ML20154D724
| ML20154D724 | |
| Person / Time | |
|---|---|
| Issue date: | 04/15/1988 |
| From: | Sheron B NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Jordan E Committee To Review Generic Requirements |
| Shared Package | |
| ML20153F028 | List: |
| References | |
| NUDOCS 8809160078 | |
| Download: ML20154D724 (13) | |
Text
_ _ _ _ _ _ _ _ _ - . ,
' ' ' " ' ' ~ '
, a g. . . .
I
+
k a UNITE 3 STATES NUCLEAR REGULATORY COMMISSION
$ : WASHING TON, D. C. 2C%4 APR 15 588 MEMORANDUM FOR: Edward L. Jordan, Chainnan Comittee to Review Generic Requirements FR04: Brian W. Sheron, Director Division of Reactor and Plant Systems Office of Nuclear Regulatory Research
SUBJECT:
SUPPLEMENTARY INFORMATION FOR THE CRGR REVIEW OF THE GENERIC LETTER FOR THE INITIATION OF THE SEVERE ACCIDENT POLICY STATEMENT
, in a rencrandum dated March 30, 1988 from Eric S. Beckjord and Thomas E. Murley to Edward L. Jordan, we forwarded for CRGR review all the documents needed to initiate the Individual Plant Examinations called for in the Severe Accident Policy Statement (50FR32138).
Attached to this remorandum is additional information which is provided as background for the April 18, 1988 reeting to review the IPE generic letter.
Attachment 1, provides the staff response to Section IV.B of the CRGR charter, Revision 4 dated April 1987. .
Attachrent 2, provides the staff response to coments cade during the CRGR meeting on November 25, 1987 on the IPE generic letter.
Attachment 3, provides changes made to page 10 of the IPE generic letter.
For furtter infonnation on this subject contact Farouk Eltawila on x23543.
3
. A Brian W. Sheron, Director Division of Reactor and Plant Systems Office of Nuclear Regulatory Research Attachrents:
As stated 4
8809160078 880505 PDR Attachment 2 to Enclosure 2 REVCP NRCCRCR NEETINQ134 PNV 1
Attachment 1 Response to CRGR Charter (Section lY B) I B(1) (11} A draft predecisional generic letter along with its attachments were transmitted in metno to E. Jordan f rom T. Nrley and E. Beckjord.
(iii) This request for the IPEs f s to implerent an existing requiretrent, i.e., the Cosatssion Severe Accident Policy (50 FR 32158), in which the Cornission stated that analyses will be r.ade of any plant that has not yet undergone an appropriate systeratic examination.
(IV) The proposed rethod of iropletrentation is a 10 CFR 50.54 (f) letter requesting the individual plant examiration from all licensees holding operating licenses and ccnstr ction permits for nuclear plants.
(V) To implerent the Severe Accident Policy Stateinent would require each plant to do an examination of its decay heat removal system and those systems used for other safety functions. The conduct of the IPE is not, in and of itself, of value in terms of reducing the prcbability cf core melt. It is of value in that it provides the ,
utility and its staf t with a better understanding of the actual state of the plant and its capability to cope with sesere acciderts. ;
The IPE ray indicate that there are vulnerabilities that could be i reduced by low-cest changes via procedures or minor design redifications.
(VI) The generic letter will be sent to all licersees holding cperating licenses ano ccnstruction permits for nuclear power plants. These utilities with an existing PFA cr similar analysis shculd submit the results of their analysis provided that they update and certify that the analysis reflect the actual design, operation, raintenance and erergency operation of the plant.
(VI'; The generic letter requires utilities to perform the IPEs and if plant specific vulnerabilities were identified, it is expectec trat the licensee will move expeditiously to correct any icentified vulnerabilities that were cetermined to be important to correct.
NRC will evaluate licensee IPE submittals to asses 3 whether the conclusions the licensee draws frcm the IPE are adequate. NRC consideratign will include both quantitative reasures anr1 non-quantitative jt.dgerent ard will use the Safety Goal Policy as one factor in the assessrent. If hRC disagrees with a licensee's disposition of certain issue, it will be bandled unoer existing precedures for issue resolutien including 50.109 (a) ensure that plant specific vulnerabilities are icentified and corrected (b) rerform an IPE, dist ,se of issue to ensure that these important vulnerabilities are corrected ,
I i
(c) Even if the potential change in the risk to the public is Icw, there is a value associated with deronstrating by the IPE, in a logical and structured manner, that potential risk is already low.
(d) Since the generic letter requests each applicant to perform a i systematic evaluation which would require a plant walkdown, the exposure to facility erployees and other onsite workers during i conduct of the IPE will be minimal because most of the walkdcwn will take place outside centainment, t (e) During the systematic examination, we don't anticipate any costs ;
incurred due to facility dcwn time or the cost of construction !
delay. <
l (f) 1he staff believes that the potential irpact of the !PE is to enhance plant, design, cperaticn, maintenance and training
. procedures. .
(g) The estinated resource burden on the NRC reviewing IPE submittals is approximately $7 million over 5 years. The ;
majority of the review would be done by contractors with NRC supervising the review prccess.
(h) Since the generic letter dces not specify a set of improvements sufficient to achieve a certain level of safety, differences in facility types ind age are net relevert to action required by the generic letter.
(VI!!),(XI) The staff has ret performed a specific value impact analysis i
for implerenting the severe accident policy statement. The ,
severe accident policy cercluded that systeratic exerninations are needed. Plant fixes ir.itiatec by the staff will undergo a value impact.
The systematic examinations ray indeed leac a utility to l conclude that there are vulnerabilities in the plant that
! should te eliminated or at least reduced. Even if no
! vulnerabilities are identified, dccuments to derenstrate reaching this conclusion are needed to confirm ccepletion of in.plementation of the Severe Accident Polic) l
+
f
, t
) f t
I i l 4
Attachment 2 Respense to CRGR Coments on IPE Generic Letter
- 1. A fundamental concern of the CPGR was whether this effert is worth the resuurces to be expended in view of: the diversity of expected licensee IPE submittals; the existing uncertainty of the risk / safety improverents expected to be rtalized from the IPE program; the lack of clear o9fini-tion of hRC's goals in undertaking the Individual Plant Examinttions; and the uncertain relationship between the IPE goals and the Comission's safety goals. i Response '
The goals and expected ber.efit of the IPE are clearly ststed in the Severe ,
Accident Policy State u nt.
... the experience of the NRC and the nuclear industry with plant-specific probabilistic risk assessrents is that each ... has exposed relatively unique vulnerabilities to severe accidents. Generally, the undesirable risk from these unique features has been reduced to an acceptable level by low-cost changes in procedures cr t.inor design mcdifications."
in additien, a discussien cf what the staff expects the t tilities to de with the results of the IPE and what the staff will do with the information submitted has been clarified in the Generic Letter (GL) starting on page 9.
The conduct of the !FE is not, in and of itself, of value in terns of reducing the probability of core relt. It is of value in that is prcvices a better understanding of the actual state cf the plant with its capability te cope with severe accidents. Utilities are expected to expediticusly correct any identified vulnerabilitics cetermined to be inportant (see additional response to Coment #7). The staff will review the submitted IPFs to obtain reasonable assurance that the licenste nas acequtely reviewed the plant design and cperation and that conclusicns regarding any needed changts are adequate. The hRC Safety Geal Policy, the Backfit Rule, and non-quantitative judstrents will be factcrs in the assessrant of this adequacy.
- 2. The proposed generic letter specifies several options that can be useC to utisfy the exarinaticn requirements. These vary from an abbreviated PRA, i.e., IDC0R IPEM, to a full scope level III PPA. In addition, an optien is piesented for licensees to use cther systematic evaluatien nethods. CRGR raised questions regardirg what scope ar.d depth of analysis is needed ard expected of licensces, and what are the differences (weak-resses and strengths) of the IDCOR rrethodology vs. that for TRA. Concerns were expressed that it was difficult to see how such a wide range in i scepe (from an abbrevie.ted level 1 PRA to a full secpe level !!! PFA) I could all te acceptable and consistent with NRC's goals, further, the !
cptions en rethcdology were of concern in the absence of staff analysis :
of the possible differarces in the ccepleteness and acceptance of the results. For exanple, there were irdications that staff ard ICCOR [
containrent analyses produce substantially different results. Further. {
there was uncertainty whether a Susquehanna-type IPE would be considerec acceptable now by the staff. If generic procedures are acceptable and in l i
fact desirable, why would generic IPEs also not be an acceptable approach? ,
Also, wculd an external (seismic) event evaluation rethodology, based en a mechanistic (rather than probabilistic) treatrent of seismic hazard, i such as the EPRI Seismic Margins Evaluation rethod, be considered acceptable to the NRC staff?
_ Response The scope and depth of the IFE process and acceptable rathods of performing the IPE is discussed in the CL. The IDCOR t'ethodology, which is one acceptable method t ar performing the frent-end cnly of the IPE, has been reviewed by the staff and the adequacy ar.o differences frcm a FRA are discussed in the draft evaluation report (included with the CRGR package).
The IDCOR source te m analyses were not judged to be acceptable by the staff and the GL gives guidance in Appendix 1 for the back-end IPE analysts acceptable to the staff.
With regard to the Susquehanna IPE, a containment evaluatien was not rade.
While many useful insights appear to have resulted from the performance of the Susquehanna IPE and follcw-up activities by the utilities, the staff has not reviewed this early IPE in detail and can not cenclude on the acceptability of the results of the IPE fcr Susquahanna at this tire.
Ceneric IPEs would oct satisfy the Ccer:ission's Severe Accident Policy which indicated the need for a syster atic examination to identify any plant-specific vulnerabilities to severe accicerts.
A discussion of external events is ircluced in the response to Corrrent #4
- 3. The staff Suidance on such subjects as containrert failure n.echanisrr ard timing is not well specified; fer example, hcw direct heating is to be treated. The CRGR expressed concern that, since the staff has not given definitive guidance, the resultirg analysis may result in different answers even fer sister plants, and the potertial lack of censistency and eniformity of the submitted analyses cculd pose a significant problem in attempting to achieve uniformity in plant procedures and plant fixes to cope with severe accidents. Further, it was rot clear to CFGR the extent to which staff guidance has been discussed with licensees, and whether the language of this guidance will be clearly understood anc, thus, correctly implerented by each licensee.
Response
The staff has revised the guidance in /prendix 1 regarding containtert failure trechaniss.s ar.d will address any specific questions that utilities ray have ir this area through a series of workshcps.
0 8
- 4. The proposed staff position was that external events, such as seismic, tornadoes and ficods are excluded from the scope of the IPE at this tire, and efforts will continue tc reach agreerrent with industry on methodology for external events. However, PRAs have been developed which include external events (e.g., Indian Point) and for IPEs. licensee's can submit ;
a full secpe FRA that includes external events. CPGR expressed concern '
that, as a result of this approach, staff resources will likely be spent reviewing some PRA with extnrnal events, but not all. It appeared that this approach of different scopes at different times could result in inefficient use of staff and licensee's resources, e.g., more than one plant walkdown may be necessary.
Response
The staff is requesting coly examirations for internally initiated events at this tine. Consideratien of external events is Deirg deferred to allcw 1) an identificaticn of which external hazards need censideration, 2) to Nrmit l
. development of siirplified examination procedures, and 3) to integrate other Comissicn programs ~ that de61 with external events with the IPE to ensure that there is no duplication of effort. To this end, the External Events Steering >
Group, consisting of senior RES, NRR, and AE00 managers has teen chartered with ccordirating this effcrt. ;
- 5. The staff prepcsal is to send the proposed generic letter to all plants, whether or not the plant has a PRA. Many PRAs have been completed, and since a nurier of these ccmpleted FRAs are lesel 3, which rr.ay be core '
in-depth and cortplete than the IDCOR IPER optien in the preposed generic letter, the CRGR raised the question whether there ray be already serre plants that have been sufficiently well analyzed in the sesere accident centext that, rather than require additional analysis, it wculd be j apprcprute to proceed to the next phase involvirs the need for plant i specific ccrrective actions, including accident ranagement planntrg, j Response l One irportant aspect of the IPE, as indicated in the CL, is utility invoherent in the process to gain the raximum benefit. For plants with an i existing FRA, effort would still be required to update the FRA, familiarize themsehes with the contents if perfortred by a contractor, and certify that ;
the input to the FRA is correct. Ve agree that plants with an existing FRA can proceed with plant corrective actions without further analyses other than I the items noted abcve. Thus, the GL indicates that licensees with an existing PRA would te e>rected te subtit their IPE en a shorter schedule. In additien, the GL specifically indicates that licensees are expected to expediticusly prcceed with corrective actions.
- 6. The staff stculd be prepared to present argurrents fer/against going ahead with the FRA/ISAP cptien exclusively. Similarly, the staff should be prepared to discuss in greater detail thy it is recessary er critical to rank'decuterit accident sequences or to treat in the IPEs dominant plant
r sequences that do not represent outliers, as indicated in the proposed sequence selection and IPE adequacy criteria. With regard specifically to the proposed guidance on containrent systems performance, how are licensees expected to apply the proposed guidance regarding redundancy and diversity for dominant accident sequences in this aspect of the IPE for their plant? If this guidance is reflected in NRC staff IPE review criteria, this would represent substantial backfits for sore plants. And it is not clear hew such backfitting would be justified, if the IPE goals are not sorehow related to the Corrnissicn's safety goals in the proposed IFE guidance package.
Response
The screening criteria are cnly used to ensure that the potentially infortant severe accident sequences have been included in each IPE. As stated in Appendix 2 to the GL, "The proposed screening criteria are not to be interpreted solely as bottom line thresholds of acceptability .. .." The adequacy of the plant will be decided ultimately by regulatery judgerent that will include, in accition tc the IPE results, an assessment of other factors such as plant cperational managerent. Separately, tra IPE results for a class or several classes of plants could be :orpared with che Consnission's Safety Goal Policy. This inferration would be used to det',rrine if deficiencies in the regulations exist. If deficiencies were identif. fed, the Safety Goal would be used to deterr.ine if redificaticns to the current regulations are needed.
- 7. There should be a clearer uncerstar. ding and discussion of the intended secpe and depth cf IMC review cf licensees' IPE submittals ard the intended schedule for ccrrpleting the as Jcciateo plant fixes that ra) fick frcm this effort. Consideration sbculd be given to au IPE implerentation approach that does rot put tiRR on the critical path fcr completing IPE implerrertatieri, e.g., the "50.59" implementation approach preposec in connection with US! A-44, Station Blackcut. It is not clear that the risk / safety irprevements expteted to te realizec frem the IPE program warrant trajcr PTR resource expenditure in reviewirg licensee's IFE submittals. (See Cement i;c. I abcVe in this centext.)
Response
A detailed IPE review plan is beir.g develeped by the staff to be completec shortly af ter issuance of the GL. This plan will be available to the utilities prior to the start of their IFEs. As indicated on page 9 of the GL, utilities are apected to correct expediticusly any identified, irportant vulnerabilities, f RC review of changes initiated by the IPE would not be required unless regulations such as 50.59 or 50.00 are (cund to apply. Past experience inoicates that utilities quickly enrrect deficiencies ar.d/or pursue safety irprovements highlighted by FFAs.
- 8. The staff 9.i.,ld be prepared to discuss whether f RC could/shculd start new on dt(s'cNent of generic Severe /ccident l'aragement procedures. Do we know enough new from sesere accident studies tc cate to begin
knowledgeably that process.
Some were of the view that the staff could have ready in Cecerber a useful compendium of severe accident r.anagement information that would effectively serve that purpose. This question should receive detailed treatment at the tire of the Comittee's formal rr. view of the IFE guidance package.
_ Response RES already has in place programs to perform detailed investigations of two major severe accident management strategies: (1) venting and (2) depressuriza-tion to avoid direct container.nt heating.
Severe Accident Fanagezent Research Program Plan.In addition,are These programs RES is developing a expected to synthesize existing severe accident knowledge into either generic strategits er criteria to review industry proposals. The staff have already ccepiled existing insights on severe accidents into reports that will be issuto with the GL and have also draf ted a shorter sumary report. Certainly, enough is alreacy known for utilities to begin effective planning to manage
, severe accidents.
- 9. The regulatory basis for the proposed
- proach, a 59.54(f) letter, should be clearly articulated in the centext of other options such as r uleraking. The staff should discuss actions $. hey would take if a licensee refused to cooperate with preposed approaches or to imple%at the fixes indic3ted by the IpE results.
Response '
The 50.54(f) approach was based on CC; guicar.ce cn hcw to irplen4nt systeratic plant examiraticns ictntified in tha Ccenissicr's % ere Accident Folicy Ststerrent. As indicated in the GL, any disagrescent with ccrrecticos to icentified vulnerabilities would be dealt t'Ln thrcugh the Eackfit F,ule. A discussion cf the cverall implenentatier Statenent, including the relaticrshie ;o thef Safety the Sesere Goal,Accident Folic) centainment initiatives, ano other Cemissicn i t.itiatives is currently teing prepared as a separate paper.
i
}
l I
l l
l
', DRAFT PREDECCIONAL CRER COPY April 15, 1988 To All Licensees Holding Operating Licenses and Ccnstruction Pemits for huclear Power Reactor Facilities
SUBJECT:
INDIVIOUAL PLANT EXANlt:ATICN FOR SEVERE ACCIDENT VULNERABILITIES - 10 CFR 550.54(f)
(Generic Letter No. 88-xx)
- 1. SUWARY In the Comission policy staterent on severe accidents in nuclear pcwer plants issued on August 2,1985 (50 FR 32138), the Ccmission concluded, based en available inforration, that existing plants pose no undue rid to the public health and safety and that there is no present basis for imediate action en generic ruitmaking or other regulatory requirerents for these plants.
Fewever, the Ccerission recocnizes. c'ased en f;RC and irdustry experience with plant-specific probabilistic risk assesstents (PRAs), that systenatic examinations are beneficial 3 identifyina plant specific vulnerabilities that could be fixed with low cost improvenents. Therefore, each existing plant should perform a, systematic examination to identify any plant-specific vulnerabilities to severe accidents and report the results to the comission.
The gereral purpose of this examination, defined as an Individual Plant Examinaticn (IPE), is for each utility (1) to develop an appreciaticn cf seve.re accident behavior, ('.) to understand the nost likely severe accident sequence that could occur at its plar.t. (3) to gain a more quantitati%e understanding of the everall pretability of core damage and, (4) if recessary, to reduce that overall pretability cf core darage by backfittirg, where Attachment 3 to Enclosure 2
DRAFT PREDECISIONAL CROR CCPY appropriate, hardware and procedures that would help prevent or mitigate sevtre accidents. It is expected that the achievement of these goals will verify that the overall industry severe core darage and large radioactive release probabilities are consistent with the Connissien's Safety Goal Policy Statement. Besides the Individual Plant Examinaticns, closure of severe accident concerns will involve future NRC and industry efforts in the areas of external events and accident management.
Therefore, consistent with the stated position of the Connission and pursuant to 10 CFR 550.54(f), you are requested to perform an Individual Plant
. Examination of your plant (s) for severe accident vulnerabilities and submit the results to the hRC.
- 2. Examination Process The maximum berefit frce the !PE would be realitti if the licensee's staff teceres lnvolved in all aspects of the examir.atien to the degree that the knowltdge gained free the examinaticn bectres an integral part cf operating precedures tnd training progran.s. Therefere, we enccurage each licensee to use its staff in ccnducting the !PE to the maxirum extent possible. An IPE shoulo include the following activities:
(1) Examine the plant design, cperations, raintenance, surveillar.ce, and emergency procecures to identify potential severe accident sequences for the plant (2) cuantify the expected sequence fregt.encies, (2) deternine the leading contributcrs to ccre darage and unusually peor centainment perforrance and deterrine their urderlying causes, (a) identify any potential Flant irprovetents for the preventien anc citigatien of severe accidents.
?
DRAFT PREDECISIGNAL CRCR CCPY or pwi containrent perfortnance with the attendart radioactive release. The determinatien of potential benefits is plant specific ard will depend on the frequency and consequence of the accident sequence leading to core dar. age and containnent failure.
- 8. Use of IPE Results
- a. Licensee Af ter each licensed utility conducts a systen.atic search for sewre accident vulnerabilities in its plant (s) and determines whether potential improverrents.
. both design and procedural, warrant irplementation, it is expected that the licensee will nove expediticusly to correct any identified vulnerabilities that the licensee determined to warrant correction. Ccrplete inferration en changes initiated by the licensee should be provided with either 50.59 or 50.90 subrittals, as required by the applicable regulaticos. Changes should also be reported in the IPE suttittal in response tc tr.is letter, as described in Appendix 4
- b. NRC The NRC will enaluate licersee !PE sutn.ittals te cbtain reasenable assurance that the licer.see has adequately analyzeo the plant design and cperatiens to ciscoser instances of particular vulnerability to core relt or urusually peor certainn.ent perforrance given a ccre relt accident. Fbrther, the FPC will assess whether the cccclusiens the licer.see draws free the IFE regarding ch- nges to the plant syster.s. corrcr.ents, cr accident raragerent precedures 9
COPY are adequate. The consideraticn will include both quantitative reatures and non-quantitative judgrent and will use the NRC Safety Goal Policy as one factor in the assessrent.
In scwe cases, the NRC may detertnine that, in addition to analytical uncertainty, human perform!rce and/or operaticnal ranagement consideraticns are such that NRC regulations for that plant can only be ret with certain safety improvements:
- 1. If hRC ccnsideratien cf all pertinent and relevant factors irdicates that the plant design or operaticn shculd be changed to meet g( regulations .
then apprcpriate functional enhancerents will be required and espected to be implerented without regard to cost except as apprcpriate tn select arcrg alternatives.
- 2. If f.RC censideration irdicates that plant design er operation cculd be enhanced by substantial additieral protection beyond !!RC C regulations, then appropriate foretional erbancerents v.ill be recorrended and supported with aralysis dercr.strating that the terefit of such enhancetent is v.crth the ccst to implenent and reintain that enharcerent, in accordance with 10 CFR 50.109,
- 3. If NPC consideration ir.dicates that the plant design and operatien nects NRC regulations, anc that further safety enhancerents _ representing substantial additional plan protection are not needed or warrar.ted unless significant rew safety ir.ferr.ation becores available, safety enhancerents kculd r,ct be suggested.
10
DRAFT PREDECC1BNAL CRC 3 COPY
- 5. A description of each functional sequence selected by the criteria of I Appendix 2 should be provided, including discussion of accident progression, tiring, specific assurptions, and human reliability, i
- 6. The estimated core damage frequency arid the likelihood or conditional probability of a large release. The timing of significant large releases for each of the leading functional sequences. A list of analysis
(
assurptions with their basis should be provided along with the source uncertainties. !
. 7. Identificatien and listing cf the main centributers to the estimated core i
)
3 damage frequency, i ~
i !
I i
F. Identification and listir.g cf the main centributcrs to any unusually peer centaircent performance.
r j
l j S. Identificatien cf the USI(s) and CSl(s), if applicable, that have been
[
essessed to esticate their centribution to the ccre carage frequercy er f i<
to urusually pect contairrtent perforrance, f l
I l
] 10. A cescription of the technical tesis fer resolvir.g any US! or GS! when l l applicable, l 1
(
l I 11. For %e functional sequences selected with the criteria cf /ppendix 0, i 1
l previce a list of any poter,tial irproverrents (including equipment changes I i
I as well as changes in raintenance, cperating ard trergency procedures, a
i l '
i l 8
1 l Appendix 4 '
2
! [
_ _ _ _ _ _ -_ .-