ML20082D178
| ML20082D178 | |
| Person / Time | |
|---|---|
| Issue date: | 06/30/1991 |
| From: | Beckner W, Chen J, Nilesh Chokshi, Jeng D, Kelly G, Kenneally R, Mccracken C, Murphy A, Reiter L NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | |
| References | |
| NUREG-1407, NUDOCS 9107230240 | |
| Download: ML20082D178 (81) | |
Text
_ _ _ _ - - _ - - -
.=
amaaaw = -
= x--:_
=;
Procecura anc Sul,mi:~:a.
Guic ance fo:r ":ae IncivicLua ly_alr: Examina~:1on or ax~:erna_
Everr:s C(PEEE) for Severe Accic. err: Vu nera xt:1es Final Report
~~
U.S. Nuclear Regulatory Commission Office of Nuclear Itegulatory llesearch J. T. Chen, N. C Chokshi, R N1. Kenneally, G. B. Kelly, W. D. lleckner, C. NicCt.teken, A. J, N1urphy, i Reiter D. Jeng p+* * %,
+
.. g
'l 9107230240 910630 07 PDR
._ _ __--~ _ _._._ ____._ __ _.___._
i
.1 t
' AVAILABILITY NOTICE Availability of Reference Materials Cited in NRC Pubhcotior.s
- Most documents cited in NRC publications will be available from one of the following sourcos-I l
1.
The NRC Pubhc Document Room, 2120 L Street, NW, Lower Level, Washington, DC i
20555 1
2.
The Superintendent of Documents, U.S. Govemment Printing O'fice. P.O. Box 37082,
']
Washington, DC 20013 7082 j
3,. The National Techn: cal information Service, Springfield, VA 22161 J
Although the listing that follows represents the majority of documents cited in NRC publica-tions, it is not intended to be exhaustive.
Referenced documents available for inspection and copying for a foo from the NRC Public
-Document Room include NRC correspondenco and infomal NRC memoranda: NRC Office of Inspection and Enforcement bulletins, circulars, information noticos, inspection and investi-gatio' notices; Licensee Event Reports; vendor reports and correspondence; Commission papers; and applicant and licensee documents and correspondence.
The following documents in the NUREG series are availablo for purchase from the GPO Sales Programi formal NRC staff and contractor reports, NRC-sponsored conferenco proceed-ings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regula-tions in the Codo of Federal Reguiations, and Nuclear Regulatory Commission issuances, Documents available from the National Technical Information Service include NUREG series reports and technical reports prepared by other fodoral egencios and reports propered by the Atomic Energy Commission, forerunner agency to the Nuclear Regulatory Commission.
Documents availablo from public and_ special technical libraries includo all open hierature items, such as books, journal and periodical articios, and transactions. f-ederal Register notices, federal and stato legislation, and congressional reports can usually be obtained from those libraries.
Documents such as theses, c ssertations, foreign rep 0rts and translations, and non-NGC confere1co proceedings'aro ava!!able !ct purchase from the organization spenscring the publication citea.
~
l g
g
- Shge copm of NGC cmu repons are acab:e fm, a the event or so;A unca maten re
-v tc tno Office ni f orman-Sn (Msaarc-n Uw ' 90% D obt' a
%-~ + <
1:
NE < > ( q u atory Cv m o 1 00 W n *;cn L i '. m l
1 k.
- A
~
'5
((l Jani '.i f preges tre % tas:-
.1t: - KC Librery j _ cu 5 %
- o. ue n. <
are 6 sabue thora f or m%na uso by t*w r -
1 jl i
'W%c~k J stW rMted cod may tw pun Msr:d imm W c;
";~*'
o' i'
,n I n eren i
f Mons Utenwds,- com thu t mann
- e cirr"rcs im+ n:m 14 D ramad %
l l
ij m v m Nv : con o
-.. - -.__,.,..-_.--_ -. ~--
- - ~ -
s
l NUREG-1407 Procedural anc Submittal Guic ance for the Individual Plant Exantination of Externa Events (IPEEE) for Severe Accident Vulnerabilities Final Report Manuscript Completed: April 1991 Date Published: June 1991 J. T. Chen. N. C. Chokshi, R. M. Kenneally, G.11. Kelly, W. D. Ileckner, C. McCracken, A. J. Murphy, l Reiter. D. Jeng Division of Safety Issue Resolution Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission
- Washington, DC 20555 p>>..m
AllSTRACT liased on a l'olicy Statement on Severe Accidents, the presents guidance for performing and reporting the re-licensee of each nuclear power plant is requested to per-suits of the individu:d plant examination of external form an individual plant examination.'lhe plant examina-events (IPl!Ill!).The guidance for reporting the results of tion systematically looks for vulnerabilities to severe acci-the individual plant examination of internal events (IPli) l dents and cost-effective safety improvements that reduce is presented in Null 110-1335.
or eliminate t he important vulnerabilities. This document I
i e
l i
iii N111(1i(i-1407
CONTENTS Page iii Abstract.........................,
ix Executive Summary..........,
xiii Acknowledgements......................
I 1 i nt rod uction........................
I 1.1 ll ac k gr o u nd.....................................................
I 1.2 I PEEE Objectives..........................
2 1.3 Putpose of Document..............
3 2 Events Evaluated for inclusion in the IPilllE...........
3 2.1 Seismic Events.............................
3 2.2 Int ernal Fires...............................
4 2.3 liigh Winds and Tornadoes........................
4 2.4 Ext ernal Fkiods................................
4 2.5 Transportation and Nearby Facility Accidents....
4 2.6 lightning...........................
5 2.7 Severe Temperature Transients (lixtreme llcat. lixtreme Cold).
5 2.8 Severe Weather Storms.........
5 2.9 External Fires (Forest Fires. Grass Fires)......
5 2.10 Extraterrestrial Activity (Meteorite Strikes. Sate!!ite Falls) 5 2.11 Volcanic Activity.......
5 2.12 Su mmary...........
6 3 Acceptable Methodologies for Performing the Seismic IPEE!!.,,,
6 3.1 Scismic PRA.............
6 3.1.1 New Scismic PR A Analysis.....
ti 3.1.1.1 General Considerations.
7 3.1.1.2 llazard Selection.............
7 3.1.1.3 Fragility Estimation 8
3.1.1.4 Scismic PRA Methodology linhancements..
8 3.1.1,5 Containment Performance 8
3.1.2 Use of an Existing PR A 10 3.2 Scismic Margin Methodologies.
10 3.2.1 General Considerations.
I1 3.2.2 Review 1.evel !!arthquake and Associated Spectral Shapes 12 3.2.3 Reduced. Scope Margins Methal..
12 3.2.4 NRC Seismic Margins Meth( dology 12 3.2.4.1 Walkdown..
12 3.2.4.2 Relay Evaluation 13 3.2.4.3 Soit Failures..
13 3.2.4.4 Screening Criteria (Use of Screening Tables).
13 3.2.4.5 Seismic Input.
13 3.2.4.6 Evaluation of Outliers-flCl.PF Calculations.
14 3.2.4.7 Nonseismic Failures and lluman Actions.
14 3.2.5 EPRI Seismic Margins Methodology 14 3.2.5.1 Selection of Alternative Success Paths 14 3.2.5.2 Walkdown 14 3.2.5.3 Relay Evaluation 14 3.2.5.4 Soit Failures 14 3.2.5.5 Screenmg Criteria (Use of Screening Tables).
NURl!G -1407 v
IJ
I i
3.2.5.6 Seismic Input 14 3.2.5.7 Evaluation of Outliers-llCLPF Calculations...
14 3.2.5.8 Nonscismic Failurcs and lluman Actions 14 3.2.6 Containment Performance..
14 3.3 Optional Methodologies...............
15 4 Acceptable Methodology for Performing the Internal Fires IPEEli.
16 4.1 New Fire PRA Analysis..
16 4.1.1 Identify Cntical Arcas of Vulnerability..
16 4.1.2 Calculate the Frequency of Fire Initiation in Each Area 16 4.1.3 Analyze for the Disabling of CrLical Safety Functions 16 4.1.4 Identify Fire. Induced initiating Events / Systems Analysis..
16 4.1.5. Perform Containment Analysis 16 4.2 Use of an Existing Fire PRA...,.
16 4.3 Optional Methodologies...
16 Acceptable Methodolo[ny for Performing the liigh Winds, Floods, and Transportation and Nearby 5
Facility Accident IPlili i.......
17 5.1 Intnx!uction.
17 5.2 Analytical Procedure 17 5.2.1 Review Plant-Specific llazard Data and Licensing !!ases 17 5.2.2 Identify Significant Changes Since OL Issuance 17 5.2.3 Determine if the Plant /Facilitics Design Meets 1975 SRP Criteria.
17 5.2.4 Determine if the Hazard Frequency is Acceptably low (Optional Step)..
17 5.2.5 Perform a llounding Analysis (Optional Step)...........
18 5.2.6 Perform a Probabilistic Risk Assessment (Optional Step) 18 5.3 Optional Methodologies 18 6 Coordination with Ongoing Programs..,.,.
19 6.1 Introduction.
19 6.2 Description of Ongoing Programs.
19 6.2.1 IPE Program Related to Internal Events 19 6.2.2 Programs Related to External I! vents 19 6.2.2.1 Scismic Programs.
19 6.2.2.2 Internal Fires Programs.
20 6.2.2.3 External Flooding Program.
20 6.3 Approach on Coordination with Ongoing Programs.
20 6.3.1 Coordination With Internal Events Program (IPE).
20 6.3.1.1 Preanalyses Planning 20 6.3.1.2 Plant Modifications.
20 6.3.1.3 Accident Management.
21 6.3.2 Coordination Among External Events Programs.,
21 6.3.3 Coordination With Seismic Programs.
21 6.3.3.1 USI A-45 and GI-131 21 6.3.3.2 The Eastern U.S. Seismicity Issue (I he Charleston liarthquake issue) 21 6.3.3.3 US! A-46 22 6.3.4 Coordination With Other Issues 22 7 Peer Review 24 7.1 Seismic Related Insights 24 NURliG-1407 vi
i 8 Docu m e n t at ion a n d ll e }.o rt in g................................................................
25 1
8.1 Infor mation Submitt ed to th e NltC.....................................................
25
)
8.? Information Itetained for Audit............................
25 9 ll e f e r e n ce s....................,...................................
26 Apperullees It eview i xve i llart h ;uake.................................................................... A-1 11 Comparison lle'wep. :. P cduced. Scope and 17ull. Scope Scismic Margins !! valuation................... 11-1 C
- 0. tai.ed Docum6.
't ion and Iteporting G uidelines............................................... C-1 D NitC ltespons to Commen ts and Ques tions.................................................. D-1
.I,igure 5.1 liccommended IPliilli A;."e.h for Winds 171oods, and Others...................................
18 Tables 3.1 Iteview lxvel liarthquake-Plant Jites liast of the llocky Moun ains...............................
9 3.2 lieview lxvel l!arthquake-Western U b States I'lant Sites..........
9 I
vii NURiiG-1407 i
l
i I
EXECUTIVE SUMM AlW llackgrounti Objectives ol' the li'EEE in the Commission puhey staternent on severe accidents
'!he rentral objectives of the IPilliare sirnitar to that of in nuclear power plants issued on August 8,1985, the the IPib that N for each heensee (1) to develop an Cornmission conduded, based on available information, appreciation of severe accident behavior, (2) to under-that existing plants pose no undue risk to the pubhc stand the most likely ses cre accider t sequences that could health and safety and that there is no present basis for occur at its plant under full. power operating conditions, immediate action on any regulatory requirements for (3)to gain a qualitative understandmg of the overalllike-these plants. Ilowever, the Commission recognited, hhood of core damage and fission pniduct releases, and based on NI(C and industry caperience with plant.
(4)if nectasary, to teduce the overall hkelihmd of core specific probabihstic risk assessments (Pl( As), that sys.
dunare and fission product releases by modifying, w here tematic examinations are beneficial in identifying plant, appropriate, hardware and procedures that would help specific vulnerabdities to severe accidentr. that could be in ennt or mitigate sever <: accidents.1 he key outcome of fned with low cost improvementt As part of the imple.
an IPli ill is the knowledre and appropriate improve-mentation of the Severe Accident Policy, the Commission ments resulting from such an examination process.The issued Genene I etter 88-20 on Nosember 23,19S8, re.
examination can be conducted using any of the acceptable questing that cath hcensee conduct an individual plant approac het examination (IPl!) for internally initiated events includ-int nternal nooding.
I(lentification of thternal Iwents i
incluttett in the ll'EEE ManyI'll Asindicate that,in some instances, the risk f om external events coeld (ontnbute significantly to core in snpporting the implernt eaahon of the Severe Accident darnage. llowever, the examination for externally initi.
Policy, a sady was performed to determine which exter-ved events is proceeding on a later scheJule to allow the nal initiators could be a potenhally important accident stad to carry out additional work to (1) identify which initiator that may pose a threat of severe core damage or exter nal harards need a systematic examination. (2)iden-of a large radioactive release to the environment. The tify acceptable examinahon methods and develop procc-external es entu onsidered, conustent with past probabil-dural and submittal ruidance, and (3) coordinate the inde istic nsk assessments (PR As), are thost events whose vidualplant examination of externaleventt(IPElil!,with cause is external to all systems used durmg normal and other ongoing external event programs. In December emerrenev operations. The external events evaluated in-1987, an !!xternal lhents Steering Group (liliSO) was ciude scismic events, internal fues, high wmds and torna-established to make recomrnendations regardmg the does, external flooJs, transportation and nemby facility scope, methoJs, and cooidination of the IPl!!!li. The accidents, lightning, severe temperalme transients (ev ITSG completed its task in M'y 1990. Based on the treme heat, cxtreme cold) severe winter stonns. external liliSG recommendations the staff prepared this report fir es (forest fires, rr ass lires), cxtrater r est nal activity (me-to provide detailed puidance to the beensees on the con-teonte 5tnkes, satelhte falls), and volcanic achvity, duct oIthe IPl!!!!! and on the structure and content of the IPl!Eli submittal. The staff issued a draft of this report liased on the results of that study, the staf f has concluded for public co-untnt in July 1990. It helJ a workshop in that five external events need to be included specifically September 1%0 to explain the IIT.l!!! process and to ir, the IPl!!T: scismic events, mternal fires, high winds, obtain specibe comments and questions on draft Supple.
Doods, and transportahon and nearby facility accidents.
ment 4 to Generic l~etter bS-20 and the draft of this llowever, hcensees should confirm that no other plant-documentt in addition to numerous comments and ques.
unique external events with potential sesere accident Lions laised donng the workshop, the staf f received wnt.
sulnerability are bemg excluded from the IPlilm ten comments from 1h oryanuations.
Eunnination Methotis This revised report renects the stalPs thomuch conud-eration of the public comments recened. It prevides.
Seismic Fxents specifically, the pmJelines defining the IPEl E objec-tives; identibes external n ents that should be mcluded in A seismic Pit A (i evei 1 plus contamment performance the IPEl;E; idenhhes acceptabte methodolopes; idente analysis) er a seismic margms methoJolory (SMM) is hes coordinahon between the IPI l li and the onpomp considered a uable appioach toi&ntify potennal sulner-Nitt ' pt oprams; and piouJes the stafPs responses to puh-ahibties. (imdance is pmuJed hir hecusees pcr for nung a lie comments and questmnt new senmic Pil A ci updatmy an eskung sennue PI( A:
is N t ;li t:(i-1407
i lixecutive Summary t
emphasis is placed on the identification and ranking of in ines easing levcis of detail, ef fort, and resolutien. Ilow-dominant plant sequences that could lead to seismically ever, the licensee may choose to bypass one or more steps induced core damage rather than on the numerical esti-so long as it identifies the vulnerahihties or demonstrates mate of absolute frequency of occurrence. Methodology that they are insi;tnificant. 'lhe screening approach con-upgrades include plant walkdowns, evaluation of relay sists of the following steps:
chatter, and evaluation of the effects of soilliquefaction.
i Guidance is also provided for licensees using either the 1.
Iteview plantapecific ha/ard data and licensing NitC or EPill4ponsored scismie margins methodology, bases.
4 The margins nieihodology sereens components according to their importance to safety and seismic capacity. Ily 2.
Identify significant changes since the operating li-design, the methodology utili/cs two review or screeninE ccnse was issued.
lesels geared to peak ground accelerations of 0.3g and 0.5g. Iteview level earthquakes (Ill.lis) were assigned 3.
Determine if the plant and facilities design rnects based on the lawrence 1.ivermore National I aboratory the 1975 Standard Iteview Plan (SitP)eriteria.
(l.i.NI.) and filectric Power itesearch Institute (liPitt) hasard ccimates, sensitivity tests, seismological and engi-If the 10b Sit P eriterla are not satisfied, orif it is known a neering judgment, and plant design considerations.1hc glori that they will not be satisfied, one or more of the use of the d3g full-scepe and focused seope 111.!! for following approaches should be taken to forther evaluate most plants in the Central and liastern United States the situat on, i
would meet IPl!Illi objectives. The level of effort in the analysis of relay chatter is the inajor difference between Optional:
these two categories. I?or some sites wher; the sebmic 4.
Determine if the hazard frequency is acceptablylow, h Tard islow, a reduced scope mar $ns mcthodology em-phasl/ing plant walkdowns is considered tiequate For i
Perfor m a bounding analysis.
site
- in the Western United States, except Cahforma coastal sites, the 0.5g Ill l! should be used. Methodology 6.
Perform a probabilistic risk assessment (Pila).
upgrades include relay chatter, liquefaction, and plant walkdown enhancements for the NitC methixh guidance g.nli Wihds on alternative access paths f or the !!Pitt method; and evaluation of nonseismic failure and human actions for The staf f recognizes t hat other methods capable of identi-both methods.
fying plant-specific vulnerabilities to severe accidents may be acceptable. A licensee may request that the staf f Internal Fires review any other systematic examination methm! to de-termine if it is acceptable for IPlil!!! purposes.
7ternal fires li'lilil' can be accomplished by per-lo. ming a irvel l fire PI( A.Those issucs identifieJ in the l' ire Risk Scoping StuJy (NUlGiG/Cl(40SS) should he U,UINUD,UU
- b OUNUIUN >NSUUd I
addressed using plantopecific data and a specially tai-Guidme is movidd on coordinating the IPlil!!! process lored walsdown procedore, wim onpig pgrm The fd lesel of comdination is among the major elements related 1o Ihe implemer.tation The cuidance does not address the fire sulnerabihty of the Severe Accident Policy, that ir, coordination among evatuation (I'lVF) methodology currently being devel' the IPlitili, the internal events IPF, containment per-oped by the Nuclear Management and itesources t ouned formnce igrovements, and accident management.The (NUM ARC) and liPRI. 'l his methodology is being re-second level of eoordination is among t he major elements
+
viewed by the staff; when the reviewys completed, the of the ll'lilili, that is, seismie evems fires, and high staff willissue an evaluation report on its acceptabihty for winds floods, and others.The third level of coordination use in the IPl!!!!"
is within each major element of the IPl!!!!1 liigh Winde Floods, and Transportation and Programs subsumed into the IPl!!!I! include the external Nearby Facility Accidents event aspect of Unresolved Safety Issue (USI) A-45 (decay heat removal), Generic issue (01)-131 (in-core The recommended overall approach consists of a pro-flux inapping system) and the !! astern U.S. Seismicity pressive screening.1hc screening criterion for reporting Wharleston car thquake) hsue. Programs that need to be potential severe accident sequences is consistent with coordinated with the IPlWE include USl A-46 (Scismic that used for internal event IPlis. The steps in the pro-Equipment Qaalification, which also covers the s ?istnie gressive sercening approach represent a series of analyses spatial systems interaction of USl A-17 and the concern NLJitliG-1107 x
lisecutive Surninary i
I I
of USI A-40 for the seismic capability of large safety.
l'Cerl(CVICW related above ground tanks), and GI-$7 (lif fects of I are l'actection Systein Actuation on Safcty itelated 1: quip.
' Die licensee should conduct a peer review by indiv;Juals who are not associated with the initial evaluation to on-rnent).
sure the accuracy of the doeurnentation package and to validate both the IPlil 11 process and its results.
I 1
xi NUltliG-1407
ACKNOWI.EDGEMENTS
'this dosurnent re!
.nts the staff position on the Indi.
contubutors to the desclopment of this guiJanee docu.
viJual Plant ihar.,... tion for severe accident s ulner.
rnent: they ar e named below. in addition. significant input abdities dec to exter nal evt nts (IPI tili). ltepresentatives was recen'ed f ram consultants and contractors to the of both the Of fice of Nuclear llegulatory itesearch and NRC, who are : b.o named below. I dward Ibil, of NitC, the Ofhee of Nuclear iteactor itefulation were active provided tecl.nical editing.
NitC Charles I:.. AJer Guy A. Arlotto William D lleckner Goutam llagthi Demetrios !.. ItasJekas I;a/irnieras h1. Carnpe T. Y. Chang John T. Chen
' thomas hl. Cheng Nilesh C. Chokshi Adel A. lil-llassioni John 11. I' lack it. Wayne llouston David C. Jer.g Glenn 11. Kelly Itoper h1. Kenneally Thomas 1 Kmg P. T. Kuo Conrad hieCracken Warren himners Jocelyn A. hiitthell Thomas li. N1orley Andrew J. h1urphy I) avid P. Notley
'l i mas hl. Novak
- 1. con lleiter Jarnes li. Ilichardson Itobert ltothman 1awrence C. Shao lirian Sheron Jack StrosmJer ites G. Wescott Consultants it. J. Iludmt/
I uture llesoure-s Ass.ocutes. Inc.
G. E. Cununings law rence 1.iscirnore National laboratory it. P. Kennedy Str uctural hiethanics Consulting, Inc.
N1. K. Itasindra EQl! 1:ngmeenny, Inc.
Contrattors and Suhtontra(tor s P. Ami.o Science Application international
- 1) l.. liernreuter I aw rence 1.isermor e National I ahora*ory hi,1.Ilohn Sandu National laboratories it. J. Iludmt/
l~uture liesources Associates, Inc.
D.11. Chung I aw rence 1.ivermore National I ahoratory
- 11. C. Davis lawrence 1 iserrnure National I ahoratory G. S. Ilardy
!!Ql! lingmeeting, Inc.
J. it. hieDonald Texas Tech Urnversity it. C. hiun ay I aw rence 1.tvermore Nanonal I aboratory A. N1. Nafday t!Qi! lingineering, Inc.
P. G. Piassmos lawrence 1.ivermore National I aboratory N1. K. Itavindra l'.Ql! !!ngineenng, Inc.
J. Savy Iawrence 1 ivermoie National Iaboratory sui NUltlE 1407
1 INTRODUCTION 1.1 lhlChgrOllild the lil SO completed its task in May 1990. Ilased on the lil!SG recommendations, the staff prepared this tcport to provide detailed guidance on the conduct of the IPlilil!
On August 8,1985, the Nuclear llegulatory Commission and on the structure and content of the IPliliti submittal.
issued a policy staternent on severe accidents (NitC, it issued a draft of this report for public comment in July 1985). 'lhe Commission concluded, based on available 1990. It conducted a workshop in September 1990 to information, that existing plants pose no undue risk to the explain the IPlilill process and to obtain specific com.
public health and safety and that therc is no present basis ments and questions on draft Supplement 4 to Generic for immediate action on any regulatory requirements for 1,ctter 88-20 and the draft of this guidance document,in these plants. However, the Commission recognizes, addition to the comments raised during the workshop, f he based ca NitC and industry expericace with plant-staff received written comments from 16 organizations.
specific probabilistic risk assessments (PRAs), that sys' This final report includes changes resulting from the reso-tematic examinatior s are beneficial in identifying plant-lution of these comments.
specific vulnerabilities to severe accidents that wuld be fixcd with low-cost improvements. As part of the imple.
],2 [p(([g[3()hjpeggyes mentatmn of the polig statement, the Commission ts-sued Generic l etter 88-20 (NitC,1988 and 1989), re-
'!he objectives of the IPlil!!!, which ate similar to the questing that each licensee conduct an individual plant objectives of the internal event IPIL are for each licensee:
examination (IPl!) for internally initiated eunts.
1.
to develop an appreciation of severe accident
- behavior, llisk assessments indicate that the risk from external events could be a significtml contributor to the core darn-2.
to understand the most hkely severe accident se-age in some instances. Ilowever, licensees were re-qucnces that could occur at the licensee's plant un-quested to proceed with the examinations only for inter-der full power operating conditions, nally initiated events (including internal flooding) in Generic 1,etter 88-20. I!xamination of severc accident 3.
to gain a qualitative understanding of the overall vulnerabilities duc to externaby imtiated es ents (IPlilTII) likelihood of core damage and fission product re-is proceeding separately and on a later schedule to allow leases, and the staff to carry out additional work (SIIC A88447) to (1) identify which external hazards need a systematic 4.
if necessary, to reduce the overalllikelihood of core examination, (2) identify examination methods and de-damage and radioactive material releases by modi-velop procedural guidmce, and (3) coordinate the Ipl:1!E lying, where appropriate, hardware and procedures with other ongoing NitC programs thai deai with various that would help prevent or mitigate severe acci-aspects of external event evaluations to ensure that thete
- dents, is no duplication of industry cf forts, llowever, the staf f recoenized at the outset that the exter.
m:I initiators could not necessarily be treated in exactly To accomplish these objectives, the staff established the the same way as internal initiators in the implementation lixternal thents Steering Group (li!!SG)in December of the Sevet.; Accident Policy because the sources and 1987 to make recommendations regarding the scope, treatmcnt of uncertainties in estimates of core damage methods, and coordination of the IPl!Eli (lleckjord, frequencies for external and internal events can be quite 1987,19S8). Specifically, the liliSG is responsible for different. in addition, some methods endorsed by the staff developing broad guidance for dealing with (1) cxternal for evaluating c:ternal hazards and identifying vulner-events on a generic basis both organizationally and tech-abilitics do not produce estimates of core damage fre-nologically and (2) the implementation of the severe acci-quency. lir example, seismic margins methods produce dent policy with respect to external events. The litiSG estimates of seismic hazard levels of high confidence-low establisned three technical subcommittees dealing with probability of failure (llCl.Pl?) for a plant rather than earthquakes (seismic events'). internal fires, and high estimates of core damage probability.
winds, floods, and "other" external events. The subcom-mittees were chartered to define the scopc of the external Therefore, the staf determined that an explicit estimate events examination, identify acceptable examination of cere damage frequency was not needed to meet the methodologies, and coordinate ongoint ssues and activi.
intent of the Severe Accident Policy and would not be a i
tics (for example, Unresolved Safety issues and Generic condition of the IPlil!!! Thus, Objecthe 3 above would l
1ssues).
lie addressed only indir ecdy by some methods acceptab!c
.1 NUltlCG-1407
- l. Introduction for use in the IPlilill Nevertheless, the key objective of state-of the art improvementsareidentifiedin Sections 3 gaining an understanding of plant behavior through the through 5. Section 3 addreses the seismic portion: Sec.
examination pmcess could be met, tion 4 the internal fires portion; and Section 5 the high winds, floods, and other portion of Ihe IPlilili. Coardina-1.3 Purpose of Documcrit tion between the Irlilill and the internai events IPli, The purpose of this document is to provide guidelines for other external events, and ongoing prograrns within each conducting the IPlil?Il and on the structure and content external event are provided in Section 6. A discussica of of the IPlil!!! submittal, it is not the intent of the peer review is provided in Section 7. A suminary of NUltl10-1407 to go beyond the infonnation request con, documentation and reporting guidelines is provided in tained in Supplement 4 to Generic Letter 88-20. 'lhe Section 8. 'lle staff's responses to public comrnents nnd external events recornmended for inclusion in the IPlil!!!
questions raised during the IPliliti workshop held in Sep-are identified in Section 2. Acceptable methodologies for tember 1990 and the written comments received soon performing an IP111!li along with upgrades to reflect afterward are given in Appendix il I
l i
l f
N Ul(I!G-1407 3
~ - -
2 EVENTS EVALUATED FOR INCLUSION IN Tile II'EEE The external events considered, consistent with past rule out scismic events on the basis of initiating probabilistic risk assessments (PilAs), are those events event frequencies.
whose cause is external to all systems used in normal operation and emergency operation situations. Internal 3.
llased primarily on their vintage, the current popu-fire and inte nal ibod are external to the " system" and lation of plants exhibit various levels of seismic de-therefore have been ctmsidered as external events in past sign requirements and margin. Some of the very Pl( As. Ilowever, internal floods ar e being considered in early plants have been backfitted under the System-the internal events IPl! process (N1(C,1988).
atic I! valuation Program to ensure certain margins for safe shutdown using criteria different from cut-In supporting the implementation of the Severe Accident rent licensing criteria.
Policy, a study of the risk of core damage to nuclear power 4.
'there have been modifications to plants since their plants in the United States due to externally initiated events was performed. 'lhe objective was to determine original designs; for instance, the reduction of snub-which caternalinitiators have the potentialof initiating an bers at some plants. 'lhese changes, in part, have accident that may lead to severe reactor core damage or eclied on existing conservatism or risk-based argu-large radioactive release to the crivironment. Scismically ments (e.g., l.OC A + SSli combinations) 'lhe sys-initiated events are investigeted in NUlti!G/rlt-5042, tematic examination of plantsby the lPl! and lPl!!!!!
Suppl.1; internal fires, high winds / tornadoes, external will give an integrated picture of plants as they exist.
floods, and transpottation accidents are investigated in it will also allow an integrated evaluation of the NUlt!!G/ Cit-5042: "other external events" are investi-eficcts of individual changes made to plants over gated in NUltliG/ Cit-5042, Suppl. 2.'lhe "other exter.
time.
nal events" covered are nearby industrial / military facility 5.
There are unresolved safety issues and generic is-accidents, on site hazardous material storage accidents, severe temperature transients, severe weather storms, sues (e.g., USl A-45. USl A-46) that are in various lightning strikes, external fires, extraterrestrial activity, stages of implementation. The IPli/IPlil!!! provides volcanic activity, carth movement, and abrasive wind-a convenient as well as meaningful fcamework for addressing many of these issues.
storms.
^*""
- C
- Y"* C9 "
C" Some external events may not pose a significant threat of in c st-effecuve plant-specific improvements.
a severe accident to all plants, some events may have been considered in the plant's design to a sufficient degree, and
'lherefore, the seismic external hazard should be in-some events may have been or will be reviewed under cluded in the IPlilili.
ongoing programs at some plants.The staff's evaluation and recommendations are contained in the followinit 2,2 111101'11111 Fires
- sections, liased upon the examination of past fire Pit As, the ( ont ri-2.1 Seismic Events bution of internal fires to the probability of core damage may be significant and is very plant specific (NUllliG/
'the following are based upon an cxamination of current Cit-5042), llowever, the numerical teruits always con-seismic design criteria, previous and ongoing seismic is-tain large uncertainties. 'lhe fire risk scoping study sues and programs, and'scismic Pil As:
(NUltliG/ Cit-5088) further confirms the following:
1.
Mean scismic core damage frequencies calculated 1.
The overall fire-induced cole damage frequency for from past PilAs (NUltliG-il50, NUltliG/ Cit-the four plants studied (Seabrook. Oconce, l.imer-5042, Suppl.1) have been found to be in the range of ick, and Indian Point)incicased from the original 10 A to 10 8 per year. Identified vulnerabilities are PI( A studies even though, for certain fire scenarios, plant specific and include yard tanks, electrical there was a net decrease.17or all plants reviewed, equipment, diesel peripherals, structural failures, fire contmues to represent a dominant risk con-and equipment anchorag's.
- tributor, 2.
New data such as the occurrences of larger than 2.
Most initiating event frequencies were increased anticipatcd earthquakes and the development of based on a much more complete data base available new hypotheses indicate that the plant-specific scis-on fire occurrences in nuclear power plants. Under mic hazard may be quite different from that envi-cmrently applied risk assessment methodologies, sioned at the time of licensing and make it difficult to this iacrease in initiating event frequency alone 3
NUlt!!G--1407
i
- 2. l! vents livaluated for IPl!!!II results m a direct increase in overall fire-induced view Plan sections, particularly Section 2.4, floods pose no core damage frequency wit. all other factors re-significant threat of a severe accident because the maining constant.
exceedance frequeng of the design basis flood, excluding floods due to failure of upstream dams,isjudged to be less 3.
Use of an expanded data base on historical fire sup-than 103 per year (Chery,1985), and the conditional c ore pression times for nuclear power plants resulted in a uamage frequency for a design basis flood is judged to be suppression probability distribution with a lower less than 10J. Thus core damage frequencies are esti-probability of suppression within a given time than mated to be less than 10 8 per year for a plant designed that assumed in the original risk assessments. Under against N!(C's cut tent criteria. Ilowever, the latest prob.
current methodologies, this again results in an in-abic maximum precipitation (PMP) criteria published by 4
i crease in fire initiated core damage frequeng with the National Weather Ser ice (NWS) call for higher rain-all other factors remaining constant.
fall intensities over shorter time intervals and smaller areas than have previously been considered; this could 4.
Updated information on the ignition and damage result in higher site flooding levels and greater roof p(md-thresholds of cabic insulation in some cases resulted ing loads than have been used in previous design bases in lower thermal damage limits. In some cases, no (01103) Licensees are requested to assess the effects of
. change in damage limits was requit ed. A decrease in applying these new criteria to their plants in terms of the assumed thermal damage limits would, in gen-onsita floodingand roof ponding. Also, some older plants c.al, be expected to lead to increased estimates of may have higher patential risk and need sptematic exami.
fire-mitiated core damage frequency.
nations for plant-specific vulnerabilities.
5.
Plant modifications made as a result of Appendix 11 M Tiaigmbtim el NemIg requirements teduced the core damage frequency at Indian Point and Limerick for the requantified areas I',llCIIIIY ACCIIICIIIS by factors of ten and three, respectively. For Seabnok, the identified Appendix 1( plant modifica,
,l'hese events consist of accidents related to transporta.
tions did not affect the requantified cme damage tion and accidents at mdustrial and military facihties.
scenarios for internal fires. The Oconce Pita had Plants designed against NR C's current critena (N URl!GI already incorporated Appendix R modifications and CRM should have no significant vulnerabihty ta se-vere accidents from these events because the imtlators no modifications subsequent to in performance were identified. llence no Appendix R impact could considered m the design should have a recurrence fre-be identified for either Seabrook or Oconce.
quency leu than 103 or base been shown through a bounding analysis not to affect the plant. llowever, 6.
A number of issues that were not adJtessed in the changes may has e occurred since the onginal design and past fire PRAs (ef fectiveness of fire brigade, effee, there inay be exceptions that need some splematic liveness of fire barrier, scismic/ fire interactions, examination. Also, some older plants may not meet the control system interactions, and effects of fue sup.
NI(C's curr ent criteria and need systematic examinations pressrmts on safety equipment) could increase the for plant specific vulnerabilities.
vulnerability to fire.
2.6 1.iglit nliig Therefore, based on the above studies. the internal fire Ij htning has been expenenced at many nuclear power hazard should be included in the IPlilili, g
plants in the United States (NURliG/ Cit-5t142, buppl. 2:
2.3 Iligli Wiil(IS allti Toritatloes AlioD.1986: ACRSs 19h9). The impact of i>tnmg on plant operation and the vulnerability of plants to a severe For plants designed against NRC's current criteria, these accident due to lightning has been examined.The major
. events pose no significant threat of a severe accident conclusion is that the primary impact of lightning on nu-because the current design criteria for wind are domi clear power plants is the loss of offsite power.The loss of nated by tornadoes having an annual frequency of offsite power is included as part of the internal events 4
exceedance of about 10J. llowever, older plants and IPli, and examination for vulnerabilities duc to this aspect some modern plants having facilities not designed against oflightning is therefore already included in the 1Pli proe-these criteria need a systematic examination to identify ess.The staff has concluded that,in general. other effects plant-specific s ulnerabilities (NURl!G/CR-5042).
of lightning on nuclear power plants are insigmficant.
Ilow ever, further examination of lightning effects may be 2,4 IUterilal lilOutls warranted for certam sites w here, based on past operating experience, lightning strd.es ate likety to cause more than l'or plants designed against current criteria as described just loss of off sit e power;1or example, they may also affect in Regulatory Guide 1.59 and applicable Standard Re-safety related instrumentation and control systems.
i NURI O 1407 4
- 2. livents livaluated for IPl!!!!!
liased on an exammation of historical data on lightning. as caused several complete and partial losses of offsite well as knowledge of plant systems the staff concludes power (NUlti!G/Cib5042, Supph 2). The potential ef-the following:
feet of iuss of offsite power and station blackout wih be addressed in the internal event IPli; thus severe weather 1.
I.ightning has typically caused partial or complete storm evaluations need not be repeated in the IPlilill loss of offsite power, which is the main impact of bghtning and which is aheady bemg examined as 2.9 IWiertial liirCS (l',orest I,. ires,arass part c' the internal events IPl!.
IIirCS) 2.
1.ightnmg is much less hkely to affect the onsite power system.
These are fires occurring outside the plant site boundary.
Pottntial effcets on the plant wuld be loss of offsite 3.
I.ightning has affected safety related equipment and power and forced isolation of the plant ventilation and has caused reactor trips, but these esents have not possible control room evacuation. Usuauy, extern;d fires been significant in terms d impact on the plant.
are unable to sptcad onsite because of site clearing during the construction stage (NUltliG/ Cit-5042, Suppl. 2).
4.
Safety systems (e.g., diesel rencrators, electrie.dly llowever, there has been one instance during which a powered pumps) are not norniauy in geration.
nearby fotest fire caused a partial loss of offsite power.
Thus, whde control systen nay be d. unaged, the
'lhe cfIcet of loss of offsite pow er will be addressed in the safety systems are less susceptible to dannpc and internal events IPli anJ need not be repeated in the may be manually activated.
IPlil!!! The other effects have been evaluated dming operating beense (OL) review against sufficiently con-5.
Redundancy of nfety systems and the capabihty for servative crituia; thus they do not need to be rea3sessed recovery of sy tenis (ieplaemp fuses ;r resettmg in the IPhlil!
breakers) further reduce the likelihood that the low h equency of lightnin;' damage events will result in a severe accident.
2.10 Intraterrestrial Activity Oleteorite Strikes, Satellite Falls)
'lhe stalf hasj. dged that the probability of a severe acci-u dent caused by lirbtning (other than one due to loss of offsite power)is relatively low and lurther consiJeration baannuial auMy n eqnWM to be natural satel-of lightning effects should be petformed only for plant huch as mmon or attiheal satellites tint enter the sites where lirhtning suikes are 1.kely to cause'more than qiW.s anno @ca hom gn kauw & probabihty just loss of of fsite power or a uram.
of a metconte stoke is very sinall (less than 103)
(N URl:G/CR-5012, huppl. 2), it can be dismissed on the 2.7 h,CVore I,clHperature,I.raHsients bws ohts low initiating esent hequency.
(lhtr01110 IIPat, lhtreille Cultl)
Severe tempemture transients may air"t nuclear Power 2.11 Volcanic Activity plants in the Umted States (NURl!G-1032t flowever, the ef fects are u :uaHy limited to r educing the capaeny of Niost nucicar power olant +ates ate too far away from the ultimate heat sink and loss of offsite power (NURI (il acove solcanacs to espect any clfect at the plant, so most CRda12. Suppl. 2) 'ihe capacity reduction of the ulu-licensees need not cons:dcr the wicame elfeels, llow-mate heat sink would be a slow process that allows plant ever. those sites in the ucinity of actn e wleanoes shaulJ operatorr suffietent time to take proper actions such as assesn oicame activines (N U RiiG /CR-5012, Su ppt 2) as reducing power output level or achiesing safe shutdown, past of the 1P1 lili procest if necessary, and maintaining the plant in a safe shutdow n condition.The other potential impact on the ptant, loss ol offsite power, will be consideied within the realm of the 2,12 Sugungary station blackout i ule tNRC, lushb) and the mter nal es ent IPl2 t herefoie, the tempeialme trainicots need not be in summaiy. based on the alue evaluation, five esents addicssed m the IPl i li need to be meluded by all heensees in the IPl 1 12 'eisnue events inte nal lues, high wmds thiods and transpoita-2.8 SCVore WCatller StorIns tam anJ naiby f acihty accidents All hcensees sumilJ conhnn, howeser that no plant-unique esternal esents Ses cre w eathei stoims bec stoon h.uMoi m. saowstoim, know n to the hcensee today with potential setet e as ci-Just stoim sandstonnwcompanied by su ony wmJs h o c dent sulnci Ahly me bent culuded fium the IPl:1~l, 5
Nt iRI G-1407
3 ACCEL'TAllLE METilODOLOGIES FOlt l'ERFORMING TIIE FEISMIC IPEEE 17or the purpose of performing an IPlilill, two method-and supports, and common.cause effects (the culling or ologies are considered acceptable to identtfy potential pruning of trees should be dont with these considerations seismic vulnetabilities at nuclear power plants. 'the first is in mind); and (4) internal event models should be devel-a *cismic probabilistic risk assessment (Pit A)(NUld!G/
oped knowing that, in the seismic analysis, the fragilities Cit-2300; NUllEG/ Cit-2815 Vol. 2: NUlGIG/ Cit-of a component ate sensitive to elevation. Also, a compo-4840). In addition, an evaluation of the reliability and nent and its peripheral equipment may have different usefulness of results and insights obtained from external fragilities. Additional discussion of this subject can be event Pila methodologies is contained in NUlt!!G/
found in NUld!G/Cib4840.
Cit-5477. The second method is one of the scismic snargins methodologies (Shihi) described in NUltEG/
PltA calculations that account for all uncertainties are Cib4334, NUluiG/ Cit-4482, and NUltliG/ Cit-5076, clearly acceptable. Ilowever, the staff believes that, for EPill NP-6041 or the reduced Shiht desenbed later in the scismic IPlili!!, it is not necessary to carry out com-this section*
plete uncertainty quantifications defming a distribution of core-damage frequencies in order to identify vulner.
In mecting the objectives of the IPliEE, the examination abilities. hican point estimation using a single hazard should focus on qualitative insights from the systematic curve (rather than a family of hazard curves) and a single plant examination rather than only on absolute core dam-fragility curve (rather than a family of fragility curves) for age frequency estimates. Guidance for performing the cach component is sufficient to get insights into potential scismic IPEEli using a Pila or margins me'hodology is seismic vulnerabilities. To highlight the most pertinent provided below.
results/ insights from the seismic portion of the IPE!!E, mean point estimates using hazard curves described in 3.1 Seismic l'ItA NUlUiG/ Cit-5250 and EPiti NP-6395D should be ob-tained. Further discussion on the use of hazard curves is This discussion deals with the use of Pita techniques in cont..ined in Section 3.1.1.2.
the scismic IPEEl! The Pil A should be at least a Level 1 plus containment performance analysis. The basic ele-The above point estimation approach is valid only be-ments are (1) hazard analysis,(2) plant system and struc-cause of the IPE!!!! objective: to identify dominant se-ture response analysis,(3) evaluation of componcot fra-quences and components and where possible rank them.
gilities and failure modes,(4) plant system and seqt ence (This point estimate should not be confused with a " Phase analysis, and (5) containment and containment system 1" type Pila analysis where point estimate calculations analysis including source terms, to identify unique seismic are used only to define scopes for more detailed Phase 11 sequences or vulnerabilities different from the internal and Phase 111 studies), Fragilities used in this point esti-event analysis. Specific guidance and enhancements are mate, where possible, should be plant specific and rigor-provided for licensees performing a new Pila or updat-ous to be able to identify dominant components and rank ing an existing seismic PltA-them. Correlations and other aspects of analysis should be treated so that, when a mean scismic hazard curve is 3.1.1 New Seismic PRA Analysis ud with the mean plant fragility curve, the resulting core damage estimate approximately represents the mean estimate that would be derived from the full uncer-3.1.1.1 General Cons,derations i
tainty analysis.
1.icensees choosing to do a seismic Pit A built on an inter-nal events Pil A should be aware of important considera.
The recommendation of performing point estimation tions that, if incorporated in the planning of the internal type calculations is made primarily to highlight insights ever,ts Pila, will rninimize their resource expenditure needed for the severe accident behavior perspective.This and speed the staff reviews. For example, (1) a welb should not be construed as de-emphasizing or ignoring organized walldown team and a properly planned walk-uncertainties. Analysts are encouraged to make careful down will enable many issues to be adJressed at the same study of the oiigins of the possible uncertaintiec,includ-time:(2) the peer review team shoulJ consider the need ing those that are hardest to quantify. hiany of the insights to review both internal and external event analyses; obtained from a PR A analysis are obtained by trying to (3) fault tree analysts for internal events should be aware gain a better understandmg of the uncertainties. Consid-of spatial mteractions (including internal flooding ef-eration of uncertainties may affect how the results of a fects), failure of panive comlunents such as structures PR A are implemented in plant changes.
NURIIG-1407 6
- - - - -~
- 3. 'lhe Seismie IPlilili i
i 3.1.1.2 llamd Selection consistent with current estimates of pround motion should be used. In the Central and liastern United Statcs.
For the Uaited States cast of the Rocky Mountains, two current spectral estimates am be found in the I.lRL and highly sophisticated seismic bazard studies were con.
I!PRI studies. Since simdar spectral shapes are obtained ducted by lawrence Iivermore National Iaboratory from LLNL and I!PRI hazard studies, separate analyses (LI AL)(NURl!O/CR-5250) and the filectric Power Re-using both spectral shapes are not needed. Median spec search Institute (liPRI)(FPRI NP-6395-D). For many tral shapes of 10,000 year return period provided in sites, these studies yield significant differences at the low-NURF.0/CR-5250 along with variability estimates are probability and high level ground motions. 'lhe initial recommended for use in the analyses. Other site 4pecific PR A* carried out using these estimates (Surry and Peach spectral shape estimat es may be proposed (that is, derived liottom in NURIIG-1150) indicate that, despite large from a suite cf appropriate recorded carthquakes). For diffmnces in absolute nurnetical estimates, the identi-the Western United States, site 4ptcific spectral shape fication, ranking, and relative contributions of the domi-should be established and used. Since only one spectra!
nant seismic sequences are sirtually the same for both shape is used for both hazard analyses two separate plant 1 I.NL and FPRI hazard estimates. This equivalence is response and fragility analyses are not needed.
apparently due to the fact that the slopes of the seismic hazard curves are not significantly different over those if an upper bound cutoff to Fround motion at less than ground motion !cvels that, in conjunction with the fragill~
1.5 g peak ground acceleration is assumed, sensitivity ties, cor. trol the relative distribution of seismically in-studies should be conducted to determine whether the duced core damage frequencies. Although these results use of this cutoff affects the delineation and ranking of are very encouraging, there is no guarantee that this will seismic sequences, be true for all sites in the Central and liastern United States.
3.1.1.3 l'ragility 1:stimation While a full seismic hazard uncestainty analysis is rmt
'the following guidance on fragi ity estimation is included t
necessary in performing a seismic PRA for the IPlil!!:,
to clarify the use of fragibty in the context of the " point the staff prefe 8 that mean (arithmetic) hazard estimates estimation" approach discussed above. Details and from both the 1.LNL and the !!PRI studies should be used methods for fragility and high confidence of low probabil-to obtain two dilferent point (mean) estimates. lf a licen-ty of failure (ilCLPF) calculations are discussed in a see elects to perforrn only one analysts, it should use the number of references, for example, NURl!G/CR-2300, higher of the two mean (arithmetic) hazard estimates.
NURl!G/CR-4334, NUREG/CR-4482, NUltliG/CR-5076, NURiiG/CR-4659, Vols.1-3,12PRI-NP 6041, and
'Ihe we of both the LLNL and IIPRI mean hazard curves NURIIG/CR-5270, it is recognbed that large uncertam-has another advantage in that the extent of uncertainty ties exist in fragilities estuaation (NURl!G/CR-5270). A will become obvious and the emphasis on the bottom line perspective on how this uncertainty af fec's the results of numbers is reduced.The use of both of these estiraates analysis (numeric 2d and other insights, for example, domi-(LIM L and I!PRI) will serve to identify differences, tf any, nant sequences Imd component 4 should be maintaineA in the delineation of dominant seismic sequences (minor variations in contributions and rankings are anticipated).
Consistent with the point estimation apptoach, or e c2m Such differences would have to be addressed by the licen-use a single mean component fragility cune for each see in its identification and listing of vulnerabilities, component and hence for sequence level and plant-level assessments. ' Itis mean curve is defmed by the median For plants in the Western United States, for which there capacity,6, and composite uncertainty, se, where sc2 =
are no counterparts to the ILNL and I!PRI studies, a pr? + su?, when pr and pu are estimated separately. pr licensee should conduct its own study to define the mean and su represent random uncertainty and modeling un-l' hazard estimate for use in the IPlilili. 'the licensee certainty, respectively, it is also acceptable to use a family shon3 also provide reasonable assurance that any signifi-of fragility curves instead of a singic curve.
cant uncertainty in those elements of hazard (for exam-plc, slope) that control the identification, ranking, and When a single mean fragility curve is available, llCLPF relative ec'tribution of seismic contributors to core dam, capacity for a component (sequence, or plant) can be age is addressed in sensitivity studies. As in the Central approximated by 2) se below the median (ip.,1% com-and liastern Umted States, the identification and listing posite probability of failure is essentially equivalent to of vulnerabilities should take this uncertainty into ac-95% confidence of less than 5% probability of failure).
count.
While developing sequence-level and plant.!cvel fragili-ties, the licensee should retain the ability to report Most scismie PRAs use peak ground acce'eration as the 1ICI.PFs with and without nonscismic failures and human hazard parameter. lf this is done, spectral shapes that are actions.
7 NUREG-1407 l
l
3, 'lhe Seismic IPlil:l!
3.1.1.4 Scismic Pila Methodolit.y 1:nhancements does not supply hcl.PF cabulations, the stall will calculate the llCl.PF values based on in'ormation i
1(eview of par t seismic l'R As indicates that certain areas provided in the ll'lil'.li subrnittel and will use thern have been treated either inconsistently or not at all.
in the evaluation of the subnuttal. Note that plant-
'lherefore, the following areas should be included:
level, sequence. level, and comp nent fragilities are 1.
Plant li'alkdmwis. Walkdowns are performed to find i
as-designed, as-built,andas-operated seismicwcak-3.1.1.5 Containment Performance nesses in plants. liach licensee should perform a walkdown consistent with the intent of the guide-
.the primaiy purpose of the containment performance lines desenbed in Sections 5 and 8 and Apoendices paluation is to identify sequenm and vulnerabilities t hat D and I of the !!PRI Seismic Margins Methodology inmtve cimtainment, contamment functions, and con-(l!PRI NP-6041).
tainment systems (e.g., igniters an ! Suppresgion pools) seismic failure modes or timing that are sigmheantly dif.
2.
Relay Charter. Relays, in this context, include com-m@on. Additional guidance is presented in Sectio ent fmm mme found in the IPliinternal esents evalu.
ponents such as electric relays, contactors, and i
switches that are prone to chatter. Additional guld-g ggg g g,g anec is given in NURl!G/CR-5499.The scope of the relay chatter evaluation should be consistent with The use of an existing seismic PRA to address the seismic the site's scismic margin review level earthquake IPlil!!! is acceptable provided the PR A reflects the cur-classification (full scope oc focused scope) as identi-rent as built and as. operated condition of the plant and i
fied in Tables 3.1 and 3.2 and divuned in See-some of the dericiencies of past PR As discussed beloware tions 3.2.4.2 and 3.2.5.3. 'lhat is, a more complete adequately addressed.
evaluation is to be performed for sites in the full-1.
Hazard Selection. For PR As at sites cast of the lhicky scope category than for sites in the focused. scope category. It is anticipated that chatter and recovery Mountains that did not use the I.l_N1. and !!PRI actions will be modeleu as necessary. The focused.
mean hazard estimates, sensitivity studies should bc l
scepe evaluation can be limited to a review of low conducted to determine if the use of these results seismic ruggedness telays for plants that are not would affect the delineation or ranking of seismic included in the USI A-46 program, sequences. For l' ras in the Western United States, Ihe sensitivity studies should be carried out to deter-Ihe examination of the relay chatter effects (for mine the effect of uncertainty in hatard on the de-example, the Ilatch margin evaluation) has resulted lineation and ranking of seismic sequences.
in large resource expenditutes in terms of staff' 2.
IMdmvns. Since a walkdown is considered to be hours. 'lherefore, careful planning and use of ge-oncof the mostimportant ingredientsof theseismic neric insights,if they are applicable to the plan 5, are IPlilil!, a supplementary walkdown in confo: tr.ance I
necessary. Additional guidance on this topic is also with the intent of the procedures desenhed in Sec-included in Section 3.2.4.2.
tions 5 and 8 and Appendices D and I of the IIPRI margin methodology (!!PRI Ni'-6041) should be 3.
Liquefaction..the potent:al for soil liquefaction and performed. It may be necessary ta amplify the car-associated effects on the plant need to be "xamined lier analysis based on the walkdown outcome.'lhese for some sites because of specific site conditions.
re3ults sihould be reported.
'lhe impact on plant operation should be assessed from the point of view of both potential for and 3.
Relay Chatter. Relay chatter effects either have not consequences of liquefaction. Procedures for assess-been considered or were assumed fully tecoverable ing soit liquefaction are described in liPRI N P4041.
in past PR As. Relays, in this context, include com-4.
ponents such as electric relays, contactors, and
_ HCLPF_ Cu/culations (Optiona/h 1.icensees can re-switches that are pmne to chatter I.icensees should port plant letel, sequence level, and component-analyic the effect of relay chatter or deterrnine that level llCl.PF-values and use this information to the type of relays used in the safety systems are not support decisions related to the identification and subject to relay chatter. 'lhe scope of the review listing of vulnerabilities. In ses cral PR As (for exam-should be consistent with the site's scismic margins plc, Millstone 3 and Diablo Canyon), HCI PF esti-review ic-cl carthquake classification (full sco,ne or mates are reported along with other PRA results.
focused scope) as identified in Tables 3.l and 3.2 and These PRAs c2m be used as guidance for deriving discussed in Sections 3.2.4.2 and 3.2.53 Add;tional 1ICLPFvalues from fragilities.11C1 PFralues are to guidance is provided in NURl!G/CR-5499. Results be reported both with and without the effects of of this effort that lead to a PR A revision or plant nonseismic failures and human actions, if a licensee fixes should be reported.
NURiiG-1407.
8 i
3 'the Seistnie ll'l!!!!!
Table 3.1 Itolew I.ott l'arthquake-Plant Sites l'.ast of the llo(Ly hlountains iteduted S(ope fl;g Itot.L Point 1)uane Arnold
- South Texas Turkey Pt.
Comanche Peak G and Gulf St.1.ucie Waterford Crystal lliver itiver llend 0.3g I'ocused Scope Arkansas #2
!) avis-llesse
!jmerick Salem licaver Valley firesden hieGuire Shoreham llellefonte lirley h1illstone Surnm er*
liraidsood iermi hionticello Sorry li t/ patrick Nine h1ile Pt.
Susquehanna llrowns i erry i
litunswick I' ort Calhoun North Anna'
'lhree hiile Island llyton Gmna Oyster Creek Vermont Yankee Callaway lladdam Neck Palisades Vogtle Calvert Chifs liarris Peath llottom Wans llar Cataw ba*
llatch Perry Wolf Creek Clinton Ilope Creek Point lleach Zion Cook 1;cwaunce Prairie IslanJ Cooper 1 aSalle Quad Cities 0.3g l'ull S(ope Atlansas #1 hiaine Yankee liobinson Yankee llowe Indian Point Oconce' Sequoyah Conimitted to Po for m a Seismic Pila l'ily nm" Seabrook
Noter
- SpCt't.nl D110 01100 lo hhallo% L,oll Condilhth$ h apploplialf Iet thCAC lW alH sus (M C SCChon 3 2 2)
'Ihlay thatter evaluahon shouki be mundar to a f ull wope itvn a.
Table 3.2 Itesiew l.esel l'.arthquake-Westet n United St.ites Plant Sites 0.5 g*
Trojan 1(ancho Seco Washmpton Nuclear Palo Verde Seismic hlargin h1(thods I)o Not Apply to the l'ollowing Sites:
liiablo Canyon San Onohe Noics:
llllCd MaICA %Ils %f tb raO!!Illh it n S. UldCV lbf llc 0llVC Ciln 450 llitillNU hit' th ilIllC %l1C ha.'Did i% hihldal lo
- Ilk!K alC4 a $11014 I!n' WO%18?l!) k I R
lhal al sales ra',1 el lhe Kirky Mountann th.il alc found m the U.h bm Changes in the retic % Icui carlbriuake tann 0 5p lo u 3g shi ull be approsCd hiloir doing sign hc.mt an >'yas.
O N L ilt!!G - 1407 A
i
- 3. 'Ihe Seismic IPl!!!!!
1 T
4.
Nonscismic Failures amt fluman Actions. In sevecal 3,2,1 Getscral Cortsideratiotti:
seismic Plt As, nonseismic failures (c.g., failures of the auxiliary feedwater system and fadure of feed
'lhe beismic margin metha!okigy is considered accept-and biced mode of core cooling, oalter) depletion, able for addressing seismic concerns in the severe acci-I power-operated relief valve failures) and hornan dent policy implementation. Two rnethodologies arc cut-actions (e.g., de!ays or failures in performing sf cel.
rently avadable: one developed under NRC sponsorship fied actions or operator misdiagnoses a situation and and the other developed under I!PRI sponsonship 'the takes ar> improper action that is not related to the staff has determined that both methods (with the noted actual, current plant situation) have been important enhancements) will adequately address IPlil!!! objec.
contributors to seismically induced core damage tives-frequencies or risk indices. Unless nonscismic fail-urcs are considered, improper decisions may be "the two ruethods use different system analpis pMaso-made regarding plant modifiuitions or procedural phics. 'lhe NRC method is based on an event / fault tree changes.
approach to delineate accident sequences. l'or example, for PWRs, two safety functions are considered to be most important to plant seismic safety; reactor suberaticality
'the licensee has the option to expand its PRA or and early emergency core cooling. If these functions are demonstrate that the exclusion of nonseismic fail-
- nsured for a given carthquake. there is high confidence uses will not sigmficantly alter the PHA results or that core damage would not occur at that levelJihe !!PRI insights.*lhe scope of nonseismic failures and hu-methodohity is based on a systems " success path" ap-man interactions that might affect seismic sequence?
proach, This approach defines and evaluates Ihe capacity should be defined by the licensee bas ed on the inter-of those components required to brmg the plant to a
- nal events analyses.
stable condition (cither hot or cold shutdown) and inain.
tain that condition for at least 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. Several possible Liquyaction. *lhe potential for soil liquefaction anJ succm pah may st. HoW k NRC and the EPRI associated c!fects on the plant need to be examined inethods were used in the trial applic,ation at the flatch for some sites because of specific site conditions, plant. Application of the NRC method was limited to a
'the impact or' plant operation should be assessed spn
- w. InQhts gamd hom the uce oMese two from the point of view of both potential for and mdo& and the ddferences between them are discussed consequences of liquefaction. Procedu res for assess-n repmW D s et at W), and %ao et ing soil liquefaction are described in IIPRI NP-6041.
"b @N liach liecnsee should examine its plant critically to enc 6.
Conruinment Performance. Iicensees should ensure that the generic insights used in margin methodology-o that the performance of containment a:. ' contain, development to identtfy critical functions, systems, and ment systems are addressed. Section 3.2.6 contains success path logie are applicable to its plant.This is par-guidance.
ticularly vital for older plants where systems and fune-tions may differ greatly from the plants considered in the development of the margins methodologies (NURilG/
7.
1/CLN' Colo,larium (Optional). I icensees can ex-CR-4334, NUREGICR-4482, NURl!G/CR-$076, and
- tract and report plant level, dominant-sequence-EPRI NP-604 th level,anddominant component lew!1ICl.PFsboth with and without the effects of nonseismic failure Based on written comments on draft of NURl!G-1407 and human actions and use this information to sup-and public meetings, the staff has defined three catego-por t decisions telated to t he identification and listim; rics of margin studies i; quiring varying levels of effort, of vulnerabilities. If the licensee does not supply
'the three categories are fulbscope, focused scope, and
' 1ICLPF calculations, Ihe staff will calculat e 1 ICl.PF reduced scope. 'lhe focused-scope category is new, w hile values based on information provided in the IPElill the fulbscope (although not identified as such in draft-submittal and will use them in the r valuation of the NUREG-1407) and reduced-scope categories are re-submittal, e
tained. 'lhe primary purpose of this further subdivision is i
10 reduce the level of effort for some plants. Liecnsees of phnts with relatively higher hazard and a lower seismic 3.2 SelSmle Marg,m Methodologies design basis wat conduct more detailed studies than will iicensees of other plants (groupiag of plants into various
'lhis discussion deals with the use of the seismic margin categories is discussed in Section 3.2.2 and Appendix A),
methodolopy in the seismie lPE E!!. Specifically, guidance and enhancements are provided for a heensee using
'ihe aamination scope in each category for NRC and either the NRC or liPRI margins methodology.
EPRI methods is discussed in detail m Sections 3.2.4 and J
- NUR FO-1407 10
d
- 3. 'lhe Seismic IPl!!il!
3.2.5. An introductory discussion of the rationale and the 3.2,2 lleview Lewl Ettriliqttake and general characterization of full 4 cope and foused-scope Associated Spectral Shapes follows.
- lhe scismic margins methodology was designed to dem-onstrate sufficient margin over SSli to ensure plant miciy and to find any " weak links" that might bmit the plant
'the rnajor difference between the full scope and the shutdown cipability to safely withstand a seismic event focused scope is the scope of relay chatter evaluation.
bigger than SSil.'lhe scismic margin method utilizes two llased on detailed studies carried out at the flatch and review or screening levels geared to peak ground accel-Diablo Canyon plants at considerable resource expendi-erations of 0.3g and 0.5g. It it the staff's judgement that ture, it appears that the problems that could be caused by the use of a 0.3g review level earthquake (Ill.I') for most relay chatter at these plants were recoverable with exist-of the nuclear power plant sites in the Central and !!ast-ing procedures, llowever, there is a concern among the ern United States (east of the Itocky Mountains) would staff and industry consultants that such conclusions can serve to meet the objectives of the IPl!!!!!. Ilowever, all not be censidered generic without some additional plant sites east of the llocky Mountains are not subject to the teviews. Additionally, both the NitC4ponsored and the same level of carthquake haeard. For some sites where
!!PRI sponsored relay tests indicate that relay perforrn-the seismic hazard is low, a reduced 4 cope margin ap-ance is very sensitive to variables such as sprmg tension, proach center ed on walkdowns is acceptable. For western orientation, and mounting. Tests further indicate that a sites other than California coastal sites, a 0.5g RLl!
significant number of relays may have capacities between should be used foi the margin approach. '1he ItL11s SSII and RLli levels. USI A-46 evaluations are to be defined for U.S. sites, as well as sites that can perform a i
perforrned only at the SSil icvels. Therefore, the staff is r educed 4 cope SMM at e presented in Tables 3.1 ar.d 3.2.
recomn'ending that the full 4 cope plants evaluate relay The seistnic margin evaluations should utilire the chatter in a manner consistent with the approach sug.
NUlt!!G/ Cit-0098 median rock or soil rpectrum gested in l! Pill NP-6041 or equivalent. Note that consid.
anchoicd at 0.3g or 0.5g depending on the g level and ernble efficiency crm be achieved using lessons learned primary condition at the site. Further discussion on the from the llatch experience (Moore et al.,1990). For review level carthquake is presented in Appendix A.
plants in *e focused scupe caterory, a low level of cffort is recommended The lessons haled from the full 4 cope Plants in the 0 3g bin are further st. livided into full.and i
reviews may necessitate tc.4amining the relay chatter focused scope categories, as discussed earlier. This cate-issuen through the generic ' issues process (i.e.,
gorization is based on consideration of harard as well as NURiiG-0933).
the seismic design basis. Additional consideration is also given to the outlier plants resulting irom the IIastern U.S.
Seismicity (the Charleston !!arthquake) Issue resolution.
..Other differences between the full-and focused scope relate to the level of effort for evaluating soil failure
'lhe ground motion should be considered at the surfacc in modes and the number of margin calculations (llCI.PFs) the free field. If secondary conditions such as ahallow soil to be reported (Reed et al.1990; Rasin,1990). 'the differ-conditions are being considered, appropriate procedures ence in the level of effort in these areas stems primarily should be used to determine the free field motion in the from a perceived need for more accurately characterizing vicinity of those affected structures and components, and plant behavior and numerical results for plants in the the capacity evaluation of structures and comp (ments full-scope category, it is also perceived that licensees of should take into account the cifects of soil-structure in.
plants in the focused-scope category will be able to iden.
teraction.
tify important vulnerabilities with more liberal use of fewer, approximate, and hounding type analyses without liccause reecnt ground motion estimates, such as those adverse impact. Ilowever, the actual !cvel of cl' ort in included in the LLNL and I!PRI hazard studies, indicate these areas is very much site and plant dependent and relatively higher ground motion at frequencies greatt, should be determined on the basis of plant specific con' than 10 lit than that shown in the NUltIIG/ Cit-0098 siderations. For example, a plant in the full-scope cate-spectrum, the margin evaluation of only nonductile com-gory that is hicated on a cock site will not perform any soit ponents (if uppropriate)-for instar.ce, rel.ys-that are failure evaluation, while a plant in the focused scope sensitive to high frequencies should be performed as dis-category that is hicated at a coastal plain site may require cussed in Section 3.2.4.1 No plant-specific response
, more sophisticated investigations. In any case, discussions analysis is anticipated to address concerns related to high-y here do not preclude the use of properly substantiated kcquency ground motion. However, if a lices.sce decides judgments to define the scope and depth of an examin-to evaluate plant response for high-frequency ground ation.
motion, the response spectral shapes derived from the 11 NURiiG-1407 L -
- 3. '!he Snmie IPlil:F.
I I
appropriate site-specific inedian uniform hatud re-l'ocused Scort, l'ull Score, aml 0.5g sponse spectra (10,000 year return perioJ) shown in
.Dw wddown Would be performed and documented in NUltliG/ Cit-52$0 anchored at 0 3g or 0.5g should be accor dance with the recommendations contained in liPitt used.
NP40R 3.2.3 lleduced Scope N1argins 51ellun!
3.2.4.2 Itelay 13 aluation l'or sites where the scismic hazard islow, a reduced scope Itelay s in this context include such components as electric seismic margins inethod emphasi/ing the walkdown is relay s, contactors, and switches that are prone to chatter.
adequate. Well-conducted, detailed walkdowns have been demonstrated to be the most important tool for W fo00 wing parymphs define the scope of the relay chatter evaluation for each of the three bins:
ident fying scistnic weak links whose correction is highly cost cIfective. Applicable sites are idenafied in Table 3.1.
Reduced Scope The initial steps of the full-scope margm methodolory up USl A-4611 ants-Completion of the USl A-46 review to and including the initial plant walkdown er e p"rfor med will satisfy the IPlilili intent, ierardless of method selected (NitC or I!Pltl). Ilasically, certinent actisities up to and including the initial plant Non A-46 Plants-No action is needed.
walkdown need to be perforrned.These activities include gathering system information, classifying front-line sys.
Focuscd Scope tems and identifying front.line components, classifyinf USl A-46 Plants -l'ollow USI A-46 procedures. If low-support systerrs and identifying support system compo-seismic.ruggednew relays are discoverca during the USl nents, and identifying plant umque f eatures*
A-4b icview, the relay review should be expanded to include relays outside the scope of liSI A-46 but within i urther guiJance on the differences between the the scope of the IPlilif t reduced. scope and full scope seismic margins methods, that is, elements preserved and clernents eliminated are Non A-46 Plants-1ucate and evaluate low-seismic-prouded in Appendix 11.
ruppeJness relays (bad actor list).
The evaluation should be documented in a walkdown full Score anJ U.5g (Including ifc3 tern U.S. flanto team report and snbjected to a peer resiew (see Sec-USI A-4611 ants-1:ollow USl A-46 procedures for the tion 7)-
resiew; review spterns within the scope of the IPl!!ili, including those that are also within the scope of l'S1 3.2.4 NitC Seismic N1argins N1etliodology A-46, unng appmpri t margin mPlu NP-om)or USl A-46 ptocedures at the RI.h I
The ruidance in NUltliG/ Cit-4334. N Ulti!G / Cit-44S2, Non A-4h Plants-Review the rehys m.al'systemswithin j
and NURiiG/Cll-5076 is supplememed by that in the We wope ohhe IPElib. usmg ap' ionnat 7 - sn (l!PRI following sections to (1) reflect the partitioning of the WhM) or USI A-46 procedu.
t ti.t 0.3g screening criteria into the reduced-scope, 0.3g focused-scope, and 0.3g null-scope bins idenufied in Ta.
.Hw method as ongmally developed o e not address ble 3.1 and (2) identify enh;mtements so that the method the relay chat ter iwue because infonnation on the !.aject can be used f >r IPFlili implementation.
was lacking. I lardy et al. (1954), summanecJ the results of several efforts in tlus area and provided ruidance to ad-3.2.4.1 Walldown drew this issue in an IPlilili context. Relay chatter analy-sis could be s esource intensive, and careful planning and Reduced Score use of renerie insights,if they are apphcaMe to the plant, are desuable. Insights and recommenJations based on See Appendix 11, Secuons 11.1 (1) through (4) for puid-the i lata esperienee are avadable in Moore e t al. (1990).
ance.
Attempts to address the concerns related to high-Emphasis on walkdow ns also applies to containment anJ fr equency pround motion by analy sis is s cry hkely to entail contamment systems (that is, containment functions re extensive ef forts, including the development of new and quired to prevent early Gulure, contamment integrity.
ruush more comples buildmp models that transmit end isolation, and prevention of hypash USl A~ 45. and
- unphfy high ficquency input and generate accurate and GI - 131.
meaninpful floor spectra at hich frequencies. Estimates NUlu!G-1407 12
- 3. The Seismic IPlil!!!
of high frequency amplification in cabinets containing (GliitS) are observed. Also, spatial mteraction evalu-relays will also have to be developed. Itatter than using ation, suc!. as assessine 'he cifects of flooding, as noted in analysis, the foDowmg approach is more suitable:
I! Pit! N1%011, is retained.
1.
Piepare a list of relays that are known to have high.
Rcduccd Scope frequency sensitivity.
Appenda 11 contains guidance.
2 Screen relays that are known to have very high 0.3g Bin-Focus.' and full Scorc llCl_PI's(that is, eliminate them from further con-sideraCon without perfortning specific response cal-The criteria in the (0.3g column of NUltl!G/Cib4334, culations).
Table 5.1, or 1: Plt! NP-6041 Tables 2-3 and 2-4, should be used.
3.
Assume that the remaining r: lap will chatter at the review level carthquake.
0.Sg Bin
'lhe enteria in the 0.3-0.5g column of Null!!G/
4.
Screen the remaining relays by showing either that CIM334, Table 5.1, or i Plt! N1%041, Tables 2-3 and the electrical circuity is insensiuve to high-frequency 2-4, should be used.
(hatter or that they can be recovered f rom changes of state and associated false alarms.
3.2.4.5 Se.ismic input 5.
I inally, replace the remaining relays with relays that nnfucnf Scope are not sensitive to high frequency (an alternative 1or the evaluation, the SSl! ground response spectra and approach is to show that the remaining relays are rugged by conducting tests).
in structure spectra should be used. New in structure response spectra, if descloped, should be mean plus one standard dcsiation to be con.sistent with the conservatism Although stated in the contest of high-frequency pround motion, this approach can be used to aJJtess the relay in the design mput. Any differences between the l'inal chatter issue.
Safety Analysislleport(1 Salt)and new response spectra
~
should be hiphliphied and discussed.
3.2.4.3 Soil Failur es py,yg,Sg,p pgf,Scyp ofg3 Soil failure analyses include an evaluation for instahdity For the evaluation, the NUlt!!G/ClbOO98 raedian rock settlement, and liquefaction, or sod spectruta (depending on prunaty condition et the site) anchored at the assigned review level carthquake Reduccd Scope sho' ld be used.
u No evaluation is necc..sary.
3.2.4.6 thaluation of Outliers-llCI PI' Calculations Focused Scope
.I.'wo approaches. f.rardity analysis (1 A) and conservative l! Plt! NP-h041 contains guidance; a review based on ap-deterministic failute margin (CD1 Mk for computing propriate design and construction records is considered component and plant ilCl.Pl s are acceptable. For the adequate. A detailed analpis. as necessary, will be per-N!!C marrins method, il the licensee imually those the formed if soil failure is deemed significant.
CDFM method to calculate component and plant 1ICl.PF values, it is possible to malse plant 11C1 PF state-Full Sco;ic and 0.Sg ments with and without the inclusion of nonseismic fail-liPlt! NP-6041 contains gui= Janet: it is anticipated that unqand human aedons by developing complete frapili-dedor Ou few components that remam m the plant-les el esisting soil test data will be adequate An evaluation of Hoo an quadons (ophonah plant site conditions using state of the-art approaches should be performed if soil f:ulure is deemed sWniheant.
A d in FPM NP+0% m M k GmW Fw meat ltureednew Spectrum (GlitS) to estimate 3.2.4.4 NCITCning Clittlia (EsV Oi SCf cening Tahics) in i_ pics should tAe into acomnt tk Llot resuhs from ongoing wor k on t he reconcmanon cf Gl'l:S and 1 M 'l Pl.
The screening guidance given in the Genene implemen-tation Procedure for Seisnue Venfication of Nuclear
. "F Power Plant ligwpment (GIP) may be used, provided a resiew is conducted at the appropriate ill la caveats m-Outliet s shoulJ he es.duated hn the pros mons m the GlP cluJed in marpms t eports arc ob>cn ed. anJ lin ntations on d the rbnt n aha m the USl A 40 Pmprom.1 or tlements the use of the peneric equipment runedness spxtrum outsiJe the IN A da wpe ru ustmes and pipmg) the 13 NUltl G-1407
3.1he Seismic Ipl!!!!!
requirements of the plant FSAlt should be used in the l'or IPl!!ill purposes,it is desirable that, to the masimum evaluation. Since the evaluation is done at the design extent possible, the alternative path involve operational level, the outliers should be addressed in accordance with sequences, systems, piping runs, and components differ-10 CFit 50.72(b).
ent from those used in the preferred path. 'the procedure used in the trial application of the !!Pitt methodology
}
M>rund Scope (l! Pill NP-6359) provides an acceptable approach for use
'the scismic capability evaluation engineers /scismic re-in selecting success paths (preferred and alternative).
view team may use judgment to rank the unscreened structures and equipment from the lowest to the highest.
3.2.5.2 Walkdown i
'the licensee should determine the number, scope, and Same as Section 3.2.4.1.
type of IICLPF analyses with the aim of identifying vul-nerabilities and ranking them. Iteed et al. (1990) and 3.2.5.3 Itelay 13aluation Itasin (1990) suggest that 11CLPF capacities should be calculated for the lowest one. third of the ranked compo.
Same as Section 3.2.4.2.
nents: the remaining coinponents should be assigned a conservative llCifF based on the highest calculated 3.2.5.4 Soil railures 1
IIClfFs. The assumed and calculated ilClfFs should be reported.
Same as Section 3.2.4.3.
Full Scope and 0.4 3.2.5.5 Screening Criteria tUse of Seitening Tables) llCiffs for unscreened structures and compoacnts Same as Section 3.2.4.4.
should be calculated as needed to accurately charactcrize plant 1ICLPFs and vulnerabilities and rank them.
3.2.5.6 Seismic input 3.2.4.7 Nonselsmic Failures and lluman Actions These activities should be included; guidance on includ.
3.2.5.7 thaluation of Outtlers-IICLPr Calculations ing nonseismic failures and human actions is provided in Same as Section 3.2.4.6.
NUltlIG/ Cit-4826 (Maine Yankee evaluation) and in two draft reports by Hudnitz (1987 and 1990).
3.2.5.8 Nonseismic railuies and flut.ian Actions 3.2.5 EPitt Seismle Margins Melliodology success paths are chosen based on a screening criterion applied to nonseismic failures and needed human actions.
The guidance provided in !! Plt! NP-6041 is supple-It is important that the failure modes and human actions mentcJ by that in the following sections to (1) n.flect the are c!carly identified and have low enough probabilities to partitioning of the 0.3g screening criteria 2nto the not affect the seisraic margins evaluation. The screening reduced-scope, 0.3g focused scope and 0.3g full scope criteria used ir. the Maine Yankee margin evaluation bins identified in Table 3.1, and (2) identity enhance.
(NUltliG/ Cit-4826) addressing both single train t ad ment'so that the method can be used for the implemen-multi train systems is an acceptable approach.1he redun-tation of the IPlilili.
dancies along a given success path should be specifically analyred and documented when they exist. (In a comple-3.2.5.1 Selection of Alternathe Success Paths mentary sense, where a single componant is truly "alone" in performing a vital function along a success path, this The !! Pill SMM as currently constituted calls for evalu-should be highlighted too).This information will serve to ation of a preferred path and an alternative path. The innate the extent to which a single faihate would or NltC panel that reviewed the I!Piti methodology recom-would not invalidate the plant's ability to respond saf ely
- mended:
to a given earthquake level.
... that a reasonaoly cornplete set of potential success 3.2.6 Containment Performance paths be set down imtially, rather than a very small num-ber, since limiting the n umber of success paths too quickly
'the primary purpose of the evaluation for a seismic cvent can rr..cnt the identification of some plant-level is to identify ' vulnerabilities that involve early failure of IICE 'F insights, and can mask plant differences regard-containment functions. These include containment ing defense-in. depth.'lhe Panel believes that preliminary integrity, containment isalation, prevention of bypass analysis to narrow the number of paths to the required functions. and some specific systems depending on a con-two or three should begin with the fuller set.and it recom-tainment design (e.g., ignitern, suppression pools, ice bas, mends that this narrowing be documented in detail."
kets). The analyses performed for internal events IPli N UT 11G-1-107 14
- ~
J
- 3. 'the Seismic IPl!!ill should be used to determine the scope of systerns for the Valves mvolved in the containtnent isolation system are examination, expected to be seismically rugged (NUltl:G/ Cit-4734). A walkdown to ensure that they are similar to test data and have known high capacities and that there are no spatial liach licensee should develop a plan to address contain-interaction issues wdl suf fice. Scistnic failures of actua.
ment performance during a seismic event consistent with tion and contr ol systems ar e more likely to cause isolation the above defmed purpose. Additional guidance (no re-nstem failures and should be included in the examin-quirements implied) on estending margin type ap-ation. I or valves relying on a backup air system, the air proaches to obtain containtnent insights is contained in system should also be included in the seismic examin-lludnitt 1991a and 1991b, and Reed, et al.,1990. Some a' tion.
general guidance is provided here based on past pita experience and mme rencric capacity estimates of typical Components of the containment heat removal / pressure components involved in containment systems. I rom a suppression functional system that are not included else-Survey of past PRAs (Amico,1989),it appears that high-where and are not known to have high capacities should consequence sequences involve pross structural failure of be exarnined. An example of such a component might be a the containment itself or failure of major equipment or fan cooler unit supported on isolator Shims. 'lhe walk-structures within the containment at veiy high accelera-down 5.hould include examination of such components tions (llCl.PF values greater than 0.$g) and isolation and their anchorages. Similarly, support systerns and failut e due to seismically induced relay chatter.
other systern interaction effects (e.g., relay chatter) should be examined as applicable.
Generally. containment penetrations are scismically tug-l'or Mark I and ice condenser containments, it is not ged; a rigorous Iragility analysis is needed only at review feasible to screen out components (e.g., torus, ice basket levels greater than OJg. but a walkdown to evaluate for support) or a genetic capacity basis. 'the potential for unusual conditions (e.g., spatial interauions, unique accident sequences initiated by a containment functional penetration conhgurations)is rewmmended. An evalu-failure should be examined.
ation of the backup air systern of the equipment hatch and 3.3 O )tioital Metliodologies personnel hxk that employ inflatable seals should be I
perforined at all review levels. Also, some penetrations need cooling, and the possibihty and consequences of a A licensee may request the staff to review any other cooling loss caused by an carthquake should be con-tysternatic examination method to determine its accept.
sidered.
ability for 11' 11!11 pur poses.
l r
15 NUlti!G-1407 i
4 ACCEPTAlli.E METilOI)OI.OGY FOlt l'EltFoltMING Tile INTEltNAI, Filt19 IPEEE 17or purposes of an IPl!EE, a Eevel 1 probabilistic risk spreadmp potential, or other idio 9ncrasics are consid-assessment (PI(A) is consiJered acceptable to identify cred. Also, the data base on hres in various areas should potential internal fire vulnen.bilities at nuclear power be coupled with location-specific information obtained plants. Some fire issues identi ~ied in the l' ire 1(isk Scop-from the plant walkdow n and other experience to account ing Study, (1) scisrnic/ fire init ractions, (2) cfIcets of fire for uncertainties.
suppressanu, on saf ety equipment, and (3) control sy stem interactions, should be addressed in the IPElill 'the 4.1.3 Analyre for the 1)isabling of Critietil walkdown procedures of the IPlilili should address the Safety I? unctions above issues and should be f pecihcally tadored to assess the pote.itial vulnerabilities related to these issuet The Determine the likelihood of equipment being disabkd by licensee should use a plantopecific data base on fire bri-a bre. The areas to be addiessed include:
pade training in the IPE!!U to assess the cifectiveness of manual fire fighting to determine the response time for 1,
I ire growth and spread, includmg the treatment of the manual bre fighters. '1he licensee should also show hot gases and smoke.
the cifectiseness of fire turriers in the IPliEE. 'the cut-rent fire Pl(A method has its limitations (NUlt!!G/ Cit-2.
De tecuon/ suppression cf fet tiveness and rehabihty.
5088,19h9) and signifkar t " engineering judgment" must 1
Component fragdity to fire and combustion prod-be brought to bear once the 11( A has been accomplished to allow for sensible aprheation of the results. 'the staf f beheves that the type of " engineering judgment" needed 4.
Probability estunates (disuibutiona for fault tree to mterpret the results ela Pl( A is fully within the compe-quantifical an.
tence of most fire-saf ety experts, including experts within the regulatory staff. Ennher, despite current hmitations 4.1.4 Identify I' ire induced initiating Esents/
in tlye methodology, a hre "vulnerabihty search" in the Systems Anidysis spirit of the Severe Accident Policy Statement and the IPli exercise is feasdie, and such a vulnerability scarch Perform the analpis to determine the frequency of fire need not wait for the completion of further methodology initiated accident sequences leadmp to core damage, development.1:irall/, in meeting the objectives of the 1PEEli, it is desirab'c to focus on relative insights rather 4.1.5. l'erform Contailuneni Analysis than on absolute core damage frequency.
Perwrm containment analpis if containment Iadure 4.1 New Fire Pila Asialysis modes ddier sigmficanity from those found in the IPE internal esents evaluation.
'!here me several different approaches for the analysis of fires (NUlti!C/ Cit-2300,1983, N Ul(l!G/ Cit-2S 15, Perform in a fashion similar to an internal-initiator Pl( A 1935, NUl(EG/ Cit-4S40,1990, and NUl(EG/ Cit-$259, 1990). Ahhough mt all fire Pl( As delineate their anal) sis 4.2 Use of ail Existiiig Fire Pita steps in exactly the same way, the followmg steps, m one fonn or another, snould be part of any analysis.
lhe use of an existing fire Pl( A for the internal fires 4.1.1 Identify Critical Areas of Vulnerability IPEl.li is acceptable prouded the PI( A reflects the cut-rent as bat and as operated status of the plant and the The criterion is whether a fire could compromise impor-licensee addresses the deficienacs of past P14 As that are tant safety ec,uipment. limohasis shoulJ be placed on iJentified in the 1 ire 1(isk Scoping Study (NUltliG/
areas where r mitiple equipment could be compromised, Cit-SOSS). Deficiencies may include the use of low condi-in partic:
'r, several trains of redundant equipment to tional failure probabilities for dampers and penetrations, perlorm e same safety function. Attention should be no consideration of damage from the ust of fire suppres-given to the potential for crossvone spread of fire and the sants, inappropriate estimates of the effectivenc>s of likelihood that transient fuels might supplement fuels manual fire fighting, and no consideration of seismic / fire already present in a zone.
mteractions, i
4.1.2 Cttleulate the l?requency of 17 ire 4.3 Optional Methodologies I
initiation in Each Area l
A licensee may request the staf f to reuew any other
'lhis caltalation is sensitive to location withm a larger systematic examination method to deter mine its accept-area, pacticularly if fuel loadmg conditions crossvone abihty for ll'lthU purposes.
N Ult t!G-1407 16
5 ACCEL'TAllLE METilODOLOGY FOlt l'EltFOllMING TilL lilGli WINDS, FLOODS, AND TitANSI'OltTATION AND NEAltllY FACILITY ACCIDENT IPEEE 1:or the purloses of an IPimli, the staff recommends a 5.2.1 Iteview I'Itmt Specilie llazant Data and progressive screening approach to idt ntify potential vul-1.leensing Iluses nerabilities at nuclear power plants due to high winds, floods, and transportation and nearby facility accident.
All licensees should review the information on plant de-
'lhe owners of Trojan and Washington Nuclear Plant 2, sign ha/ard and the licensing bases,includmg the resolu-who are requested to evaluate the effects of volcanic tion of each event.
vtivities in asses *ing severe accident vulnerabilities, 5.2.2 Identif Signliicant Clumges Since 01, mld determine if the recommended screening ap-3 pioath is applicable to their unique situation.
IWmuce All licensees should review the site for any sigmficant canm Una ik operaung igense was issued with re-5.1 liitiocitictioit socet to (l) mihtary and mdustu:d faedities within 5 miles of the site,(2) onsite storage or other activities mvolving hazardous materiah, (3) transportation, or (4) develop-it is assumed that the IPl! for internal esents will be in ments that could affect the onginal design conditions.
progress or completed when the portion of the Ip!II:
pertaining to high vinds, floods and transportation and 5.2.3 lietermine if the I'lant/l'acilities 1)esign neatby facihty accident is being performed. Some exter-nal events wdl be addressed in the internal events IPli Meets 1975 Sill' Criteria nnalyses (e.g., the prunary effect of lightning is loss of All licensees should compare the information obtained oilsite power, which is included in the internal events from the review discussed in Sections 5.21 and 5.22 for analyses); other external events will have been screened conformanec to 1975 SitP etiteiia and per fonn a confirm-from further eonsideration by the staff. l'or those external atorv walkdown of the plant. The walldown would con-events not in eithet of these categories, further considera.
centrate on outdoor facdities that could be affected bj non using the progtes.sise screenmg approath shown in high winds, onsite storare of ha/ardous matetials, and i igure 5.1 is n commended.
of fsite developments. If the companson indicates that the plant conforms to the 1975 Sl(P criteria and the walk-dow n t eveals no potential vulnerabihties not included in b.3 Aritil,t.3 ictil I,roceittiiT the ooginal desicn basis analysis, it is judged that the contribution from that hazard to core damage frequency is less than 10 0 per year and the IPl lili screening crite-
.the steps shog n m ligure 51 represent a series of analy-rion is met.
ses m mereasmg level ol detail, cfloit, and resolution.
Ilowever, the ik emee may (hoose to bypass one or more Otheru ne or i, a licensee knows that the 1975 SitP clite of the opthmal steps as long as the 1975 Standard lles k w ria will not be met, it shon'd take one or mote of the ilan (Sl(I) (NUl(th75/Os7) enteria are inet or the optional steps given in Secuons 5.2.4, '.2.5, and 52.6 to potential vulnerabilities are either identified or demon-further evaluate the situation.
stra;ed to be insigmheant.
5.2.4 I)elermine if the llazard l'tequency is Acceptably 1.ow (Optional Step)
In general, the containment structute, equipment hatch, peisonnel air lock, and other penetrations alc desy'ned if the origmal design basis does not meet cunent regula-and constructed to have high capacities in resnting the tory icymrements the licensee may thoose to demon-effects of high wmds,11ood% and overpressut e mduced by strate that the miginal design bavs 6 sufhciently low-tramportation or nearby facihty accidents. Therefore, no that is less than 103 per year, and the conditional core adJihonal containmero perfor mance assessment (heyond danupe fiequency it judged to be less than 10A that thscussed for the seismie poiton, of the IPElf in Sections 3.1.1.5,3.1.2, and 32 A)is needed unless a heen-If the ourmal deurn heis ha/ aid combined with the see prethets or idennhes plant umque accident sequences conditional core damare htqueng is not sulhe.ently low (hlferent hom those deletmmed by the inteinal esents o c., less than the scieening enterion of 103 per year),
i li'll adJitional analysn nuy he needed.
17 NUlt! 41407
i
- 5. 'the Accident IPl?IIIIs I
(1) Review Plant Specific Hazard Data and Licensing Bases (FSAR)
V (2) Identify Significant Changes, if any, since OL lssuance Y
NO (3) Does Plant / Facilities Design Meet 1975 SRP Criteria?
YES s
m (Quick Screening & Walkdown)
OR > (4) is the Hazard Frequency Acceptably Low?
YES Nof OR m (5) Bounding Analysis
_YES m
(Response / Consequence)
NOf OR > (6) PRA (7) Documentation m
(Incl. Identified Reportable items and Proposed improvements '
Figure 5.1 Itceommended IPIII!!! Approach for Winds,17kxxis, and Others 5.2.5 Perform a Bounding Analysis lowing key elements: hazard analysis, fragility evaluation, (Optional Slep) plant systems and accident analysis (event / fault trees),
and radioactive material release analysis 'lhe detailed This analysis is intended to provide a conservative calcula.
proecdure is described in NUR!!G/ Cit-2300, NUltlIG/
tion showing that either the hazard would not result in Cit-2815, anr1 NUltl!O/ Cit-5259. If the core damage core damage or the core damage frequency is below the f requency is less than 10.e perycar, the event need not be reporting eriterion. The level of detail is t hat level needed considered further. The level of detail is that level needed to demonstrate the point; judgment is needed for deter.
to conclude that the core damage frequency is low or to mining the proper level of detail and needed effort.
find vulnerabilities.
5.2.6 - Perform a Probabilistic Risk 5.3 Optional Methodologies-Assessment (Optional Step)
A licensec may request the staff to review any other systematie examination method to determine its accept-A probabilistic risk assessment (Pita) consists of the fol-ability for IPlilill purposes.
NUltl1G-1407 18
i 6 COORDINATION WITil ONGOING PitOGRAMS 6.1 Illtroducil011 staff identified alternative approaches to certain de-sign procedur es and mothfications to the NI(C crite-If unnecessary duplication of effort is to be avoided, coor-ria in the Standard Iteview Plan to reffect the cur-dination with ongoing programs is necessary. 'the first rent state of the art and industry practice. 'lhe level of coordination consistsof the three major elements (oncern for the scistnic capacity of safety related related to the implementation of the Severe Accident above. ground tanks (at the SSl!)is included in USI Policy, that is, coordination of the IPl!!!!! with the inter.
A-46.
nal events IPl! and accident management. 'ihe second level of coordination consists of the threc tnajor elements 3.
USl A-45, " Shutdown Decay lleat itemoval of the IPlitill, that is, seismic events, internal fires, and llequirements," has the objective of determining high winds, floods, and others. 'the third level of coordi-whether the decay heat removal function at operat.
nation consists of each major element of the IPl!!!!!, for mg plants isadequate nnd if cost beneficialimprove-example, seismic events, and the ongoing prograrns re-mentcobidbeidentified.USI A-45wassubsumedin lated to that element, the IPl! (Gl. 88-20); therefore, the external event aspects including the seismic adequacy of the decay 6.2 I)CSCriptio!) of' Oligollig l'rogrtistis heat removal system should be included in the IPlil!!!.
6.2.1 II'E l'rogram itelated to Internal 4.
USI A-46, " verification of Seismic Adequacy of Events liquipment m, Operating Plants" has developed an in Generic Letter 88-20, the NitC requested that the alternative method and acceptance criteria (to cur-licensee of each plant to perform a systematic exam.
tent licensing requirements) to venfy the seismic ination to identify any plant-specific vulnerabilities to aglequacy of eqtupment in some operating plants severe accidents and to report the results to the staff.'the with construction permit applications docketed Sc-process was defined as an individual plant examination fore about 1972. All these plants will be reviewed to (IPli). I.icensees were requested to proceed with the ex.
the existing safe shutdown earthquake (SSli). 'the aminations for internally initiated events only (including scope of USl A-46 has been expanded to cover the internal flooding). Lamination of externally initiated seismic spatial system interaction of USl A-17 and events would proceed separately and on a later schedule, the concern of USI A-40 for the scismic capability of flowever, while performing the IPli for internally initi.
Ir.rge safety related above ground tanks, ated events, licensees were advised t, document and re-tain plant-specific data relevant to extenal events so that 5.
GI-131, " Potential Seismic Interaction involving they can be readily retrieveu in a convoie t form when the Movable In-Core Plux Mapping System Used in needed for later external event analyses Westinghouse Plants." was identified because por-tions of the in core flux mapping system that have 6,2,2 l'rogrants Related to Extors.al Events not been seismically analyzed are h>cated directly above the seal table. l'ailure of this equipment dur-6.2.2.1 Scistnic Programs ing a seismic event could cause multiple failures at the seal table and could produce an equivalent The followmg is a brief description of the programs re-small break 1 OCA.
lated to seismic events:
6.
'ihe "liastern U.S. Seisn icity issue" (formerly the 1.
US! A-17. "Systern Interactions in Nuclear Power Charleston lianhquake issue)came about as a re-Plants" addresses NitC's concerns regaiding the sult of a U.S. Geological Survey letter in 1982 that interaction of various systems with regard to pointed out the possibihty that large, damaging whether actions or consequences could adversely earthquakes have some likelihood of occurring at affect the redundancy and independence of safety kications that had not been considered in licensing spterns. 'the evaluation of system interactions re-decisions. 'lhe staff.'nitiated the Seismic Ilazard lated to internal events ano internal floods is in-Characterization Project at LLNI, which provided cluded in the IPl! (GL 88-20). 'lhe evaluation of protubilisti, cismic hazard estimates for all nuclear spatial system interaction under seismic conditions power plant sites cast of the llocky Mountains. A (the SSli)is included in USI A-46.
similar project was carried out by IIPitt for the elec-tric utilityindustiy.'Ihe staff's purpose in evaluating 2.
USI A-40, "Scismic Design Criteria," investigates the probabilistic studies has been to identify plants selected areas of the seistnie design process. 'ihe in the Central and liastern United States w here past l
l 19 NUltliG-1407
- 6. Coordination with Ongoing Programs licensing decisions may have resulted in their beinE (i.3 A})l)roaCll on C00rdilmilon With outliers with respect to seismic hatard, that is, the likelihood cf exceeding their design bases. As a re-Onpig I's n ms sult of the probabilistic analyses performed, eight if duplication of effort by the staff and licensees is to be plants at five Eastern U.S. sites were identified as avoided,it is isnportant that the above onge;ng programs being outliersflhe IPElill will provide a resolution be coordinated, for the outlier plants with no need for additional analyses or documentation from the beensees.
6.3.1 Coordination Mhh Internal Events 6.2.2.2 Internal rires Program $
l'rogram (II'E)
'the following is a brief description of programs related to internal fires:
,lhe coordination between the internal events IPl! and he IPlillE can be categorized into three phases:
1.
NUREO/CR-5088, " Fire Risk Scoping Study,"
preanalyses planning, plant modifications, and accident identifies some fire issues that had not previously rnanagement, been addressed in the fire PRAs: fire growth code, seismic / fire interaction, fire barrier effectiveness, 63.L1 Picanalyses Planning manual fire fighting effectiveness cifects of fire sup.
Considerations include (1) definition of elements and pressants on safety equipmtat. and control rjstem their boundaries, (2) walkdown procedures and spatial interactions. A plant. specific analpis (meluding a interactions, and (3) composition of the peer reviety specifically tailored walkdown) should be performed group. It is likely that the IPE will precede the IPl!EE.
to assess the actual risk impact of these issues at a Careful planning, taking into account the above consid-plant.
erations, will avoid a duplication of effort by the licensee.
2.
GI-57,* Effects of I ire Protection System Actuation 6.3.1.2 Plant Modifications on Safety llelated Equipment," assesses the impact of inadvenent actuation of fire protection systems Since the IPli and the IPREE are likely to be performed on safety systems ~1his is one of the issues identified separately,it isimperative to examine the irnpact of modi-in the Fire Risk Scoping Study. *lhe industry, fications identified during the 1PE on external events and through l!PRI, has a program collecting data on the vice versa.'the staff examined several PR As that included effects of suppress:mts on the safety equipment.
both internal events and external events (llohn,1989) to identify possible interactions. liighlights of the major 3.
USI A-45, " Shutdown Decay licat Removal Re-findings m the seismic area (which to some extent are quirements," was initiated to deterrnine if the decay applicable to fire), are the following:
heat removal function at operating plants is ade-1.
In general, the modifications proposed as a result of quate and if cost beneficial improvernent could be identified.'lhe US! A-45 was subsumed in the IPl; the mternal events analysis would not adversely (GL 88-20): therefore, the adequacy of the decay affect the seismic or fire risk, provided the modifica-heat removal system under internal fires should be tions do not become wak links.
adJressed in the fire IPEl!!!.
2.
In general, the modifications made could potentially con ute to an inmase in M at W pant in k 6.2.2.3 External Flooding Program following ways..
GI 103, " Design for Probable Maximum Precipitation (PMP)," is a related external flooding issue 'the staff a.
Many of the modifications proposed may in-provided the resolution of this issue to all licensees in volve adding valves or suction lines to existing Generic Letter 89-22, dated Oct. 19,1989. Specifically, systems. 'lhus, the possibility' of violating the the NRC requested that future plants be designed against pressure boundary and creating a potential di-a new PMP criterion. For existing plants, the NRC recom.
version exists if the modification were to lail mended that licensees review the information contained during an carthquake. Also, modifications may in Generie Letter 89-22 and determine if they need to involve routing different trains of electrical take additional action. For the IPEEE, the severe acci-power or power from adjacent unitsJihe posd-dent risk frorn PMP should be assessed. The licensees bility exists that the circuitry could be designed should assess the effects of applying this new PMP crite-m such a way that failure of non safety-related rion to their plants in terms of onsite flooding and roof electrical components could act ually defeat the l
ponding to determine whether that would lead to severe circuitry that was desired to provide redun-accidents, dancy, and NUREG-1407 20 m.
--m w
e-mm.---w-w.u,av---mm's e
er e vmw w-w+++-
-+rm-
--a-m
- 6. Coordmatmn with Ongoing Programs b.
The possibihty exists that inadequate anchor-tilied during the scismic walldown to grasp the issues age could defeat the planned redundancy dur.
from the seismic-capacity point of view.
ing a seismic event.
6.3.3 Coordination With Seismic l'rograms 3.
'the potential adverse effects of the modifications include.
A numbcM progwnuelaW to sdsmhntuequiring licensee action have been identified. hiany of these pro-grams have arisen as a result of the changing perception a.
Poor accessibility for maintenance, of hazards and revisions in the design and qualification criteria 'ihere are two categories of seismie programs as b.
Sttffening of systems leading to higher stress they relate to the scismic IPlilil!. 'lhe first category in.
due to thermal gcles durmg normal plant op-volves programs, e.g., USI A-45,"Shutdow n Decay lleat cration.
Itemoval l<equirements",01-131. " Potential Seismic In-teraction involving the hiovable in. Cure Flus hiapping The cited study (llohn,1989) provides specific examples Systern Used in Westinghouse Plants," and "the !! astern of modificatmns and their effects on other initiating U.S. Scismicity issue," that have been subsumed into the es ents.
II'l!/IPlilil!. USI A-15 and GI-131 should be specifically addressed as part of the seismic IPlilili,'lhe liastern U.S.
6.3.1.3 Accident hianagement Seismicity issue requires no additionallicensee actions or e second category involves programs (e.g.,
relming. y' VeGfication of Scismic Adequacy of Equip-Guidance on the integration of fmdings from the IPhlili W A-46, and accident management is being developed (Sl:CY-89-308, Oct.1989).
rnent in Operating I'lants,,,) that can be coordinated with the seismic IPlilili. 'lhe coordination of these programs with the scismic IPlilil! ir most beneficial in reducing the 6.3.2 Coordlitation Among thternal Even!S resources expended by the licensee and the staf f.
I'rograms 6.3.3.1 USI A-45 and GI-131
'the issue of integration betw een external events primar-
.lhe nddology used in the seismic IPl!!!!! can also be ily mvolves mteractions between scismic events and fues and scismic events and floods. S,etsmically mduced hres us d to address USI A-45 and GI-131. *lhe systems and components for addressing USl A-45 will have been de-and fhiods are to be addressed as part of the IPliiili.fthe termined by the internal events IPli, and the purpose of effects of seisnucally m, duced fires and the impact of mad.
vertent actuation of fire protection systems on safety sys' the seismic IPlil!!! is to identify any significant and tems should be addressed. ihe effects of seisnucally in-an que scismic vulnerabilities in the decay heat removal function. In addition, the scismic lPl!!!!! will evaluate the duced external flooding and internal flooding on plant potential seismic interaction of the movable in-core flux safety should be included.,lhe scope of the evaluation of scismically induced fhxds,in addition to that of the ester-mapping system used in Westinghouse plants.
nal sources of water (e.g., tanks, upstream dams), should Capacities of decay heat removal components can be include the evaluation of some internal fheding consis-established using either the fragility analysis (FA) or tent with the discussion in Appendix 1 of !! Pit! NP-6041-conservative deterministic failure margin (Cl)l'ht)
The coordination between the seismic and the fire or approaches depending upon the methodology chosen to flood analysts should be based on the following:
implement the seismic It'lilili. 'lhus resolution of these issues can be casily accomplished during the scismic 1.
'lhe seismic analysts should generally search for and IPl!!!!! evaluation, identify the initiating events (certain specific seis.
mically initiated failures of equipment or structures)
'lhe potentialinteraction between the seal table and non-that can cause fires or floods and Category I seismic systems associated with the movable imcore flux mapping system can be identified during the 2.
the seismic and fire or flood analysts should also seismic walkdown of the IPliEI! If needed, the compo-discuss other concurrent seismically induced Iml-nent capacities or consequences of component failure can urcs or possibic effects on human actions and then, be cuduated using the same procedures that are used in proceed to complete the rest of the IPEl I! analysis.
the seismic IPPl!!i.
The coordination should include a meeting, prior to seis.
633.2 The l' aster n U.S. Seismicity issue (The mic walkdown, in w hich the fire and flooJ analysts discuss ON icd"" Ed' *l"dk' M"*I the key issues, how the analysis will be done, and what to As a result of work car ried out to resolve the Eastern U.S.
look for.'the hre or flooJ analpt may need to pm ticipate Seismicity luue(Charleston Earthquake loue), pr obabil-in parts of the sciunic walkdown or resivt the areas iden-otic sennuc hatard estimates esist for all nuclear powei 21 NURI G-1407
- 6. Coordmation with Onromp Programs plants cast of the 1(ocky Mountainsflhese should be used mie margms methoJology is utilized to implement the diteetly by any hecnsee,n that region opting to satisfy the seismic IPl!!!!! (since it is quite similar to the USl A-46 nistnic IPlil:1: with a sco,mic Pit AJihe hazard estimates evaluation), it would need additional equipment to be also played a Ley role in determining the review level resiewed over thal tequired for implementing USI A-46.
carthquake used in the seismic margins methodology op-tion. The 111 I li will provide a resolution for this tssue lhird, the levels of resiew and walkdown are different.
without tequiring adJitional arudyses or doeurnentation lhe Seismic Quahfication Utility Group (SQUG) and from licensec*
EPiti have developed a detaded Generic implementa-tion Procedure (GIP) for the USl A-46 review and walk-63.33 USI A-46 down that was reviewed by the NI(C staff, and a Safety 1". valuation lleport (SI:lt) was issued.The GIP should be implernentation of the USI A-46 progtam involves plants followed in performing the USI A-46 review and walk-with construction permit appheations docketed befere down. 'the puidelines associated with the seismic Pil A or about 1072.The USl A-46 plants thus form a subset of all seismic margins methoJology are not as specific as those the nuclear power plants in the U.S. that are requested 10 m the GIP. To illustrate this point, in the walkdown re-perform the seisnue 11'111!.
view of cyansion anchor bolts, GlP calls for the use of a wrench test for the bolt tightness check, whereas the
'the tnost etheient way to address the oty:omg seismie margins walkdown ensures only that the anchor bolts are proprams for USI A-46 plants is to conduct the A-46 adequate to hold down the equipment as designed with no tevmw and walkdown to gather telesant information for specific testing r equirements to confirm anchor capacity.
the seismic IPlilili. In oider to facditate this appio.nh,
'the completion of the scismie lPlil!!! does not automati-the aettvitiesof LISI A-46 and the seismic IPI!!!!! need to cally mean that the 1151 A-46 review is sctisfactority com-be coordmated, and the plant walkdown needs to be w cll pleted.
planned Severalinherent dif ferences between the A-46 piogram and the seismic IPlil:1!!hould be noted at the There rnay be overlaps or ddferences in the equipment outset befoie attemptmg to coordmate t he tw o piogr ams' scope for USI A-46 and the seisnoe ll'lilill i or equip-ment that is withm the scope of USI A-16 or the seismic l'iist, the objectnes are quite difletent, The USl A-4b iPlil!!! only, it is clear that either (ilP or IPlilili guide-program has beensmg implications on plant operation, hnes, respectively, should apply. l'or the overlapping this prortam will assess and ensure the seismic rupred.
equipment, the cificient approach is to use the GIF for ness of safety-iclated equipment m a plant to withstand both walkdowns; however, the IPlilill should use the the SSl! The scismic 11'111!, on the other hand, gener-review les el carthquake. Caveats and mteraction (such as ally tries to identdy plant vulnerabihties when subjected noodmp) prousions of l'14(I NP4011 should be ob-to carthquake lesels higher than the SSl! design basis.
served.
Second, the scope of the revit as are different. USI A-40 in summary,it is recommended that beensees wordmate is wneetned with only one success path (with some re-the information collection for the USl A-46 and seismic quirement on equipment redundancy) of equipment IPlil'li resiew and walkdown in erder to minimi/c or needed to brmy the plant to safe shutdown in the es ent of avoid duphtation of elfort by the licensees ar.d staff. Care an carthquake and maintain it there for at least 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />.
shoulJ be esenised in the coordmation to ensure that the The scenano conuders an catthquake of the SSl!lesci objecoves of both piograms are fulfined. Coordmation of with a pow:ble low of of fsite power because of this carth-the two programs has been shown to be feauble in the trial evaluanon of he Ilatch plant usmp the i Pitt seismic quake. The ptobaNhties of a seismically mduced 1 OC A t
omall or latpe) and a hirh enetpy hoe break (lli 1.10 raatpins methodolopy.
occut t mg are juJped to be low enough that their consiJ-eration at this caithquake lesel is not watianted. Pipmp, tubinp, and situetutes wdl be esanuned durmg a walk-6.3.4 ('oanlillatioll Willi Ollier 155110N down onl3 if they base the potential to cause seismic interaeuon unh the equipmt nt resiewed and caue t'am-In addition to the specif e USls and Gls discussed above, age to tho equipment.The icuew of abusetrounJ tanks if, during its IPI:lil' a licensee (1) dneosers a notable las part of USl A 40)is an ek ephon.The se.smie llTil sulneraNiny that is topicahy awociated with any other o concemed wah the s ulneraNhues of the whole p ant.
t 'S1 ot GI and piopmes measures to dapose of the spe-not just the egmpment. Also, evaluauons are generaH) uhc safety i 'ue or p) contludes that no vulnerabdity maJe at les els ahos e the dcQn h.on. At ihn les el of esnts at its p! ant that is topeally.moe tied with any t !Sl catthquake, sennucally inducea ! OCAs at e consil teJ, or r if, the tah a M wnudet the USl or GI resobed for a and mit7itmp systenn anJ egmpmt,t to aJshew thn Hant upon resa anJ Au;wnee of the resuhs of the initiator are ses iew ed. t ha c!o:e. esen d the 1 Pl(I sen-IPl 1 l' lhe f,G uing should be dm u sed:
N Uld - G-1407 22 i
- 6. Coordination with Ongoing l'rograms
'the ability el the methodology to identify vulner-ance, including sources of uncertainty when Pit A is a.
abilities associated with the USI or GI being ad-used.
- dressed, b.
'lhe contribution of each USl or G1 to core damage c.
'Ihc technical basis for resolving the issue.
frequency or unusually poor containment perforrn-23 NUllEG-1407
7 PEER REYlEW L
In Supplement 4 to Generic 1.ctter 88-20, the staff re-and peer review, taken as a group, provide lvth a cadre of quests that cach licensee conduct a peer review by indi-utility personnel to facilitate the contmued use of the viduals wno are not associated with the initial evaluation results and the expertise in the methods to ensure that the to ensure the accuracy of the documentation and to vali-techniques have been correctly applied.The staff expects date both the IPEEli process and its results. The submit.
all utilities to have in-house personnel w ho have the most tal should contain, as a minimum, a description of the expert knowledge of their plants, system configurations, review performed, the results of the review tearl's evalu-and oper.iting practices and procedures.
ation, and a list of the review team members-s The staff recommends that the peer review team have j
che purpose of the ccer a..cw is twofold.ne first pur-c mbined experience in the areas of systems engineering mse is to provide quality control and quality assurance to and spenfic external events. : or exan.ple, the seismic the IPlil!E process. Intiependence of % review team is peer redew team shoumammbmed experience in the desirable bec.,ase it reflecu a quality control and quality reas of syterm engineering, seismic capacity engineering, a
mc As or seismic margins methodhlogies, assurance ittitude. This does not in'1 ply that toe staff is seeking a review or document control similar to that specified in Appe ndix il to 10 CFil Paa 50.The staff oes 7.l b,0lSmle Reimed ins. iits ig seek to ensure 1 hat the IPliEli process produces reliable, firee trial plant appheations o! the seismic margins
. actual information. If it is necessc ry to use a reviewe[
methodologies have demonstrated that consiJerable who is not totally removed from the plant-specific IPElil:
process, the licensee should oc confident that the re-judgment is involved n applying these methoJs, and that viewer can be objective and capable of providmg a critical peer review groups have considerably aided thisjudgme nt review. I he "in-house team can be su/pkmemed t y om and have also enhanced the overali credibility ot the stuJ-b @vi W90b side consultante as determined appropriate by the been-t see.
A meeting of the peer review team with the scismic re-view team prior to the plart walkdown will provide im The second pur =
' ne review rdates to the impor-sights into the appropriateness of the proposed plant tance of havinr. tN sonnel cognizant of the IPEEl!.
walkdowns. Also, the peer review team's endorsement of The maximurn ' x. a the utihty vauld occur it the the final results will add credibility to the 1ICl.PF cote.hmation of ans i..volved ia the origmal ana;ysis estimates.
K K
w
[
r NUliliG -! 10 /
24 r
l l
8 DOCUMENTATION AND REPOltT1NG
'lhe lPEEE should be documeated in a traceable manner 8,1 Inf0rination Submitted to the NltC to g 9 vide the basis for the fir, dings.'lhis can be dealt with mo. uficiently by a twodier approach. 'lle first tier A detailed list ofinformation to be submitted to the NitC consists of the results of the exhmination, which will be is provided in Appendix C.
reported to the NilC for review,'lhe second tier is tbc documentation of the examination itself, which should be 8,2 1nf0rniation Retailled for Audit retained by the licensee for the duration of the license
(
unless superseded, lle'ained documentation should include applicable uvent trees and fault trees, current,crsions of the t.yetem note-books (if applicable), walkdown reports, and the results of the examination, in general. n!; documents essential for a
'lhe information submitted to the NitC should be organ-practitioner in the fic!d to understand what was done in ucd and presented in accerdance 1.ith Appendix C. '1he the IPl!Illi sho ;;o be retained, in addition, the manner in submittal may enable many issues to be dealt with in the which tb; validity of these documents has been ensured IPEEE review. Pertinent issues ate discussed in Section 6.
should be documented. If credit is allowed in the IPEEE For some issues, for exan., le, USI A-46, a detailed docu-for r iy actions taken by the operators, the licensee should mentation requirement exists, and it should be follond have established plant procedures to be used by the plant in the broad framework of IPf!EE submittalc 5pecific staff responsible for managing a severe accident should information relevant to particular issues ;.g., USis and one occur. Procedures should provide assurance that the Gis, should he identified.
operators can and will take the proper action.
a C
25 NUlti!G-1407
n
,e 1
9 REFERENCES Amico, P.," Containment Considerations for the Use of
-, EPRI NP-6359," Seismic Margin Assessment of the the Seismic Margins Methods for Risk Screening " I etter Catawba Nuclear Station," Vols. I and 2, April 1989 Report, Science Applica6ionsInternationalCorporation, June 30,1989.
-, liPRI NP-6395-D, "Probabilistic Seismic Hazard Evaluation at Nuclear Plant Sites in the Central and East.
Iledjord, ii.'(NRC), memorandum to 1. Shao (NhC),
ern United States: Resolution of tne Charlo.Jon issue "
Subject:
lixernal livents Stecting Group, December 21, April 1989.
1987.
Moore, D., et al., "Results of the Seismic Margin Assess-
--, memorandum to 1 Shao (NRC),
Subject:
lixternal ment of f latch Nuclear Power Plant," Proceedings of Third Events Stecting Group, May 31,1988.
Symposium on Current issues Related to Nuclear Power Plent Structures, Equipment and Piping, Orlando, Florida, Hohn, M.," Status Report on Issues Related to Internal /
December 1990.
External Event Interaction and Decay lleat Reraoval Requirements for IPEs " Draft, Sandia Nationallabora.
Orvis, D., et al,"Scismic Margin Review of Plant 1:atch tory, May 1989.
I L* 1: System Analysis," LLNL Report No. UI Rie Ch-104834, August 1990.
Hudnitz, R., et al., " Extending a llCLPF-Hased Seismic Margin Review to Analyze the Potential for large Radim Rasm., W, (NUM ARC), letter to W. Minners (NRC),
logical Releases and the Ir.;portance of Iluman Factors
Subject:
1 mal Industry Lomments on Draft Generic and Non-Scismic Failures " Draft, Future Resources As.
Letter 88-20, hupplement 4, "Individt.al Plant Exami-sociates, Inc., March 1987.
nation f Ihternal Events (IPEEE) for Severe Accident Vulnerabilhies," and Draft NUREG-.1407, " Procedural Hudnitz, R., letter to C. McCracken (NRC),
Subject:
and Submittal Guidance for the IPliEE," October 10, Modification to Menorandum of 12 July 19S8, January 3, 1990.
1989' Reed, J., et al. " Recommended Seismic IPli Resolution
" Enhancing it.c NRC and EPRI Seismic Margin Procedure " Proceedings of T/drd Symposium on Current Review Methodologies to Analyze the importance of issues Related to Nuclear Power Plant Structures, Equip-Non-Scismic Failures, Iluman Errors, Opportunities for ment and Piping, Orlando, Florida, December 1990.
Recovery, and large Radiolopcal "eleases, Draft 2, Sep' tember 1990' Shao,1, et al., " Consideration of External I! vents in Severe Accidents," Proceedings of Third Symposium on Chery, D., (NRC), memornaum to D. Moeller and D.
Current issues Related to Nuclear Power Plant Structures.
Okrent (ACRS),
Subject:
Hydrologic Engineering Pres-Equipment and Piping Orlande, I f orida, December 1990, entation to Combined Meeting of the ACRS Subcommit-tecs on Site Evaluation and Ihtreme I! vents Phenomena, USNRC, AEOD lingineering Evaluation Report," Light-October 9,1985.
ning Events at Nuclear Power Plants," AEOD-l!605,
, +
- April 1986.
Code cf Federal Regulations, Title 10 " Energy" (19 CFR),
-, Generic l.etter 88-20, " Individual Plant lixamin-U.S. Government Printing Office, Washingtoi., D.C., re-ation for Severe Accident Vulnerabilities-10 CFR vised periodically.
50.54(f)," November 23, 1988.
Daily, C. (NRC), memorandum to R. Savio and M. Stella
--, Generic Letter 88-20, Supplement No 1. "Initia-(ACH S),
Subject:
Assessment of issue Concerning Oper-tion of the Individual Plant Examination for Severe Acci-ating Reactors: Lightning Induced Reactor Events, dent Vulnerabilities-10 CFR 50.54(f)," August 29, August 2,1989.
1989.
Davis, P.,"A Peer Review of Two Seismic Margin Assess-
-, Generic Letter 88-20, Supplement No. 4,"Individ-ments as Applied to the flatch Nuclear Power Plant",
ual Plant lixamination of External livents (IPEliE) for l
Proceedings of Third Symposium on Current issues Related Severe Accident Vulnerabilities-10 CFR 50.54(f)."
to Nuclear Poner Plant Structures Equi;> ment and Piping, draft for comment, July 23,1990.
Orlando, Florida, December 1990.
, Generic Letter SS-20, Supplement No. 4 "Individ.
lilectric Power Research Institute. EPRI NP-6041, "A ual Plant hxamination of External Events (IPEEE) for Methodology for Assessment of Nuclear Power Plant Severe Accident Vulnerabilities-10 CFR 50.54(f),"
Seismic Margin," October 1988.
final, April,1991.
NUREG-1407 26
- 9. References i
I
, Generic Letter 89-22, " Resolution of Generic
-, NUREG/CR-4840, " Recommended Procedures Safety issue No.103, ' Design for Probable Maximum for the Simplified External Event Risk Analysis for Precipitation'," October 19, 1989.
NUREG-1150," Sandia National Iaboratory, Septem-ber 1989.
, NUREG-1032, " Evaluation of Station Blackout Accidents at Nuclear Power Plants," June 1988
-, NUREG/CR-5042, " Evaluation of External llaz-ards to Nuclear Power Plants ' i the United States," De-
, NUREG-1150, " Severe Accident Risks: An As-cembu 1987, t.cssment for Five U.S. Nuclear Power Plants," Vols. I and 2, June 1989.
-, NUREG/CR-5042, Supplemerit 1, " Evaluation of
, NUREG-1335, " Individual Plant Examination:
External Ilazards to Nuclear Power Plants in the United Submittal Guidance," fina' report, August 1989.
States-Seismic lla7ard " April 1988.
, NUREG-75/087, " Standard Review Plan for the
-, NUREG/CR-5042, Supplement 2," Evaluation of Review of Safety Analysis Report for Nuclear Power External Ha7ards to Nuclear Power Plants in the United Plants," 1.WR cdition, De,tember 1975.
States-Other External Events," February 1989.
-, NUREG/CR-0098, "Developtrent of Criteria for
-, NUREG/CR-50 /6,"An Approach to the Quantifi-Seismic Review of Selected Nuclear Pe ;er Plants," May cation of Seismic Margins in Nuclear Power Plants: The 1978.
Impor',cc of thVR Plai.t Systems and Functions to Seis-mic Margins," h.ay 1988.
-, NUREGICR-2300, "PRA Procedures Guide,"
January 1983,
-, NUREG/CR-5088, " Fire Risk Scoping Study,"
-, NUREG/CR-2815, "Probabilistic Safety Analysis January 1989.
Procedures Guide," Vols. I and 2, August 1985.
-, NUREG/CR-5250,"Scismic Hazard Characteriza-
-, NUREG/CR-4334,"An Approach to the Quantifi-tion of 69 Nucler.r Power Plant Sites East of the Rocky cation of Seismic Margins in Nuclear Power P' ants,"
Mountains," Vols.1-8, January 19b9.
August 1985.
EG/CR-5259," Individual Plant Examination
-, NUREG/CR-4482," Recommendations to the Nu-
, n s:
m and I rocedures, draft, clear Regulatory Commission on Trial Guidelines for 3'
Seismic Margin Reviews of Nuclear Power Plants,"
March 1986,
-, NUREG/CR-5270, " Assessment of Seismic Mar-
-, NUREG/CR-4659," Seismic Fragility of Nuclear gin Calculation Methods," March 1989.
Power Plant Components, Phase 1," Vol.1, June 1986.
-, NUREG/CR-5477,"An Evaluation of the Reliabil-
-, NUREG/CR-4659, " Seismic Fragi!ity of Nuclear ity and Usefulness of Externa 11nitiator PRA Methodolo-Power Ph.nt Components, Phase 11, Motor Control Cen-gies," January 1990.
ter, Switchboard, Panclboard and Power Supply," Vol. 2, December 1987.
-, NUREG/CR-5499, " Guidance on Relay Chatter
-, NUREG/CR-4659, "SSsmic Fragility of Nuclear Power Plant Components, Phase 11, Switchgear, I&C
-, NUREG/CR-5501, " Selection of Review Level Panels (NSSS) and Relays," Vol. 3, February 1990.
Ear 1hquak for Seismic Margin Studies Using Scisr.iic PRA Results," June 1991.
-, NUREG/CR-4734, "Scismic Testing of Typical Containment Piping Penetration Systems," December
-," Policy Statement on Severe Reactor Accidents,"
1986.
Federal Reginer Vol. 50, p. 32138, August 8,1983.
-, NUREGICR-4826,"Scismic Margin Review of the Maine Yankee Atomic Power Station," Vols.1-3, March
-, SECY 88-147, " Integration Plans for Closure of 1987.
Severe Accident issues." May 25,1988.
27 NU REG-1407
APPENDhX A REVIEW LEVEL EARTilQUAKE
APPENDIX A REVIEW LEVil EARTHQUAKE
'the seismic margins methodology was designed to dem-review. The results of the 'mming for the plants in the onstrate sufficient margin over the Safe shutdown !!arth-Western United States rre presented in Table 32.
quake (SSli) to er.sure plant safety and to find any " weak links" that raight limit the plant shutdown capability to
'the ratismale for the selection of the review level earth-safely withstand a scismic event larger than the SSl! or quakes (It!.lis)and the grouping of the plantseast of the lead to seism cally induced core damage. The methodol.
Rocky Mountaias is discussed below.
i ogy involves the screening of components based on their irrportance to safety and sci mic capacity. '!he seismic A.I Introductiori margins method utihzes two review or screening levels
..he specification of a review level carthquake (111 Ji) for s
geared to peak ground accelerations of 03g and b.5g. In use in carrying out an individual plant examination for areas of low to moderate teismie hazard, most plants that have been evaluated using Pit As or margins studies have gemal events (IPlilili) has been a complex problem been shown to have llCLPFs at or below 03g. Past expe-involvmg th" search for consistency, it would be prefer-rience indicates that, at the 03g screening level, a small able if the selection of the RI.F.s were completely consis-number of " weak links" are likely to be identified, effi.
tent with the mdividual plant examination (IPli) for mter-ciently defining the dorninant contributors to seismically nal events and the mherent strengths of the scismje mar-induced core damage, it is the staff's judgment that the gins methmlologies, but it is very difficult to satisty all of use of a 03g review level earthquake for most of the these elements m any ngorous quanutative sense. Ihus, miclear power plant sites in the Centrat and Fastern for exampk auempung to equate the review level carth-United States (cant of the llocky Mountains) would serve quake to the reporting criteria in the IPli(mean sequence to meet the objectives of the IPliliti.
pquency kading tc core 6 mage of 103 per year) is iraught with difficultiesbecause of the large uncertamties in numerical estimates of sciamically induced core dam-All sites cast of the llocky Mountains, however, are not age, the inappropriateness of a comparison between m'-
subject to the same level of earthquake hatard.The re-rnerical estimates of scismically and internally induced cent studies by ILNI. (NURiiG/CR-5250) and I!PRI core dans ge(the source and treatmeat of uncertainty can (I!PRI N P4395-1)) show significant differences depend' be quite diifernt), and the innerent difficulties in rel it-ing on location and specific site conditions. Ilecause the ing the output of a seismic margins study (IICI.PF) to two studies do not necessarily agree with each other, it estunates of cue damage fregeenev. For somc of the was dccmed necessary to use both studies in determining same reasons, it was recognized that external initiators, which ren.ew Icvel earthquake should be assigned to each including carthquakes, need not necessarily be treated in site. Ilazard compansons were made using the median' the same manner as internal initiators in implementing 85th percentile, and me:m from the site-specific results the Seycre Accident Policy It abould be noted that the sensitivity tests and engmeerml.PRI studies. Ilased o(1 the prwided by the 1J.NI.and 1 RI.ll defines a reporting level, A IICI.PF value lower g and seismological judg-than ha li does not necessarily represent a riant vulner-ment, the staff has defmed the review level earthquake Wihty. However, the licensee should assess the signifi.
for each site (03g,0.5g, or reduced scope) m l'able 3.1. A cance of IICLPF values lower than RIJi and take any cecond criterion, plant design basis, was uset to subdivide necessary actions and make 01her improvements that are the 03g bin.The subdivision, based on a composite condi' deemed appropriate by the licensee.
tional probability of exceeding the SSli for each nuclear power plant, resulted in plants withm the 03g bin being A.2 General Iw, aluat,on Procedure i
assigned a full-scope,or focused.se review.
A.2.1 Dala ICyaluated
.the sites in the Western Um.ted States (west of the Rocky Mountains) are treated differently. Those sites in coastal The staff has recommended three review level carth-California where the seismic hazard is much higher and quakes to be used when applying the scismic margins the resulting design bases are greater than 0.5g cannot methmlology to nuclear power plants cast of the Rocky make use of the margins methodology. The other plant Mountains for the IPlilitiJlhe review levels or " bins" ar e sites in the West should use n 0.5g review level carth-0.5g. 0.3g, and a reduced scope level. The basic informa-quake unless it can be demonstrated that the seismic tion used was the lawrence 1.ivermore National labora-ha/ard level at a particular plant site is consistent with the tory (ILNI.) hazard study (NURiiG!CR-5250) and the scismic hazard at thc 0.3g hin plant sites cast of the Rocky filectric Power ller.carch Institute (l!PRI) hazard study Mountains. Western sites that show such a consistei,cy in (liPhi NP4395-1)). These 'tudies represent state +f-seismic hazard wi!! conduct the full-scope OJg margins the-art estimates of seistnic ha/ard Ilecause the two A~l NURIG1407
)
Appendix A studies do not necessarily agree with each other, it was response spectra and PG A. 'lhe likelihoods of exceeding deemed necessary to use them 'acth in deiermining which spectral response accelerations in the 2.5 to 10 Hz range bin a particular site belonged m.
were examined because these frequencies are more closely related to the types of motion that could cause in the LLNL study (NUREG/CR-5250), it was noted damage at nuclear power plants. Unit weights (2 nth that, for some sites, the mean estimates of seismic hazard each)were assigned to the likelihoods of exceeding spec-were dominated by the inpar of one ground motion expert tral response ordinates at 2.5,5, and 10 IIz. One-half unit (No. 5). This, dominance vas caused by the low attenu.
weight (Inth)was assigned to the likelihood of exceeding ation, high uncertainty, and relatively high motion on rock the PGA.
found in this expert's input. *This input has received a great deal of attention, and some have argued that it is A.2.4 Ranking Criteria inconsistent with the data. The staff requested LLNL(as a sensitivity study) to calculate the hazard at nuclear Emphasis was placed on the relative ranking of sites with power plant sites cast of the Rocky Mountains leavhg out respect to other sites using the same setsmic hazard study, the input of this expert.
sta.istic, and ground motion measures. Extensive use was made of a clustering methodology developed by LIEL Data from the Saguenay Event in Quebec, Canada (No-for the NRC (llernreuter et al.,1989a,1989b). For a given vember 1988), the largest earthquake in castern North hazard study, statistic, ground motion measur a and refez-America in 50 years, appears to be quite different from ence level, this methodology divides the ensemble of sitej nt groups so that the sites in any one group are "close previous data sets and has not helped to resolve the con, troversy. At this time, in order to avoid relying exclusively to each other with respect to seismic hazard I or example, on the LLNL results that include the input of expert No, the sites may be divided into groups based on mean esto 5, the staff is treatmg the LLNL hazard estimates based mates of exceeding 0.5g PGA from the EPRI study or median estimates of exceedm.g the 2.5 Hz spectral re-on the other four ground motion experts as a separate study when binning nuclear power plant sites for IPEEE.
sp nse (associated with the NUREG/CR-0098 response spectrum anchored at 0.3g) from the LLNL five-expert study. Although there were a fixed number of groups, no A.2.2 Comparison Procedure minimum number of. '.tes were in a group, and indeed
- E*"E' #
" U Hazard comparisons were made using the mean, median, and 85th percentile from the site-specific results provided A.2.5 Spectral Shape I
by the LLNL and EPRI studies. Each of these pieces of information represents a different way of characterizing The spectral shape associated with the 0.3g screening the distribution of seismic hazard estimates at each site as level was assumed to be the median NUREG/CR-0098 i
l determined by a particular study.
spectrum anchored at 0.3g. There has been some discus-l-
sion that the screening level should actually be associated Meaflhe mean is a commonly used statistic that can be with a somewhat higher ground motion (the Seismic assigned actuarial significance. However, becuse of the Qualification Utility G rou p (SQUG) bounding spectrum) skewed natta e of the distribution, it is also a highly unsta-but in th's relative comparison, the use of this alternative ble (with respect to methodology and input assumptions) spectrum v old make little or no difference.
view of hazard.The mean is highly sensitive to the charac.
terization of the extremes of the distribution.
A.3 Specific Binning Procedtire Mcdian:'the median is more stabic than the mean and A.3.1 Initial [linning Evaluation shows the greatest agreemerit between the LLNL and EPRI studies. However, it is only the 50th percuulle of As the first step, sites that consistently fell into the group the hazard and is insensitive to the extent of uncertainty.
that had the highest likelihood of exceeding the 0.3g NUREG/CR-0098 5% damped median spectrum were 85th Percentile An c! tentative candidate to the mean is conditionally assigned to the 0.5g bin. Sites that fell into the 85th percentile. i. mflects uncerhinty in that it indi, the group that had the lowest likelihood of exceeding the cates the breadth of N distribu tion, but it is less sensitive 0.3g NUREG/CR-0098 5% damped spectrum were as-to cereme outliers.
signed to the reduced. scope bm.
The ground motion measure compared was the weighted A.2.3 Weigltting Criteria combination of 2.5 Hz,5 Hi,10 Hz and PG AJlhe individ.
In tk mst, great emphasis has been placed on the likeli.
h ad ot excee@ peak ground acceleration (PGA). In 1.
Agreement among the LLNI. five-expert, I.LNL this evaluation, siac hazard comparisons were made using four-expert, and EPRI studies, and NUREG-1407 A-2
Appendix A I
l 2.
Agreement between the median and cPher mean or
'the staff's resolution of the Eastern U.S. Seisraicity Issue l
85th percentile statistics.
(Charleston Earthquake Issue) has identified eight plants I
at five sites as outliers.The staff determined that these plants should be assigned to the full-scope category.This This resulted in a comparison of r separate hazard action added a single additional plant, Arkansas Nuclear groupings (three pieces of informath a for each of the One, Unit 1, to the list derived on the basu of seismic three studies).
hazard and ruistaic design.
For example, if a particular site fellin the top group (0.5g
'The candidates 'or the 0.5g and reduced-scope bins were bin) for all of the criteria except the EPRI median, it then subjected to additional evaluation by the staff.
remained in the 0.3g bin. 'lhe conclusions must be sup-ported by all the haza:d studies. On the other hand, if a A.3.3 Subsequent Binning Evaluations particular site fell m the bottom group for all of the criteria except for the LLNL four-and 11NL five-expert lhe candidates for the 0.5g bin were first examined to mean estimates, it was included in the reduced-scope bin, provide some assurance that, although the hazard w,s Only one measure of uncertainty, mean or 85th percen-relatively high, it was high enough to warrant inclusioi.a tile, needs to be satisfied.
this bin.
As a test, it w s c nsid red appr pri t that a site be-A.3.2 0.3g Ilin Subdivision longed m the 0.5g bm if a hypothetical nuclear power plant at that site was assumed to have a liCLPF of 0.3g
,rhe staff investigated the potentini for using the scismic and the mean annual core damage frequency associated design basis as a parameter for makmg the truttal binning with that hypothetical plant was 103 or higher. The work ass:gnments. there war insufficient technicalbasis for is cited in N U REG /CR-5501 showed that the mean annual use; thus it was not used for the mitial binning. Ilowever, core damage frequency was roughly an order of magni-when combmed with Lazard and engineering judgment, tude lower than the mean annuallikelihood of exceeding the use of the seismic design provided a basis for an the plant llCLPF and very roughly equal to the median overall cost-effective reduction in the scope af the 0.3g annual likelihood of execeding the plant ilCLPF.
margins review. The staff repeated the process that was used to obtain the initial binningwith 'he sole change that llased on these estimates, the staff assumed that inclusion msteau of factoring in only the seismic hazard, the seismic in the 0.5g bin would be supported if:
hazard and the seismic design basis were used.
1.
The mean or 85th percentile annual likelihood of Composite conditional probabilities were obtained for exceeding the 0.3g spectrum from all three studies the three seismic hazard curves (EPRI, LLNL with four was 103 or greater, and experts, and LLN L with five experts) and the three statis-tical measures of the hazard curves (mean, median, and 2.
The median annual likelihood of exceeding the 0.3g 85%)-nine separate probabilities for each site. A com-spectrum from all three studies was 10 5 or greater.
jmsite conditional probability was formed by adding the weighted conditional probabilities of exceeding the uni-This evaluation should be viewed as n " sanity check"; it form hazard spectra at a particular mound motion fre-should not be viewed as a plant-specifie statement on core quency;i.e., the irersection of the plant-specific seismic damage frequencies. The reasons are:
design spectrum for the particular frequencies with the 1.
The uncertainty and generic nature associated with uniform hazard spectra yields the conditional probability.
the correlation in NUREG/CR-5501, The frequencies were those used for the initial btnnmg (2.511z,5Hz,10liz, and PGA). The were also weighted 2.
The use of spectral estimates rather than peak the same (2n, b7,2n, and 10).
ground acceleration, g
Using the same agreement crit as in the initial bin-3.
The inclusion of the 85th perem%omates. and ning, six sites were identified, i.t
,nsistently fell into the top group. These are listed u Table 3.1 to do the 4.
All the previously mentioned problems associated full-scope 0.3g seismic margms review.
w th bonom line numbers.
As a " sanity check" of this approach, the 'ist derived from Finally, the staff examined the cimdidates for the 0,5g and this appnuch was compared to the list denved from scis-reduced-scope bins to assure itself that the classification mic hazard alone.The six full-scope plat. ; were among made good seismologictd sense and t5cre was no need to the top ten seismic hazard sites, include additional sites in these bins. In conjunction with A-3 NUREG-1407
Appendix A this examination, limited sensitivity tests were alst car-A.43 03g Ilin ried out to determine the impact of slight relaxations in the consistency criteria.
All sites not identified as belonging in the 0.5g or reduced-scope bins were assigned to the 03g bm.
AA Ilinmng of Sites-Results A.4.1 Reduced Scope Margins Methodology Ilin A.4.4 Other Considerations The cons:stency criteria outlined in Section A3.1 were slightly modified to identify sites for the reduced-scope The grouping was maJe assuming that each I'v.ation was bin. The two bottom median groups were included rather associated with one site coridition (rock or vuying depths j
than the bottom group alone. When this was done, five of soil). Some twelve plant sites cast of the Rocky Moun-sites (South Texas, Cornanche Peak, Waterford, River tains whose main Category I structures are located on llend, and Crystal River) were identified as belonging to rock also have some Category I structures or components the redaced scope bin.
located on shallow or intermediate depths of soil. Since shallow soil, less than about 80, feet thick, c:m significantly Also added to this bin were several sites for which no arnplify ground motion, these sites should perform soil EPRI calculations were available but were in the bottom amplification studies to determine the effect.
groups in both the ILNL four-and five-expert studies.
They are Duane Arnold. Big Rock Point, Grand Gulf, St.
Lucie, and Turkey Point. The ten candidate sites in the In particular, for four of the sit:s included in the 03g bin reduced-scope bin lie in areas of low scismic hazard along (on the basis of their primary site conditions), the ha7;ird or near the Gulf and Florida coasts and in the upper for structures or components on the secondary site condi-Midwe? t.
tions is equal to or higher than the hazard associated with those plants in the 0.5g bin, Licensees should, if the A.4.2 0.5g Ilin site specific analysis indicates, use the 0.5g screening ta-b!cs for elements affected by soil amplification. Similarly, As a result of the evaluations cited above, two sites (Pil-for one site in the reduced scope bin, site-specific analysis grim and Seabrook) were identified as belonging in the should be carried out to determine the effects on those 0.5g bin.
elements affected by soil amplification.
l-i l
l l
NUREG-1407 A-4
APPENOlXIl COMPARISON [lETWEEN A REDUCED-SCOPE AND FULL-SCOPE SEISMIC MARGINS EVALUATION
APPENDlX 11 COMPARISON llETWEEN A REDUCED-SCOPE AND FULleSCOPE SEISMIC MARGINS EVALUATION Thete are differences bctween the reduced scope and 3.
The reduced-sspe evaluation should be identicalin full scope margir.s evaluation both in the extent of the quality and effort to that as for the full-scope mar-systems analysis and in the amount of quantification of gins methodology. One crucial feature is that it 1 ICLPF values for eqaipment identified in the walkdown.
should involve interactions among scistnic capability The comparison is presented in Table 11.1. The emphaais evaluation engineers, systems engincars, and the on walidown and not on quantification also applies to the licensee's plant operations personnel. 'the walk-pctformance of containment and containment syvems down team should visually inspect pertinent struc-(that is containment performance analysis should concen-turcs, equipment, and anchorages consistent with trate on identifying seismically induced vulnerabilities the full-scope NitC or liPill methodology, If poten.
and sequences different from those obtained from the tially vulnerable components are found during the IPli), US! A-45 (Decay lleat itemovat Itequirements),
walkdown, a capacity check may be necessary using and G1-131 (In-Core I' lux Mapping System)-
the applicable SSli ground response spectra.These results should be documented. Data sheets similar to those found in Appendix 1 of liPit! NP-6041 11.1 Eteliteitts Preserved should be used to document the walkdown. A t eview of construction drawings for structurai details that The following elements of the mismic margins methodol-can not be seen m the field should be performed.
ogy mast be preserved; that is, they must be identical in the reduced-scope and full-scope evaluation:
4.
While the post-walkdown assessment effort for a mduced-scope evaluation should be identical in 1.
For either the NitC or !!Piti methodology, the sys-quality to that in the full-scope margins methodol-tems engineers must perform significant pre-ogy its thrust and level of effort are different be-walkdown work that should be preserved m a eaase sequence level (NRC) or success path-level reduced scope evaluauon. In the NitC methodol-(l! Pill) llCl.PFr will not be computed. Instead, its ogy, this myolves defining initiating events, defining emphasis should be on identifying possible weak-eveat trees and the safety functirns involved, and I nk items that may need strengthenmg.
identifying systems and components necessary to carry out these functions. In the !! Pill methodology, this involves defining success paths (primary and 11.2 RedUet,l0IIS alternative) and the systems and components in-
.lhe following, although needed in the full scope margins volved in these paths. l'or both methodologtes. the methodology, are not needed in a reduced-scope margins thrust of this work is to narrow the scope and focus evaluation the effort of the key element of the review, the walkdown' 11.2.1 NitC Sdsmic Margins Metliodology 2.
For either the NitC or the liPRI methodology, the 1.
The systems engineers need not prepare or quantify seismic capability evaluation engineers must per-fault trees and Ikiolcan expressions representing form significant pre-walkdown wort that should be accident sequences. Also, since fault trees will not preserved in the reduced. scope evaluation. In each be developed, these engineers need not combine methodology, this involves developing an under-nonseismic failure basic events with seismically initi-standing of the seismic input to the plant and the ated failures in any rigorous fashion, although the scismic design basis and realistic ground and fhior existence of those non-seismic failures,ifidentified, response spectra. It also involves pre-walkdown should be noted and their importance assessed in screeningef he keyrystemsandcordonentsidenti-the course of the margin evaluation.
t ficd by the systems engineers so as to make the walkdown itself most cificient. The thrust of the 2.
The seismic capability evaluation engineers need screening is to identify items thought to have very not developilCl PFeapacityvaluesforallof the key high IICl.PF values, items suspected of having low equTment items that would be represented on the llCLPF values, and therefore lists of items to be sequence level llooleans (which will not be devel.
examined at various levels of detail during t he walk-oped), It follows that they can not develop a plant-down.
level lit 't.1 F capacity value.
H-1 N URIG 1407
_.~.
Appendix IF B.2.2 EPIU Seismic Margins Methodology items found on the success paths (primary and alterna-tive) being studied. it follows that they can not develop The seismic capability evaluation engineers need not de-any success-path-level 11CifF capacity values that would velop iICifFcapacity values for all of the key equipment be taken as representmg the plant-level IIClfF capacity.
B Table B,1 Reduced-Scope Margins Method Based on NRC Selsmic Margins Methodology (NUREG/CR-4482, Chapter 4)
Step No.
Description in Reduced Program?
1 Selection of Earthquake Review Level Not applicable, NRC designates sites that qualify 2
Initial Systems Review Yes, in entirety 3
Initial Component HCLIF Categoriation Yes, in entirety 4
First Plant Walkdown Yes, in entirety 5
Systems Modeling hnalize Event Trees:
Yes Fault Tree Development:
No 6
Second Plant Walkdown Only as needed 7
Systems Model Development No 8
Margin Evaluation of Components, Plant No Based on EPRI Seismic Margins Methodology (EPRI NP-6041, Chapter 2)
Step No.
Description in Reduced Program?
1 Selection of Seis;nic Margins Earthquake Not applicabb, NRC designates sites that qualify 2
Selection of AssessmentTeam Yes, in entirety 3
Pre-Walkdown Preparation Work Yes, in entirety 4
Systems and Element Selection Walkdown Yes, in entirety 5
Seismic Capacity Walkdown Yes, in entirety 6
Subsequent Walkdowns Only as needed l
7 Seismic Margin Assessment Work No I
i f
l-NUllEG-1407 B-2
APPENDIX C DETAILED DOCUMENTATION AND REPORTING GUIDELINES
APPENDIX C DETAILED DOCUMENTATION AIND REPORTING GUIDELINES
'lhis appendix provides the guidelines for detailed docu-1.
Plant layout and containment building information mentation and reporting format and content for the not contained in the Final Safety Analysis lleport IPEi!E submittals. 'lhe major parts of this appendix are (l'S All).
the guidclines for scismic analysis (Section C.2), internal fire analysis (Section C.3), other analyses (Section C.4),
2.
A concise description of plant documentation used specific safety features and plant improvements (Section in the IPEEli,(e.g., the 1 S AR; system descriptions, C.I.4), and the licensee review team (Section C.l.5).The procedries, and licensec event reports); and a con-licensees are requested to submit their IPEEE rcports cise discussion, of the process used to confirm that using the standard table of contents given in Table C.1 or the IPllEE represents the as-built, as-operated provide a cross reference.This will facilitate review by the plant. The intent of such a confirmation is not to NRC ar.d prornote consistency among various submittals, propose new design reverification efforts on the part The contents of the elements of this table are discussed in of the licensees but to accmmt for the impact of sections below, previous plant modifications or mod:fications con-ducted within the IPElill framework.
'the level of detail needed in the documentation should 3.
A description of the coordination activitics of the be sufficient to enable N RC to understand and determine the validity of all input data and calculation models used IPliEE teams among the external events (e.g., for c
seismically mduced fires),
to essess the sensitivity of the results to all key aspects c the analysis, and to audit any calculation. It is not necca sary to submit all the Acumentation needed for such an C.l.4 Subnu.ttal of Spee.Ge Safety Features NRC review. Itclevant documentation should be cited in and Potential Plant improvements the IPEEli submittal, and be available'm easily retriev-The licensec should proviac a discussion of the criteria able form. The guideline for judging the adequacy of used to define vulnerabilities for each ex'ernal event retained documentation is that independent expert ana-evaluated.The licensee r.hould list any potential improve-lysts should be able to reproduce any portion of the ments (including couipment changes as well as changes in results of the calculations m a straight fory.ard, unambi-maintenance, operating and emergency procedures, sur-guous manner. I'o the extent possible, the retained docu-veillance, staffing, and training programs) that have been mentation should be organized almag the hnes identified selected for implementation based on tne IPEfiE (c m the areasof review. Any mformation that is comparable schedule for implementation should be provided)or that to that provided under the IPE for mternal events can be have already been implemented. A discussion of antici-
~
pated benefits in terms of averted potential risk or in-creased plant scismic capacity as well as drawbacks to any C.1 General improvements should be provided.Those improvements that have been taken credit for in the analysis and have C.1.1 Conformance with Generic Letter and not yet been implemented at the plant, should be specifi-Supporting Material cally highlighted in the submittal.
Certification that an IPEEE has been completed and C.I.5 IPEEE Team and Peer Review documented as requested by Generic Letter 88-20, Sup-plement 4. The certification should also 'dentify the
,lhe basts for requiring the m.volvement of the licensee's staff in the IPliEli review is the belief that the maximum measures taken to ensure'the technical adequacy of the IPEEE and the validation of results, including any uncer-benefit from the performance of an IPEEE would be tainty, sensitivity, and importance analyses.
reahzed if the heensee's staff were involved in all aspects of the examination and that mvolvement would facilitate integration of the knowledge gained from the exam-C.1,2 General Methodology ination into operating procedures and training programs.
Thus the submittal should describe licensee staff partici-Provide an overview description of the methodology em' pation and the extent to which the licensee was involved ph>yed in the IPEEE foi cach external event.
in all aspects of the program.
C.I.3 Information Assembly The submittal should also contain a description of the peer review performed, the same type of review as re-Reporting guidelines include:
quested for the internal event IPli, the results of the C-1 NURiiG-1407 v.....
Appendix C review team's evaluation, and a list of the review team damage, including a qualitative discussion of uncer-members.
tainties and how they might affect the final results, and contributions of different ground motions to C.2 Seismic Events core damage frequencies.
Section C.2.1 describes submittal information geldelines 7.
Any seismically induced containment failures and for licensees who choose the seismic PRA for seismic other containment performance insights. Particu-IPEliE, whilq section C.2.2 describes information guide-larly, vulnerabilities found in the systems / functions lines for licensees who choose the scismic margin method which will lead to early containment failure that for the seismic IPEEE, *lho submittal should be pre.
might result in high consequences. 'this includes:
sented in conformance veith Table C.l.
isolation, bypass, integrity, and systems (e.g., ig-niters) required to prevent early failure. 'lhe com.
C.2.1 Seismic PRA Methodology puted fragilities of containment components, systems, and functions as applicable should be pro-
'the fallowing information on the seismic IPEEE should vided. The licensee may submit competed llCLPFs be documented and submitted to the NRC:
associated with containment periormance (Op-
- tional).
1.
A description of the methodology and key assump-tions used in performing the seismic IPEEE.
8.
A table of fragilities, bc ;h generic and plant-specific, used for screening as well as in the quantification.
2.
De h.vard curve (s)(or table of hazard values) used The estimated fragilities for the plant, dominant and the associated spectral shape used in the analy-sequences, and domina it components should be re-sis. Also,if an upper bound cutoff to ground motion ported. (Optional:'lhe estimated IICLPF for the of less than 1.5g peak ground acceleration is as-plant, daminant sequences, and components with sumed, the results of sensitivity studies to determine and without nonscismic failures and human actions whether the cutoff affected the overall results and may be submitted by the licensee.)
the delineation anu ranking of seismic sequen:es.
9.
Documentation with regard to other scismic issues 3.
A summary of the walkdown findings and a concise (Section 6) addressed by the submittal, the basis and description of the walkdown team and the proced.
assumptions used to address these issues, and a dis-cussion of the findings and conclusions. Evaluation ures used.
results and potential improvements associated with 4.
All functional / systemic seismic event trees as well as the decay heat rernoval function and movable in-data (including origin and method of analysis). Ad.
core flux mapping system (for Westinghouse plants) dress to what extent the recommended enhance.
should be specifically highlighted.
ments have been incorporated in the IPEEE. A de-
- 10. A Discussion of nonscismic failures and human ac-scription of how nonseismic failures, human actions, tions that are significant contributors, or have tm.
dependencies, relay chatter, soil liquefaction, and seismically induced floods / fires are accounted for.
pacts on results.
Also, a list of important nonseismic failures with a t 1. When an existing PR Ais used to address the seismic rationale for the assumed failure rate given a setsmic IPEEE, the licensee should describe sensitivity stud-ies related to the use of the initial hazard curves, supplemental plant walk 60wn results and subse-5.
A description of dominant functional / systemic se-quent waluations, and rclay-chatter evaluations.
quences leading to core damage along with their
.the licensee should examm.e the above list to fill in frequencies and perccatage contribution to overall those items missed m the existing seismic I R A (See seismic core darnage frequencies (for both LLNI.
Section 3.1.2).
and EPRI hazard curves if used). Sequence selection l
criteria are provided in GL 88-20 and NUREG-C.2.2 Se.ismic Margins Methodology 1335. If either hazard curve causes a sequence to meet these criteria, that sequence should be in-The following information on the seismic IPEEfi should cluded. The descriptior. of the sequences should be documented and submitted to the NRC for a full-include n discussion of *.gecific assumptions and hu-scope or a focused-scope SMM review:
man recover actions.
1.
A description of the methodology and a list of im-6.
The estimated core damage frequency (for both the portant assumptions. including their basis, used in LLNL and EPRI hazard curves if used) and plant performing the seismic IPEEli. Address the extent damage state frequencies, the timing of the core to which the folicwing were taken into accou it:
NUREG-1407 C-2
Appendix C nonseismic failures, human actions, dependencica, L
A description of the procedures used to identify relay chatter, soil liquefaction, and seismically in-systems and componenti for the wal!. dawn in per-duced floodc! fires. Also, a list of important nonscis-forming the seismic IPEllli, mic failures with a rationale for the assumed failure rate given a seismic event.
2.
A summary of the walkdown fMings and a concise description of the walkdo".a team and proceQres used.
2.
A summary of the walkdown results and a concise description of the walkdown team and procedures 3.
A discussion and the results of a j specific compo-used.
nent capacity evaluations performed, the methods 3.
All functional / systemic scismic event trees data (in-ciuding origin and method of analysis) when NRC 4.
Documentation with regard to other seis'nic issues-SMM is used.
(Section 6) addressed by the submittal, the basis and
{
assumptions used to address these issues, and a dis-4.
A description of the most important sequences and cussion of the findings and conclusions,livaluation important minimal cutsets (for both seismic and results and potential improver.ents associated with i
nonscismic failures) leading to core damage (NRC the decay heat removal function and movable m-method) or a description of the success paths and core flux mapping system (for Westinghouse plants) 1 procedures used for their selection and of each com-should be specifically highlighted, ponent in the controlling success path (EPRI method).
C,3 Internal h.res j
The information on the internal fires IPlilili identified 5.
Any seismically induced containment failures and below should be docur..ented and submitted to the NRC, other containment performance insights. Particu-larly, vulnerabilities found in the systems / functions 1.
A description of the methodology and key assump-which will lead to early containment failure and high tions used in performing the fire IPEEli and a d scussion of the status of Appendix R modifica-consequences. This includes: isolation, bypass, con.
j tainment integrity and systems (e.g., igniters) re-tions.
quired to prevent early failure. Also, computed fra-2.
A summary of the walkdown findings and a concise gilities (if used) r.nd ilCLPFs of containment components, systems, and functions as applicable, description of the walkdown team and th. proced-urcs used. This should include a description of the efforts to ensure that cable routing used in the analy-6.
A table of frag.dities (if used) and IICLPFs, both sis represents as-built information and the treat generic and plant-specific, used for screenint; as well ment of any existing dependence between waote as in the quantification. The estimated fragdities (if shutdown and control room circuitry, used) and IICl.PFs for the plant, dominant se-quences, and dominant components should be re-3.
A discussion of the criteria used to ids dy crinal ported-fire areas and a list of criticalareas,incir ta) sin gic area in which equipment fat!ures i cM 7.
Documentation with regard to uther seismic issues serious crosion of safety nargin,N Waa e(a).
(Section 6) addressed by the r bmittal,the basis and but for double or multiple areas imm co.ncun assumptions used to addren these issues, and a dis.
barriers, penetration seals, ilVAC 'actm;;. ctc.
cussion of the findings and conclusions. Evaluation results and potentialimprovemeawsociated with 4.
A discussion of the criteria vm f'r fire size and the decay heat removal function and movable in.
durationandthe treatmentof t, "v fhu spread core flux mapping system (for Westinghouse plants) and associated major a urr' ti As should be specifically highlighted.
5.
A discussion of thw d.c uN. :on data base, includ-ing the plant-sg. ~ c ca' aase sed. Provide docu-8.
For the NRC method, provide a discussion of non-mentation in each t - c whe' - the plant.specifie data seismic failures and human actions that are signifi-used is less const vmive than the data base used in cant contributors to or have an impact on the results.
the approved fire vulnerability methodologies. De-scribe data handling method, including major as-The following is the information that should be docu-sumptions, the role of expert judgment, and the mented and submitted to the NRC for a reduced-scope identification and evaluation of sources of data un-SMM review:
certainties.
C-3 NUREG-1407
Appendix C 6.
A discussion of the treatment of fire growth and ies related to the use of the initial hazard, supple-spread, the spread of hot gases and smoke, and the mental plant walkdown results and subsequent analysis of detection and suppression and their asso-evaluations.The licensee should examine the above cuted assumptions, includirig the treatment of list to fill in those items missed in the existing, fire suppression-induced damage to equipment.
PRA.
7.
A discussion of fire damage modeling, including th CA Ili h Winds' Floods' and Others E
definition of fire-mduced failures related to fire bar-riers and control systems and fire induced damage to The fellowing informatior. on the high winds, floods, and cabinets. A discussion of how human intervention is others ponien of the IPI!!ill should be documented and treated and how fire-induced and non-fire-induced submitted to the NRC:
failures are combined. Identify recovery actions and types of fire mitigating actions taken credit for in 1.
A desesiption of the methodologies used in the these seq'Jences.
examination.
8.
Discuss the treatment of detection and suppression.
2.
Information on plant-specific hazard data and li-meluding fire fighting procedures, fire bngade train-ing and adequacy of existing fire brigade equipment, censing bases, and treatment of access routes versus existing barriers.
3.
Identified significant changes not reported per 10 CFR 50.71(c)(See Section 5.2.2) if any, since OL 9.
All functionalhystemic event trees associated with issuance with respect to high winds, floods, and fire-tnitiated sequences.
other external events.
- 10. A description of dominant functional / systemic sc-4.
Results of plant / facility design review to determine quences leading to core damage along with their their robustness in relation to NRC's 1975 SRP frequencies and percentage contribution to ovemp criteria.
core damage frequencies due to fire. Sequence se-lection criteria are provided in Gl. 88-20 and 5.
Results of the asselsment of the hazard frequency NURIIG-1335. 'lhe description o the sequences and the associated conditional core damage fre-r should include a discussion of specific assumptions quency if step 4 of Figure 5.1 is used.
and human recovery action.
6.
Re ;lts of the bounding analysis if step 5 of Figure
- 11. The estimated core damage frequency, the timing of 5.1 is used.
the associr.ted core damage, a list of analytical as-sumptions including their bases, and the sources of 7.
All functional event trees, including origin and uncertainties.
method of analysis (PRA only).
- 12. Any fire induced containment failures identified as 8.
A description of each functional sequence selected, being different from those identified in the internal events analysis.
tncluding discussion of specific assumptions and hu-man recovery action (PRA only).
- 13. Documentation with regard to the decay heat removal function and Fire Risk Scopmg Study issues ao.
9.
'the estimatcd core damage frequency, the timing of dressed by the submittal, the basis and assumptions the associated core damage, a list of analytical as-used to address these issues, and a discussion of the sumptions incl;2 ding their bases, and the sources of findings and conclusions. livaluation results and po.
uncertainties, if applicable (PRA only).
tential improvements should be specifically high-
- lighted,
- 10. A certification that no other plant-unique external event is known that poses any significant threat of
- 14. When an existing PRA is used to address the fire severe accident within the context of the screening IPlilili, the licensee should describe sensitivity stud-approach for "lligh Winds. Flmds, and Others "
\\
l l
NURI!G-1407 C-4
Appendix C Table C.1 Standard Table of Contents for IPLEE Submittal 1.
Executive Surr. mary 1.1 Background and Objectives 1.2 Plant Familiarization 1.3 Overall hiethodology 1.4 Summary of hiajor l'indings 2.
Examinatica Description 2.1 Introduction 2.2 Conformance with Generic 1.etter and Supporting htaterial 2.3 General hiethodology 2.4 Information Assembly 3.
Seismic Analysis 3.0 hiethodology Selection 3.la Seismic PRA 3.1.1 Ilazard Analysis 3.1.2 Iteview of Plant Information and Walkdown 3.1.3 Analysis of Plant System and Structure llesponse 3.1.4 livaluation of Comp (ment Fragilities and Failure Modes 3.1.5 Analysis of Plant Systems and Sequences 3.1.6 Analysis of Containment Performance 3.lb Seismic hiargins Method (SMM)(NRC, !!PRI, or Reduced SMM) 3.1.1 Review of Plant Information, Screening, and Walkdown 3.1.2 System Analysis 3.1.3 Analysis of Structure Response 3.1.4 livaluation of Seismic Capacities of Components and Plant 3.1.5 Analysis of Containment Performance 3.2 USI A-45, GI-131 and Other Seismic Safety Issues 4.
Internal 1; ires Analysis 4.0 Methodology Selection 4.1 Fire llazard Analysis 4.2 Review of Hant Information and Walkdown 4.3 Fire Growth and Propagation 4.4 Evaluation of Component Fragilities and Failure modes 4.5 Fire Detection and Suppression 4.6 Analysis of Plant Systems, Sequences, and Plant Response 4.7 Analysis of Containment Performance (If Applicable) 4.8 Treatment of Fire Risk Scop;ng Study 1 ucs 4.4 USI A-45 and other Safety Issues 5.
1ligh Winds, Fhmds, and Others 5.1 Iligh Winds 5.2 Floods 5.3 Transportatian and Nearby Facility Accidents 5.4 Others 6.
1.icensee Participation and Internal Review Team 6.1 IPl!!!E Program Organization 6.2 Composition of Independent Review Team 6.3 Areas of Review and Major Comments 6.4 Resolution of Comments 7.
Plant improvements and Unique Safety Features 8.
Summary and Conclusions (including proposed resolution of LISIs and Gis)
C-5 NU111h 1407
i API'ENDIX D NRC RESPONSE TO COMMENTS AND QUESTIONS i
f c'-
. +
.~.-
~_
Al'I'ENDIX 1)
NRC lilCSI'ONSl? TO COMMl?NTS ANI) QUESTIONS D.1 Intrmluction anti Susimiary Sie The sialf denved its estimates of cosi and resoun e requis ements to perform an 11'l!!3! florn the netual
'the NitC staff conducted an IPElil! Workshop on Sep.
costs spent on two NUltliG-Il50 pt.mts and the tember 11-13. 1990 at the littsbtegh Ibiton in Pitts-cost spent on Ihe i latch seismic r eview extrapolated busgh, Pennsylv nia. 'lhe objectives of the workshop to the IPlili'! scope. At the 11'111111 won kshop, cer-a w es c lu discuss t he lPhlili process and io solicit questions tain industry estimates were piesented that were and comments on the guidance for performing the eiths r less than or comparable to the staff's esti-IPlil;li and for reporting the results of the review,'the matesJihe wiff recogni/cs that ther e are imcer tain-schcdule of the ll'lilil! Workshop was announced in the ties in the costs because thes e are unter tainties asso.
Federal llegister ($$ l'It 30332) July 25,1990, and a ciated with tfic analysis of external events. llowever, prehminary agenda of the workshop was published on the staff beheves that ther e me ways to keep the cost August 10,1990, in the l'rderal Regis; r (551:1132712).
under control, If mhlitional questions mise regard-ing the IPliEli n ocess and the associated All the questions and eorn ments raised al the wor kshop or quMon, the sud mu med wMi kenwn and sulnnitted after the workshop were categorvel into sev-rojxmd to unu qunnons cral major subject areas /lhis appendix summar ves Ihese
. questions and comments and the NltC staff resgumses
- 3) Schedule and Rewarce Aiadabil.ty: Schcdule ami re-(Sit) te them.'lhe most signific,nt comments, concerns, murce arailabihty for performing the #' HEE are of and questions, together with staff i esponse (Sit), ar e sum.
concun marized below. 'lhls summary also serves to highlight the major changes made in going from the proposed to the git. lhe methods identified in this report are not new final documents.
and have been used and discossed extensively in the past. Probabihstic rn.k assesstnent (Pit A) proced.
1)
Hacifit Analysic Should a rrgulatory backf;I unulyss of urcs for assessing the risk associated with external the proposed frEEE cffort be performed prior to inu, events have been used since the late 1970's. Tne uncr of the /PEEE generic / citer?
NitC seismic inargins inethod was published in 1985; the !!!cetric l'ower itesearch Institute (IIPill)
SR: '15e staf f does not believe, as a legal matter, that a 10 seistnic marpms method was published in 1988.
.these methods wcre derived from the insights CPit 50J6 type backfit analysis is needed for the pined from available seismic Pit As. *lhey werc IPl!!!Ii genet ic letter (mernorandum from W. Pailer widely discussed at many conferences and work 4 to Commissioner s, dated July 27,1990). Ilecause the shops and were used at three plant sites. Proceduies request to perform the IPlilil! is considered to be a for the scismic v alkdown, one of the most important request under 10 Cl It 50.54(l), the staff has per' ingredients in the seismic IPlilli!, are similar to formed a 50.54(f) analysis, which is included as Ap-those that will be used in the implementation of pendix 5 ty Supplement 4 of Generic 1 citer 88-20.
Unresolved Safety Issue (USI) A-46,"Venfication 1lowever, m view of the sigmheunt hcensee resource of Seismic Auequacy of 1:.quipment in Operating commitment required to respond to this mformation I'lanW' A number of trial walkdown trainingwork-regnest and,m the mterest of a prudent policy, the shops with a number of participants from the utili-staff has completed a value-impact analysis and has ties were conducted la the past by the Scismie Quali-meluded it as an attachinent to this appendix. Ihis fication Utihty Group (SQUG), which developed analysis shows that, based on previous experience the walkdown procedures for USI A 46. limally,the with the evaluation of severe accidents imtiated by event trees and fault trees developed for the inter-external events, the IPl!!ili has the potenJa! to nal event IPli, which was initiated ab mt 18 rnonths identify items that, if coriceted, would result in before the IPhl!!!, wdl be available for use in the substantial incretses in safety, and that the cost of IPEEE Therefore, the staff believes that a large corrections, mcludmg the cost of the litiEh,, would pool of talent is available (as evidenced by the num-be comnyensurate with averted potential risk to se-her [approximat ely 25] orconsultants and consulting vere accidents.
firms ts presented at t he Ipi: Ell wor kshop). nnd that withm the 3 year period to perform the IPEE!!,
2)- Cost Estun :tes and Rewurce Requurment.t the over-licensees can develop or obtam the necessa,y exper-all cost of the IPEEE was underestimated tise to conduct the IPElih flowever, as with the D-1 NIiltl!G-1407
[
Appendix D internal events IPli, the staff will consider extending mendation by limiting the assessment o the effects r
this date on a case-by-case basis.
to ensite fhioding and roof ponding.
N Tire Evaluation
- 4) Licensee Response Time: The initial response time of 60 days to identify the methodologies for completing hl) Emeditious NRC review of the NUMARClEPHIalter-II'EEE is too short.
natepre-evaluation methodology was requested SR: The staf f believes that it is appropriate to extend th SR: The staff had previously committed to review licensee's imtial response time from 60 days to 180 an alternate methodology being developed by days to allow for some essential preparatory work NUh1 ARC and I!PRI for evaluating fires. Cur-(t.c., the processing of bids, completion of the devel-rently, the staff is reviewing a NUh1 ARC document opment of the alternative fire evaluation methodol-describing the methodology and is waiting for more ogy by the Nuclear hianagement and Resource g.
EPRI and NUh1 ARC on the re-Council (NUh1 ARC) and lil RI, and the staff s re-sults of demonstration applications of this method-view and assess cent of this methodology). One hun-dp two nuclear plants (the staff anticipates this dred eighty days was selected in consideration of the M
M 199 Wllowmg receipt of this current schedule for NUh1 ARC /lil RI to complete information, the staff plans to compat:its review, the development and verification of their alternate neluding discussions with the AC RS and the fire evaluation methodology, and for 1he subsequent NUN 1 ARC. As stated in Response 4, these activ ties staff review.These activities are expected to be com-are not npected to be completed until July l >91.
pleted in July 1991 (see Response 6.1).
lherefore, the staff plans to cespond separately to the NUh1 ARC,liPRI alternate methodohigy, so as
- 5) Inclusion ofIssues: The m.clusion of (a) lightning (b) not to delay issuing the IPlilill generic letter and volcanic activitics, and (c) Genene issue (Gl) 103l.
guidance document. llowever, as discussed in Re-7Designfor Probab!c Maximum Precipitation (PMP) sponse 4, additional time was given to licensees, so m the JPEEE was questioned.
that they have the results of the staff review before they commit to a fire evaluation methodology in SR: 1.icensees need to coofirm that lightning or volcanic their IPliEli submittal plans.
activity is not a dominant contributor to severe-accident risk at their nuclear power plant sites.The 6 2) GI-5" *Effccts of nre Protection System Actuation on determination should be based on plant specific ex-Safety-Related Equipment:" The relationship of this GI perience. The concern related to lightning (as to the IPEEE was questioned.
pointed out by the ACRS)is thatlightning strikca in addition to causing loss-of-offsite power, may dam-SR: 'lhe effect of fire suppressants on safety equipment age instrumentation and control systems. If this had is one of the safety issues identified in the Fire Risk happened before at a site, the staff would expect the Scoping Study, (NURiiG/CR-50SS) and may be a IPlilili for that specific plant to address this con-si;;nificant conoilwor to risk. Accordingly, it was cern. In regard to volcanic activity, only two sites raised as a geocric safety issue and was also included would be af fected. In either case, a simple discussica in the IPlilill The staff expects that if a licensee will be sufficient far those plants not affected by discovers a significant vulnerability in this area these events. For plants that may be affected, a through the IPliliE, the Fcensee would address the success screening process, such as the one described problem and not awtut the GI-57 resolution. During in this report can be used.
the walkdown, licensees can collect relevant infor-mation on whether actea ed fire protection systems With regard to GI-103, the NRC acknowledged the would spray safety-rela'.ed equipment. and can insti-importa.nce of this new Ph1P criterion in Generic tute some protective measures to prevent the safety Letter 89-22 by tequiring that future plants be de-equipment from being sprayed by fire suppressants, signed against it (i.e., design basis). For existing
'lhe additional effort t collect this information dur-plants, the NRC recommended that licensees re-ing the walkdown should not be a burden. Ilowever, view the material contained in GL 89-22 to deter-the formal resolution of G -57 does not have to be a mine whether they believe additional action is nec-part of the IPf!Eli essarj; however, licensees need 1,ot change their design bases. For the 1PEEli, the staff believes that
- 7) Seismic Events: Trcatment ofscumic events needs clari-assessing the potential for a Ph1P to cause a severe pration.
accident is justified, since the National Weather Service PMP data are being applied to future de-7.1) The need to use both the LLNL and EPRI scismic signs. Ilowever, the staff has clarified it's recom-hazard curves was questioned.
NURiiG-1407 D -2
Appendix D Sit: The staff considered the difference between the c 7.4) 7hc scope rf the tclay chatter evaluation was ques-two curves in specifying the cahancements for 'hc tioned seismic margins methods and the seismic PR A.
!!cwever, based on the available information to SR: Detailed rela) chatter studies conducted at the date, the staff is unable to dispute the merit of cither Ilatch and Diablo Canyon plants showed that con-curve and sonsiders both of them to be valid. The siderable resources were expcnded to perform the staff also believes that the added cost of using two relay chatter review; aad using existmg procedures, curves should not be burdensome, based on what operators could solm the ; clay chatter problems was spent on two NURIIG-1150 plants A utiht) identified at these plants. ' dowever, the staf f cJ may choose to use only a single curve, provided the industry consultants ar e conserned that such conclu-higher one of the two curves is chosen.
sions cannot be consiJered generic without some additional plant reviews.
7.2) The use of a site specific scistnic curve (in licu of the Therefore, as decned above, the staff is recom-LLNL and EPRI curves) was sugested-mending that the o ' g bin be subdivided mio two categories, a fullwge and a focused scope cate-gory. I or plants in 1he full-scope cateeory, licensees SR: The l.l.NI. and 11PRi curves are " site-specific" scis' wiu have to evaluete the relay chatter consistent true curves. Eac,n used its standard methodology and w th the approach discussed in llPRI NP-6041 or its uniform mterpretation of data bases to calculate the equivalent. l'or reduced. scope review, the imple-seismic ha/ards for power plant sites in the liastern mentation of USI A ~46 program wdl provide infor-United States. l'he use of other site-specific seismic mation for satisfymg the IPlilili provismns. Note hazard curves is an acceptable option subject to re-that licensees can perform the IPlilil! with consider.
~
view and acceptance by the staff. llowever, the staff able efficiencv,if they take advantage of the lessons believes that the cost associated with the develop-learned from'the llatch and Diablo Canyon relay ment of new site-specific seismic ha/ard cunes chatter evaluations. llor plants in the focused-scope ceuld be very high and time consummg.
category, a lower level of effort is recommended; this would entail hioking for and addressing low-
'" P"
'Y ""
- 73) The use of plant desu;n bases in the scismic binmng process was sumested.
SR: ne staff investigated the potential of using the scis-mic design basis as a parameter for making the initial This section paraphrases, summarizes. and categori/es binning assign,nents. llecause there was insufficient into subject areas, questions and comments either raised technical basis for its use, it was not used initially, at the workshop or receised by the staff (16 parties sub-Ilowever, when considered in conjunction with the mitted written comments, sce sources of comments).The seHnic ha/ard, the use nf the seismic design bases NRC staff response is also provided. Table 1).1 contams a provided a means for reducing the scope of the 0.3p listing cf the subject areas dtscussed in this section. The
.nargins review. Specifically, plant sites in the 0.3g workshop transcr pt an.1 a copy of the comments received bin we? e assigned to a full-scope oc a newly defined are available in the NRC Puhhc Document Room, focused-scope category. The full-scope category is essentially the review specifica in the dratt generic Table D-1 Categorization of Question and Comments
!ctter and guidance document that were distributed for public comment, whereas the focused-scope re-1.
IPlilili, IPli, CPI, and Accident Management view represents a reduced scope review. The pri-mary purpose of this further subdivision is to reduce 2.
llackfit analysis the level of review effort, mainly in the relav chatter
~
3.
Cost estintates and resource requirements area, for plants with a lower hazard or higher design basis Plants with a relatisely higher hazard and 4.
Schedule and response time lower seismic desigr, basis should perform a more detailed study than the other plants. (Grouping of 5.
t hgh winds, floods, and transportation and plants into various categories is discussed in Section nearby facihty accidents 3 and Appendix A cf this report.) Of the 56 sites
(>.
Internal fires originally assigned to the 0.3g category, / remam m the full-scope category, the remainder moved to the 7.
Setsmic events focused-scope eategory.
D-3 NURI G-la07
~.m
- i Appendix D 1.
IPEEF, IPE, CPI, and Arcident Management and will be addressed in a future generic letter on accident management.
1.1 Ilow do the IPE, IPEEli, and the accident manage-ment all relate to the design bases of the plant in 1.3 The Severe Accident Policy Statement is silent on terms of identified plant vulnerabilhics, improve-external events, thus, there is no need to do much ments, and potential increase in risk? Ilow do the beyond what is already done for internal event IPEs.
plant operators make the day to-day decis n when (Ref. D.16, p. 53)
PRA insights and Tech Specs are in con
,? (Ref.
' P' #)
SR: The Commission Policy Statement identified the need to seek vulnerabilities systematically at all op-erating plants. It didn't distinguish between internal SR: The thrust of the whole severe accident program is
- and external events, llowever, PRA studies have to reduce the likelihood of severe ccidents and shown that external events in particular scismic and their consequences. As such, they are looking at f re, are principal contributors to overall risk. Ac-accident scenanos beyond the traditional design ba-cordingly, the staff recommended to the Commis-sts envelope If a vulnerability is identified and a fix is sion in SECY-86-162, dated May 22,1986, that proposed,it is important to make sure that the pro-external events be included in implementation of posed fix has no adverse effect on the plant. If a the Severe Accident Policy Statement.
licensee makes modifications to the plant, which resulted in a change of the plant design basis, then 1.4 IPEEE, basically, is an evaluation looking at a point that must be documented, tracked, and accounted in time, a snapshot in time, is there an intent of for, in accordance with the provisions of 10 CFR keeping it living? (Ref. D.16, p. 78) 50.59. If the PR A identifies a conflict with the Tech-nical Specifications or operating procedures, the li-SR: The staffis treating the IPEEE as a one time evalu-censee should examine the reason for the conflict. It ation. There is no requirement to keep it living.
is important that the licensee not make plant or licwever, based on utihties' experience, once one procedural modifications without understanding the has gone through the process and invested the re.
basis behind the PRA conclusions. For exampic, if sources and constructed the PRA or equivalent, it the underlying model in the PRA was developed would be usefulla keep it up to date. Ilowever,it is with simnlified assumptions and modeling tech-really up to tN utilitj whether or not to keep it niques,it might be prudent to perform a more realis, living.
tic evaluation to assure that the rnodeling assump-tions have not biased the results in an inappropriate 1.5 is it correct to assume that there is no requirement
- manner, for the pedigree of the program, that it can be basi.
cally be a study without a O A type of pedigree? (Ref.
1.2 After the utilities had factored vulnerabilities identi-D.16, p. 80) fied through IPE and IPEEE in their emergency operating procedures (EOPs) and the ' lech Specs SR There is no requirement for an Appendix H-type that support the EOPs, what else would the utilities QA program to check the IPE or IPliliE. The licen-be reqmred to do for the accident management?
sees should perform an ongoing internal quality as-surance effort to ensure that the results of the IPE Any srecific example of a guideline that the staff
[
would be putting forth as part of accident manage-and IPEi!E are factual and represent the as-built, ment? (Ref. D.16, p. 46) as-operated plant. Typically, licensees will define
" pinch points"in their performance of a risk assess.
}
ment to stop and assess the progress and quality of SR: It is important to recognize that accident manage-their effort to date. As in the mternal event IPE,'the j.
ment responses are not just limited to emergency staff is asking for a peer review as part of the lPEEE.
l procedure guidelines (EPGs) or EOPs. Technical The peer review provides a type of OA function.
- support and the kind of guidance and personnel training that are needed should be part of accident 1.6 In the area of other caternal events that are not
. management. lPE and IPl!El! results should be con-included in the IPEEE, does the utility need to de-L sidered as an information tource which provides in-velop a hazard curve associated with that particular puts to training programs and to the development of event? (Ref. D.16, p. 81) emergency preparedness exercises. Accident man-agement takes that information and uses it in the SR: No. The staff is not asking the utilities to justify not planning, train;ng, exercises, and to establish the including thosc events. llowever if a utility knowsof communication and the feedback mechanism at the a particular hazard that is greater th-an what the staff utilities. Specific guidance and examples are still has considered in the generic study. the utility being developed by industry (NUM ARC) and N RC, should consider including it in the IPEEE.
NURl!G-1407 lu4
i.
l l;
Appendix D 1
1.7 is sabotage included in the IPliEE? (Ref. D.16, p.
dominant components as well as assess the findings 82) against the reporting criteria.
SRf Sabotage is not included.
1.12 In the draft generic letter, how will this information be used is rather generalit seems that it is up to the utilities to determine what to do with the results, 1.8 l'lorida Power & l.ight has a policy that upon ap-
. proaching hurricanes in S.171orida, the unit will be how they see fit; make the change or justify not shut down and the unit will go to a Mode 3 or a Mode making the change. Ilowever, m the generic letter it 4 in advance of the hurricane, does that'mean the also says that the NRC will assess whether the con-'
hurricane need not be considere<l in the IPEEll?
clusions the h,censee draws from the IPl!EE regard.
(Ref. D.16, p. 82) ina changes to the plant systems or components are -
adequate. May be you should include some kind of 1 SR: In general,- the shutdown mode is outside of the n example of either a positive or a negative finding IPHEE scope. Ilowever, in a case like this, the licen-that you,ve made that can provide some sort of gu de
. see should rnake sure that the plant can be shutdown as to what is adequate and what is net ndequate, and maintained in a safe shutdown condition (USl (Ref. DJ6, p. 92)
A-45 requirementr). In other words, the combined sri If the staff disagrees with what a licensec did, the frequency of the hurricane and failure to shutdown -
options that are open to the staff are contained in and to mamtam the plant in a safe shutdown condi-the regulations. 'lhe staff may icquest additional tion needs to be assessed.
nformation via questions pertaining to the submit-tal, or may impose plant modifications via the backfit 1.9 17 ced with large uncertainties, how are risk, human rule.
a reliability, operational reliability, maintenance reli.
ability, etc. associated with external events to be 1.13 Vulnerabilities need to be tied to core damage risk.-
quantified? (Ref, D.16, p. 84)
If you can't tie a vulnerability to a core damage risk, then it's not a vulnerability. 'Ihe Severe Accident SRL The staff recognizes that there are significant uncer-Policy Statement sa, s that if you identified a vulner-taintics in quantifying risks associated with both in-ability, then what you do next is to see if by sixing that ternal and external events. Thus, the staff has deem-vulnerabilityyou can reduce the risk of core damage.
phasized the bottomt!ne numbers in both the IPH, So it seems that you have to have a quantitative and more importantly, for the IPERE.
number for core damage before and afteryou fix the identified vulnerability. (Ref. D.16, p. 94) 1.10 When should the improvements resulting from the IPE procas be carried out, right after the IPE or There are several stages in the process: (1) To iden-wait until the IPEEE is completed? (Ref. D.16, p.
tify a vulnerability; (2)To identify fixes for that vul-87) nerabilityt (3) To actermir.c if it's substantial; and
- (4)To determine if it's cost-beneficial. The cost-SR: The stalf has looked at the interaction between the beneficial side of it has to be determined by the
-iaternal and external events. 'The staff has con.
reduction in risk to the population outside the olant, cluded that it is unlikely that the cost effective im.
which, in effect, requires a containment failure.
provements based on internal IPEs would have a
. negative impact on safety for external events (see So we have'to identify the vulnerability list and then Section 6.3.1.2). The Generic 1.etter states that it follow these steIis, lic.ause the IPEEE implements expects each licensee to move expeditiously to cor, severe accident policy, we all have to make sure that rect any vulnerabilities that it deterrnines warrants we're implementmg it in the way miendcd.
correction.
SR* The process described is what the staff would go through if we chose to backfit a plant. It's not neces-1 11 Since the purpose of the IPHEH is to gain a quahta-sarily what the utilities might do in their plant. We 1
ttve underst mdmg of core damage frequency, not leave it up to the utilities to decide what process they quantitative, is it good enough for scismic PR As to would itse and how they would defiac a vulnerability.
Just report scismic risk in terms of high, medium, or low, instead of putting in numbers? (Ref. D.16, p.
89) 1.14 Clarify the IPl.EE Objectives. (Ref. D.1)
SR:- No. Core damage frequencies are requested from SR: The purpose of the IPEllE is to gain a qualitative the scismic PRA so as to obtain insights and the understanding of core damage frequency, not quan-relative ranking of the. accident sequences, and
.titative. Section 1.2 provides the discussion of this D-5 NUREG-1407
1 Appendix D aspect and also points out that sorne methods have staff can seek to have the fix implemented,if neces-been developed for evaluating external cvent harard sary by Order, if the staff determines that a fix de-and identifymg vulnerabilities that do not produce sired by the staff is beyond the design bases or Regu-j estimates of damage frequency Scismic margins lations of the Commission, the staff must prepare a method is cited as a specific exampic. 'Ihe objectives backfit analysis and submit it to the Commission. A were reworded to emphasize this point.
backfit rgnalysis does not have to include a probabilis.
tic risk evaluation, but can instead be presented
.2.
Itackfit Analysis relying primarily on engineering judgment.
2.1 'lhe staff has stated that the Office of the General 2.5 'Ihe Severe Accident Policy Statement doesn't give Counsel is kioking into whether the request for an you any probabilistic numbers that you can really IPfilill should be under 50.54(f) or the backfit rule.
work with. Ilowever, in June 1990, the Commission if the Office of the General Counsel does determine directed the stiff to consider 10A core melt fre-that the backfit rule applies, would it be correct to quency as safe enough as in the Safety Goal Policy assume that this supplement to the generic letter Statement.
will be issued similar to Generic Letter 89-16, the Ilardened Vent of the Mark Ps, where doing the We propose to link "how safe is safe enough" with action or performing t he IPilf!I! would be voluntary, the Safety Goal Policy Statement bout undue risk and for those utilities who do not volunteer to per.
in severe accident with adequate protection in the form it, plant. specific backfit analyses would be per-backfit rule. We could make a very good case that formed using plant-specific values and criteria in plants currently are safe enough especially if we use terms of our own resources required and the scope that number on the individual plant level. (Ref.
of the analyses tha' an individual plant would have to D.16, p.156) perform? (Refs. D.13 and D.16, pp. 90-91) gg,lhe Commission Safety Goalis not just 10A peryear SR: The staff has determined that a backfit analysis is not f r core melt. The Commission safety goals ma the needed for the IPl!!!!i.
quantitative health objectives for fatalities. Ihc staff had recommended that subsidiary objectives (e.g.,
- 2.2 ~ -If the General Counsel does determine that this can 10A CDF) be established as a way to implement the go ahead under a 50.54(f) request for information, S fety Goals in a practical manner, will the utilities have an opportunity to kick at that for themselves and perhaps appeal that decision or In addition, the Safety Goals are not to be used to judge individual plants. They are to be used tojudge do their own analyses, whether we feel that backfit the acceptability of the NRC regulations.
rule actually applies or not? (Ref. D.16, p. 92) 2.6 There is a concern about the closuie process, par-SR: All utilities will be required to respond to the IPiilili ticularly in relation to the fact that both the llPRI 50.54(f) request.
and the Livermore hazard curves will be used in a 1
seismic PRA.
2.3 Without some kind of numerical or specific figure of merit, how can one reaUy say that it's cost. effective What's going to happen is ultimately, if a seismic to implement one type of fix over another, or even to vulnerability would be expensive to repair,you are make a fix at all? (Ref. D 16, p. 92) forced into doing some probabilistic type cost bene-fit analysis. Ilven though you may have done a scis-SR: The staff traditionally uses a cost benefit figure cf mie margins assessment to identify that vulnerabil-merit of $1000/ man. rem in imposing new require-ity, I see no way that you are able to avoid 'not ments.
reverting to using the Livermore hazard curves in 1
your ultimate decision-making process;I think thisis That does not constrau a licensee to_ use the same the time that there ought to be some effort to re-cost benefit as their criterion for what to fix or what
. solve t ae difference between these two curves. (Ref.
~
not to fix. Any criterion that the licensee chooses to D.16, pp.162-164) use should be justified.
SR: When we do backfit analysis and regulatory analysis, 2.4 What are the criteria that NRC will use in determin-it's not uncommon to have areas of great uncer-ing what to fix? (Ref. D.16, p.154) tainty, even as large as the difference between these two curves. The I!PRI and 1.LNL hazard curves dis.
SR: If the staff determines that a fix is required to bring play a level of that uncertainty. Some backfit evalu-the plant into conformance with the regulations, no ations are evaluated without having core damage /
l cost. benefit or other analysis is required, and the risk salues, but rather rely on engineeringjudgment.
NURIiG-1407 D-6
~
4)
. qfe ~As
% @o %"39:,
IMAGE EVALUATION
^
j///A-
((f
- ghD.
e
//
O g'ce',
TEST TARGET (MT-3) g
'(
///
f4+
%<e<>
1.0 f:e en a m p422:22 b gg m llf 2.0 1:
l.l
[
im li L8 e m f.25 l4
"\\
l.6
\\
=
4 150mm 4
6" as
- 4%g k
h
.Oh 5
'c')
s n ks,o,
,p
//,,,
sy g, _ ;-
,jt,s o
,-3,p j
.4&
o
.p,,, -
o Ohb
..t
-._,I _ ;
w
Appeodix D ne Commission makes use of all the information 5.
Iligh Winds, Flood?, and Transportation ar.a available in making its decisions. HCLPF s may alto Nearby Facility Accidents be an important consideration in making decisions regarding backfit.
5.1 The flooding criterion screens at a frequency of 10 5 peryear, but the wind screens at 10 8 peryear, why is If a licensee prefers to use a single curve m. the there an inconsistency between flooding and wind?
seismic review, that is also acceptable provided the (Ref. D.16, pp.144 & 150) higher of the two curves i3 chosen. See resnonse 7.8.
SR: The screening criteria are essentially a reporting 3.
Cost Est.: mates and Resource Requ.irements criteria, which are consistent for all external events and internal events.
3.1 The st4f's rtource estimate for IPEEE ts low.
(Refs. DS, D.13, and D.16, p. 42)
We're not using 10 5 per year flood frequency alone r screen mt floods. Based on a number of flood SR: The staff believes that the cost estimate forIPEEEs studies, the j,udgment was made that the probable v as developed conservadvely. Obviously, there are maxm.um flo d has a 10 5/yr. or less frequency. But uncettainties in the costs because there are uncer-that alone does not screen out the flood concern.
tainties in how a licensee will choose to analyze Judgment was made that the conditional failure external events. The staff has used previous utility prol ability for a plant designed against the probable 7
and NRC experience in conducting external event m x1 mum fl od is somewhere around 10s, so that analysis in arriving at the cost estimate and has at-the combined probability is 10 8 or less per year. A tempted to clarify the scope of the IPEEE where stm ! r approach was used for the wind, where the questions on scope were raised. However, whe':
design basis wind was usually selected to have a questions still exist, licensees should come in and probability of less than 10 6 per year.
talk tc the staff to make sure that they have a clear understanding of the IPEEE process.
.'.2 The item 2 on the flowchart for screening external 3.2 Six personyears was the estimate for the 1PEEE cost events, identifying significant changes, does that re-fer ta the hazard at the site or does that refer to the or manpower resource requirement, what is the ba.
sis? (Ref. D.16, p.155) design of a plant? (Ref. D.16, p.147)
SR: The basis is discussed in Appendix 5 to Generic SR: It refers to the hazard on the site and the land use of Letter 88-20, Supplement 4 Basically what the staff the genent vicinity, for example, a new airport built did was to estimate the overall effort required to nearby.
perform the IPEEE and to use the cost spent for NUREG-1150 plants and the Hatch seismic review 5.3 For item 3, review plant apinst the current standard to estimate the IPEEE cost. We also received esti-review plan critena, why do older plants need to do mates from a few PRA companies. We do recognize that? Obviously, some older plants were not de.
that there are some costs that licensees will have to signed us, current methods and codes. (Ref. D.16, spend over and above just what it will cost the PRA p,347) company. We do believe that the staff's estimates are reason 21e. K. Fleming of PLG Inc., whose firm SR: Item 3 is a screening criterion only. If you know has done most industry-sponsored PRAs, and D.
already that your plant does not meet the 1975 Stan-Dube of NE Utilities provided estimates that are dard Keview Plan (SRP) criteria, you should move either less than or comparable to the staff's esti-on to the next step in the evaluation process.
mates. Tal,lc D.2 pmvides a compartson of the staff's and the industry resource estimates.
5.4 The inclusion of lightning and volcanic activity is 4.
Schedule and Response Time questioned. (Refs. D.1,8, & 11) 4.1 What is the schedule for IPEEE, staff review, and SR: The confirmation that lightning or volcanic activity is fixes? (Ref. D.16, p. 50) not a significant contributor to severe accidents at the nuclear p)wer plant, should be assessed to the SR: The staff intends to complete closure of severe acci-satisfaction of the utilities. A relatively simple dis-dent issues in 1995. Accordingly, r three year com -
cussion by the licensee to state why a licensee be-pietion period for the IPEEE is specJied so as to lieves these issues to be unimportant. will be suffi-give the staff time to review the submittals.
cient for these issues for most plants.
D-7 NURl!G-1407
Appendix D Table D.2 IPETE Resource Estimates (Manhours) llatchi Future Pl.G Task NUMARC2 Actual Tod a>
- SMM+
SPRM 1.
Select SME 0
200 200 0-200 2.
Select success paths 600 850 700 600-950 3.
Mod. scismic bld'g model 600 430
- 30 0--950 4.
Perform SSI & Dev. FRS 2000 1020 1020 0-2000 5.
Soil lig. evaluation 1000 500 500 0-1000 6.
Relay chatter evaltiation 2000 2800 1700 1400-2100 7.
Pre-walkdown 160 160 100-200 8.
Walkdown preparation 200 200 100-300 9.
Seismic walkdown 900 900 400 600-1400
- 10. SMM of outliers 1800 2680 2000 1200-2700
- 11. Report & documentation 700 1320 1000 9S0-1200
- 12. Walkdown travel expense 500
- 13. Containment review isolation.
t>ypass, etc 800 Long term mitigation 400
- 14. Misc. cost (startup, plant support, training peer resiew.
NRC interaction) 1200 Y tal Seismic IPEEE 13000 11060
% 10 4980 n000 1700-2700 Surry Peach litm PRA Pl G Seismic event 1400 1320 1500-3000 1100-1500 Internal fires 350 450 750 900-1500 High winds 0-200 External floods 150 250 0-300 Others 0-650 QA 450 450 200-500 Plant support 1200 1300 1300-2500 1000-2000 Soerces:
" + 11 P. Moore of Southern Company Senices. lac.
8NUM ARC cost estimate of foliacog SNt%1 >f Na 13.199n
'Pl.G letter to NRC. dated Oct. 9.1%0 5.5 The requirement for assessmg GI 103. Design for new PMP cnterun m penenc letter 69-22 dated Probable Maximum Precipitation (PMP) and lack of Oct. 19. 1950, by requinng that futur, plants he specific guid mce in the generie letter are ques.
designed ap.iinst th:s new entenon. l'or existmg tioned. (Refs. D 1. 8, 11. & 13) plants, NRC recommended that the bcensees ie-view the matenal contamed in (il. 89-22 to deter-SR: IPl!Eli iruposes no requiremer.ts. With regard to mine w hcow t% behese additional aettoa en their the PMP, NRC acknow! edged the importance of the NU RiiG-140 7 D -S
e Appendix D part i.; necessary. However, this previous review was planned as part of the seismic walk-downs ihat directed toward assessing the adequacy of the design would specifically h>ok for the scismit 5duced fire basis, whereas the IPEliE is directed toward severe vulnerability issues.'lhe idea is to first idenufy those accidents. Therefore, the staff is requesting that areas that could be vulnerable so that they can be PMP be looked at to assess the effects on plants in brought into focus during the walkdown.
terms of onsite flooding and roof ponding to deter-mine whether ihat would lead to a severe accident.
For example, tf a plant didn't have its diesel fuel tank his is consistent with the staff's request that licen-strapped down properly one could postulate a large sees confirm that nn plant unique external events fuel source for fire as a result of a seismic esent.
with the potential to initiate severe accidents have Other similar seismic / fire interactions were summa-been excluded from the IPEEE, as stated in Section rized in Section 7 of NUREG/CR-50SS.
3 of the Generic Letter 88-20, Supplement 4. The 6.3 If the utilities had already assessed the safc shut-general procedurc ean be found in section D. 2.4 0f SRP and section 11.4 of NUREG/CR-2300. The down capability, are spot checks acceptable for the staff believes that this information is readily avail, cable routing verification? (Ref. D.16, p.130) able per GL 89-22 recommenJation.
SR: Licensees should rely on previous assessments for IPEEE informatton, provided the information is up-6.
Internal Fires to-date. The licensee does not have to perform any 6.1 The generic letter doesn't state whether the Fire design verification, or retrace cablesif that had been Vulnerability Evaluation (FIVE) methodology and donc previously.
the associated database, being develo[v.:d by 6.4 Will the fire database be updated and when willit be NUM ARC and EPRI, are acceptabic or not, either available? (Ref. D.16. p.132) for meeting the IPEEE requirements or for satisfy-ing the issues in the Sand:a fire risk scoping study.
SR: EPRI is updating the fire database through 1988. It is expected to be available to staff 30metime in the Are you plarming to put that in the final genen.e December-Januarv timeframe.
letter and guidance document? Since we don't know whether the FIVE methodology is acceptable or not, 6.5 How is safety svstem separation to be assessed in the we can't really make our decision on what method to fire analysis? (Ref. D.16, p.132) use and respond in the 60-day timeframe as re-quested in the generic letter. (Ref. D.16, p. 96)
SR: Separanon should be modeled as it exists and the fuel sources as they exist in order to understand, SR: De development of FIVE has been acknowledged using the codes (propagation analysts), the effects of in the generic letter supplement. However, the staff fire on redundant trains of equipment.
is unable at this time to endorse the FIVE method for the use in IPEEE, because NUM ARC and EPRI 6.6 in treating of transient combustibles for fire, would have not completed its validation and documenta-it be sufficient in the IPEEE fire analysis to state tion. The staff intends to review the NUMARC/
that administrative control for transient combus-EPRI met hodo:ogy and, if it is acceptable, endorse it tibles takes care of this or would additional quantita-as an acceptabic way to deal with fires under the tivc analysis be needed? (Ref. D.16, p.148)
IPEEE. However, final review will not be completed in time to incorporate a final position on FIVE in'o SR: Transient combustibles cannot be ignored. If they're the generic letter, therefore, the staff will address procedurally controlled, a fire protection engineer the acceptability of the FIVE methodologyin a upa.
should be involved in doing the examination, so that rate document.
a determination can be made as to whether procc-dural control will really limit significant transient With regara to the 60. day init Ed sea n % tnat combustibles m a given area.
has been extended to 180 days after the issuance of the final generic letter, primanly io allow time for ba N he requirement stated in Sec. 4.3 Item 5," provide completion of the FIVE methodology.
documentation in each case where the plant-specific data used is less conservative than the approved data 6.2 What are the procedurally directed walk. downs in base" is a disincentive to use plant specific fire terms of addressing scismic-fire interaction. Do initiation data. The IPEEE results will capture ge.
they pertain to walk downs for the fire or walk-nenc vulnerabdities instead of plant specific fire downs for the seismie IPEEE. (Ref. D.16, p.130) vulnerabihties by preferring generie data (Ref. D.8)
SR: The procedurally directed walk-downs associated SR: In most cases. plant-specific data are rather limited.
with internal fires vulnerability evaluation can be
't he use of generic data. which usually have a much D-9 N U REG-1407
l Appendix D broader base, will provide valuable insights about 6.10 'Ihe inclusion of GI-57, " Effects of Fire Protection what could happen at a plant.The awareness of this Systen Actuation on Safety-Related Equipment,"
kind information is very important to the plant oper-was questioned. (Ref. D.1) ating staff.This is consistent with the purpose of the SR: The effect of fire suppressants on safety equipment j
Commission's Severe Accident Policy, "to under-stand the most likely severe accident sequences that is one of the safety issues identtfied in the hre Risk could occur at its plant."
Scoping Study, NUREG/CR-50S8. Relevant mfor-l mation on whether the actuated fire protection sys-6.8 Water as a suppression agent may cause potential tems would spray safety related equipment or not damage to safe shutdown compon:nts. What about and some protective measures to prevent the safety CO and Halon? The Sandia study states that there equipment from being sprayed by fire suppressants, 2
are no data available to quantify damage from these can be ccHected during the walkdown. The addi-sources. Is a simple statement or identification as tional effort to collect this information during the
- 1 potential damage adequate? (Ref. D.16, p.149)
IPEEE walkdown should be minin al. We also want to point out that the resolution of this issue does not SR: In the majority of instances seen to date m power have to be a part of the IPEEE.
p! ants, water caused the damage.,Iherefore, the staff believes water damage is the most probable. If 7
Seismic Events there is an area where suppression damage from 7.1 What is the j.ustification for includmg Earthquakes CO or Halon is likely, however, one should not 2
ignore it.
m WEEE? (Ref. D.7) 5.9 A major emphasis of the IPE/IPEEE is to have li-SR: B sed up n the examination of NRC's and indus-censee staffinvolved to help ensure the most bene, tty s pl nt-specific probabilistic risk assessments l
fit. NUREG-1407 states that should involve engi, (PR As), the mean core damage frequencies at some neering judgt-ents of the fire safety experts, Does it plants could be relatively high, m the range of 1 E-.1 t IE-6 per year (Table D.3). Many cost-effective mean a fire protection engineer is adequate? (Ref.
D.16, p.166'1 improvements that would reduce the potential risk were identified from these PR As; many were imple-1 SR: The staff would consider a fire protection engineer mented at plants as discussed in Appendix 5 of Sup-of a utility to be a fire safety ex;3ert.
piement 4 to Generic Letter 88-20. This findmg is Table D.3 Summary of PRA Results of Core Damage Frequency (IE-5)
Total Total Int'l Extn't Ifigh Light-Plant Total Int'l -
External Seismic Fire Flood Floods Winds ning PWR i
Pt. Beach 31.3 13.9 17.4 6.1 3.3 7.7 0.4 0.006 j
Turkey Pt.
23.6 7.1 16.51
.7 7.5 4.6 2.4 0.26 St. Lucie 7.44 1.4 6.04 1.3 4A 0.32 0.02 i
ANO1 17.9 8.8 9.15 7.3 0.58 0.72 0.53 0.02 IP2 43.5 6.0 37.5 14.0 19.2 4.3 IP3 15.7 9.0 6.7 0.31 6.3 0.13 l
Zion 34-40 34.2 0.1-6
<0.1-6 MS3 15-23 14.7 0.8-8Est i
Oconce 15-28
-7.4 8-21 6.0 1.0 0
L3 l
10.0 (NRC) 2.5 (NRC) 2.3 (NRC) -
BWR Quad City 19.7 9.9 9.8 S.3 1.3 0.01 0.01 0.2 l
Cooper 43.7 28.9 14.8 8.1 L1 5.
0.4 0.2 l
I Limerick 9.2 8.4 0.8 0.5 0.3 Shoreham 7.4 5.4 2.0 2.0 (NRC) -
IP2/IP3 -Indian Pomt 2/Indan Point 3 hfS3-Afdistone 3 l
1 NUREG-1407 D-10
Appendix D consistent with the statement in the Commission's SR: References to probabihties are now climinated from Severe Accident Policy that systematic examinations the objectives. The IPlilili objective is to gain a are beneficial in identifying plant-specific vulner-quahtauve understanding of the overall hkchhood abihties to severe accidents that could be itxed with of core damage and fission product releases. This is low-cost improvements. References D.2, D.4, D.6, different from that of the IPli, wheie quantitative and D.11 all acknewledged that systematic examina-understancing is emphasized. In Section 1.2 of this tions are valuable tools for gaining insights into the report, the staff has acknowledged specificallj that plant operation and identifying cost ef fective plant "some methods has e been developed for evaluating improvements.
external hazards and identifying vulnerabihties that do not produce estimates of core damage frequency.
Another indication that carthquakes can be impor-
.Thus, objective 3 above would be addressed only tant risk contributors can be seen from the carth-indirectly for some methods to be used in the quake experience on foreign nuclear power plants.
IPl!Eli" It should be noted that the seismic margin 1 or example, (1) On Apnl 22,14S7, I ukushima methods were derived from scismic PRAs.The ob-units 1,3, and 5 in Japm, tripped as a result of an jectis es have been reworded to emphasi/c this point.
carthquake with a magnitude of 64; and (2) On March 4,1977, one steam generator at the Kozloduy Also, see the staff response to items 1.11 and 1.14.
nuclear power plant in Bulgaria was displaced by 5 7.5 Scope and Depth of Assessrnent: "Ihe scope of inches.This earthquake experience caused a ma'jor naluations requested for seismie lPl. is morc exten.
overhaul in the seismic design of latei Russtan sive than needed to satisfy the Severe Accident Pol-plants. It also resulted in a major backfit at the cy Statement." (Ref. D.1)
Armenia unit 2 in Soviet Russia.
SR: The staff, based on a NUMARC recommendation.
Therefore, based on risk considerations and the po-has &fineJ three review categories with varying lev-tential for identifying cost-effective improvements, els of effort. 'this approach leads to an overall reduc-the NRC concludes that seismic even,s should be tion in the effort tocam out the exammation. Addi-included in the IPEEE.
tional details are discus' sed in response to comment 7.2 The use of both Seismie llafard Cun es (ILNI.and EPRI Seismie llazard Curves) was questioned 7.6 Seismically mduced floods are men toned for the (Refs. D.1,6,7,11, & 20) first ome in draft NURI G-1407, Secuan 6.3.2 and not in Generie 1 etter 8S-20 Supplement 4. We un-SR: There is not enough ea,thquake data at this time to derstand that the scope of review for seismically determine precisely the vahdity of a smgle cune. In induced external fhedmg is limited to a reuew of other words, there is no way to conclustvely validate external sources of water (e.g, tanks, upstream or dispute either curve. Therefore, both the i LNI.
dams, or other significant sinictures) and not inter-and EPRI seismie hazard curves are recommended clearly stated in Genene 1.ctter 8'g. This should be nal water sources such as pipin for use in the seismic PR As This is consistent with S-20, Su pplement the recommendation of the NUREG-1150 peer r -
a, in ' order to asoid possible confusion in future view group. An acceptable alternatise for hcensees mterpretations. (Ref. D.1) choosmg to perform only one analysis is to use the higher cune. Also see response 7.8.
SR: the scope of the sci mically induced floods, in adde tion to the external sources, includes the e atuauon 7.3 Are extensne margin calculations needed for all of seme internal flooshng c( nststent with the discus-components? (Ref. D.7) ston in Appendtx 1, Check I.ists and Walkdow n Data Sheets, of EPRI NP-6041. Section 6.3.2 wdl be SR: Extensive calculations of IICI.PI s are not needed m modified to include refeience to EPRI NP +041. In order to achieve the NRC's goals lor the senmte addinon, the genene letter has been modif ed.
IPEEli. Refer to rcsps.nse 7.17 for the swpc of margins evaluations. Also. See revised Sections
?"
Draft (ienene letter SS - 20, Supplement 4, Secuon 5 32 A.6 and 3.2.5.7 of this report.
hsts the three related programs subsumed in the IPEEE:(1) the aternalevent portion of USl A 45, 7.4 IPEEli Objectnes and Methotk '
t ecommend
" Shutdown Uccay lleat Remo.a! RequirementsJ that the objectnes of the IPl:1 F be modified to (h Gl-131. "Potennal Srnnne lnteracuan lin olvmg better dehneate IPl;EE objecoves for each of thc the Nhnahie infore I hn Mappmg System Used m accepted methodolopes, or other non-probabilntte Westmphouse l'i m t s.'
and O) the 1 ' hat leston methads that may be proposed by becmecs (Ref I'ar thquake 1+ue ' Whu the IPH:E n uuh/ed for D.1 )
closme of a subsumed mue, we understand that no I )-- I I N UR E( L l407
Appendix D special evaluation, documentation, or reporting will two separate plant response and fragilityanaly-be needed beyond the program defined by IPEEE ses are not needed." The additional effort to (Ref. D.1).
generate results from an additional hazard curve is relatively trivial, and requit es convolu-SR: It ts generally true that evaluation and reporting tion of a hazard curve with the existing plant beyond that identified in this report should not be lesci and sequence icvel fragility curves.
required for subsumed issues except for any addi-tional in, formation that may be needed as a part of b.
Licensees east of the Rocky Mountains usmg a the normal staff review of the IPEEE submittal.
seismic PRA for the IPEEE examination are Note though, that the IPEEE submittal is to address requested to use the results of two seismic haz-specifically USI A-45 and GI-131. Closure of the ard studies. These studies, conducted by the IPEEE also means closure for these issues. No spe-lawrence Liverme re National laboratory cific ceporting requirements are identified for the (LINL)and the Electric Power Research Insti-Eastern U.S. Seismicity issue, formerly identified as tute (EPRI), represent a state-of-the-art icvel-the Charleston Earthquake issue.
opmental effort. Ilowever, for reasons associ-ated with methodology development, these two 7.8 Seismic hazard related comments from Ref. D.1, p studies can produce significantly ddferent scis-8, Comment 1, are summarued as follows:
rnic hazard cun'es.
a.
Use of two hazard curves is not needed. Addi-The uncertainty associated with seismic hazard tional expenses for two analyses not justified.
assessment is clearly demonstrated when one ampares the vast differences between the b.
'Ihe EPRI methodology has been reviewed and rr :an, median,15th, and 85th percentile estt-accepted by the USGS.The I.I.NL results are o ates associated with one curve. The differ-not realistic.
( ecs between the IJ.NL and EPRI harard et tres further demonstrate the large unect-SR: a.
The staff has reused its position regarding the ta nty use of hazard curves for a PR A analysis in re-sponse to this and other similar comments.The in an attempt to resche the dtfferences be-staff still prefers that both 11.NI. and 1.PRI tween the two curves the staff requested assis-hazard curves be used in an anal) sis as thts w dl tance from the National Academy of Sciences.
serve to highlight uncertainties m the bottom The Academy has enteria it uses to esaluate line aumbers as well as robustness m the scientific and technical approaches used m re-Menttfication of vulnerabt'ities. Ilowever, an search. It concluded that both studies followed option of usmg one hazard curve is now in-good scientific pnvedures and practices, and
~
. iuded provided the higher ha/ard curve of the therefore both studies are credible. Therefore, two is used. The reasons for using the highet the staff is encouragmg licensees to use both hazard cune are twofold. One, as discussed m LI.NI and EPRI seismic hazard cunes in tbc more detail in b. below, is that the validity of IPEEE evaluation. Iloweser, if only one cun e one curve oser the other has not been deter-is used, it should be the higher one. As stated nuned yet. The second is that the use of the earlier, the use of both wdl serve to identify higher hazard cune wi!1 ensure that all poten-differences if any, m the delineation of domi-tial seismic sequences are identified.
nant sennue sequences. Ahhough NURI G-1150 stud:es did not identify any signifteant dif-Comments related to expenses appear to re-ferences in ranking anJ contnbutions, vari-sult, to a certain extent. from a misunderstand-ations m contributions and rankings could oc-mg of the scope of analyses required to obtain cur when curves has e markedly different slope results usmg tw > ha/ard cunes. Sinular com-characteristics.Taken together, these pieces of ments were also made at the workshop specifo data (contribunons and rankmgs from I oth cally suggesting that two hazard curves wdl ne-hazard estimates) give a complete representa-eessitate two separate plant response and tion of the senmic event. These data will be fragility analyses. Iloweser. the stait neser in extremely useful to the hcensec in identifying tended that two separate response or fragilh>
plant vulnerahthtics and decidmg if ; ! ant moth-analyses would be needed and therclore spect-heauons are warranted fied only one spectral shape. This point o fur-ther emphastred m Sectum 3.1.1.2 of th:s re-The NUNI ARC cl.um about the renew and accept-port by cateponcaHy stating %nte only one ance of the 1;PRI method 4y by i N iS appears to spettral shape is used for hath ha/ard analyses.
m nch sten /c the estent and mient of the liS(iS NUREWlE W 12
i Appendix D review. Certain observations need to be made te the staff intends to use HCLPF or margin related garding the USGS review of the EPRI methodology, insights in the evaluatism of the IPEEE submittals. It First of all, the USGS reviewed the EPRI methodel.
should be further noted that it is a current practice ogy as a staff contractor and their findings were to include 1101 PF mformation in PR A submittals.
incorporated in the staff Sdety Evaluation Report (SER) addressing this methodology. Second, the re-7.11 Draft Generic Letter 88-20, Snpplement A Appen.
view was limited to the methodology and did not dix 4, Section 4.2.2, item 6, for SMA methud, calls include geology, tectonics, ground motion, or site for calculation of HCLPF values "with or wahout nonseismic failures and human actions.""Ihis iten l
specific resuhs. The staff SER (Richardsm 1988) cicady stated that acceptance was limited to the should be clarified to state that it does not app!y to i
methodology, and any application to regulatory is.
the EPRI SMA methodology. We understand that suc~.msnotpartof theapproval Finally,therewere was the intent because, in the EPHI SM A method, a number of caveats in both the USGS and the staff success paths are chosen avoidmg unreliable equip-evaluations and neither indicated a blanket accept.
ment and unrealistic human recoscry actions. (Ref.
ance of the EPRI results. 'the staff conditioned its D.1) approvel by noting several areas in which problems SR: The staff agrees and will clarify the noted item. Also, may crise if certain precautions were not observed see response to item 7.6 of Section D.1 of thi.s Ap-based on the USGS review.The staff concluded that pend m.
. the staff intends to use seismic hazard calcula-tions resulting from the application of the SOG!
7.12 With regard to the contamment performance evalu.
EPRI methodology in conjunction with similar ation (Ref. D.W l
tesults obtained from LLNL Seismic Hazard Char-acterization Program (SHCP). If significant differ-a.
!! she,1J he clarified that only systems required ences are observed that can cot be resolved, the to pr even' early containment fadure need to be NRC staff will use the two sets of calculations to assesscJ-define the rance cf scismic hazard to be used in the decision making process, in any case, these uncer-b.
For reduced scope plants, we recommend no tainties are such that the specific calculation of seis.
contatnment performance evaluation.
mic hazard, be it that obtained by EPRI or LLNL, should be viewed with some cautiim. The staff finds SR: a.
.Ihc statf has now mcluded this clanfication that seismic hazard calculations are better used for (see the staff response to item 7a of acetion making relative comparisons than for placing reli-D.1 o,' this Appendix).
ance upon the specific numerical estimates.'
h The staff is still recommendmg retention of the 7.9 'lhe two Eastern U.S. plants that were placed in the amment pm naepay b w
wn Mmnt ea@ fanures baauw & waMan 0.5g RLE bin need not be resiewed at that level.
will identify anchorage and spanal interaction Recommend adding a footnote that should read
. indicates an Eastern U.S. Site whose RLE is F "
tainment is cor.sistent with the defense-in-greater than 0.3g unless the itcensee can demon-depth nhilosophv adopted in other parts of this strate on site specific examinanon that the plant s i
seismic exposure is similar to, or less than, those plants assigned to 0.3g RLE". (Refs. D.1,10 and 14) 7.13 The draft generic letter states that the Charleston earthquake issue is subsumed in the IPEEE and that SR Since the publication of the draft penene letter and comp!ction of the IPEEE willconstitute a resolution
(
the guidance document, both plants placed in the of the Charleston carthquake issue. We behese the 0.5g RLE bin have committed to either enhance the Char!cston earthquake issue should he closed based existing PRA (Ref. D.2)or to perform a new seismic on the information contained in EPRI Report, PRA (Ref. D.19); therefore, they have been re-N P--6395-D. (Ref D.1, & Ref. D.5, Attachment 1) moved from this category.
SR: The issue of the 1886 Charleston earthquake has 7.10 Reporting of HCLPF values for components. se-been resolved. The issue of eight outlier plants iJen-quences, and the plant for both new and existing tified thru the Eastern U.S. Seisnucity program has i
PRAs should not be required (Ref. D 1; been subsumed in the IPEEF..nd no specific report-ing is required to close t!ns usue. The staff will SR: The staff has accepted this comment and the report-resiew the iPEEE results for the affected plants.
mg of hcl PF values for licensees using PR A meth-ods is now optional.The staf f has a!so indicated that 114 Ntedian esumates of seismic ha/ard curses should he j
this information is readily avadable from PR A.s. and used rather thaa mean values. Nican values are D -l3 N U R EG - 1407
i Appendix D 4
unduly influenced by outlier experts and thus are 7.17 Summary of NUMARC recommendations for thc unstable. Median values are less affected by the implementation of the seismic aspects of Generic extreme estimates and thus provide more stability.
Letter 88-20. Supplement 4. (Ref. D.1, Attachment (Ref. D.1) 1)
SR: The staff is recommending the use of mean hazard sp mes Mew summagemmMa-curves for the following reasons:(1)The use of mean
" " "## E#
D hazard curves and mean fragtlity curves will lead I with Attachment 1 of Ref. D.L Many of these rec-ommendations have been discussed earlier. De-approvimate mean IcVel frequencies for core dam-age. No statistical meanmg can be attached to a tailed discussion is irMded only when the staff does point estimate obtained through use of median haz.
not agree totally with a recommendation.
ard cutves. (2) The instability and uncertainties are better prcsented by displaymg mean results from Selection of Full, Focused-and Reduced-Scope PlaMs two hazard studies as recommended by the staff.The use of median curves is tantamount to ignoring un-certamties and some expert opinions without an Reduced-scope, 03g RLE Review (Full-scope and focused-scope SM/0:
adequate techm, cal basis.
The staff has accepted the NUMARC suggestion of 7.15 Scope of relay chatter evaluation. (Ref. D.1) creating full-and focused-scope categories in the 03g bin.The 0.3g bin is subdivided into the full and focused scope based on the NUMARC suggested SR: See the staff response to item 7.17 of this sect;on, approach of using both hazard and seismic design response to Attachment 1 of Ibf. D.I.
basis as parameters. Additional consideration was also given to the identification of outlier plants re-7.16 If the intent for the NRC SMA method is t( require sulting from resolution of the Charleston carth-the development of level 1 and 2 functi. mal se, quake issue (see Appendix A of this report for more quences from event trees, the cost of SMA using the discussion).
NRC method would be substantially mereased.
(Ref. D.1)
.1 0.5g RLE Rcriew: Provide opportunity to two East-ern U.S. plants in the 0.5g bin to submit site-specific justification for a binning change from 0.5g to 0.3g, SR: The staff does not require the development similar to consideration given to Western US plants.
of functional event trees beyond that contained in NUREG/CR-4334, NUREG/CR-4482, and SR: See response to item 7.9 of this section.
NUREG/CR-5076, to addrecs containment per-formance issues when the margin approach is used.
.2 Multiple Units at a Site: Lessons learned in evaluating i
(Note that for the PRA appreach the tie-in between the first unit may be used in examining the other Level 1 and Level 2 is quite clear, and this should not unit (s); in particular, any areas of concern that may be an issue). As stated in this report, the licensee be identified during the evaluation of the first unit should develop its own containment performance would be examined in the other unit (s). Otherwise,
.i plans based on the IPE results. What is required is to the scope of review for the other unit (s) can be examine containment functions (regardi ss of the reduced accordmgly.
plant damage states that may be indicated by Level 1 margin sequences or success paths) required to pre-SR: he staff agrees'that results and findings from the vent early failures and report HCLPFs for these first unit should be used to help in the evaluation of j
functions and components if below the RLE. Obvi-other units, provided appropriate similarities exist.
ously, a licensee has an option to develop Level 2 However, such judgements can be made only on a trees at its discretion. Discussions of various ways to case-by-case basis. This report has been revised to extend Level 1 margin analysis to Level 2 are con-include a statement to this effect. In any event, walk-tained in Budnitz 1991a and 1991b for both NRC downs of all units will have to take place to ensure and EPRI methods. (This is suggested for general similarities. It is very likely that the greatest reduc-guidance, no specific requirements based on these tion would be achievable in analytical effort.
references are implied.) A success rath oriented approach is also discussed in Ref. D.1 and Reed, et 3 Scope of Deterministic Scismic Reviewc Identification al.,1990. In summary, the cost of the NRC SMA of Success Path Elements. For all three types of method need not be greater than the EPRI SMA review (full, focused. and ; educed-scope), proced-method.
ures for identifying structures and equipment to be I
NUREG-1407 D - 14
l Appendix D retiewed are the same and are based on the recom-Table 7.17.2 NRC Recommendations for Relay Chatter mendations in EPRI Report NP.-6041.
Review Resiew Type Plant Type Recommended Review SR: The staff agrees and has referenced EPRI NP-6041 Full-scope A 46 Follow A-46 procedures as the primary document for the EPRI success path for A-46 review. Expand methodology. However, due consideration should scope to include IPEEli be given to supplemental comments made in Sec-systems using appro-tions 3.2.5.1 and 3.2.5.8 of this report regarding the priate margin or A-46 selection of success paths.
procedure. Review at assigned RLE.
Non A-46 Review all IPEEE
.4 Containment Review: The full-and focused-scope systems using appro-SM A reviews should be limited to evaluation of only priatc procedures at
- NIE, those functions that are necessary to prevent early Focused-scope A-46 Same as NUM ARC containment failure.
recommendation.
Non A-46 Same as NUM ARC recommendation-SR: See the staff response to item 7.12 of this section.
Reduced-scope A-46 Same as NUMARC recommendation.
Non A-46 Same as NUMARC
.5 Relay Evaluation The following table outlines the NUM ARC recommended position on relay chatter evaluation.This table is based on presentation made Comparison between Tables 7.17.1 and 7.17.2 indi-to the staff by NUMARC on November 29,1990' cates that the staffis in agreement with NUMARC on the focused-and reduced-scope categories en-compassing the majority of plants. Reasons for dif-Table 7.17.1 NUMARC Recommendations for Relay ferences in the full-scope review are discussed in Chatter Review Section 3.21 of this report. It should be noted that, for plants performing PRAs. the scope of the relay
]
Review Type Plant Type Recommended Review review is also defined by the above table.
]
l
.6 Soil Failure Investigation: For plants in the focused-Full-scope A-46 Evaluate A-46 per A-46..
scope SMA category, a review based on the design For relays within IPEEfi and construction record is considered adequate. A (not in A-46), perform a review of soil failure should not be required for bad actors review.
plants in the reduced-scope bin, i
i Non A-46 Perform a bad actors i
review for all relays SR: The staff has adopted both recommendations. For within IPEEE.
the focused-scope category, the use of design and Focused-scope A-46 Evaluate A-46 relays per construction records is considered adequate pro-A-46 (SSE). If bad actors vided appropriate data is available. A detailed analy-are found. expand scope sis will be performed at the licensee's discretion if to include IPEEE relays, soil failure is found to be significant. For the Non A-46 Perform a bad actors reduced-scope bin, no soil evaluation is required.
review for all relays flowever, it should be noted that the need and the within IPEEE.
effort required to evaluate soil failure is site specific Reduced-scope A-46 Perform A-46 review. No and should be determined on a case-by-case basis.
additional review for
.7 Screening Criteria: Tables 2-3 and 2-4 of I!PRI IPEEE relays.
NP-6041 can be used.The A-46 screeningguidance Non A-46 No relay evaluation.
given in the Genene implementation Procedure (GIP) may also be used.
SR: The staff recommended relay chatter evaluation is SR: The staff has adopted this recommendation. It outlined in the following table.
should be noted that all caveats given in the margin D-15 NURI!G-1407
I Appendix D methodology as well as limitations on the use of a.
Ilazard results presented in EPRI Report GERs should be observed. and the IPEEE review is NP-6395-D can be used in performing the to be performed at the assigned RLE. Spatial inter.
SPRA.
action issues, such as flooding discussed in EPRI NP-6041, most be addressed.
b.
NRC should allow licensees an opportunity to pctform site specific studies in order to develop
.8 Evaluation of Outliers: For both full-and focused-new, more realistic scismic hazard data.
scope SMA reviews, HCLPFs should be determined SR: a. See the staff response to item 7.8 in this section.
for elements not screened out during a walkdown.
For focused-scope reviews, it is recommended that b.
Licensees always have the option to conduct judgement be used to rank the capacities of the outliers from the lowest to the highest. HCLPF ca.
additional studies they deem necessary and pacities should be calculated as necessary for some present them to the staff for review. However, 4
the new hazard should not be used in lieu of the components, other cornponents should be assigned conservative HCLPFs. For reduced-scope plants, ILNL hazard, but it can be used to provide outliers should be evaluated according to the plant additional msight into uncertainties.
FSAR.
.13 Fragility Calculations: Mean fragility curves are ade-quate h haghy caWadons.
SR: The staff has adopted these recommendations. See Sections 3.2.4.6 and 3.2.5.5 of this report.
SR: This is identical to the staff position.
.9 Scirmic Input: For full-and focused-scope SM A re-
.14 Relay Chatter: Consideration of relays in a seismic views, use NUREG-0098 median spectra anchored PRA should be limited to relays with low seismic to the RLE for the plant. For reduced-scope re-ruggedness.
views, use spectra developed for the SSE ground response spectrum.
SR: The staff preference is that the relay chatter review scope be defined by the plant categorization used in SR: The NUMARC recommendation of the use of the margin resiew.That is, for a plant identified in NUREG-0098 is consistent with the staff recom-the full-scope category, if the licensee chooses to mendation.'the suggested recommendation for the conduct a seismic PRA, the relay review is to be i -
reduced-scope review is accepted with a caveat that done as outlined for that category. Relay fragilities
+
any difference between FSAR and new response and recovery actions should be modeled in the PRA spectra should be highlighted and discussed.
as appropriate.
.10 Review Documentation: The documentation of the
.15 llCLPF Calculations: IlCLPF calculations should IPEEE for the full. and focused scope SM A review not be required for a SPRA.
should follow the guidance outlined in EPR! Report NP-604L The report for the reduced-scope review SR: The staff has made this an optional recommenda-should be concit.,c.
tion. See response to item 7.10 of this section.
SR: The staff expects that information outlined m Ap_
7.18 The New Hampshire Yankee and Hostor. lidison pendix C will be included in the IPEEE submittals.
requested that the NRC recognize their use of PR A for performing the seismic portion of the IPEEE, in both NUREG-1407 and GL 88-20. Supplement 4.
Integration ofIPEEE and A-46 Reviews:
pnor to final issuance. (Refs. D.2 & 19)
,11 It is recommended that IPEEE and USI A-46 re' i
SR: The staff has modified GL 88-20, Supplement 4, I
views be conducted concurrently and that the review and this report to acknowledge the licensee's com-tasks be combmed whenever possible' mitment to use PRA.
SR: The staff welcomes the NUMARC emphasis on in-7.19 An alternate binning approach, using both hazard tegrating these two major seismic efforts: this rec-and seismic design basis considerations, should be ommendation is consistent with the staff philosophy considered. (Ref. D.4) discussed in Section 6 of this report.
SR: From several suggestions regardmg the binning Scope of Seismic Review Using SPRA Approach:
process, the staff has accepted the binning process reanmended by the NUMARC to further suldi-
.12 use of Scismic Ila:ard Resultr vide the 0 3g bin plants into the full-and focused-NUREG-1407 D - 16
Appendix D scopes. hiany of the individual utilities have en-nations of failures can induce sequences that are dorsed the NUh1 ARC comments, different from those found in the IPE internal event evaluation.
7.20 Utilities should be given the option of using either LLNL or EPRI hazard results. (Ref. D.5, Attach-7.27 The Charleston issue should be closed for a majority ment 1) of the Eastern U.S. plants; for the outliers, the issue can be subsumed thmugh the IPEEE.
SR: See the staff response to item 7.8 of this section.
SR: The staff agrees. See response to item 7.13 of this 7.21 The IPEEE should not be required for closure of section.
Charleston for every plant. (Ref. D.5) 7.28 Does Section 6.3.2 imply that scismic event success SR: See the staff response to item 7.13 of this section.
paths must also be simultaneously protected from postulated fire / floods? The sentence "The effects of 7.22 The NRC should consider modtfying the bin catego.
seismically induced external flooding and internal ries based on the design hazard concept proposed flooding on plant safety should be included" is not elcar. (Ref. D.5) jointly by NUh! ARC /EPRI. An alternate binning scheme is suggested. (Ref. D.5)
SR: With regard to floods, see the comment and the staff SR: The staff has considered the NUh1 ARC /EPRI ap-response to item 7.6 of this section.
proach in further subdividing the 0.3g bin. The staff With regard to fire, see the staff response to item 6.2 binning approach is more consistent with the of this section.
NUh1 ARC suggested approach in Ref. D.l.
7.29 The sentence "lloweser, the beensee should assess 7.23 Clarify the extent of peer review for the seismic the significance of IICLPF values lower than RLE IPEEE. (Ref. D.5) and take any necessary actions and make other im.
"IFN"" N k
SR: The staff intent is now clarified in Sections 3 and 7 of E"'*N" # "
C8F this report and the generic letter, the extent of the peer review should be consistent with the IPE guid-ance as provided in NUREG-1335.
gg. The judgements about the significance of findings can be made only when findings are available; there-7.24 The use of two hazard curves is dlogical. Allow the fore, more specific guidance is difficult to give at this use of EPRI hazard data. (Ref. D.5) time, and such attempts may create confusion. Ilow-ever, the intent of the statement is to limit the scope SR: See the staff response to item 7.8 of this section.
of evaluation for which significance needs to be as-7.25 Recommend deleting the contamment walkdown for reduced-scope studies. (Ref. D.5) 7.30 'the staff btnning process only recognizes hazard and SR: See the staff response to item 7.12 of this section.
7.26 In Section 3.2.6 there is some confusion. If you uti-lize the IPE to identify " success paths," then you 7.31 Provide additional guidance about some twelve would not identify sequences and seismic failure plant sites cast of the Rocky hlountains whose main modes that are significtmtly different from those Category I structures are h)cated on rock, and also found in the IPE internal event evaluation. (Ref.
have some Category I structures or components k>-
D.5) cated on shallow or intermediate depths of soil.
(Ref. D.5)
SR: Even if the IPE is used to identify success paths, failure modes such as passive failures, structural SR: The RLE assignment has been made considermg failures, and spatial interaction failures are gener-toit conditions where the main plant structures are ally not considered in an internal event IPE. Addi-hicated, namely, at rock level for the above cases. As tionally, the " common cause" effect created by a noted in this report, significant amplification may seismic event is unique in that the entire p! ant is occur through the soil layers above the rock, and, subject to the ground motton causing combinations hence, plant structures founded on soil may of failures that may not be manifested in an internal experience much Ingher motion than the rock-event IPE. Thus different failure modes and combi-founded structures. In such cases, the use of screen-D-17 NUREG-1407
~_.
l Appendix D ing tables based on the RLE assignment may not be well as margin methodologies as demonstrated in appropriate for the soil-founded structures and trial applications at Maine Yankee (NUREG/
components. The licensee should investigate this CP -4826) and Hatch (Davis,1990). nis is particu-i soil amplification phenomenon using any suitable larly important now that the " focused-scope" cate-means (e.g., analytical studies, comparisons with gory has been introduced requiring more use of other appropriate studies) to determine how to judgement. The composition of the in-house team evaluate the soil founded structures and compo-should therefore strike a balance so that sufficient nents.
expertise is available to ensure that Ihe m ethodology is properly implemented while utilizing in-house 7.32 Suggests an alternate approach related to seismic staff as much as possible.
binning. (Refs. D.5 and 10) 7.37 Based on lessons learned from PRA and Margin SR: See the staff responses to items 7.17 and 7.19 of this evaluations a simplified walkdown procedure for a section.
majority of plants should be developed. (Ref. D.9) 7.33 he tie-in of seismic margin to a Level 2 PRA is not SR: The staff has revised the scope of the scismic exam-defined. Ako, clartfy "all HCLPFs related to... con-ination in line with NUMARC's suggestions with tainment performance" (Ref. D.6, Enclosure) some exceptions.
SR: See the staff responn to item 7.16.
7.38 He scope and objective of the walkdown are not sufficiently described. The "40 person months per 7.34 Does resolution of USI A-45 have scope implica-unit" required for a walkdown i2 excessive. (Ref.
tions for seismic margin options. (Ref. D.6, Enclo-9,13) sure)
SR: No changes to Appendix 1 of the Generic Letterare SR: Yes, as noted in Section 6.3.3.1, functions and sys-required.His report was revised to read as follows:
tems for addressing USI A-45 should be same as those identified in the internal event IPE. Other-wise, some of these functions ano systems may not Perform a walkdown consistent with the intent of necessarily be included in a margin evaluation. A the guidelines described in Sections 5 and 8, and specific reporting provision is also called out in item Appendices D anct I of the EPRI Seismic.
7 of Section C.2.2, Appendix C. of this report for this USI.
The 40 person month estimate indicatWn the com-ment is not consistent with past experience at sev.
7.35 On page 24, Section 4.2.2, item 2 0f the draft generic eral plants. For example, EPRI NP-6359 (Scismic letter, replace the term " findings" with "results."
Margin Assessment of the Catawba Nuclear Sta-(Ref. D.8) tion) states that the total technical manpower ex-pended was 39.7 man months: approximately 16 man months associated with walkdown preparation and
- SR: The staff accepts this comment: appropriate sec_
tions are revised to reflect this change.
the walkdown. 2L5 man months for evaluating un-screened components, and 2.0 man months for re-7.36 A peer review implying the um of external" experts porting. Southern Company Services, Inc. (Hatch in the professional field" for a review of the method-Plant) estimates that they expended 8 staff months ology chosen and its application is not necessary.
for walkdown planning and the walkdown.
in-house review is more appropriate. (Ref. D.R)
The level of effort and resource allocation are justi-SR: The staff has now clarified peer review discussions fled based on approximately 20 seismic PRAs and 3 in the generic letter and this report to be consistent seismic margins evaluations. Both evaluation meth-with the internal event IPE guidance ods have demonstrated that thorough walkdowns (NUREG-1335), which emphasizes in-house re-are one of the most important tools for identifying view.However, the staff has also recognized that a scismic weak Imks.
j licensee may not have in-house expertise in all areas i
of the external events and an in-house team can be The importance of a detailed walkdown is also sup-supplemented by outside experts.
ported by NUM ARC. In fact, when NUM ARC pro-posed the Reduced Scope Program for sites in low Furthermore, it should be recognized that substan-seismic areas, a plant walkdown identical to the Full tial judgement is involved in applications of PRA as Scope cvaluation was recommended.
NUREG-1407 D-18
Appendix D 7.39 An analysis using both LLNL and EPRI hacard 7.44 Reporting of " functional sequences" may not be curves is unnecessary. There is inconsistency be-possible if only systemic sequences are generated for tween the ger.cric letter and NUREG regarding use a PRA or the EPRI success path approach is used for of both curves in an existing PRA. (Ref. D.18) a SMA. (Ref. D.11)
SR: True. For those cases, the reporting criteria are de-SR: See the staff response to item 7.8 of this section.The generic letter and NUREG have been revised to be scribed tn Appeendix 3 of Supplement 4 to GL SS-20.
consistent.
7.45 Reporting HCIfFs with and without non.scis'nic 7.40 To avoid a different interpretation, define HCLPF failures and human actions does not contribute to for the sequences and plant for the PRA and rnar-the four stated purposes of the IPEEE. (Ref. D.11) gins methodologies. (Ref. D.18)
SR: The fourth purpose of the IPEEE is to reduce the SR: The term HCLPFin the context of the margin meth-overall likehhood of core damage and radioactive ods is clearly defined in both NUREG/CR-4334 and material release by modifying hardware and procc-EPRI NP-6041. Examples of how plant level and dures. Cost-effective decisions can be made only if sequence level HCLPFs are detcrmined can be seen both scismic and non-seismic failures are included in in NUREG/CR-4826, the margin evaluation for the the licensee's decision making process. If the ran-Maine Yankee plant. The mathematical definition dom failure probability of a diesel generator to start of HCLPF for both the PRA and margin methods is is high, no scismic fix to the plant is likely to signifi-the sune. Examples of determining sequence and cantly reduce the frequency of the sequence.
n, ant. level HCLPFs are also described in NUR EG!
'CH-4334. Section 3.1.1.3 of this report gives further 7.46 Using the SM A tt is difficult to perform a nonseismic guidance on how to determine component, c-failure and human action evaluation. (Ref. D.ll) quence, and plant level fragilities when only mean SR: Guidance is noted in Sections 3.2.4.7 and 3.2.5.8 en fragdity curves are used.
how to address the above evaluations for various methods. Such evaluations have been made and re-7.41 Soit liquefaction computations to the level of deta:1 ported in the trial plant applications. AdJitional recommended in EPRI NP-6041 are not necessary guidance on this issue is availabic in (Budnitz 1987 to obtain a qualitative understanding of the overall and 1990). The intent is to help the licensee make probability of core damage and radioactive material the right decisions regardmg plant modifications.
release.
See response to item 7.46 above.
What is appropriate and reasonable is an assessment 7.47 Use spectral shapes consistent with the 1.LNL and as to whether the site is suscep:ible to liquefaction EPRI hazard studies. Does this mean uniform haz-behavior. (Ref. D.18) ard spectra? Is this consistent with the NUREG/
CR-0098 spectral shape? (Ref. D.11)
SR: See the aff response to iten 7.17, Soil Failure SR:.Ihis report recommends Ihe use of the median spec-Investigation.The staff has accepted the NUM ARC recomrnendations in this area, tral shape for a 10,000 year ret urn pcriod provided in NUREG/CR-5250 or a site specific spectral shape based on a suite of appropriate records for pet form-7.42 Put Farley in the Reduced Scope Program bm. (Ref.
ing PR As. This is different from the NUREGl
)
CR-0098 spectral shape recommended for evalu-ations associated with the SM A methodologies.1he SR: Farley is now assigned to the Focused Scope pin.
reasons for this difference is: PRA takes into ac-Plant binning was accomplished by companng nme count the full range of the hazard requiring use of a separate pieces of information related to seismic reahstic description of ground motion as much as hazard groupings and enginecting judgement.
possible whereas margin evaluations are only con-ducted at one carthquake level, and the screening 7.43 Required use of LLNLand EPRI hazard curves adds tables used in the margin methods are developed significant expense for no significant benefit. The from earthquake experience data more compatible four purposes of the IPEEE can be met using EPRI with the ground motion represented by the hazard curves.The EPRI method has been review ed NUREGCR-009S spectral shape.
andaccepted by USGS and NRC; the LI NL method has not. (Ref. D 11) 7.48 Allow jus fication other than expensive sensdisity studies for use of a cutoff other than 1.5g. (Ref.
SR: See the staff response to !!em 7.8 of this section.
I'.Il)
DR9 N U RI.( L l 407
m Appendix D SR: Such sensitivity studies are routinely performed to to propose an alternative position and submit infor-ensure that an adequate range of integration has mation to justify it.
been defined. These stu(.es are not expensive. In any event, licensees have the option to propose al.
7.57 Assignment of the RLE should be allowed to be ternative methods. These will be reviewed by the based upon complete site-specific evaluation of the staff on a case-by-case basis.
- p. gical and seismological data for the site. (Ref.
D.14, Attachment A) 7.49 A mean component fragility curve is defined y the median capacity, A, and composite uncertainty.. Is SR: See the staff response to 7.56 abose.
this a correct statement? (Ref. D.11) 7.58 'lhe cost for Pilgrim will be higher because the staff SR: Yes, the statement is correct.
has characterized the Pilgrim site as a high hazard site. (Ref. D.14 Attachment A) 7.50 Can SOUG GIP walkdown guidelines be used in lieu of EPRI NP-6041? (Ref. D.11)
SR: The cost estimates provided in the draft genenc 1ctter are generic. Clearly, some licensees may SR: See the staff response to item 7.17.7, Screenine Cri.
spend more because more detailed analyses are teria, of this section.
needed. For mstance, containrr.ent and containment system performance evaluations for Mark I and Ice Condenser containments will be somewhat more 7.51 Reportinn HCLPFs with and without nonseismic failures a'nd human actions does not contribute to expensive since generic capacity data on the systems the four stated purposes of the IPEEE.
are lacking. Licensees have the option of performmg a seismic PRA instead of the margins method. Staff SR: See the staff response to items 7.10 and 7/5 of this consultants have indicated that licensees assigned to section.
the 0.5g bm should senously consider the PR A ep-tion as a means of controlhng expenses.
7.52 The licensee should have the option of usme"heavs datv" experts in lieu of a " peer review" in their Ay we dw na4 msponw to hem,a. of das SM'A. (Ref. D.11) wcuon.
7.59 Coordinatin With Other External tivent Procrams.
SR: See the staff response to item /.00 of this section.
m
~
. hsumption of the C,harlestor t:arthquake issue e rates a "de fatto" provi. son fot IPEEE implemen-7.53.rhe C,harleston h,arthquake Issue does not need to
.aion bv pre-supposing that the h.ewsce wd. l 1,e be subsumed into the IPEEE. (Ref. D.1l) perform'ing a I,RA or accepts the NRC,. s seismic hazard estimates in determming the RI IL SR: See the staff response to item 7.13 of this section.
7.54 Include more Midwest plants in the Reduced-Scope Program bin. (Ref. D.11) 740 The draft Generic l.etter should address the future requirements for maintenance of the seismic marnin SR: See the staff response to item 7.42 of this section.
ident fied by the IPEEE.(Ref.D.14, Attachment A)
~.55 No basis is provided for future pbnt modifications to SR: See the staff response to item 8 of Section D.I.
maintain the plant marnm iJentified from the IPEEE. (Ref. D.14) 7Al SMA and I'RA relay chatter enhancement has not been shown to be cost effectise, particularly in licht SR: See the staff response in item 8 of Section D.1 of this of the A-46 resolution. WE endorses the focused Appenda.
SM A approach for all plants performmg a SM A or PR A. (Ref. D.ll) 7.56 There is no specific prousion in the Genenc l citer to aHow a licensee to make their own detaded esalu-SR: The relay chatter ev duation has been changed; see ation to determine which reuew level carthquake response to NUM ARG comments m item 7.17 of hin they should be awgned. (Ref. D.14) this secuon.
SR: The assignment of the reuew lesel earthquake was 7A2 It is not cost benefiaal to melude lon^ tam coohng based on both state-ol.the-art 1.I.NI. anJ EPHI and prenure suppression syuems, the cont;un-studies. Howner, the beensee alwass has an opuon m< nt performua e evaluanon t Ret i LII)
N U R EG -1407 D-20
Appendix D i
SR: The staff has revised the scope of the containment 1.etter 88-20, Supplement 4. " Individual Plant I!xamin-review to focus on the early failure modes. See the ation of External !! vents (IPElili) for Severe Accident staff response to item 7.17.4 in this section.
Vulnerabihties," and Draf t NUREG-1407, " Procedural and Submittal Guidance for the IPElili," October 10, D.3 References Reed, J., et al. " Recommended Seismic IPl! Resolution 13udnitz, R., et al., "lixtending a llCl.PF-llased Seismic 1 weedure," Proceedmgs of Third Symposium on Current Margin Review to Analy7e the Potemtal for I arge Radio-Issues Related to Nuclear Power Plant Structures, Equip-logical Releases and the Importance of Iluman Factors ment and Piping, Orlando, l~loride, December 1990.
and Non-Setsmie Failures, Draft, March 1987.
USNRC, Generie Letter 88-20 " Individual Plant Exa.
" Enhancing the NRC and EPRI Seismic Margin mination for Severe Accident Vulnerabihtics-10 CFR Review hiethodologies to Analy7e the importance of 50.54(f)," Nos ember 23,1988.
Non-Seismic Failures, Human Errors, Opportunities for
, Generic 1.etter 8S-20, Supplement No.1. "Initiat-Recovery, and I arge Radiological Releases, Draf t 2, Sep-ion of the Individual Plant lixammation for Severe Acci-tember 1990.
dent Vulnerabilities-10 CFR 50.54(I)." August 29,1989 Chilk, S. (NRC), Memorandum to J. Taylor (NRC), and
-, Generic 1 etter 88-20, Supplement No. 4 "Individ.
W.
Parler (N RC), dated July 17, 1990,
Subject:
ual Plant lixamination of External Events (IPliEli) for SECY-90-192-Individual Plant Examination for $cvere S vere Accident Vulnerabdities-10 CFR 50.54(f),"
Accident Vulnerabilities due to External I! vents draft for comment, July 23,1990.
(IPEEE).
, Gencric 1.etter 8%20, Supplement No. 4,"Individ-Code of Federal Regulations Title 10 "linerry" (10 ual Plant Examination of Rxternal Events (IPEEE) for CFR), U.S. Government Printing Office, Washmgton, Severe Accident Vulnerabilities-10 CFR 50.54(f)J fi-D.C., revised periodically, nal, J un e,1991.
Davis, P.," A Peer Review of Two Seismic Margin Assess-
, Genene I.etter 89-22 "ltesolution of Generie ments as Applied to the Hatch Nuclear Power Plant",
Safety issue No.103, ' Design for Probable Maximum Proceedings of Third Symposmm on Current Issues Re-Precipitation," October 19, 1989, lated to Nuclear Power Plant Structures, Equipment and Pipmg, Orlando, Florida, December 1990.
-, NURl!G-1032, "livaluation of Station illackout Acciden's at Nuclear Power Plants," June 1988 Electric Power Research Institute EPRI NP-t041, "A Methodology for Assessment of Nucicar Power Plant
-, NUREG-Il50, " Severe Accident Risks: An As.
Seismic Margin, October 198S.
for Five U.S. Nuclear Power Plants," Vols.1
""d
""C
-, liPRI NP-6359, "Scismic Margin Assessment of the Catawba Nuclear Station," Vols. I and 2, April 1989.
-, NUREG-1335, " Individual Plant Examination:
-, EPRI NP-6395-D, "Pmbabilistic Seismic Ilarard Evaluation at Nuclear Plant Sites in the Central and
-, NUREG-75/OS7, " Standard Review Plan for the liastern United States: Resolution of the Charleston Review of Safety Analysis Report for Nuclear Pov,er issue," April 19S9.
Plan's," 1.WR cdition, December 1975.
-. NUREG/CR-0098, " Development of Criteria for Moore, D., et al.,"Results of the Seismic Margin Assess.
Seismic Review of Selected Nuclear Power Plants," May ment of Hatch Nuc! car Power Plant." Proceedmo of 1978-Third Symposium on Current Issues Related to Nu'elear Power Plant Structures, Equipment and Piping, Orlando, NUEGCR-2300. PRA Procedures Guide "
Florida, December 1990.
January 1983.
Orvis, D., et al. "Scon.ic Margm Review of Plant Hatch
-~, NURl!G/CR-2S15 "Probabilistic Safety Ana ysis l
Unit h System Analpis7 1.1.NI. Report No. ULRI e Pmeedures Guide," Vols, I and 2, August 19S5.
CR404834, August 1990.
-, nub'EG/CR-4334."An Approach to the Quantifi-Rasin W. (NUMARC). letter to W. Mmners (NRC),
cation of Setsnue Margms m Nuclear Power PlantsJ S6 ject: Emal Industry Comments on Draft Generie August lusi.
D-21 N U R EG - 1407
1 Appendix D.
-, NUREG/CR-4482," Recommendations to the Nu-D.6 Letter from J. Garrick (PLG Inc.) to J. T. Chen cleai Regulatory Commission on Trial Guidelines for (NRC), dated October 9,1990.
Seismic Margin Reviews of Nuclear Power Plants,"
.. March 1986.
D.7 Letter from P. Smith (The Readiness Operation)
-,NUREG/CR-4826,"SeismicMargin Reviewof the Maine Yankee Atomic Power Station," Vols.1-3, LLNL D.8 Letter from G. Mueneh (Entergy Operation. Inc.)
March 1987, to S. Chilk (NRC), dated October 5,1990.
-,NUREG/CR-5076,"An Approach to the Quantifi.
D.9 Letter from J. Skolds (South Carolina Electric &
cation of Seismic Margms m, Nuclear Power Plants:' Die Gas Co.) to J. T. Chen (NRC), dated October 10, importance of BWR Plant Systems and Functions to Seis-1990.
mic Margins," May 1988.
D.10 Letter from W.Hairston,Ill(Alabama Power Co.)
-, NUREG/CR-5088, " Fire Risk Scoping Study."
t W. Minners (NRC), dated October 12,1990.
January 1989.
D.11 Letter from C. Fay (Wisconsin Electric Power Co.)
-, NUREG/CR-5250," Seismic Hazard Characteriza-t E. Beckjord (NRC), dated October 5,1990.
tion of 69 Nuclear Power Plant Sites East of the Rocky Mountains " Vols.1-8, January 1989.
D.12 Letter from G. Sorensen (Washington Pubh.c
-, Policy Statement on Severe Reactor Accidents-Power Supply System) to Document Control Desk Federal Register, Vol. 50, p. 32138, August 6.1985.
(NRC), dated October 5.1990.
-,- SECY 88-147, " Integration Plans for Closure of D.13 Letter from N. Reynolds (Winston & Strawn) to E.
Severe Accident Issues, May 25,1988.
Beckjord (NRC), dated October 9,1990.
D.4 References for Sources of D.14 Letter from R. Bird (Boston Edison) to Document Comments Control Desk (NRC), dated Octobu 12, IM D.15 Letter from D. Shelton (Centerior Energy) to D.1 Letter from W. Rasin (NUMARC) to W. Minners Document Control Desk (NRC). dated Octo-(NRC), dated October 10,1990.
ber 16,1990.
D.2 Letter from T. Feigenbaum (New Hampshire Yankee) to Document Control Desk (NRC), dated D.16 USNRC Transcript on IPEEE Workshop, Pitts-October 3,1990.
burgh, Pennsylvania, September 11-13,1990.
D.3 Letter from W. Orser (Detroit Edison) to W. Min.
D 17 Letter from G. Goering of Northern States Power ners (NRC), dated October 8,1090.
Company to F. Miraglia of NRC, dated June 10, 1985.
f D.4 Letterfrom R.Wheaton(R.D.Wheaton Asso.)to J. T. Chen (NIIC), dated October 6,1990.
D.18 Letter from II. Tucker (Duke Power Co.) to E.
D.5 Letter from M. Lyster (Centerior Energy) to Document Control Desk (NRC), dated October 8.
D.19 Letter from G. Davis (Boston Edison) to Docu-1990.
ment Control Desk (NRC), dated March 1,1991.
i NURLIG-1407 D-22
T ATI'ACIIMENT TO APPENDIX D VALUE/ IMPACT ANALYSIS FOR TIIE IMPLEMENTATION OF INDIVIDUAL PLANT EXAMINATION OF EXTERNAL EVENTS 1.
Introduction external events being in the range from 6E-5 to 3.8E-4 per reactor year. More recently, NUREG-1150 analyses The primary objective of this attachment is to provide a for Surry (NUREG /CR-4551, Volume 3) and Peach llot-value/ impact analysis to support the issuance of a supple-tom (NUREG/CR-4551 Volume 4) have indicated that ment to Generic Letter 88-20 (Ref.1) requesting an In-the mean core damage frequency from fires is in the range 4
dividua' lam Examination of External Events (IPEEE) of 10 5/RY and from Seismic events 10.s_ loggy, 3
from ai ensees holding operating licensees for nuclear power plants. Implementation of the IPEEE program is A common finding from these PRAs is that support sys-consistent with the Commission's Severe Accident Policy tem failures have been identified as significant contribu-
- (50 FR 32138) dated August 8,1985 (Ref. 2). The imple-tors to the probability of core melt. At the support system
- mentation of the IPEEE program wdl provide the utilities level, there is often sharing and interconnection between
)
and the NRC staff with a better understanding of the ac-redundant trains, questionable separation and spacialin-
. tual state of the plant ard its capability to cope with se-dependence between trains, and poor overall general ar-vere accidents. The IPEEE, program may reveal external rangement of equipment from a safety viewpoint. For ex-event vulnerabilities that could be reduced by procedure ample, many plants have redundant trains of equipment i
changes or hardware modifications to upgrade te sitting side by side in a common area and adequate pnysi-frontline and support safety systems.
cal separation and protection of redundant safeguard trays is la&ng. M type q,ginemi armngement of In general, ia performing a value/ impact analysis (Ref,3) equipment creates vulnerabihttes m that smgle events the staff would (1) identify potenG1 external event vul-such as a fire or a flood can disable multiple trains of nerabilities to severe accidents in operating light water 8*Iet related equipment resulting in an inability to cool reactor (LWR) power plants. (2) identify modifications cE that could reduce plant risk from these vulnerabilities, (3) determine the safety benefit of these modifications' TM provides a list of the specificvulnerabilities found and (4) assess the net cost of the modifications. Ilowever' from rame of these studies and potemial modifications to m this study, we do not know what the utilities will find address these vulnerabilities. In general, external event from their IPEEE programs. Also, we do not know what Mies were identified in:
fixes the utdtues wdl propose. Therefore, for this study we have used data from published probabilistic risk as-a.
Electrical switchgear/ battery failures due to seismic sessments (PRAs) to identify potential vulnerabilities excitation.
that could be identified in an IPEEE and compared the benefit of fixing such vulnerabilities to the cost of doing the IPEEE as well as the cost of the fixes themselves.
b.
Water storage tank (30, HWST) bihnes due to seismic excitation.
i 2.
PRA Findings c.
Pump and valve common-mode failurcs (AFW, Although the Commission has concluded that existing CCW, SWS, HPIS. LPIS, etc.).
plants pose no undue risk to the public, the Commission emphasized that systematic examinations of existing d.
Fires m cable spreading rooms, switchgear rooms, or j
plants are needed to confirm the absence of any plant common cable run areas (HWR/PWR).
j unique vulnerabilities to severe accidents. This conclu-sion was based on the fact that previous plant-specific Modifications include both hardware and procedural J
(PRAs) have typically revealed valuable insights on plant components which, if implemented, would serve to
]
specific vulnerabilities to severe accidents, reduce the estimated core damage frequency. Table 3 provides an example list of the plant. specific modtfica-Table I summarizes previous PRA results in terms of core lions that have been made. These modifications were damage frequencies (CDFs) due to internal and external made based on the msights gained from plant-specific events from 13 PR As (Refs. 4 & 5) that are availabic to PR As. Ilowever, many fixes are made in the course of do-NRC. 'llese results indicate that the mean value of the ing a PRA which are never quantified or reported. For CDF, for these plants is in the range of IE-4 to 4.4E-4 example, deficient equipment anchorages were found at per reactor-year, with the cwe darnage frequencies due to almost every plant during seismic walkdowns.
D-23 NUREG-1407
- Appendix D Attachment Table 1 Summary of PRA Results of Core Damage Frequency (IE-5)
Total Total Int'l Extn't liigh 1.ight.
Plant Total Int'l External Seismie Fire Flood Floods Winds ning PWR Pt. Beach 313 13.9 17.4 6.1 33 7.7 0.4 0.0%
Turkey Pt.
- 23.6 7.1 16.51
.7 7.5 4.6 2.4 0.26 St. Lucic 7.44 1.4 6.04 13 4.4 032 0.02 ANO1 17.9 8.8 9.15 73 0.58 0.72 0.53 0.02 Mean 20.1 7.8 123 IP2 43.5 6.0 37.5 14.0 19.2 43 IP3 15.7 9.0 6.7 031 63 0.13 Zion 34-40 34.2 0.1-6
<0.1-6 MS3 15-23 14.7 0.8-SEst -
Oconee 15-28 7.4 8-21 6.0 1.0 0
13 10.0 (NRC) 2.5 (NRC) 23 (NRC) -
Mean 25-30 14 3 11-16 BWR Quad Cities 19.7 9.9 9.8 83 1.3 0.01 0.01 0.2 Cooper 43.7 28.9 14.8 8.1 1.1 5.
0.4 0.2 Limerick 9.2 8.4 0.8 0.5 03 Shoreham 7.4 5.4 2.0 2.0 (NRC) -
Mean (4) 27 20.2 6.8 IP2/IP3-Indian Point 2/ Indian Point 3 MS3-Millstone 3 usually strengthened, however, and are rarely reported butions from those excluded events would result in higher specifically in the PRAs in terms of the impact on CDF or estimated core damage frequencies.
averted risk. Based on the insights gained over the last ten years, almost every systematic examination has resulted For this study, the following inodifications v6ere consid-in plant-specific insights, that in conjunction with the cred as possible means of reducing the vulnerabilities plant specific evaluation of risk reduction options, would which would most likely be uncovered in a plant specific always result in identifying cost effective remedies.
IPEEE.They were used here for the purpose of assessing the value-impact of the IPEEE program; however, they 3.
\\,alue-Impact Assessment do not necessarily represent the only means for reducing The analyses performed for the resolution of USI A-45, plant risk to severe accidents.-
Decay Heat Removal (DHR) Requirements (Ref. 6),
were used to make reasonable estimates of the value of (1) Seismic Resistance of Batteries and Switchgear conducting the IPEEE. Specifically, reduction in core damage frequency resulting from proposed modifications a.
Ensure that battery installation racks meet cur-and the cost of those modifications were evaluated. It rent seismic requirements. All racks should be should be noted, however, that the purpose of the USI steel with appropriate tiedowns to prevent mo-A-45 program was to evaluate the adequacy of the DHR tion under seismic excitation.
t enction only; accident sequences that did not involve this function are not included in the analyses. These excluded b.
Provide additional ties to floor for electrical sequences involving large LOCAs, reactor vessel rup-equipment (transformers, switchgear, buses, tures, the pressurized thermal shock sequence, interfac-battery chargers, and motor control centers) ing system LOCAs and anticipated transients without for anchorages to prevent cabinet motion scram (ATWS). Thus, the core damage frequencies de-during seismic acceleration. For tall cabinets, rived under that nrogram do not represent the total fre-provide additional restraints to prevent top-quencies for those operating plants. Includep the contn-pling.
NUF iG-1407 D-24
Appendix D Attachment i
Table 2 Modification Options identified for the Case Studies Plant Vulnerability Modification Pt. lleach RWSI' failures and electric switchgear Provide water from spent fuel pool and add restraints failures from seismic events to switchgear and battenes Service water pumps lost from failure Install shield wall to protect pump motors due to spray loss of safety systems due to fire in CSR Install added fire suppression and Al'W rooms Turkey Pt.
Surge floods safety systems increase height of existing flood wall I oss of cooling due to loss of water tanks increase strength of tanks and heat exchanger and CCW heat exchangers from seismic supports event loss of safety systems due to fire in CSR Install additional suppression in CSR St. Lucie loss of safety systems due to CSR fire Enclose one train of safety related cables in fire barrier
- 1. css of cooling due to loss of water tank
- h. crease strength of tanks with addition of external supports A501 loss of cooling due to fa. lure of lilVS Install provisions to power auxiliary feed pump pump and to take water from CSl' from Class th bus lou of safety systems due to fire in CSR Add redundant deluge valve with separate sensing and control Inss of cooling due to loss of tanks and Strengthen tanks with external supports and anchor emergency electric power due to seismic switchgear event Quad Cities
- 1. css of decay heat removal due to ^res in linhance cperating ocedures for the safe-CR or CSR shutdown pump loss of electric power due to seismic esents Upgrade battery racks and add restraints to SWGR and l'uses Cooper loss of safety systems due to fire in CSR Add fire barrier around IiPCI and RilSW power cables loss of cooling due to failure to tanks and Install added anchorage or tanks and heat heat exchangers from seismic event exchangers loss of e nergency electric power due to Add supports and tiedowns to switchgear and seismic events transformers Cooper Alt.
Ioss of cooling due to scismic cvents Strengthen I fI'l!X mounts, valve, CST, and transformer tiedowns loss of decay heat removal due to fhiods Develop procedures for safe shutdown in high llood crests (2) Seimsie Resistance of Tanks (3) Fire Protection Upgrade anchorages and walls for water Morage tanks Where safety-related cahhng is concentrated, ensure that (RWST and CST) designed using the procedures of TID adequate fire protection is provided by mstallation of ad-7024 and with II/D ratios greater than 1.
ditional suppression systems, ther mal protection. ctc. and D-25 NURl!G-1407
~
l i
Appendix D Attachment reliable alternate shutdown capability is available. Re-low-cost fixes may be found to reduce risk and may e
view all procedures to ensure that minimal quantities of be cost effective fuels are present in fire-susceptible areas (control rooms and cable spreading rooms in particular).
for those fixes that were made at specific plants, the e
averted risk was significant.
3.1 Analysis of Specific Modifications in many cases even if the cost of doing the IPEEE Table 3 summarizes for selected plants the value-impact (estimated to be as much as $1 million at the upper analyses resulting from application of the specified bound)were added to the cost of doing the modifica-I system modifications described above. Besides value-tions, the modifications might still be cost effective.
i npact, core damage probability, population and occupa-tional doses, and costs are shown explicitly in the tabic. As Thus, the staff concludes that there is a high likelihood expected, the value (Col. 4,5, & 6)and the impact (Col. 7 that conduct of the IPEEli will result in the identification
& 8) of any given modification are plant and site depend-of vulnerabilities that, if fixed, would re ; ult in a substan-ent. None of the suggested modifications is cost effective tial increase in safety and that could be fixed in cost-
)
(Col. 9) based on avertible offsite costs alone. However, effective manner. Accordingly, the systematic examin-
)
some modifications may be cost effective if onsite costs ation of each operating nuclear plant could provide the
]
are included (Col.10).
most complete compilation of data and analysis available to develop an integrated perspective on risk from external 3.2 Plant Specific Value-ImEact mnts. R wuld Ise identify human, pnndural, design.
and operation vulnerabilities and could provide practical Analyses means to explore and select cost-effective alternate solu-1 Table 4 summarizes the results of the plant-specific ti ns to plant vulnerabihties. Therefore, a plant-specific value-impact analyses performed for USI A-45. Various exammation, like IPE!!E, conducted by analysts with ac-combinations of modifications were evaluated for each cess to plant data and procedures, could better establish plant. Besides value-impact, core damage probability, the level of risk and identtfy cost-benefit improvements at population and occupational doses, and costs are shodn a particular site.
explicitly in the table. As expected, the value (Col. 3,4, &
- 5) and the impact (Col. 6 & 7) of any given modification o.
References are plant and site dependent. None of the modifications is 1.
Generic Letter 8d-20 Supplement 4, Individual cost effective (Col. 8) based on avertible offsite costs alone. However, some modifications are cost effective if Plant Examination of Eaternal Events (IPEEE)for Severe Accident Vulnerabilities, Draft for Com-onsite costs are included (Col. 9).
ment, July 23,1990.
4.
DISCUSSION and Conclusion 2.
" Policy Statement on Severe Accidents " U.S. Nu-c! car Regulatory Commission FederalRegister, Vol.
As can be seen from Table 1, external events can be sig-50,32135, Augest 8.1985.
nificant contributors to overall rist-from a nuclear power plant. Previous risk analysis of external events have al-3.
NUREG/CR-3568, A Handbook for Value-Impact ways uncovered items which were modified to reduce risk Assessment, PNL-4646, dated December 1983.
(Table 3). In many cases the reduction in risk resulting from these modifications was not quantified and thus 4.
NUREG/CR-5042, Evaluation of External Hazards value-impacts were not calculated. However, from the to Nucicar Power Plants in the United States, data available from the A-45 analysis (Fables 4 and 5)and Lt.NI, December 1987.
l Table 3 the followirig conclusions can be drawn-l 5.
NUREG/CR-5042, Supplement 1, " Evaluation of the cost of the modifications considered may range External Hazards to Nuclear Power Plants in the e
from approximately 50K to 24M dollars per plant United States, Seismic Hazards" LLNI, April 1988.
the risk-averted (on site and off-site) may range 6.
NUREG-12X9, " Regulatory and llackfit Analysis:
e from approximately 50 person-rem to 2600 person-Unresolved Safety issue A-45, Shutdown Decay rem per plant over the life of the plant Heat Removal Requirements, November 1988.
NUREG-1407 D-26
Appendix D Attachnient Table 3 Examples of Averted Risk from PRA Experience Plant Description of hiodifications Reference Oconce Changes to turbine bldg., control NSAC PRA room, turbine bldg. eq., and procedures mods to reduce plant vulnerability to internal floods (CCW)
Yankee Establish risk lusis for external Chapman Rowe event requirement resolution Tornado /high wind requirement Seismic design changes Indian 2 hiod, of structural design of IP2 PRA control room hiillstone 3 R eplace diesel gene:ator oil hts 3 PRA cooler anchor bolts (seismic)
Conn Yankee App. R hiod.
Tornado /high wind mod.
Pt. Heach Add additional fire suppression Table 4 Appiication of Specified System Modifications; Results of Value.lmpact Analyses for Specific Plants Aserted Base dp(CDF)
Aserted Dose impact Onsite Value impact Var.
p(CDF) w Var.
(person. rem)
(Gross)
Costs ss/ person. rem)
Plant No.
(per r-yr)
(per r-yr)
Offsite Net Onsite ($xE6)
($xE6)
GrossNet (1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
Pt. Heath 3
3.13 E-4 1.2E-5 36 15 0.99 0.26 2.SE4 1.4E4 lb 1.5E-5 45 18 0.24 0.33 5.3E3
=0 Turkey Pt. 3 2.36E-4 7.2E-5 535 81 3.10 2.33 5.SE3 1.3E3 2
1.3E-5 99 15 0.91 0.42 9.2E3 4.3E3 St. Lucie 3
7.44 E-5 2.9E-5 100 37 0.60 1.05 1.05E4
=0 2
1.2E-5 42 15 0.052 0.44 1.2E3
=0 ANO1 lb/2 1.79E-4 6.4 E-5 84 71 0.131 1.97 1.6113
=0 Notes:
Column 2.
Modifications are as desenbed in Sectior. 3.
Column 6 = Averted Onsite Dose-Installation Dose.
Column 7 = Present Worth of Installation Costs + Operation and Maintenance Costs + Replacement Power Costs Dunng Installation 4 Cost of 1.imited Scope PRA Column 8 = Present Worth of Replacement Power Costs + Ims of investment + Cleanup Costs.
Column 9 = Col 7/ Col 5.
Column 10 = (Col 7 - Col 8)/(Col 6 + Col 5) l D-27 NUREG-1407
Appendix D Attachment Table 5 Modifications Based on Limited-Scope PRA, Results of Valuc. impact Analyses for specific plants Averted p(CDF) dp(CDF)
Averted Dose Impact Onsite Value. Impact Base w Var.
(person. rem)
(Gross) Costs
($/ person rem)
Plant (per r-yr)
(per r-yr)
Offsite Net Onsite
($xE6)
($xE6)
Gross Net (1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
Pt. Beach 3.13E-4.
2.7E-5 81 33 1.23 0.59 1.5E4 5.6E3 Turkey Pt.
2.36E-4
. 8.5E-5 634 96 4.0 2.75 63E3 1.7E3 St. Lucie 7.4E-5 4.1E-5 144 51 0.65 1.49 4.5E3
=0 ANO1 1.79E-4 6.4E-5 84 71 0.131 1.97 1.55E3
=0 i
Quad Cities 1.97E-4 9.11E-5 2521 103 5.94 2.72 2.4E3 1.2E3 Cooper 437E-4 3.01E-4 2295 278 24.3 6.58 1.1 E4 6.9E3 C (alt.)
2.95E-4 2241 271 3.19 6.42 1.4E3
=0 Notes:
Column 5 = Averted Onsite Dose -Installation Dose Column 6 = Installation Costs + Operation and hiaintenance Costs + Replacement Power Costs During Installation in 1985 Dollars Column 7 = Present Worth of Replacement Power Costs + Inss of Investment + Cleanup Costs Column 8 =Cc! &Coi 4 -
l Column 9 =(Col C - Coi rWCol 5 + Co! 4) l i
NUREG-1407 D-28
NRC FORM 335 U.S. NUCLr AR REGULATORY COMMISS*ON I 1-REPORr NUMBER
( Assigned by NRC, Add Vol, (2-39)
NnCu tio2, supp, Rev,.nd Adoendum num-t*'*-
" erl 320t 32c2 BIBLIOGRAPHIC DATA SHEET l
)
(See Instructions on the tovese)
NUmhWD
- 2. Ti1LE AND SUSTliLE 3 OAT L RLPOH I PUUUSHLD Procedural and Submittal Guidance for Individual Plant Examination of External go,,
I yng Events (IPEEE) for Severe Accident Vulnerabilities i
June 1991
"^Nr NuMac R Final Report
- b. AUI P104(SJ
- 6. f Ytt OF HLPOP T J. T. Chen, N. C. Chokshi, R. M. Kenneally, G. H. Kelly, W. D. Eleckner, C. McCracken, A. J. Murphy, L Reiter, D. Jeng
- 7. swoo COvtsco enciusme Doo
- 8. PtRFORMING OHGANIZAllON - NAME AND ADORESS Ut NHO. prov+oe Di.rs,on, Orhce or Hege, U S Nuc ed* Hegu;ator y Comemsvon, and m mng address; n conPactor, pro-de name.nd m mns addr.ssa Division of Safety Issue Resolution Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, DC 20555 9 SPONSORING ORGANIZATION - NAME AND ADORE SS pf NRC, type
- Same as anove', if contractor, provide NHC Dms=on. Office or negion, U. S. Nuclear Regulatory Commissm, and maihng address.)
Same as atxwe to. SUPPLEMLNr ARY NOlLS
- 11. ABSrAACT (200 words or less)
Based on a Policy Statement on Severe Accidents, the licensee of each nuclear power plant is requested to perform an individual plant examination. The plant examination systematically kioks for vulnerabilities to sesere accidents and cest-effective safety improvements that reduce or climinate the important vulnerabilities. This document pre-sents guidance for performing and reporting the results of the individual plant examination of external events. The guidance for reporting the results of the individual plant examination of internal events (IPE) is presented in NUREG-1335.
13 AV A'L AB; LIT Y S T ATLVF NT 12 NEY WOROS/DCSCHiPTORS (List words or phrases that wm assist reseschers m iccabr'g tt e reint 1 Unlimited 14 SECU50TY CL AS5f tC AT SON S,evere Accu. lents Policy Statement Individual Plant Examination of Externa! Events (IPEEE)
Unclassified Vulnerabilities
""""'"")
IPEEE Guidance Unclassified is NuvatHv P m 16 6%CL NROFORM 335 (2-89)
r p,...,
< ~.. -.. - -
n-
+
u,u :
-a
-,., sc
.. -.,. ~...a..
a-.n..-
x.>
.~,,a 1
a h
d d
8 t
i u
4 4
e I
t:
l I
I r.
l.
I i
THIS DOCUMENT WAS PRINTED USING RECYCLED PAPER
3 l
UNITED STATES i
,,n cas, y m 7
- NUCLEAR REGULATORY COMMISSION postact i, nre raio C
WASHINGTON, D.C. 20555 9
,,.,,~.u.,
OfflCtAL BUSINESS g
0 PENALTY FOR PR!VATE USE, $300 w
!?O5551395,u37 I IAN flj,kh$f.]p[y
- i C-Oc TP LICATIONs gyg3
-271 13 S m 1 sa c,7 a g DC
- 2095, On M
ccC M e#
M47 M>
N[TO z
~
C' bY M
73 MJ V- >~ -
C t' '
2O M9 D0 h7n A M Cn m MM e w hN.
%=4-M=
>M n-n7O w-1, =
"C c' M M-2A M$
_=M M 'A U>
=7 ZU r.c Z>H E
~
Z O
-r:
6 C
7, M'
_____ ____ _____