would need to implement computerized spare parts management.
However, the costs should still be negligible compared to other cost categorie:
Therefore, to upper estimate is given for this cost element.
Procedures h It was assumed that 5 plants would need to extensively overhaul their maintenance procedures at a cost of $2.7 million per plant, for a total cost of $13.5 million. As an upper estimate, the number of affected plants might be significantly larger, perhaps by a factor of 5 (i.e., 25 plants). The total cost would then be 567.5 million.
As a lower estimate, it may be argued that industry initiatives have reduced the number of plants needing procedures upgrades to 2, rather than
the 5 assumed in the base case. The total cost would then be roughly $5.4 million.-
@ Staffine to Reduce Overtime Th regulatory analysis provided a range of $40 million to $60 million, with 40 million as the nominal or base case estimate.
Reheed Risks of Onsite Procerty Damace The regulatory analysis estimated a range of $26 million to $39 million.
~$33 million was used as the base case.
@ Cost Savines Due to imoroved Availability The base case estimate assumed that 20 to 30 affected plants (with substantial room for improvements in maintenance) would achieve a 5 percentagi point improvement in capacity factor. This resulted in cost savings ranging from $1.5 billion to $2.3 billion. The mid-point of the range, $1.9 billion.
was used as the base case.
go 9
s
)
1 I
/p. nn s%'e UNITE D IT ATE s
- f , e NUCLE AR RE GUL ATORY COYMISSION g ADVis0RY COYViTTE! Oi, RE ACTOR sAFIGUARDs
. u s mectoi..o.c.em ss 9.....
Septe-ter 13, 1956 The Hcnerable Lando W. Zech.- Jr.
Chairr.an U.S. No: lear Regulatcry Comission Washington, D.C. 2D555
Dear Cha'irtan Zech:
SU$JICT: FROP03ED RAIVA,1N3 RELA 1tD TO MA1NithAh;E OF NUCLEAR FO *IR PLAh15 Du*irg the 342st ceeting cf the Adviscry Co-rittee ce Peacter Safe-5 seds. Se;te-ter E.30, 25!! we discossed the pre;ned a erdrent to 30 C;R Fa' 50 related te weirterarce of nuclear p:=er pla ts. We had preiicLsly disc &ssed tMs tegic tith the hKC staff during our 30th.
reeting. A.;sst 21-D. IllE. std darin; reetings cf cur maintenance Fractices and Fre:ers-es 5.t:o--ittee on June 15 and Sette .ber 7. .
25!!. Ds'de; t*ese reet rgs. we had the bete'it of co re'is by a 4
re;-esertative ci tte ha:1 ear Feta;e,ent att Fescurces Ccungil
( fi.$.a :.; ) . We alst had the tetefit cf tre dc:u ett referenced.
Ve carrLt t'de'se t>c p*c;: sal te estat11st a maintenance rule. The p-c;:sel a;; ears te be base u;:r. the ic110=in5 Fypetheses:
- 2. Fa'rtenarce cf ry:1 ear r:=er riarts, as non perferred. p:ses a 54g'ificant rist tc tutlic healit and safety.
- 2. Ttt esistemte cf a traittetance rule w:L1d reds:e this risk.
There is 50Te evic'ette to su;;:'t the first of these hy;:t* eses, altt:,;t there is tc dire:t cvartitative irficatcr. There is anec-d:tal evicet:e that p::r fra'etera*.:e has cettrib.ted to platt pr-estilattlity and tal enen led 10 the esiste'ct cf platt states that a'e inte';'eted as 4:cidtrt pre:Wes:rs. Co r.:t setse su;;ests that a well-mainteired platt reses less risk than one pc:rly staintained.
So e eperating plarts have better staintetance pro;rar.s than others, ar.d so.e have pecgrams that sto;1d be ir.; roved.
ke have seen no evidence to su; pert the secced hyp: thesis. The regulatery analysis provided by the staff stakes the arbitrary assump-tion that a red.sction in risk will occur as a result of the rule, and bases its test benefit conclusions en a 9.ress about the 4* cutt of risk reduction ez;ected. Nor have we seen eniden:e that the existence cf a rule w:uld not stake things worse. Indeed there are characteristics of regulations, and espe:ially the way in wHth they are typically l enicrced, that lead us to believe that, under a rule. a c;ve toward utiferrity wevlt occur, and this is likely tc de:resse the effe:tive-ness cf so .e of the better existing p.regrams.
i l
l
.=
w, The Honorable Lente W. Zech, Jr. -2* 5eptember 13, 1988 Finally, it appears to us that's.tintenance practices in the industry are11rt; roving and that a rule F.ay be disruptive to the substantial industry initiatives that have been developed to accomplish this l imprcrement. (
Attitional corrrents by ACP5 sember Harold W.' Lewis are' pres'ented below.
.- Sincere 1y, W. Kerr Ct.a i rra n A::stic e1 cc erts ty A!*$ re-ter Faccle V. te is a
I wart -to tale tris es;:-tur.ity to ctserve that it is ect re:essarily' true that rcre r. air.ter.ance is better rainter.ar.ce -. a substantial re ter cf eserts are iritiate by testirg and reintenance operations.
Tr.ere esists a well.tesele;eg t*ecry of reliability which deals with such retters as tre c;tirui level c' caittenette -- there are becks on t'e s',tf ect - art it veult te i;seful it brirs some es;ertise ir.to the s'a1 3sis cf this e,.,estien.
Ee##'e' e:
Le*;e-r cate! As5' St 29,193i fro- Bill F.. FCfris, Directer Divisice ef Feit.latcry A;;14catices, Office ci h. clear Reg'1 . story Researct, to Ray :'c F. Fraley, Ese:utive tire::ct, Ac.isery Cc rittee et Reacter Sa'e;.ards, ercicsing Dra't Co- issic Paper fer hetite of Frepesed i F.1e tiir; f er F.aittenan:e .c' h. clear Fc.ee Flonts
~
1
}
u 1
i
t .y
/ x
]Mf. 2 t .
- NUCLIAR MAN AGIMENT AND '4$0VRCf 5 COUNCIL h !.e 1*ee *, e !a 3; ,6:e-yc- :C ;;;3 244 12:2, l'2.";!;
' October 28, 1988 .
The Honorable Lando W.' Zech,.Jr.
Chairman U. S. Nuclear Regulatory Commission WashinSten,'D.C. 20556
Dear Chairman Zech:
Ve are writing to you to be sure you are aware of our deep-concern fer the present course of action being taken by the NRC-to develop a rule on maintenance. We also want to assure you cf our complete cor=itment, as the top ranaEers of the industry, to the goal cf achieving improved reliability and safety through better maintenance.
Since joining the Commission over four years ago,'you .
have made it clear that you believed the maintenance programs in our nuclear stations were not achieving results at a rate commensurate with our overall goal of achieving operational excellence. That position has been supported and strengthened by each of the present Commissioners. The industry agreed with your assesstent and, in 1964, began an-extensive effort to analyze.
maintenance problems and to develop overall guidelines for the industry that identify the essential elements needed.in maintenance programs to reach our stated ob,4ective. Through EUMARC, INPO and EFRI and utilizing the maintenance expertisc within each of our companies, we then narrowed our focus to place core erphasis on specific areas of need.
These efferts led to the development of, and a ecmtitment by the industry to, a number of very deportant initiatives aired at improving maintenance at each of our stations. These include the accreditation of training programs for maintenance personnel, the commitment to the INPO Maintenance i '
Guidelines, a self-assessment by each plant of its maintenance programs against the INPO Maintenance Guidelines and, most importantly, focusing'manageeent attentien to correct identified weaknesses in each of our programs. Special INPO maintenance review teams were formed to assistd with the latter effort. These i the regular-evaluations,.
teams,now have whepe worksixteen visited is abovesites. and beyon,have Ve supperted the formation of special Centers, manaEed by the Electric Power Research
- Institute, to assist us with maintenance-related problers and applications specifically, the Nuclear Maintenance Assistance O
l !
i l
l.
i C _ _ _ _ _ ___ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ ______ _ _ _ _ __ __ _ _ _ _ __
- ' e,
.l ;
.{.;
The Honorable Lando W. Zech, Jr.
- Page 2 October 28, 1988
- Center (NKAC), the Maintenance Equipment Applications Center' (MEAC) with its associated Non-Destructive Evaluation Center; (NDEC), and the Component Monitoring and Diagnostic Technology Transfer Center (MEDC). Additionally, we have supported-industry-wide initiatives focused on specific problem issues such_
as motor-operated valves and check valves. We are gratified by your acknowledgment and that of the other Commissioners that these efforts are prodpcing results.- We ask your continued
- ' oversight and encouragement of these very important efforts.
It is clear, and.you have stated frequently,'that there is still core to be done. We agree. And we are'cocmitted'to continue our present efforts and to implement new efforts that prove to be beneficial. We have participated in every forum available to keep the NRC and the~public inforced of-our efforts, initiatives and co==itment. The industry had extensive .
participation in the NRC public workshop held in July, made a specific public briefing to the Com:ission in August on our.
initiatives and progress, appear'ed before the ACRS Subcommittee and full Committee several times, and has had' numerous interactions with the NRC Staff. With the exception nf the Commission itself, everyone involved in this ratter seems to agree with us that a rule on maintenance will not improve maintenance in our plants or improve the reliability or safety of our plants 4 In fact, many believe, as.we do, that proceeding with rulemaking at this time will have an overall adverse effect.
We ask fer your personal support and assistance in setting the cliente within the NRC to give proper credit to the many beneficial initiatives and positive results you have inspired, and net allow our efforts to be diluted or redirected by proceeding with a rulemaking on maintenance. We sincerely-believe that to do otherwise vill have an overall negative inpact en the industry.
We are ce=mitted to continue to work with'the NRC to review the existinr, programs as well as the need for additional programs or specific ef. orts we both agree vill assist in reaching cur mutual goal of enhanced reliability and safety. In that regard, we have established NUMARC as the focal point for wide coordination efforts and communication to improve with the NRC maintenance. NUMARC of industryly will work close with EFRI, national standards developing erganitations and others to ensure that our programs and activities are properly inteFrated and achieving our desired gosis. We*have requented NUMARC to periodically report to each of us en the status of these efforts and the progress we are achieving and to keep the Cernissien briefed en these activities on a periodic basis. NUMARC will
[ M.- ^
D L
i l
l:
'The Honorable Lando W. Zech, Jr.
Page 3
- l October 28, 1968 l
l, continue to work with INPO and the NRC to develop meaningful-ways to assess the effectiveness of our efforts. We have also asked NUMARC, INPO and EPRI to work together to determine'the feasibility of a program for a graded approach to maintenance to-allocate appropriate priority to components and systems that are important-to plant reliability and to the reduction in the number of challenges imposed upon safety-related equipment and systems.
We are receptive to constructive criticism and to new.
ideas. We ask you to maintain the present climate that encourages achievement and allows the industry and the staff to work together to reach the high levels of performance we all desire. We have the will and the co==itment. We need only the opportunity to continue en the course that has been set. To show the extent of our co==ittent, I have attached a list of'the-
- Members who have endorsed this letter.
Sincerely,
[ :.
Robert N< Campbel Chairman of the Board NUMARC Attachment Ccpies to:
Commissioner Themas M. Roberts Cec =issioner Kenneth M. Carr Co=tissioner Kenneth C. . Rogers Co= is sioner James R. Curtiss
I'
[
t~
a G T V eJ l 51gnecure .
- ,,,2t[(M/ M -
R. Patrick Mcdonald Milton P. Alexich Executive President Vice President' Nuclear Generation Nuclear Operations Alabama Power Company American Electric Power Service
.Corporat, ion l
l
+r s / y,,- nW M nacurg <
na:pe Donald B. Karner - Jerry L. Maulden Executive Vice President - ANFP *
. President & Chief Executive Offi Arizona Public Service Com;any Arkansas Power & Light Com;any
^
4 441d 1T. "
Eigns:ure g 51gnacute Charles W. Pryor, Jr. Joseph A. Tiernan Vice President & General Mar.ager Vice President Nuclear Po.er Division Nuclear Energy Babcock & Wilcox Cor;any Baltimore Gas & Electric Compar.)
rd Iga t' ' zignatuM Adrian 2 ca ,, Ralph G. Bird
- Sericr Vice President . Senior Vice President - Nuclear Bechtel Power Corporation Boston Edison Company A- 1 Wf Eagnacurs 8 5/pcune S. T d rewer I. E. Utley SM1by T. Brewer Senior Executive Vice President Executive Vice President Power Surely / Engineering & Const. Nuclear Power Systems Carolina Power & Light Company Combustion Engineering, Inc.
l
, ~ Q. .-
_2 5tgnature-T TA f.
msgyture.
Corcell Reed Eugene R. McGrath Senior Vice Presider.t : Executive Vice President Commor.wesith Edison Company Consolidated Edison Company of.
Nem York
$4a-5'. gna t ur e
<s Jn))k h asatwre Frederick W. Buckman Warren H. Owen President & Chief Operating Officer Executive Vice President Consumers Power Company ,- Duke Power Company
$.N
^
^
f b ,-
Eigt.a t ur e
- ssgnature Wesley W. von Schack C. O. Woody Chairman, President and Chief Executive Executive Vice Preiident Officer Florida Power & Light Company Duquesne Light Com;any
/ Ys f * -
n gna t ur e '
// h&.ature [/
Billy L. Griffin Eertram Wolfe Executive Vice President Vice President and General Mana; Engineering & Construction GE huclear Energy Florida fewer Corporation
~
ucu bigtsture Philip R. Clark seMC Eignature C JrI R. Patrick Mcdonald President & Chief Executive Officer Execu'tive Vice President GPU huclear Corporation Nuclear Operations Georgia Power Company
- g. ,
- h. .
31gnature I
ty L/
!".gnacure dik i.
E. Linn Draper Jr. 'J. H. Goldberg Chairman, President & Chief Executive Group Vice Presitent t - Nuclear Officer Houston Lighting A Power Company Gulf States Utilities Y) '
31gnature 51gr.ature [/'i/
Donald P. Hall Richard W. McCaughy Vice President Senior Vice President Illinois Power Company <
Production lowa [lectric Light E Power Compt
~
( A, r Eignature rj j- 51gr.e lpt r e Russell Yourgahl / James M. Cain President & Chief Operating Officer President Long Island Lighting Com;any Louisiana Power 8 Light Company x Yet+:%# -
fEt't;;'
h)f signature
);.
m' n
John B. Randarra Den E. Schaufelber ./
President President & Chief xecutive' Offi.
Maine Yankee Atomic Power Company Nebraska Pubitt Power District y:
k AhtA U. EA h M g ~
Erghatura A Charles V. Mangan M John C. Brons Executive Vice President - Senior Vice President Nuclear Generation Niagara Mohawk Power Corporation New York Power Authority l
1
\. _ _ _ _ _ _ _ _ _ - - _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ - _
._. ]
o cm ame v John F. Opeka I
,s. a....
Charles E. Larson Czecutive Vice President Vice President . Nuclear Generat Engineering and Operations Northern states Power Company Northeast Utilities NigT.a cur e
. , .W -M '
WA p / gigt.s tpe Wtiltam C. Jones V George A. Maneatis-Senior Vice President President Omaha Public Power District .- Pacific Gas & Electric Company
--,,s,--31 5sgr.ature r s' g;gn ggr, ,
Robert K. Ca*pbell . Corbin A. McNeill Chairman, President & Chief Executive Executive Vice President Officer Nuclear Pennsylvania Power & Light Cor;any Philadelphia Electric Company A 515:=.
YQ , '!A4 sapature j William J. Lindblad De1*in D. Hock President Generating Division President & Chief Executive Offs Portland General Electric Company Pu:,lic Service Company of Colora a
.1 .. cur.
h A.4
,s .. .r.
Ed.ard A. Brown Steven E. Miltenberger President & Chief Executive Officer vice President Public Service Coe;any of New Ham; shire Nuclear Operations
. New Hampshire Yankee Division Public Service Electric & Gas Cc 6
., 7_..
=-A zog l fignature ptr.st ur e Joseph F. Firlit
- j. Roger w. Kcber l Ser.ior Vice President Chief Executive Officer . Nuclear Production & Engineering Sacramento Municipal Utility Dis' l
Rochester Gas & Electric Corporation i
iyatun N
- wu~
ngnatur e Dan A. Nauman David J. Fogarty .
Senior Vice President Executive Vice President Power Production ." Southern California Edison Compa South Carolina Electric & Cas Company
._ y E -- *M
,y For.acure V [,j Sagt.s gre ' //
William J. L. Venr.edy V Willian Cavanaugh, !!!
Senier Vice President President & Chief Executive Offi Stone & Webster Engineering Corporation System Energy Resources, Inc.
1 4u
' gr.atura / FiEn T /\
Charles H. Fox, Jr.
J. 5.yc e varrington U Vice President & Nuclear Technical Chairman & Chief Executive Offic Director Texas Utilities Electric Com;an)
Tennessee Valley Authority
&t l S irs
- v. ,, .
. .... , f Lyfan C. Phillips B. Ralph Sylvia President Senior Vice President The Cleveland Electric Illuminating Company Nuclear Organization The Detroit Edison Company
)
i
_ - a
l
\:'
I
. BacartTrtpo 9 to 198~ Darkfttung.
taWil backfitting is defined as the i
modaficauon of or addluon to systems.
j structures, components, or design of a j- facDity; er the design approval er .,
manufseturing license for a faculty; or -
the procedures or orgardantion re-guired to design oonstruct or operate a faculty; any of which may result from a Des or sapended prortston in .
the Comstuasion rules er the imposF taon of a regulatory staff postuon in-terpretmg the Commuasion rules that is etther met or di!!erent from a prevn- ,
ously applicable stari postuon after:
til The date of Insunnoe of the non-struetton perupt for the faculty for fa.
stuuss having construeuen perumsts laved after October St.1945;or 39 CFA Ot.8 (I.148 Edie6en) gung ,g ,g g g , , ,g Will Sta months before the date of such documented evaluation shall to- fel No hornslag action a doctetmg of the operstirit lleense ask - clude a statement of the oejecuves of bold during the pendency pilostion for the faculty for facillues and reasons for the modification and analyses seguired by the Cc hating opnstruction permJts tasued the basis for trivoking the esception. rules.
before October St.1985. or (b) Progre.ph ist of this section gel afhe Misseuuve Direct (
tilal The data of lasuance of the op- shall not apply to beeLfits imposed atters AhaU be responsIDie ersting hernae for the facility for fa- prior to Octoteer 31.1946. mentation of ty.ls secuon a '
c111tlas harms operstme lleernes, or gelIn reuhang the determination re-Elv) The date of tasusnee of the ' autred by p6tsgraph tal of this tec- pues required by this secu desten approval under Appendia 3d. M uon the Cornmission otti corsider approved by the ERecuth or O of this part. hos the DacLfit should be pnoritund for Operspora er his destgr (3) The Commission shall recuire a and scheduled in light of other regula- 100 FDL aslit. aspt 30. 8e663
, systematic and documented analysts tory activities ongoing at the f6cally pursuant to paragraph tel of this nee' and. in addition, s'tu consider mforma-tion for bact.ftt.s afuch it seeks to tion kvallatie concerning any of the isepose.1spostiton of a backfit pursu- folles Ang factors as may be approprl-arnt to D&rtgraph daH4 kill of this seC* sie and arly other information rele-uen shaU Det relieve the Commlaston vant and material to the propeasd of perforgning art anaJysis after the jboCLf8t' fact to document the e&fety signif8' til Statement of the specific objet-emnee and appropriateness of the tives that the proposed backfit is de-
. action taken. signed to achieve; III The Comminston shall require (3) General desertation of the activt-the backfitsing of a faculty only Shen ty that would be resulted by the 11-it determines. based en the e.nalysis censee of app!! cant in order to sem-described in ptftgraph (t) of this ase" plete the backfit; tion. that there is a substantial in- (3) Potential change in the rtst to crease in the etert!! protection of the the public from sne accidental off site public health and atfety or the release of radioactive material.
eDfrJBon defense hnd securtty to be de* tel Potenus) impact on rod 201ogical thed from the backfit and that the esposure of factitty employees.
direct and indirect costs of implemen- (5) Installation and continuing costa nation Ihr that f&cility &re justtfled in gasociated l'tth the backfit incluoms etet of this increased protection. the cost of faculty donTitame er the 4 448 The provtsions of paragraphs cost of construction delay; tsWS) and tam 3) of this section are arb g8) The potenual safety impact of aps,licable and, therefore, backfit anaJ. changes in plant er operations! com-ysis as riot required and the standard plexity, includmg the relationship to does not apply there the staff ftnds proposed and entstmg regulatory st-and declares. with appropriate docu* guirements'.
mented tra!Usuon for its I1radms- ('!! The estimated resource burden either; en the 3FFLC associated with the pro-
- til That 6 modification is necessary posed bacLfit and the avausbuity of to bring a faculty into templiance such resourtes; 4rtth a lleense er the rules or orders of gS) '!he poteDtial impact of differ-
, the Commission. or into conformance enees in faculty type. design or age on Irlth Wrltten sommltments by the 18' the relevancy and pracucality of the eernsee; Or proposed backfit; (U) That an linmedjately effetuve (D) Whether the proposed bacLftt is regulatory action is necessary to gnterta er fmal and, if interna. the ensure that the faculty poses Do justific6 tion for impostng the pro-undue rtat to the public health and posed bact.fts on an interum beast anf tty.'
autros. theme analyent hooever eheute met epor these medfriestions which are to anvolve omst consleerations eserpt erur anser WtsIAFT that the rattllt3 panes no undue PtAs far es east sentributes to seiertins the solu-to the oubut heaJth and safetr and thach uen arnens earlous aserplable alternauves eM not eseIDod le reautre isMDedianetr ef- go ensurlas no undue ths Is publat henJth focuve feeunalerp erualL aneJyees am fe' and taletr
NUCLEAR MANAGEMENT AND RESOURCES COUNCIL 1776 Eye Street, N W.
- Suite 300
- Wostwngton. DC 20o06-2496 (202) 872-1280 syron Lee. .u. November 28, 1988 Prescent & Chef 1
, Executme Ottcer L
l The Honorable. Lando W. Zech, 'Jr.
Chairman U.S. Nuclear Regulatory Commission Washington, DC 20555
Dear Chairman Zech:
As published in the Federal Register dated November 28, 1988, the Notice of Proposed Rulemaking relative to maintenance requires public comment response within 60 days.
- NUMARC will coordinate and provide coments to the proposed rule reflecting inp2 from all of the U.S. nuclear utilities. Many of those companies will also be providing additional comments reflecting their own individual circumstances.
In order to provide an effective response, we request that the coment period be extended to 120 days. The extension is requested for the following reasons:
- 1. The NRC proposed rule, supporting data, and requested responses in an area as complex as maintenance requires extensive consideration and coordination of respon:;es and recommendations. .
i
- 2. A single, cohesive industry response that addresses the majority of the proposed rule will significantly reduce the effort of the NRC Staff and Commission in the consideration and disposition of public coments.
- 3. The advent of the holiday season will directly affect our ability to develop, have reviewed within the industry, and submit detailed comments on the proposed rule.
We sincerely want to provide constructive comments that reflect the extensive industry knowledge and experience. Our experience on the proposed rule on Fitness-for-Duty indicates that a comment period of 60 days is inadequate for us to ensure that our comments incorporate the input of our rnemmbers. We respectfully request your serious consideration of this extension. Please call me or Joe Colvin if you have any questions.
bincerely, O
mn -
1 ByHon Lee, Jr BL:spg cc: Commissioner Roberts Carr Rogers Curtiss Mr. Victor Stello, Jr.
-100-1
r a 3, .. .._. ,..
7g . --s ,
y&
.e
- L W.
a PG&E
,P , ^ PROPOSED MAINTENANCE RULEMAKING COMMENTS" AB' The NRC.has continuously failed tod'efine " effective maintenance"..
'AH The staff states that analysis of operational events. indicates that
. maintenance is not being. performed solas'to ensure a high. degree'of reliability. This statement is based on a 1985 NRC analysis of the fraction'of-significant events caused by maintenance. :The NRC-
' considers the maintenance percentage'too'high. .Since, however, maintenance-constitutes '2/3;of a plants activities, finding.that 1/3.
of the significant events are caused by .it should not be surprising..
1The real question is whether or'not the frequency of challenges and; safety system. availabilities are acceptable. If for example the
< industry only incurred one significant event per year but the: root
'cause'was maintenance, is maintenance =a significant' contributor to
. plant reliability and hence, of safety concern?
If there is'a wide variation in the effectiveness of maintenance programs,, the NRC should concentrate on those continuously.on the bottom of the distribution (Maintenance SALP 3's for example)'and not
. regulate the' entire industry.
o i- Al BOP.is beyond the scope.of NRC jurisdiction.
. Define effectively maintained.
~The NRC has suggested that BOP will be up to and including the protected area fence-(e.g. security). If they are serious about addressing all interfacing systems which could challenge plant safety features, then we should logically be required to extend the scope to include the offsite transmission. system; a major contributor-to
-transient csa11enges to safety systems.
b . Additionally if BOP reliability is truly a safety concern,~ force of logic will eventually require consideration of application of the full QA program to BOP. None of this is included in costs identified in the regulatory analysis.
l i
L AJ Maintenance process MPI's are generally of limited value. Such protecs
" criented indicators are hard to define consistently between utilities and'are generally susceptible to manipulation. General plant reliability measures such as safety system availability, ESF actuations, automatic scrams and equivalent ' availability, while not solely maintenance indicators, can provide significant insight.
Clearly, while poor plant availability can be attributable to many causes, prolonged high availability necessarily implies effective maintenance.
(MAINTEN.CMT) u
A0 We agree with these objectives, however, the conclusion ( AP) the Commission draws seems contradictory. If we want to
- 1) encourage industry initiatives
- 2) eliminate prescriptive rulemaking and
- 3) Identify expectations or standards of performance we wish to meet, then we should not develop a prescriptive, process based rule - even if it is developed by the industry itself.
AT If a programmatic standard must be developed we would prefer that the industry develop it.
AU. What is the nexus between a maintenance process standard and the NRC's objective of effective maintenance? The bottom line is safety system availability and minimizing challenges not whether we can weave an intricate web of maintenance processes. Are the processes effective?
AV BOP.is outside the NRC's jurisdiction. If BOP must be considered, it should be on a graded approach depending on a given BOP system's potential impact on safety functions.
AW We assume RCM is a sufficient basis on which to determine maintenance
( activities and requirements.
AX Clear and specific programmatic requirements do not necessarily imply high reliability.
BA To allow such flexibility is inconsistent with the basic objective of g
~
the rule - to diminish the wide industry variation in maintenance effectiveness. A maintenance standard will result in conformity at the j expense of those striving to excel. j BC The Japanese experience looks more like an integrated living schedule j approach than a prescribed programmatic maintenance standard. The *
{
Japanese do not seem to have programmatic standards, cnly PM scheduling requirements for major equipment. ,
BQ The maintenance rule could result in emasculation of good programs in i the attempt to address the poor. j Why not concentrate on the SALP Maintenance 3's? This would seem to meet the rule objectives of raising the level of performance of the outliers while still allowing industry and individual pirnt initiatives and flexibility.
[
(MAINTEN.CMT)
- - _ _ - _ _ _ _ _ - _ _ _ _ - _ _ _ _ _ _ - _ _ _ - _ _ - _ _ _ _ _ _ _ _ _ - _ - _ - _ _ _ _ - _ _ _ _ _ - _ - _ _ _ _ _ _ _ _ _ _ _ . _ _ _ - _. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___-__________________a
m ,
4 o
4
[
.We totally. agree with c. ' Regulation should address the outcome rather
-<i~ than'the-process. There is no assurance that any program will create.)
< the results desired. ,It is best to specify the' expected results and.
X let each utility modify its programs asLit see's best and'as.best fit-within'its particular cultureiand organization to achieve the desired-outcome.
- BR Component. reliability.would not be a relevant measureiof maintenance effectiveness. It does not provide credit.for system designs incorporating: redundancy to achieve.high' reliability of-function. The-NRC should keep its eye on the' bottom.line of.. availability of safety, functions. ..
BS The only realistic choice for such a role would be INPO.
1 BU. Assuming that a maintenance rule is ne'cessary it would.be more appropriate for.the. industry to. develop a standard. .The.NRC has indicated.in a!recent meeting that the Commission, with the help.from . '
consultants, will. develop such a standard if industry is not willing'. to -
do~so. A: standard developed by the NRC may misfocus important issues
-or.may be too restr.ictive. In fact, INPO 85-038, Guidelines for the:
Conduct of Maintenance at Nuclear Power Stations,.is the de facto-industry standard, t
BV During a recent meeting, the Commission indicated that~a standard for specific components is not intended and the INPO guidelines could be the basis of a general programmatic standard. In this regard the level of detail included in the INPO guidelines may be appropriate.
BW. Provided that both NRC and industry agreed to use the INPO guidelines as the basis for the required standard, two years may be considered as adequate to develop such a standard. In terms of implementing such a
- standard, different utilities would have different obstacles'. A total time of two years for developing AND implementing such a standard may or may not be reasonable depending primarily on the extent to which BOP must be incorporated.
BX A third party to certify plant maintenance programs might make sense with 1NPO being the most eligible candidate. However, under a rule, INPO would become an extension of the regulatory body, an unacceptable situation. ;
I (MAINTEN.CMT)
k 1
, j BY -It should be noted that NUMARC has not committed to develop a
' maintenance standard although they will work with NRC to improve the maintenance of the industry. However, if INPO 85-038 can be considered j as the industry standard or most of the standard there should not be
, any difficulties to meet the schedule.
BZ Application of 10CFR 50.109 a4 sets an extremely dangerous precedent.
Is the NRC claiming that current maintenance practices pose undue risk to public health & safety?
The use of adequate public protection as the reason for exemption from backfit analysis is questionable. If the proposed rule is needed to adequately protect the public, it implies that the plants are currently operating under unsafe conditions. Therefore we feel the backfit analysis should not be exempted.
CA BOP should be excluded from the scope of the rule. However, if required, all BOP components, systems and structures in this program should be evaluated for their impact on safety concerns. BOP equipment not related to safety will be addressed for other reasons (such as economics) at the discretion of the licensee.
CB Work accountability is an issue that we support in spirit. It is integral to the type of professional values that we want to instill in l
( all workers at the plant that they own their own work. They should think beyond the literal of procedures and feel accountable for the bigger picture. This includes feeling accountable for the quality and systemic implications of their work However, accountability may raise the issue of discipline. This is a difficult issue in that often the rank-and-file tend to look for the specter of punishment rather than for the reward in such proposals.
Worker accountability should have a positive side to it as well as merely being a catch-phrase for holding an offender " accountable" for
- errors. Tl.is should be a program to recognize good work -- the positive side of accountability -- and not indicate a focus on that punitive side.
CC This is a critical question. Until such time as the NRC develops quantifiable, objective standards that define a sufficient level of performance (acceptable risk), we will continue in this mode of forever ratcheting and never knowing when we're there.
.... Me agree that general measures of plant safety performance have improved over the last 5 years but it is our judgement that the rate of improvement is not sufficient. What rate would be considered acceptable? We're not sure".....(staff comments paraphrased)
(
(MAINTEN.CMT)
~
- l. The "w' hen enough is enough" criteria should be in the form of goals
~-7 associated with general plant safety performance. measures. If E
performance is consistently acceptable then programs and processes, whatever they are, must be working. For consistently unacceptable performance, the NRC must have license to require programmatic fixes.
CD . General plant performance indicators that are currently used by the industry are not totally appropriate for maintenance effectiveness feedback due to the number of non-maintenance factors included in them.
.For instance, an operational problem can contribute'to poor performance .
yet the maintenance practice may be acceptable. As stated previously, however, consistent superior performance relative to these indicators does reflect positively on maintenance effectiveness. Maintenance process indicators are not appropriate. See BR relative to component reliability.
CE A system similar to NpRDS, or NPRDS should prove to be beneficial to industry.in order to share the generic maintenance experience.
However, if the maintenance rule includes all components in the plant, it might not be feasible to have a system capacity large enough for this purpose. The use of.this system to monitor maintenance effectiveness by way of component reliability.is not desirable. If this system is. designed for free exchange of information, a regulatory function imposed on it may inhibit dialogue.
'(
CF We concur with Commissioner Roberts' comments.
CI Industry standards may well be tailored to the lowest common denominator. Once codified, it will be increasingly difficult for nuclear management to argue for enhancements when the current program meets requirements. Additionally, inspectors will likely have difficulty allowing for different approaches which, while they may be superior, do not conform with an established inspection guideline.
CW The Commission has been able to develop a comprehensive maintenance inspection module based on existing regulations. This module has been used during maintenance team inspections and NOV's issued based on nonconformance. The NRC has had no problems in citing utilities for program standards based on existing regulation.
Relationship is not developed - concentrate on quantifiable objectives.
Allow the utility the flexibility to design programs in ways which best fit their culture and particular circumstance.
CX See AH i
(MAINTEN.CMT)
y.
CY ,This connection has not been demonstrated. -There are a number of.
- i systems approaches and varying levels of formality which can all result-i- in " effective maintenance". What works best may well depend.on specific' plant cultures, organization, contracted orLin-house L- . engineering, size, license vintage etc.-
.CZ, Not necessarily. See ACRS's letter (Lewis' comments). Proactive maintenance seems to imply to the NRC more PM. Additional PM does not' necessarily result in.less CM. Every time a piece of equipment is touched that.is not'in a degraded condition, there is a probability that the PM.itself,will result-in CM if..not performed properly. There is also an increased probability of creating more;ESF challenges with more PM. .It's a fine-line and must be analyzed carefully to result.in a net reliability. benefit.
DC The relationship between maintenance effectiveness.and program requirements has not been demonstrated.
4 DD 1) We agree, if performance based.
DE' The standard will not be written to require cadillac maintenance systems and may degrade good plants at the cost of improving the poor performers. The average.may not change.
(
DF DF and DE seem inconsistent. The NRC wants uniformity and flexibility.
Minimum standards will soon define the standard for excellence.
El See DE EJ See DF FA Industry initiatives are given credit for 1/2 the public health &
safety and capacity factor benefits. There is currently a de facto.
industry standard which the commission claims is 90% of what they feel would be required of a maintenance standard. -Most plants are in conformance with this standard already. It is difficult to see how the rule'will bring any significant additional benefits. On this basis more credit should be given to the industry program.
t (MAINTEN.CMT)
e .
FD In all this analysis there is a basic assumption that a high I maintenance SALP rating is based on good maintenance processes. This relationship is not obvious. The NRC may place heavy waiting on bottom line plant performance e.g. availability, scrams, ESF actuations rather than process when assigning SALP ratings. The connections between safety system availability, capacity factor and process are not demonstrated. I l
5 FE I do not believe that Option 1,2,3 and 4 would be any different relative to reporting requirements. The NRC has made it clear that they expect an effectiveness measurement / feedback system as part of options 3/4.
l 1
GL Is it the same plant population year after year and safety indicator ;
after indicator that perform poorly? Is plant age a significant factor l relative to performance?
I GO Major INPO initiative on procedures since then.
GP Major INPO initiative on procedures since then.
- GQ This is a rather self-evident conclusion. Any function that comprises 70% of plant activity is likely to be responsible for a high fraction ;
of plant problems. The issue should not be whether maintenance is i responsible for a large fraction of safety system unavailability but ,
rather whether safety system unavailability is unacceptable relative to design basis or some other quantifiable risk related criteria.
10 SALP is not a measure of program comprehensiveness and/or formality and thus will provide little or no correlation with the impact of the rule.
IS What all this analysis may be uncovering is the fact that bottom line plant performance probably weighs very heavily when the NRC sits down to develop SALP ratings. It says nothing about correlations between highly formalized process oriented programs and performance.
IV French and German regulators do not impose programmatic requirements.
It is not clear that the Japanese do either.
IW See IS.
(
(MAINTEN.CMT)
KT ,The NRC must assume that a MPI'. system would be a necessary part of any.
industry ' standard. This is missing' from their.0ption 3 analysis and
- -I skews the results away from"the performance based . rule making options (l'& 2).
MG We feel that the extension of the maintenance standard to BOP sill-
= require a procedure upgrade program at most rather than 5 plants.
Additionally, we estimate the cost to be more like $10-20 million rather than.$2-3 million per. plant.
NH The basic assumption on which the 5% increase in capacity factor is based is a 1 point improvement in maintenance SALP. ratings. On what' basis does the NRC. assume that'the' rule will result in 1 point SALP increase?
L1
' 9 '
(
l-(MAINTEN.CMT)
____-__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - __ ._____ _ _ --_ .. . _ _ - - _ _ _ _ . _ _ _.