ML20207B040

Requests Commission Approval for Publication in Fr of Final Amends Revising ECCS Rule Contained in App K & Section 50.46 of 10CFR50

ML20207B040
Person / Time
Issue date:	06/09/1988
From:	Stello V NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO)
To:
References
FRN-52FR6334, RULE-PR-50, TASK-RIA, TASK-SE AC44-2-14, SECY-88-162, NUDOCS 8806210296
Download: ML20207B040 (289)

Text

-

O pa ss A A

,1 N...../

June 9, 1988 RULEMAKING ISSUEmecy_se_162 For_: The Comissioners From: Victor Stello Jr.

~

Executive Director for Operations

Subject:

REVISION OF THE ECCS RULE CONTAlkED IN APPLt:D:X K AND SECTION 50.46 0F 10 CFR PART 50

Purpose:

To obtain Comission approval for publication in the Federal Register of final arrendments revising the ECCS rule contained in Appendix K and Section 50.46 of 10 CFR Part 50.

Category: This paper covers a major policy ouestion.

Issue: Should the final text of the proposed ECCS rule be approveo by the Comission. The propoced amendments woulo:

a. allow the use of best-estimate evaluation m9thods through the utilization of more recent information gaineo on the performance of ECC systems,

b. relax certain reanalysis requirements that do not contribute to safety,

c. pemit the continued use of current methods for those licensees and applicants wishing to do so, and

d. delete s from Appendix K the reference to the Dougall-Rohsenow heat transfer correlation as an acceptable model.

~Suma ry- Section 50.46 of 10 CFR Part 50 requires that calculatior.s te l

perfonned to show that the emergency core cooling systems I (ECCS) will acequately cool the reactor in the event of a loss-uf-coolant accident (LCCA). Appencix K sets forth certain required and acceptable features that the evaluation mooels, used to perform these calculations, must contain.

Contact:

Louis ti. Shotkin, RES 49-23530 g (o 6.:2 / d 2.7 $

s l The Comissioners 2 i

The results of these calculations are useo to deternine the acceptability of the ECCS performance. In many instances, these calculations result in technical soecification limits on reactor operation (e.g., peak local power) in order to comply with ti,e 2200'F cladding temperature limit and other ,.

limits of 150.46. These limits may restrict the total power  :

t output and optimal o Westinghouse plants)peration of many in terms of efficient reactors (e.g., most fuel utilization, maneuvering capability and surveillance requirements.

Removing unnecessary restrictions on operation will all w increased U. S. electricity production, worth several huncred ,

nillion dollars a year, without loss of benefit to the public  :

haalth and safety.

NRC,00E(includingAECandERDA),U.S.nuclearindustryano foreign research on ECCS performance since the present ECCS rule was issued provides a technical understanding which shows that the existing ECCS rule restrictions are more stringent-than necessary for safety. Thus, the staff recomends that the ECCS rule be amended to reflect this more realistic safety assessment and to remove unnecessary operating restrictions. This is consistent with the 1973 Comissicn opinion published with the existing rule. As a result of the large body of research into the behavior of emergency core cooling systems during a loss-of-coolant accident, the Nuclear Regulatory Comission has proposed to amend its requirements (52 FR 6334) to allow licensees and applicants to use best-estimate calculations cecompanied by an uncertainty quantification to demonstrate compliance with the acceptance criteria specified in 10 CFR 50.46(b). The rule changes were published as a proposed rule on March 3, 1987. Based on the generally favorable public response to the proposed rule, the staff recommends that the Comission i approve the final rule for publication with no changes.

Discussion: On March 3,1987, the Nuclear Regulatory Comission published in the Federal Resister proposed amendments (52 FR 6334) to -

10 CFR Part 50 anc Appendix K (Enclosure A). These proposed amendments were motivated by the fact that since the promulgation of Section 50.46 of 10 CFR Part 50, "Acceptance Criteria for Emergency Core Cooling Systems (ECCS) in Light Water Power Reactors", and the'*cceptable and requireo ,

features and models specified ia Appendix K to 10 CFR Part SC ,

for performing ECCS performance analyses, considerable  ;

research hos been performed that has greatly increased the understanding of ECCS performance during a LOCA. We have ncw confimed that the methods specifico in Appendix K, combined with other analysis methods currently in use, are highly j conservative and that the actual claading temperatures which I

The Comissioners 3 l

would occur during a postulated LOCA would be much lower than i those calculated using Appendix K methods. In addition, the large body of research available has provided a method to both estimate the degree of conservatism in Appendix K calculations and, to determine to a reasonable extent, the uncertainty associated with that estimate. In light of these factors, the Comirsion approved the publication of the proposeo rule which would permit licensees and applicants to make realistic calculations of ECCS performance during a LOCA in the regulatory process, as well as to require an estimate of the uncertainty of the calculation to assure that there is a high probability that the acceptance criteric in 10 CFR 50.46 (b) (e.g., calculated peak cladding temperature shall not exceed 2200'F) would not be exceeded. The Comission paper which transmitted the proposed rule (SECY 86-310) contained a discussion of the Staff's perception of the inherent conservatism in the acceptance criteria in 150.46.

Enclosure J provides further developnient of the Staff's views concerning the amount of margin in these criteria.

In considering the staff's recomendation to approve publica-tion of the proposed rule, the Comission directed the staff to subject the methodology for evaluating the uncertainty in NRC codes that was developed by.NRC to both peer and ACRS review and to solicit public coment on several specific questions (Enclosure B). The staff was directed to seek public coment on an ACRS question concerning the indefinite grandfathering of plants with acceptable Appendix K models.

The staff was also instructed to submit three cuestions pcsed by Comissicner Asselstine for public coment. fiamely:

1. Should this rule change include an explicit degree of conservatism that must be applied to the evaluation models?

2. Should this rule change explicitly prohibit any increase in approved power levels until all severe accident issues and unresolved safety issues are resolved?

3. Should the technical basis for this proposed rule change be reviewed by an independent group; such as the American Physical Society?

In accordance with the Comission's request, the ACRS reviewed the NRC methcdology for ECCS code uncertainty evaluation on September 10, 1967 and again on May 5-7, 1986.

(EnclosureC). In adottion, review of the NRC methodology by an independent group of experts was also conducteo in January 1988. The sumary report of this independent panel,

The Comissioners 4 which was chaired by Professor Todreas from MIT, has been

• received and is also provided in Enclosure C. Both the ACRS and the expert'5 panel made generally favorable coments concerning the methodology; however, both groups recognized that a complete demonstration (i.e., application to small break LOCA and the reflood portion of large break LOCA) has not yet been accomplished and certain reviewers questioned whether such a demonstration could be performed successfully.

The only objectives of the NRC methodology demonstration are to demonstrate feasibility, to develop an audit tool, and to provide the necessary experience to audit licensee submittals. The staff does not believe that NRC demonstration of the methodology is a prerequisite to this rulemaking. Licensees wishing to adopt the best estim6te appro6ch permitted as a result of this rule are neither required to use this methodology nor to model their own methodologies on it. The NRC has detemined through twenty years of experience that independent analysis with independent methodologies is the most offective way to intelligently review new vendor or licensee methodologies.

It is therefore appropriate that this new methodology be subjected to stringent technical scrutiny, as directed by the Comission. The NRC St6ff is comitted to completing this ,

demonstration by the time that it will be needed to review licensee submittals and is confident that such a demonstration will be successful.

The coment ptriod for the proposed rule revision expired on July 1, 1987. Twenty-seven letters addressino the proposed rule change were received. In addition, six letters comenting on the proposed rule were received af ter the expiration of the coment period. These were also considered by the staff to the extent that new and substantial coments were provided. A list of the rule comenters, ano a sumary of the coment letters are provided in Enclosure D.

The bulk of the comments received on the rule came from the nuclear industry and were largely suppcrtive of the actiun proposto. These coments suggested a number of minor revisions to the rule, which are not discusseo here, but can be found in Enclosure E. One negative comment was received from an anonymous comenter but specific objections were not provided. Two comenters recomended that the rule not be implemented until other safety issues are resolveo. A detailed analysis of the public coments may be found in the Federal Register Notice (Enclosure E) including proposed responses to the public coments on questions posed by the

he C <nissioners 5 f

ACRS and Comissioner Asselstine, as well as several other <

coments that recomnended major changes, In conjunction with the publication of the propost;o dmendments to 10 CFR 50.46 and Appendix K, the NRC staff prepared a regulatory guide to set forth the staff's views concerning acceptable procedures for compliance. This regulatory guide, entitled "Best Estimate Calculations of Emergency Core Cooling System Parformance," was also released for public coment to assist the public in understanding the proposed revisions and to allow public participation in their development. This guide describes features that a realistic ECCS evaluation model should contain and guidance on performing the uncertainty evaluation. The regulatory guide lists a number of models and corresponding experimental data that are considered suitable for use. The comenters on the regulatory guide largely supported the inclusion of experim. ital data that was acceptable and models and com elations that fit the data acceptably well. Therefore, these references have been included in the final guide with the statement that other mooels and correlations will be considered but must be justified with appropriate experimental data. This regulatory guide is provided as Enclosure F and a list of comenters on the guide and a paraphrased sumary of guide coments is provided in Enclosure G.

In order to apprise licensees and applicants of the large body of research which supports realistic calculations of l ECCS performance, the staff has prepared hUREG-1230, "Compendium of ECCS Research for Realistic LOCA Analysis." l This report identifies the relevant ECCS research perfomeG t and describes the NRC developed methodology for the estimation of the uncertainty of thermal-hydraulic safety analysis codes. This report also ccntains documentation of NRC studies of effects of reactor power increases on risk, l background infomation on the ECCS rule, and a description of i the NRC develeped methodology for estimating L thermal-hydraulic code uncertainty. The staff feels that this document will provide valuable guidance tu licensees and applicants wishing to avail themselves of the benefits of i perfoming realistic LOCA analyses to evaluate ECCS c

! performance.

Baseo on the coments received and considered and 1he large experimental date base available at this time, the staff recomends that the Comission approve for publication the j l

final rule as contained in the Federal h gister Notice t (Enclosure E). This final rule does not alter the [

r l

o l -

The Comissioners 6 1

' s requirements found in the notice of proposed rulemaking (EnclosureA). A regulatory analysis of the final rule is provided in Enclosure !.

Resource The NRC staff resources to implement this rule are thought Estimates: to be negligible under the assumption that no unusual or special rulemaking procedures (e.g., adjudicatory hearings) will be established by the Comission. If the Comission chooses to hold hearings, resources would have to be divertea from other high priority activities.

The major staff resources required under the rule change will be to review the realistic models and uncertainty analysis that may be submitted by licensees or applicants wishino to utilize the revised ECCS Rule. Based on previous experience with the General Electric Co. SAFER model and the learning that has resulted from this effort, it is estimated that approximately one staff year will be required to review each generic model submitted. There are four major reactor vendors (GE already has a revised evaluation model approved under the existing Appendix K for both jet-pump and non-jet pump plants that is consistent with SECY 83-472 and mt.ets the requirements of the revised ECCS rule proposed herein) and several fuel suppliers and utilities which perform their own analyses and potentially might submit generic models for review. However, it is expected that only 3 or 4 generic models would be subNitted since not all plants would benefit from the rule. Thus, approximately three to four staff years will be required to review the expected generic models. Once a generic mocel is approved, the plant specific review is expected to be very straightforward.

Reccmendations: That the Comission:

1. Approve the publication of final amendments, as set forth in Enclosure E, which would permit the use of realistic evaluation models and the accompanying estimate of the uncertainty of the calculation to demonstrate that the acceptance criteria contained in Section (b) of 10 CFR 50.46 are not exceeded.

2. Note:

a. That the notice of final rulemaking in Enclosure E will be published in the Federal Register to be effective 30 cays after publication,

b. That the regulatory guide in Enclosure F will be published concurrent with this rule.

The Connissioners 7

c. That pursuant to il 51.21 and 51.31 of 10 CFR Ptrt 51 of the Comission's regulations, an environmental assessment and finding of no significant impact is attached as Enclosure H.

d. That the reporting requirements in connection with analyses required by the rule (Enclosure E) impose information collection requirements that are subject to the Paperwork Reduction Act. The requirements were approveo by OMB.

e. That pursuant to the Regulatory Flexibility Act of 1980 the rule contains a statement that the Comission certifies that the rule will not, if promulgated, have a significant economic impact upon a substantial number of small entities and a copy of this certification will be forwarded to the Chief Counsel for Advocacy SBA by the Division of Rules and Records, ADM.

f. That the Subcommittee on Nuclear Regulation of the Senate Comittee on Environnient and Public Works, the Subcommittee on Energy and the Envirunment uf the House Comittee on Interior and Insular Affairs, the Subconnittee on Energy Conservation and Commerce, and the Subccm ittee on Environrent.

Energy and Natural Resources of the House Committee n Government Operations will be informed,

g. That a Regulatory Analysis is attached as Enclosure 1.

h. That a public announcement will tie issued,

i. That copies of the flotice of Final Rulemaking will be distributed by T!DC, ADM to each affected licensee and other interested parties.

j. That the staff recomends the paper be placed in the PDR. .

k. That this paper has been reviewed with the ACRS Subcomittee on Them41-Hycraulic Phenomena on April 21, 1988, and they have indicated that they have no objection to thie, rule. The ACRS has indicated that it wishes to be apprised of the progress made on th3 demonstration of the uncertainty methodology. The staff will continue

1 The Commissioners 8~ ]

I l

l 4

to brief the ACRS until the demonstration is complete.

1. That this proposed rule change has been concurred in by the Offices of NRR and ARM.

m. 0GC has reviewed the rule and' regulatory guide end has no legal objections.

Schedulino: The staff recommends that this action be affirmed through a notation vote. No specific circumstance is known to the staff which would require Commission action by any particular date in the near future, f

./

I

[ (f/,V ictor Stelldh' r.

Executive Director for Operations

Enclosures:

A. Notice of Proposed Rulemaking B. Memorandum Chilk to Stello, dtd 1/9/87 C. ACRS and Peer Review Sumnaries D. Summary of Public Comment E. Notice of Final Rulemaking F. Draft Regulatory Guide G. Summary of Public Conment on Guide H. F.nvironmental Assessment I. Regulato"y Ane. lysis J. Matlin Iriherent in 2200'F Limit C

4 9

f

o 9

Commissioners' comments should be provided directly to the Office of the Secretary by c.o.b. Tuesday, June 28, 1988.

Commission Staff Office comments, if any, should be submitted I to the Commissioners NLT Monday, June 20, 1988, with an infor-mation copy to the Office of the Secretary. If the paper is of such a nature that it requires additional time for analytical review and comment, the Commissioners and the Secretariat should be apprised of when comments may be expected.

This paper is tentatively scheduled for affirmation at an Open l Meeting during the Week of June 27, 1988. Please refer to the l appropriate Weekly Commission Schedule, when published, for a specific date and time.

DISTRIBUTION:

l Commissioners OGC OI OIA GPA REGIONAL OFFICES EDO ACRS ASLBP ASLAP SECY

y:www.rc2.. :a- _ _ = ., .. ... . ... . _ _ : . : . . . ... . . .; . . . . . . . . .. . a.e .

l I

1 I

e ENCLOSURE A L

I h

- - . ,z w.n m.mw m.m. =.= =. . .-  ;

eu e

Proposed Rules r-2 =+=

Vol. 52. No. et Tuesday. March 3. tes?

Ths sectaon of the FEDERAL REGISTER ADOREssts: Submit wntten comments ECCS performance during a !$OCA.The conta,ns notices to the put*c of the to the Secreisry of the Commisalon. U.S. methods specified in Appendix K.

proposed asuance cd ruks and Nuclear Regulatory Commission, combined with other snelysis methods reguladons. Trw poose of mese nouce$ Washington. DC 20555. Attention currently in use. ore now known to be o to y etnsted persons an Docketing and Service Branch. Hand highly conservative: that is, tne tctual copo%ng a pamcca4 M me W

***9"*"*** deliver cornments to Room 1121.1717 H temperatures during a LOCA would be

[** ' P" ' '

Street NW., Washington. DC betwe".

8:15 a.m. and 5:00 p.m. Examine much less than the temperatures calculated using Appendix X methods.

comments received, the environmental The ECCS research he: gone beyond NUCLEAR REGULATORY assessment and finding of no signihcan' confirming that Appendix X is COMMISSION impact, and the regulatory analysis at conservative;it has allowed the Cornmission's Public Document quantification of that conservatism.ne 10 CFR Part 50 Room at 1717 H Street. NW- results of experiments, computer code Washington. DC. Obtain single copies of development, and code a;sessment now Emergency Core Ccoling Systems; the environmental assessment and Rev!alone to Acceptance Crtteria allow more realistic calculaUons of finding of no significant impact and the ECCS perform 3nce during a 1,0CA.

regulatory analysis from LM. Shotkin, along with reasonable estimates of ActucY: Nuclear Regu;atory Office of Nuclear Regulatory Research. uncertainty, than ti possible using Commission. Washington. DC 20555, telephone (301) current evaluation models.

Actiow: Proposed rule. 443-7825. It is also known that some plan:s are Foo FunTHan INFOAseAT)oN C0wTACT: now restricted in operating flexibility by sUWhsARY:The Nuclear Reguj. e to 7 LM. Shotkin. Office of Nuclear limits resulting from conservative Commitsion (NRC)is proposing en Regulatory Research. U.S. Nuclear calculationa using current models and amendment that would allow the use of Regula tory Commission. Wa shington. Appendix K requirements. nese alternative rnethods to demonstrate that DC 20555, telephone (301) 443-7825. restnctions may be preventing optimal the emergency core cooling system suppt.ansawfAny impont Arton: operation of these plants. Based on (ECCS) would proted the nuclear research performed. It is now known

. reactor core during a postulated design Background that these restrictions can be relaxed basis loss of coolant accident (LOCA). Section 50.46 of 10 CFR Part 50 because ofimproved knowledge of The amendment is proposed because provides Acceptance Cntena for safety margins.

research, performed since the current On December 8.1978, the NRC Emergency Core Cooling Systems rule was wntten. has shown that (ECCS)in Light Water Nuclear Power published an advance notice of calculations performed vems current proposed rulemaking (43 FR 57157)

Reactors." This section requires that methods and in acccrdance with the calling for a two phese speroach to the current requirements result in estiinates calculations ofloss of coolant accidents (LOCA1 be performed to shew that the revision of to CFR Part 50 and Appendi x o; cooling system performance that am K. The firs: sep would have been to ECCF mti r.mintain claddin4 sigt.ificantly rnors eservative than temperatures, cladding oxidition and tede procedural changes red to pwmf; es'imates based o. tre improved hydrogen generation to wAjn certain nisor 2 chnical changes which

• ould knowledge gained Imn this researcit in specif nes have red aced the crmers atism add. tion. the operataca cf some nucleir g,.ied.,g, hmits. lt alsoLa cy, pog,./y teWra maintair.ed that a :entainedin Ap enoix K.7he secor,d meters is being unnecessaril; sed that long term decay heat removal phase would hav;e made f.trett restricted by the rule, resalung in be pr<rvided. Append x K to 10 CFR Part technical changes based oc resea cb increased costa of eisetricity genentics. 50 sets forth certam required and results and t perating experimee.

The proposed rule, wtule continuirg t; sco2ptable features of the modals t. sed NRC actfuty on tiie ECC3 rulemsking allow the use of cwremt methods and ta perictm th-se calcuitione, The was severely curtsited as a result of the requirements, would also allow de ute criteria of to CR 50 43 and the bid. pr;ority efforts required by the cf more recent informat,on and calculauenal methodt specified in TM1-2 sceident.nja rulemaking was knowledge to demonstrate that th' Appendix X were fcriaally invued in dormant until [uly 1981

• bet. it was ECCS would preted the reacter darity a lanuary 1974 after extensive rulemal!ng revtved in the context nf si:rpilfyir3 and LOCA. The proposed amendmern, vrac hearings and are based cn tts sireomli ting the regulatory process.

would apply to all applict.nts for er.d understanding of FCC3 performance %e NRC has reviewed the comrnenta bolders of construction permits or ava lable at that time, made by outside organizations on the operating licenses Icelight water In the thirteen years fol'cwing the advance notice of proposed rulemaking reactors, would also relax rnuirements rulemaking, the NRC, the Department of as well as a number of other comments '

for certsin resnelysse which do not Energy (including the Atomic Energy received since that time. In general the l

- contnbute to safery. Commission and the Energy Research commenters support a rule change that l l oAfts: Comment period axpires July 1. and Development Agency). U.S. nuclear would permit greater flexibility in

1987. Comments r2cshed after that date industry and foreign resestchers have meeting the regulations and would will be considered ifis is practical to do obtained considerable information on incorporate the use of presently so, but assurance d ecessideration ECCS performance.The majority of this available research information. Many

. cannot be given excee as to comments LOCA research is complete and has felt that the Phase 1 scope should be l received on or before ibat date, peedy tmproved of understanding of expanded to allow the use of additiorial

0

- - - - . - , , . , . , , , . - - - , - - - - - , , - , - - , , , , - , ,. . . - , ,-y , - , , . .

,. _ o. ..._.m.m. _~ _ _ _ _ _ _ _ _ - _ _

Federd Register ] Vol. S2. Ns. 41 / Tuesday, MaJch 3,1987 / Proposed Rules $335 -

information available from completed would not be exceeded. N stan requirements of ll 50 46(a)(1) (i) and (ii).

ECCS research. expects the realistic evaluation model to Because of thi.s revision, the last Becarse of the delay la changing the retain a degree of conservatism seatence of the existing 150.48(a)(1) has ECCS rule, the NRC has used an interim consistent with the uncertainty of the been deleted in the redesignated approach. descnbed in SECY-83-472.' calculation. The proposed rule would $ 50.46(a)(1)(i).

t3 accommodate requests for improved not specifically prescribe the analytical ' & current rule contains no explicit evaluation modela, submitted by reactor methods or uncertainty evaluation requiremects concerning reporting and vendors, for the purpose of reducing techniques to be used. However. reanalysis wheo errors inirvaluation ..

reactor oparating restrictions. This guidance would be provided in the form models are diamvered or changes are interim approach requires a reallatic of a Regulatory Culde.'It should be made to evaluation models. However, calculation, with an evaluation of the noted, as discussed in SECY-83-472 current practice has required reporting j uncert:Inty in the calculation, to that the NRC has,in the past, found of errors and changes.& proposed .

I demonstrate that the improved acceptable a method for e:Urnating the

' rule would exphcity set forth l eval:adon model maintains an edequato uncertainty that was judged to be at reqairements to be followed in the event conservattaa of sefety factor. leset at the 95% probability level. nis The NRC has dedded to pnmed with of errors or changes. N definition of a probability levef of 95% is considered by significant dange la curnntly taken the rulemaking. but in the form of a the staff to meet the high level of more comprdensive amendment besad from Appndix K. Sectico Ill.b whid probability nquired by the rule. It is defines a sisnificant change as one en (1) the casamenta meened since the slao reccTnfred that the probability which changes calculated cladding pubucatica of the 1974 notice of cannot be dete-tned usirg totally proposed rih % (2) the additional temperuure by more than 20* F. The rigoroes mathematmal memods due to research ooeducted and experance revised 150.46(a)(3) would state specific the complexity of the enenle* tons.

gelned since the 1978 monce, and (3) requirements for reportna and However, the staff expe 1s that any reanalyses when errors in evaluaticc recent cxpenence applymg the methods simplifying assumptforr will be stated cf SECY-43.-472, models are discovered or changes are so that the sis # may evaluate them to made to evajuauon models. it would A report. "Compendium of ECCS ensure that they are reesonable.

Research for Reeltstic LOCA Analysla." require that all changes or errors in Appendix K. Section Il "Required ap roved evaluation models be reported NUREC 1230 la being prepared. It Documentation." would remain summ:rizes the extensive ECCS

,, feest annually and would not r= quire generally applicable, with only minor research that has been conducted. A revleions made to be consistent with the any the further action by the beensee until error is reported. Dereafter.

%N Jraft version of this report will be amended rule. *

- available et the NRC Public Document la d A new 150 46(a)(1L?t would be added y'j16 "f,c*,,[,"1 of uch o7r r, any Room.1717 H Street NW., Washington, to allow the features of Section I of DC 20555. 30 days foBowing publication Appendix K to be used in evaluation '"bs ' I" f

• E p u res da 5 of this proposed rule ti the Faderal models as an alternative to perfonning Regist:r. the uncertainty evaluation specified in nor A modeJ withgauch e.g.g. gg ,,g, and anyg,9 prior Summary of Proposed Rule Changes the amended i 50 46(a)(1)i). This method appsd d even h em or I would remain acceptable because Secrson M46 Accept.2ence Critere for /

Appendix K is conservative with respect changes in order to enswe that j they y Eawymcy Corv Cochnt Systxr.s for to the realistic method proposed in the * *i0 g PP; Ligh WcterReccaces ment j 'he d 5 amended i So.46(a)(1)(il. This would f Sectk>n 5040(a)(1) would be amended a!!ow both current and futua spplicants error or change and to insintain a and licemeef. to uw esisting evaluatior. l' snow edp ni mo6fications i arid redsogna4d 8 $0 46(a)it)(1) to dcice the requk ememt that the features models if they did not r-ced or desir' made sicce staff.ruew cf the l i'#

of Secticu t of Ap,>endix K to 10 CFR relief from current cperating r striedons. evolustion model. However. past la I $Ct44 paragraphs (a)(2) sad (3) experience has shown that many orrors part M be used to develop the or changes to evaluation mode!s are svaluation modet. %is section would would be totally rev sed to eliminate Y very minor and the burden of domedate a require that t.n acceptable evaluation medit have sufficier , supporting poM a pjp p f ,7g reporting cannot be justified for s.u.h minor er ors because they do act affect j

justification to show that the analytical c rde Th Wdb g g dte ed the o$. wing the immediate safety o7 operation of the -

technique realistically descrbes the pala N staff has tbsteiore prepwed behavior of the rea<. tor system during a h P"l"gP bA&(2) n wou,d be revised to Peri:6e reponing ic atisfy the n2ed to  ;

LOCA.The staff expects that the indicate that tenrictiol.a on reacto, de epprised of changes or errors witnout 4 enti tlest tecJ nique will. to the extent f

practicable, utilize realistic methods and opera 4ca w W imposed by the ;rovidtog u-due burden on the opphcant l be ksed upon applicable expenmental Director of Nuclear Reactor Regulation or licenses. Such report is to be ftled within one year of discovery of the error j

data. The amended rule would also if the ECCS cooling pe forniance ,

evaluations are not consi. stent with the and shall be reposed each year {

require that the uncertainty of the calculation be estimated and accounted thereafter until a revised essJustion j e a, arvcm w ima model or a revised evaluation conectmg for when campermg the results of the cate*uons of r.a useecy cor, coouns s,w,u such errors is approved by the NRC {

s calcu!ation to the temperatute limite and p ,c ,mnce .,n we ed io m.cen u and taff. .

other ettteria of 150 46(b) so that there

=sti be eM.Me for meee' ion si the comm.wan i d Significant errors may requirt more is a high probabihty that the er,teria gg*y** g1,]p,,Qc*,,,,,, ,, timely attention smce tr.e> rnay be im.p.opo divi,in u. r.de.Ja sw ..a. wen important to the safe operation of the a Vrt.al-4 *t. "Ymyevu y Core Onuhat % vem for s..A cop es of the dr.ft so.de. whch man > be plgnt, & pro poged ggjg rgyjgion w oyld wi.. wnode N..mo, u inn i ....w. repreired should be mede in = riss to ni+ t's r.,, . asp,c.na .ad eern.ar , . r. .i m. Nxin, amtewy car.atue w..mnr% oc define a si8nificanI error or ch n8e ae es, w n.w tu.,w anoe .a mi sa mud, A%.uos Direcsw. LAsion of Tu.hal one which results in a calculated peak v Nw .w h.ne,.a 'w. Inform.aon and twe..cas coaN fuel cladding temperature ddlerent by

.-m. ,. -- -

c.,.---

w. . :.: a :.s . . .:3. CazM.: a W.a.o .: x a. w. ~. L. w .. .~ --.. - a....

a 6336 Federtl Register / Vol. 52, No. 41/ Tuesday, March 3.1987 / Proposed Rules _-

calculations using NRC*a best estimate Safeguards requests the public's more than 50' F. or an accumulation of computer codes.ne cost of revising the comments on whether the existing rule errors and changes such that the sum of should be "grandf athered" indefintely.

the absolute magnitude of the evaluation models would be high with no real benefit to safety.Thus requiring That is:

temperature cr.anges is greater than 50' 1. Should the conservative ECCS .

F. Mere timely reporting (30 days) would that the applicants and licensees remove evaluation method of Appendix K be be required for such errors or changes. the Dougall.Rohsenow conelation frorn their evaluation models could not be permitted indefinitely or should this This definition of a significant change is aspect of the ECCS rule be phased out based on staff judgement concerning the justified on a cost. benefit basis. after some period of timet importance of errors and changes A new Section I.C.5.c would be added J Further. Commissioner Aaselstine typically reported to the staff in the past. to Appendix X to state the requests the public's comments on the The proposed rule revision would also Commission's requirements regarding following:

allow the staff to determine the schedule continued use of the Dougall.Rohsenow 2. Should this rule change include en for reanalysis based on the importance correlation in existing' evaluation models. Evaluation models which make explicit degree of conservatism that to safety relative to other applicant or liceane requirements. Errors or changes use of the Dougall.Rohsenow correlation must be applied to the evaluation that would result in the calculated plant and have been approved prior to the modelst effective date of thl proposed rule 3.nis rule char.ge would allow a 5 to perfonrance exceeding any of the to percent increase in the fission revision may contiene to use this en'4ria d 150.46(b) would mean that correlation as long m no changes are product inventory that could be released tl e plant is not operating within the from any core meltdown scenario.

requirements of the regulations and made to the evaluate model which significantly reduce E a current overall Should this rule change explicity would require i;amediate reporting as prohibit any increase in approved power tewred by 150.55(e). I 50J2 and $ 50.73 conservatism of the evaluation model.lf ao immediate steps to bnna the plant the applicant orlicensee submits levels untti all severe accident issues proposed changes to an approved and unresolved safety issues are irno compliance with l 50 46.

Appendix K ECCS Evaluation Models. evaluation rsodel, or submits corrections resolved?

to errors in tl.e evaluauon model which 4. Should the technical basis for this Amendments would be made to significantly reduce the existing overall proposed rule change be reviewel by an Appendix K,Section I.C.5 b to modify cor.servatism of the model, continued independent group such as the post.CHF beat transfer correlations use of the Dougall.Rohsenow correlation American Physical Society?

listed'as acceptable.The "McDonough" reference would be replaced with a later under conditions where nonconservative finding Of No SignlAcant Envirammaatal paper which is more generally available heat transfer coefficients result would Impact Availability and which includes additional data, no longer be acceptable. For this purpose, a significant reduction in overall ne Commission has determined The heat transfer comlation of under the National Environmental Policy Dougall and Rohsenow. listed as an conservausm has been defined ss a "net" reduction in calculated peak clad Act of 1969, as amended, and the acceptable heat transfer correlation in Commission's regulations in Subpart A temperature of at least 50' F from that Appendix K. paragraph 1.C.5 b. would be which would have been calculated using of to CFR Part 51, that this rule,if removed under the proposed rule adopted, would not be a major Federal e revision. Research performed since exf sting evaluation models A reduction in calculated peak clad temperature ac. ions significantl y affecting the e,ualjty  ;

Appendix K was written has shown that could potentiaUy result in an increaseofinthe human environmer.t and therefore this correlation overpredicts heat an environtacntal impact itaterr.ent is  :

the actual allowed peak power in the I transfer coefficients under certe.in plant. An increase in allowed plant peak not tequired.The primary affect of the condillons and therefore can produce power with a known nonconservatism in rule would be to allow an incrvese in t e norconse.vativ s results. Strice the peak local pewer in the reactor. %is a the analys:s would be unacceptable.

Dougall.Rohsenow corretnion is now nie defintior., of a signiEes at redation couk be used to either isdor the power known 'o be nonconservauve under I')

in overall conae votism la based on staff abspe withit the rose:ot er increue the '

certain conditions it is appropriate to no judgement ressrding the size of the total power. Chanpng the pewar shapa 1:nger reference it as a generally existing overall conservatism in without changing the tots! power would acceptable corre:stion. A number of evaluation model calculations relative to have a negligible adochin the ,'

I applicants and liwneees currently use the conservatism required to account for environ.sentM imoect %e total power I the Devgall.Rohsenow correlation in cou!d also be inenssed but weuld be approved evaluation models. Because of overall uncertainties in the calculations. expected to be increawd by to more  ;

this, the staff has considered how this Appendix K, Section U.1.b. would be removed since this requirernent would than about 54 to los due to hardware change should be implemerted.There is limitations in existing plants.%!s 5% to ns justification on grounds of safety for be clanfied under A arsehded i 50.56(a)(3).1.ikewise. Apper. dix K, 10% power increese is not expected to requiring that applicant and lice. Jes Section ILS, would be amended to cause diffiedry in meeting the existing making use of Dougall.Rohsenow rerse environmentallimita.h only change in account for the fact that not all their evaluation models at this time. evaluation models will be required to non.rsdiological waste would be an his is appropiate (even though part of increase in waste heat rejection the approved evaluaron model.Dougall. 1.These use theminor features of Appendix !l. Sectioncomrnensurate with any increase in changes to Appendix X Rohsenow. is now kun to be will not affect any existing approved power.For stations operating with an ,

nonconservative) because the existing evaluation models since the changes are open (once through) cooling system. this evaluation models are known to contam either "housekeeping" in nature o; are additional best would be directed to a a large degree of overall conservatism surface water body. Discharge of this ,

evm while using the Dougall.Rohnenow changes to "acceptable featuree not "required features." heat la regulated under the Clean Water i correlation.%is large overall .

Act administered by the U.S. i With respect to the proposed rule i conservatism ha s been demcnstrated Environmental Protection Agency (EPA) through comparisons between changes idenufied in this summary, the Adivsory Committee on Reactor or designated state agencies. It is not evaluation model calculations and

. '.. & -. .u.= ..:.a a J.a.. d I.. JM.E..k u u'&. . a .

5,&a...-

Federal Register / Vol. 52. No. 41/ Tuesday. March 3.1987 / Proposed Rules $337 o  %.

intended that NRC approval of not fall witain the purview of the Act.

Increased power level affect In any way fission prodneta could be released  !-

the responsibl!!ty of the licensee to Bedm 61ysis during cors melt acenarios and h; comply with the requirements of the potentially released to the environment Although a backfit analysia la not during severs accidenta.

Clean Water Act.The environmental required by 10 CFR 50.109 because the t ne proposed rule would still require D casessment and finding of no significant proposed rule does not reqrire impact on which this determination is that fuel rod peak c.! adding tempera ture >

applicants or licensees to make a based an seaflable for inspection at the change but only offers additional (PCT) remain below 220Q'F. Because i NRC Pubile Documeni Room.1717 H options, the factoss in to CFR 50.109(c) research indicates that significant ioel f damage will not occur until 2600*F. a Street.NW. Washington. DC. Single have been analyzed as indicated below.

copies of the environmental assesstnen! 400*F safety margin will remain. .

More detailed information relevant to However, reactore choosing to increase and the finding of no significant impact this backfit analysis may be found in the power by ove to ten percent would be  ;

, an available from LR Shotk;n. Office regulatory analysta referenced above.

of Nuclear Result o y Research. U.S. operating with less margin between the Nuclese Roqulatc>ry Commission. f,grat,,,3fofgA,,p,effje,3j,etjy,, PCT and the 2200*F limit than -

that the previoualy. The increased risk Washington DC.2601 telephone (301) oc3j,v, proposed backfit i.s designed to represented by this decrea se in margin 44 mal 5.

and increase in fission product We ob4ecun of th propoeed rule 15 inventory is nsgligible and falls within Paperwoes W~lwh Act Stalaammal to saodify 10 CFR 50.48 and Appendix X to partnP the use of realastic ECCS the uncertai ties of PRA risk estimates.

Als propoeed rule amends in addiUon, other safety limita, such as infirmadon collection requirements that evaluation models. Mars reshatic departure from nucleate boiling (DNB).

are subject to the Peperwork Reduction estimates of ECCS performance, based and operational hmits, such as turbine Act o(19eo (44 U.S C. 3501 et seq.). This on the impmved kmwledge sained from design. would limit the amount of rule b=a been sabeltted to the Office of recerrt research on ECCS performance. margin reduction permitted under the Management and Budget for review and Wou M88" natrictions. ' ut"""fy CP'ratin8 revised rule.ne proposed rule could e*

approval of the paperwork also potentially reduce the risk from requirements. 2 Cenerc/ description of the activity pressurized thermal shock by allowing that would be required by the licensee the reactor to be operated in a manner Regulatoey Analysis or applicoat in order to cornp/ete ib, which mduces the neutron Auence te the ne Omn=1uton haa prepared a draft bocaf,, the vessel.  ;

• regulat:ry analysis for this proposed The proposed amendment would ( Potectio/ impact on radic/ogicc/

regulatirn. The *nalysis examines the cost and benents of the alternatives allow alternative methods to be used to exposure to facility employees j demor.ztrate that the ECCS would considered by the Commission. De protect the nuclear reector core during a Smee the pnmary effect of the dral ter.datory analysis is available foe proposed rule invoh es the calculational 1 portulated design basis losa-of-coolant insp< ction and copying for a fee at the accident (l.OCA). While continuing t methods to be used in determining the d NRC bblic Document Room.1717 H Street NW. Waah'.cston, DC Single allow the cae of current Appendix X ECC cociing performsnce,it is espected metnods an i ter,u raments, the proposed that there will be an insignific. ant irnpact j

copies of the analysis may be obtained rule would also aUow the use of more on the radiological exposure to faci!!!y I fron LM. Shotkin. Office of Nrclear recent information and xoowledge er.iployees.Because of the educed I Reguletory Reseerd. Washington DC currentiv available to demonatrat: that LOCA restrictions resulths from the l 20555, tele,pbone (301) 44M82.5. the ECCS would perform ita safety new cdculam.u a is psamle for the k ne Comadaaioo aquesta pubbc plant to schlevt more efficien* operation ]

i fuoedoo during a LOCA. If an appliragt and irnproved fuel utilization with -

l com: rant on the draft aralysia. r licens.ce ele .ted to use a new *ea!is*ic I Ccarnenta on the drafi ansiysis may bc Inodel they would have to prnvide improv.t. it is conceiv sble that thereed maneuvering

,,,,g, f

! suberJtted to the NRC as indicated urder erif.cient sapportin41ua@ cation i 2 the Acosissees heading. vehdate the cMel and include cotJd be a reduction in radiological comparisons to expenmental da ta and exposure if the fuel reloads can be Regulatory Flexibility Certifica' ion estiinarea of ecertainty. lti ecceenug reoured.nis effect is not expected to k' for the uncertainty. the analysis would be vsr> significant-As requ! red by the Regulatory a

( have to show, with a high Ic$el of Flexibility Act of 1950. 5 U.S C. 605(b). I lastollation and conimuing ecsts i l

prabability, that the ECCS performance th) Commissicm certifies that this rule. if entsria are not exceeded.

o,,oc ated with the boekfit, includmy j the cost offocility down ' tames or the 9 n mic pa a su s 1 3 Psitrattalchange in risa to thepublic cost of constrtiction de/cy number of sinall antities. This proposed /mm the ocodento/ o// site release of LOCA considerations resulting from l rub afbets only the licensing and radioccisre materiola  ;

the present rule are restricttrg the >

operation of nuclear power plants. The The proposed rule could result in optimum production of nuclear eleuric i companies that own these plants do not increased local power within the reactor power in some plants. Dese resinctions S l fall within the scope of the definition of core and possibly incresses in total can be placed into the followinit three .

"small entities" set forth in the power Power increases on the order of categones:

Regulatory Flexibihty Act or the Small 5-10% will have an insignificant effect III Mastmum plant operatirg pow er.

Business Size Standards set out in on nsk. One effect of increased power (2) Operational flexibilay and r;gulations issued by the Small Business would be to meresse the fission product operational efficiency of the plant.

- Administrauon in 13 CFR Part 221. Since inventery. A five percent power increase and I these companies are dominant in their would result in a five percent increase m l3) Asaildluhty of manpower to work on sernce creas, thia proposed rule does fission products. Thua, five percent more oiker actmties.

t

6338 rcd:ral Regist:r / Vol. 52. No. 41 / Tuesday, March 3.1987 / Proposed Rules l

estimated that approximately one staff a WhethertheproposedboeAfitis

> The effect of the proposed rule will vary from plant to plant. Some plants year would be required to review each interim orfinolandifinterim. the may realize savings of several million generic model submitted There are four justificction for imposing the pmposed ,

{

dollars per s ear in fuel and operating rnajor reactor vendors (CE already has a bocAfit on on interim basis costs. Signdicantly greater economic revised evaluation model approved The proposed rule, when made f benefit would t.e reahzed by plants able under the existing Appendix K for jet effective, would be done so in final form /

to increase total power as a result of the pump plants but is currently working on and not on an interim basis. it would ]

proposed rule.The regulatory analysis a new evaluation model for non. jet continue to permit the performance of . ,

cited above indicates that the total pump plants and may update their ECCS cooling calculations using either present value of the energy replacement methodology under a new rule) and realistic models or models in accord cost savings for a five percent power several fuel suppliers and utilities which with Appendix K. =

upgrade would vary between 10 and 150 f'"" h I' d List of Subjects la 10 CTR Part 50 million dollars depending on the plant n dly i ni generic models Additional information concernmg these for review. However. It is expected that Antitrust. Cla ssified informa"on. Fire potential cost savings are included in only 3 or 4 generic models would be prevention incorporation by raerence, the regu alory analpis. submitted since not all plants would Intergovernmental relations. Nuclear benefit from the rule change.Thus, power plants and reactors. Penalty.

6. The potentic/ sofety impoet of Radiation protection. Reactor siting changes in plant or operational about 3-4 staff years would be required

'enteria Reporting and Recordkeepmg complexity including the c'fect on other to review the expected genene models.

requirementa.

proposed and existing regulatory Once a generic modelis approved. the requirements plant specific review is very short. In For the reasons set out la the addition, several vendors are currently preamble and under the authority of the There are safety benefits denvable Atomic Energy Act of 1954, as amended, from alternative fuel management planning to submit realistic models in conjunction with the use of SECY the Energy Reorganization Act of1974.

schemes that could be utilized if the 472.Therefore, ataff resources would be as amended, and 5 J.S.C. 553. the NRC proposed changes were implemented.

The higher power peaking factors that expended to review these models in any la proposing to adopt the following event. Since these models would not amendments to to CFR Part 50.

would be allowed wi'h the revised rule could provide greater flexibility for fuel change as a result of the revised ECCS PART 50-DOMESTIC UCENSING OF designers when attempting to reduce rule there should be no net increase in PftODUCT10N AND UTluZATION neutron flux at the vessel. This can resources required over that already FACluTIES result in a corresponding reduction in planned to be expended. in summary.

while it is difficult to accurately 1. The authority citat'on for Part 60 ia e r du ed claddi g i mpera rs estimate. it is expected that the revised to ned as follows:

that would be calculated under the proposed rule change willhave a small gggg, proposed rule offers the possibility of overallimpact on NRC resources. 3g ,3,,,,3,ggg g37,gg g gg, other design and operational changes es4. 955, om as amended. Sec c4. as Stat.

that could result from the lower d. The potentialimpoet of differences in

• calculated temperatutes. ECCS facility type. design or age on the ""uf$**j$$$,'j equipment numbers, sizes or relevoney cadprocticolity of the an. ea amended. 2c2. 20s, as Stat.1242. as serveillance requirements might be proposed bocAfit an ended. un u48. (u u1C s441. 6a42.

reduced and still meet the ECCS dengn g criteria (il not required to meet other The degree to which the proposed rule section 50 7 also tr,saed at: der Pub.L 36-licrnsing requirements). Another option would effect a particular plant depends sot we tc. 92 Stat. 2nst tu UAC sast).

on how li:nited the plant is by tne LOCA secuen so to 61so teceed under seer. tct, tas.

may be to increese the diesel / generator es Stat. %e. us. es unended 147 U.S C ** tt. .

start time duration. rettrictror.a.The Babcock and Wilcox 2mL uc. to1 Puo. L 91-193. L Stat c3 (42 in summary, the effect of the proposed (B&W) and Combustion Engineering USC us213ectav.c.2A nu to.n So/A rule on safety would have both prential (CE) companies have informally *l** I"'ed ade, wu as M ess us positive and negative aspects.The indicated that they do not feel that the i potential for reduction of ECC systems ritets which they design are limited by p nd "$ s'io s7 PA in existing or new plants is present. LOCA and. therefore. B&W and CE L st-no, sa seat. au (u UAC 4322).

However, several positive aspects may plants weuld not be a81ected. Gener?! Secuons so2a and n54 t tao lasved urter elso be realized ur. der the propoud rule. Electric Co. (CE) plants do tend to be sac. ax. se stst.1243 (<2 UAC 644 i

While the ret effect on safety would be limited m operauon by LOCA sections nm sost and E92 also .ssued j sader Pub. s97-415. 96 bn **72 (42 USC.

plant specit'ic, the effect is believed to restrictions and would benefit from d*'

be small, relief from LOCA restrictions. However. $8NY[M3$'M","Sect $ '*

this reliefis elready available for most scJo-Est also waved under asc.164. es Stat.

7. The estimated resource burden on (At e64, as ame.sded U2 USC 2m4). Section NRC associated with (Ae proposed GE plants through the recently approved SAFER evaluation medel. Any 50.103 also leeued under sec. tea, es stat. s38.

backfit; and the oveilobility of such as smended (42 U AC 2134). Appendia F also resources additional relief due to a rule change The major staff resources required would be oflittle further benefit. L*88,ed under sec.1sf. es Sin 955 (u U.S C under the proposed rule change would Westinghouse (W) plants would appear g,, ,, ,g ,,, g g gg g to directly benefit from relaxation of s mended (u USC 2m), il so.to(s). Ibl.

be to review the realistic models and LOCA limits. W plants represent the uncertainty analysis required by the and (e). son so 4e. so sa. so_n and so sota) largest number of piants being m inued unde

• sec.1etb. 64 Stat. 944. se revised ECCS Ru!a. Based or, previous constructed. WIndicates that most of amended (u uSC ucitblh I so.to(b) and experience with the General Electric Co. (c) and 50.54 are issued under sec. tett. 64 SAFER model and the learning that has these plants are limited by LOCA Stat. 949. as amended (42 USC 22n(i)h and resulted from these efforts. it is considerations.

".;2M. ?.L.C..u. :.M ' ' ~ ' ' ' - -

. ' x.w G. . k .5 d..%.'.v .a asi. .L d.h :. u.u I

Federal Reg' ster / Vol. 52, No. 41/ Tuesday, March 3,1987 / Proposed Rules 6339 o

t il 50.5He),30.so(b). 50.70. 50.71. 50.72. 30.73.

(2) Restrictions on reactor operation Appendix K-ECCS Evolution Models and 503s are tuued u ider we.1sto. as Stat-550, as amended (42 U.S.C 22ot(oll. may be imposed by the Director of . . . * .

Nuclear Reactor Regulation ifit is found H 50.2, 50.10, 50.21, 50.22, 50.23, 50.30, that the evsluations of ECCS cooling  !

50.33, 50.33a, 50.34, 50.34, 50.37, 50.35, performance aubmitted are not ['8jt 9udj,CC*P, table Tuturn of the .

50.41, 50.a2, 50.43, 50.44, 50.47, 50.52, 50.54, consistent with paragrsphs (a)(1)(l) and 50.55, 50.55a, 50.54, 50.70, 50.50, 50.103 C R/owdown Phenomeno' '

• and Appwus6ees A, g, F, t., and 0 (ii) of this section. 5. Post-CHPHect Tromfer, IAmenced) (3)(i) Each applicant for or holder of Correlations ' '

• an operating IIcense or constsuction b. m Croeneveld flow furn boilms

2. The authority citatione foUowing comladon (equation 5.7 of D. C Gioeneveld.

ll 50.2. 50.10,50.21,50.2.:. 50.23,50.30, permit shall estimate the effect of any "An Investigation of Heat Transfer in the 50.33. 50.33a. 50.34. 50.35, 50.37, 50.38. change to or error in an acceptable IJguid Deficient Regime's." AECL-3241.

50.41. 50.42, 50.43. 50.44. 50.47, 50.53, evaluation model or in the appliestion of revised December isee) aod the 50.54, 50.55, 50.55a. 50.56. 50.70. 50.80, such a model to determine if the change Westin5 house correlauon of steady. state 50.103, and Appendices A, E F. I. and Q or error la significant For this purpose, a transition Rebuttal Tes boiling ("Proprietary Redirect /

are rem:ved. signiMcant change or error is one which e,,,,,,,3,n,tunony of westinghouse Dectric USNRC Dodet RM4o-1. page

3. In i 50.46. paragraph (a) is revised to results in a calculated peak fuel 25-1. October 2a.1s72) are acceptable for use 4, read as foDows: cladding temperature different by more in the post-CHT boiling regimaa. In add; tion than 50*F froni the temperature the tranaition boiling comleuon of 9 50.C3 Aeooptance arttwta foe omweeney calculated for the limiting transient McDonoush. Milich. and King (J. B.

core scoung systems foe tent watw using the last acceptable model, or is a McDonough. W. Milich. E.C King. "An nucteer posee roectors. cumulation of changes and errore such Expenmental Study of Partial Film Boilms (s)(1)(i) Each boiling and pressurized that the sum of Ge absoIute magnitudes Region with Water at Devated Pressures m a light water nuclear power reactor fueled of the respective temperature chan,qes is RW veh Th chemical Engmeenna with uranium oxide peUets within greater than 50 F. Progress Symposium Senes. Vol 57. No. 32.

pasu 197-2ce. (19e1) is suitable for un */

cylindrical Ziracioy cladding shau be (ii) For each change to or error prcvided with an emergency core discovered in an acceptable evolution between nucime and film beildms Un of cooling system (ECCS) which shall be au these comlanons shau be restncted as [y model er in the application of such a follo* 8' t8 designed such that its calculated cooling model which affects the temperature * * * *

• performance following postulated loss- calculation, the applicant orlicensee

~ cf coolant accidents conforms to the shaU report the nature of the change or

c. Evaluauon models approved after (effective date of rule) which make un of the

}a criteria set forth in paragraph (b) of this j error and its estmated effect on the Dougall.Rohunow flow film boilms

"' ' d section. ECCS cooling performance shall limiting ECCS analysis to the fM*n$',",I.RpSm Bo n o s t e be calculated in accordance with an Commission at least aanually as de of acceptable evaluation model and shall Vemeal Tubes with Upwe'd Flow of Flu:d aI specified in i 50.4. If the change or error to. (A, hn,.. M;T Report Number 90*9 26.

be calculated for a number of postulated la significant, the applicant or licensee a' has ef coolant accidents of different Cambndge Massachusetts. September 1963) i shall provide this report within 30 days shall not un this comlation under condihons sizes. locat!ons. and other propertjes 4 sufficient to provide assurance that the and include with the report a proposed where nonconuruhve predictions cf heat schedule for providinir a reanalysis or  !

most severe pcstulated loss-of-coolant transfer result. Evaluahon models which taking wh other action as may be ake un of the hugaMohunow q<

accidents are calculated. Except as com?aton and wue approud pne to needed to show compliance with ! 50 46 providtd in paragraph (s)(1)(ii) of this

• $t s ch' t J

section. the evaluation model shall requirements Jhls schedu:e may be I'"',*'l,' '

i include sufficient supportmg developed using an integrated

,c , ble made to, or en error is cometed in. the afa' chanse is scheduling system previously approved  !

justification to show that the analytical culuanon model that results m a signiricant  ;

for the facility by the NRC. For those technique realistically describes the facilities not using an NRC approved reduccon in the overall conserva tism i.s the j behavior of the reactor system during a enivenen modet At th4: tine contmued use  ;

loss of-coolant accident. Comparisons to integrated schedulmg systerr, a scheduls of the Dougall.Rohunow cortr'stion onder a epplicable experimental data shall be will be established by the NRC steJf condiuons where nonconursanu i within 60 daye of recei;:t of the proposed O" d ''' 'I h "' t#" ' '" " '"I' "'"I d "

made and uncertainties in the snalysts l method and inputs shall be identified schedule AnI chan8e or error cortection 8'gnificant ongn be acceptaW fw thu ,wu a ,

reduction m the ca!alsted peak and assessed so that the uncerta nty g that resul'ts in a calculated ECCS ru.) ci,ddmg t,rnperatst, of et i, si so*F 7' the calculated results can be estimated. performance that does not conform to from that whu would have been calculated 1 This uncertainty shall be accounted for the criteria set forth in paragraph (b) of on (effecnve date of rule) due either to e so that, when the calculated ECCS this section is a reportable event as individval changes or error comenons or tha described in 150.55(e). I 50.n and net eneet of an accumulanon of changes or 3

coolms pirformance is compared to the critens set forth in aragraph (b) of this 150.n. The affected applicant or ear comenons. {

licenses shall propose immediste steps II. Required Documanlation b 11 y t ten uld not to demonstrate compliance or bring )

i plant design or operation into 14**'  ;

e exceeded. Appendix K. Part II. b. A ccmplete listing of each ecmputet

' Re9uired Documentation. sets forth the compliance with ! 50.46 requirements. program, m the same form as used m the }

documentation requirements for each . . . . .

evaluishon model shall be furnished to the .

evaluation model. 4. In 10 CFR Part 50 Appendix K. Nuclear Regulatory Comm.ssion upon '

(ii) Altematively, an ECCS evaluation ,, p , , o paragraph !!.1.b is removed. paragraph . . . . .

model may be developed in II.1.c is redesignated Ill.b. the text of conformance with the required and paragraph 1.C.S.b and paragraph !!.1 b 5 General standuds for Acceptabaty-acceptable features of Appendix K. and 11.5 are revised, and a new section glement, of n,to,non moe,,, ,,uem ed at ECCS Evaluation Models, include the technical adewacy of the l.C.5 e added to read as follows: calevlanonal methods. meluding for models 1

',

• 6340 Federal Regist:r / Vol. 52. No. 41 / Tuesday, March 3.19T / Proposed Rules ce cred by 1 So 40111)(4 coinphance with 37. All submissions will be available for and government securities desier 8 da regared features of Section Icf this public inspection at the Commission a not include sny person registered with Appendix K. and, f or models cosered by the Commodity Futures Trading Public Reference Sec 'n.450 Fifth Street. N.W., Wa shmgton, DC 20549. Comrnission ("CFTC"), any contra et frha i 5: the pe anc iterN df market des;gnated by the CFIC.,such a i so 4 cl w owld not be esceeded roR rvRmsR IN/OAu ATION CONTACT' .

Dated at Washington. DC this :sth dry of Lynne C. Masters. Esq. at (202) 272- "[ih$*[s),*h'o t 8 d[r on eisruary 1 .. 2848. Division of Market Regulation. such a contnet mar et (hereinsfter For the Nudear Regd4 tory Commission. Secunties and Exchange Commission, collectively nferred to as "CFTC-Sarnuel J. Chitk. 450 Fifth Street. NW., Washington. DC regulated petsons,") solely because such Secretar) of tAe Commasson- 20549. pan dects trauactions in lFR Doc. a7-4431 Filed 5 2-47; a 45 am] suPPLsusNTARY INFORMAT10N: governrnent securities that the

• * 'so coot tsm-et-" . Commission. after consultation with the Introduction CFTC. has determined to be incidental On Octobu 28.1986, the Securidu to sd penod Murea-nlated SECURITIES AND EXCHANGE uchange Act of1934 W 'T.xchenge MMHO COMMISSION The Goumment Securities Act does Act") was amended by Pub. L No. 99- l 571 the Govemment Secunties Act of not set forth standards the Commission 17 CFR Part 240 is to apply in making its determination.

1986 (the "Covernrnent Securities Act").

[ReL No. 34 24135. File No. 57 5-47] The Covemment Secunties Act provides Accordingly, the standards for e

for the regulation of govemrnent Request for Comments on Proposed secundes brokers and govemment e]s d of c' 'e Rules 3a43-1 and 3a44-1 Act apply.These sections authorize

' 8 rulemaking necessary rad appropriate A G E t#CY: Secunties and Exchange l*yf,"m es he u to implement the prtnrisions of the Commission- Covemment Securities Act la a limited Excharge Act.includi the ACTsoN Proposed Rulemaking. one in that it does not provide for Govemment Secuntie Act, and regulation that would sIIect particular definitions of terms consistent wMh the suuuARY:The Commission is publishing for comment proposed rules transactions in govemment secunties, provisions and purposes of the to implement provisions of the e g4 margtn or stutability regulation. Exchange Act.The purposes o(the Covemment Securines Act of 1966 that Instead. the Government Securitses Act Covemment Secunties Act are set out in authonte the Commission after requires the Secretary of the Treasury findings in section 1(b) of the consultation with the Commodity (the Treasury") to adopt rules Govemment Secunties Act.They Futures Trading Commission. to except can'cerning the financial responsibility, include "to impose adequate regulatioR" from the definitions of govemment protection of secunties and funds, and ~"appropnate Snancsal responsibility secunties broker and Fovemment recordkeeping, reporting. and audit of securities dealer certain persons directly govemment secunties brokers and ise,, c,n s. oq ,re,et w,.r, ,an or indirect!y regulated by the government securities dealers. The denau e somment a.nm d in e.

Commodity futures TradinA Corrunission is to preside for the say pmos meneed a se knew atimes and Commission w hose gos ernment registration of govemment secunties "M,,aernaunin==auahwllu*=nue==at

, ,, ,, g, , ,.

secunties activities are incidental to brokers end govemment securities wwe-.

their futures business. The first dealers that are not financialinstitutions (M uy pman smoter a be hers or atti nch proposed rule dafinee ae i.ncidental or registered broker-desters. Fitiancial maa f* ** *=* ===a' **" ma id.e ur w certain transactions for custoners by institutions and registend broker- ',",[g**,Id;**7 "'** * "" M" d' futures com nission merchants, primarily dealers are required to Sie notice of ts) ecy exp nues b an6 ef =>d m as agent. subject to conditions designed M r p a d m e W ce m*ment e s'idern,,a sa WW G

• i to assure that custner funds and g gg g g p.,,e og arj er e,e b sect.rities are safeguarded ar.d that suchtheir p approWe ndatory estacy te m ybua..mi enbeatm aa pdim m customer transaetions are not boninae at berue med wiumne.aru.est Undey the Gosemnert Securides Act, wansee ser us ows 'manei eder maa is e

! advertised er sclicite A The second baarf en arm @ s brder or oer tes or the terme government ser.arWes broker PtoPCsed rule cefino as incMental (D) eny nas resisted =to me Co mery urtsin prtncipal transacticos by CFTC. p.,, r,e,a,,3 cm .,r ====m s=='*et regulated persons iticluding transactiona Weswas try a Ceam#ery rome finen l

hafures contrscis, exchange ' *" "*** 8(

• M *3I d *'8 *^" '9 ' ^'8 C8 Jame.s., a sh e.a.erm martan's as.a.ai.d din"nes a sowr,maru seramwas bra'es ** claaruyr wo a.er .apeer moder en ch vrsasre r physiceto,5 trensactK'ns W13 oger any pem mM egued a h ha,neas d i

. CFTC.regulatru ysons, and certain .a,curg tnu.etwee to sesernen.t wcucum for a m unr W 4. h ad pm m M nn.uaane m pe.or w m arieem se investment trsns4tions and proprietary de suoet et wm W dose nos anchde-- , Commenen anse wan.hWPG M8 DF Csammary (N any musa es naarm et dd m retum Tredire Cama.ema, has determ.n d by hedging an j arbitrage *.ransactions with 80%ernment securilles brokers, 8'""U""' "'s"]"o"f

" '"d" '** *"8"# % " lCI this subascu'an, or

"'le a order to be tendental w e zh pman e future *4,4eted beeiness.

of paragraph (4 government seCuntie: dealtis, and. iD @ py person retsimd mG b Coastodity e Marc h,,en q,iTrod g sad Merkes us rmm Tnens Corrummon. eny contna market edvued te Camrumma s seas' met ,anny some cases. baDks.

des;srated tn the Carood ry resures Tradms conunodiues sheenas carparst,ome are set ieenity DATE: Comments should be submitted Coeurussat such cor. tract martet's ambatad controdar4 contreued by. or uswser common oceusr.)

Dy APMII 1987- c)esnag orgsnitariort er arry haar treder ce circh enth sesp.ated contnct markets (*sse.harees*1 Aponessas: All comments should be contact marset souly beenu suca perupa ettees Numeena. in ihm centesi, se Cas== nee tismcuma is sontunni mim est Se propows to toterpest es tern 4Shnad cegryg submitted in inphcate to Jonathan C. Corrunissaos. siter corudauce mia se Commodaty erserusatum" to include those cJaaraa Xatz. Secretary. Securities and rer res Tr.ers Co nrweioet u dmrr ey Fxchange Com=ission. Washingion. DC n,i, ,, ,,qe, io be vie,oc.,i ,, ,,cb ,m.,, , erse - s,,a.i

.m r,e.,,,,, u ,,,,,

es,,, ,o ese m inior

,,,,ntr, u ,,6,4 -o.ry 20549 and should refer to Ftle No. S7 farn.ntad b new perucatar fem nemage

__,m_ _ _ _ _ _ _ _ _ _ _ _ _ _ _

W e's ,'e .. . -.4 *. . . .e s ~e . m e.% 4. , 2;..... - . . - - s. . . ,

e e

e ENCLOSURE B e

'- od UNIT ED ST ATES .

/ 'o, NUCLE AR REGULATORY COMMISSION D* MIO >

.!' . ' 'o I wasHin ct on. o.c. 2osss Roe l 2

Rehm j

( f Snlezek l

g o..../ January 9, 1987 Denton CFFK:E CF THE Murray stcasiany LShotkin Shelton Lesar Victor Stello, Jr.

MEMORANDUM FOR: Executive Director for Operations FROM: .I. M amuel J. Chilk, Secretary

SUBJECT:

SECY-86-318 - REVISION OF THE ECCS RULE CONTAINED IN APPENDIX K AND SECTION 50.46 OF 10 CFR PART 50 This is to advise you that the Commission (with all Commissioners agreeing) has approved publication of the proposed revision to the ECCS rule to in the10following:

CFR 50.46 and Appendix K to 10CFR 50 subject

a. the methodology used to evaluate uncertainty should be subject to peer and ACRS review.

the ACRS question of whether the current ECCS b.

rule should be grandfathered indefinitely and should be submitted to the public for comments,

c. the FRN should note that Commissioner Asselstine requests public comments on the following:

(1) Should this rule change include an explicit l

degree of conservatism that must be applied

' t.o the evaluation models?

(2) This rule change would allow a 5 to 10 l percent increase in the fission product inventory that could be released from any l core meltdown scenario. Should this rule change explicitly prohibit any increase in approved pow *er levels until all severe accident issues and unresolved safety issues l

I are resolved?

(3) Should the technical basis for this proposed l rule change be reviewed by an independent group such as the American Physical Society?

': c c '. .'  ?.

. . .$.' . I..W b ] _

. 'l @: *t. . _

o'aa'e

..... . _ . . ~ . - . . . . . - - . ~ . - . - - _ ~- - .-.

7 - ..

t

t The proposed rule should be modified to incorporate the above items and forwarded for signature and publication in the Federal Register.

-(600)- (SECY SOSPEllSE: 1/30/87)

(RES) cc: Chairman Zech Commissioner Robertu Commissioner Asselstine Commissioner Bernthal Commissioner Carr OGC - H Street i

i l

I

-- --- -- . . ~ . . . ~ ,

W v ~

p',r

,- "c m om

,U Ja e w .

/6eose v ,ae s de 8 ? .e. a 4

9 l

)

ENCLOSURE C O

et i

t 4

p. .wa ,. ...x..... .a e :...- n. . ; z.? . s c.. . :. ..z . .. t. n . .- . ,.: . . .

l

1. o,, UNITED STATES

! g NUCLEAR REGULATORY COMMISSION

{ ;r ADVISORY COMMITTEE ON REACTOR SAFEGUARDS .

O,

$ ,o g WASHINGTON, D. C. 20555 9*****

September 16, 1987 The Honorable Lando W. Zech, Jr.

Chairman U.S. Nuclear Regulatory Comission Washington, D.C. 20555

SUBJECT:

ACRS COMMENTS ON CODE SCALING, APPLICABILITY AND UNCERTAINTY METHODOLOGY FOR DETERMINATION OF UNCERTAINTY ASSOCIATED WITH THE USE OF REALISTIC ECCS EVALUATION MODELS During the 329th meeting of the Advisory Committee on Reactor Safe-guards, September 10-12, 1987, we reviewed the methodology developed by.

the NRC Officc of Nuclear Regulatory Research for determination of the overall uncertainty associated with the use of realistic models, includ-ing related computer codes, for the calculation of thermal-hydraulic phenomena associated with loss of coolant accidents (LOCAs). In our review, we had the benefit of discussions with representatives of the Office of Nuclear Regulatory Research (RES) and the Office of Nuclear Peactor Regulation (NRR). Subcommittee meetings during which this topic was discussed were held on April 29-30, 1986, August 28, 1986, April 29-30, 1987, and August 4, 1987. We also had the benefit of the docu-

• ments referenced.

A recently proposed revision to the ECCS Rule (10 CFR 50.46 and Appendix X) will permit use of realistic or "best estimate" methods in demon-strating that a peak cladding temperature (PCT) of 2200'F will not be exceeded during a LOCA. This is in contrast to the original version of the rule which insisted on the use of a number of conservative assump-tions which were believed to provide an overestimate of PCT large enough to account for uncertainties. With the new rule change, a licensee may demonstrate that the calculated PCT, when adjusted with an appropriate allowance for overall uncertainty, has an estimated 95% probability of not exceeding 2200*F. In our September 16, 1986 letter to you coment-ing on the proposed ECCS Rule, we noted the following:

"The acceptability of realistic evaluation models rests on the development of satisfactory methodology for determination of the overall uncertainty, Most of the development work needed here is either ongoing or planned by the Office of Nuclear Regulatory Research. We recomend that the methodology used to evaluate uncertainty be subjected to peer review. We also wish to review this work."

[ . . . . . .a '. - J . . _ ._ . . .2 ~ vm.J . ~ .. ,

u .> ,

The Honorable Lando W. Zech, Jr. September 16, 1987 RES has developed a n . hod for quantifying uncertainty in PCT which it refers to as the Coce Scaling, Applicability and Uncertainty (CSAU) evaluation methodology. CSAU is . designed to address uncertainties in the capability of a code to extrapolate small-scale test data to full scale, to correctly assess a particular sequence of events, and to account for variability in important parameters. The focus of CSAU is on the important thermal-hydraulic processes with detailed attention given only to those processes which contribute importantly to overall uncertainty. The end product of the CSAU method is an estimate of the total uncertainty associated with the calculation of a key parameter (e.g.,, PCT) by a given realistic thermal-hydraulic code for a particular plant and a particular accident transient.

It must be recognized that absent an abundance of full-scale LWR plant transient data, it is necessary to rely substantially on engineering judgement in lieu of a rigorous statistical analysis. The CSAU meth-odology systematizes the application of this judgment for the derivation of a quantitative allowance for uncertainty.

We believe that the CSAU method proposed by RES offers an acceptable means to estimate uncertainty associated with the use of realistic codes. However, we wish to note the following:

• The CSAU methodology has not yet been tested over a wide range of applications. Currently, RES is in the process of demonstrating the applicability of the method by using it to determine the uncer-tainties resulting from a large break LOCA calculation using the TRAC PF1/ MOD-1 code. While it appears that CSAU will be success-fully applied to TRAC, we re omend that RES complete an adequate evaluation before the methodclogy is judged acceptable for use in regulatory actions.

• Before CSAU can be applied to a given code, complete documentation (e.g., code manual, model and correlation quality assurance docu-ment, and assessment reports) is necesssry. In the past, such thorough doevmentation has not always teen Ovailable for licensing codes. We recomend that steps be taken to ensure that future development of codes for licsnsing activities be perfonned in a manner that ensures completion and availability of needed docu-mentation before the code is released.

• The codes used to analyze thennal-hydraulic behavior are very large and complex. Validity of calculated results is dependent on the competence of the code user and the way in which the code is used.

For CSAU to be effective, the code developers, assessors and users must use the code consistently. We recomend the NRC Staff take the necessary steps to ensure that proper controls are established.

C.~ ' f

1. .. C. .~ . .- . . :i .m A..n- ,.,w . . .. ' w.

O The hionorable Lando W. Zach, Jr. September 16, 1987

• In order to ensure the ultimate success of the method, we telieve it is necessary for RES to direct its experimental thennal-hydraulic programs appropriately to the needs of CSAU. These experimental programs include the MIST, 20/3D, and ROSA-!Y coopera-tive efforts.

• We wish to caution that use of the CSAU method for regulatory applications will require the maintenance of an ongoing high level of competence and experience on the part of the NRC Staff members.

We suggest that the NRR call upon RES for such support as neces-sary.

We are encouraged by the move toward the use of realistic calculations for ECCS/LOCA phenomena. We intend to follow the progress of this effort closely, and we wish to be kept informed.

Additional comments by ACRS Member Harold W. Lewis are presented below.

Sincerely, William Kerr Chairman Additional Conraents by ACRS Member Harold W. Lewis I support the Committee's letter, but do wish to add some cautionary notes about the misuse of some familiar words, which can lead to po-tential misuse of the CSAU-(so-called) methodology.

To begin with, I support the move to "realistic" evaluations, since I believe that all evaluations should be ude as honestly and realistical-ly as possible, after which regulatory conservatism can be applied cleanly and openly. That is the thrt.st of this effort, and is fine.

Unfortunately, however, the words "best estimate" are often used inter-changeably with "realistic" to describe calculational technicues, and that is an error. To a statistician, a best estimate is an estimate taken from the top of a probability distribution, and that is simply a different idea. This is not sophistry, since the misunderstanding of words that have established technical meanings can lead to incorrect calculations. To call an apple an orange does not make it one.

We were also briefed about a set of calculations in which parameters and assumptions were varied to provide a feel for the sensitivity of the

w..

.. ':::. x . ..; r ...z. ..:~..~... . .- . = . . ..- .

The Honorable Lando W. Zech, Jr. -4 September 16, 1987 results to the specific assumptions made. That is a reasonable way to learn about the sensitivity, but is not a way to learn about the "uncer-tainty" in the result, as any statistician would understand the word uncertainty. Statistical uncertainty in its simplest form is based on the concept of random sampling from a population of known characteris-tics but unknown parameters. In that case, one can learn the unceo-tainty in an estimate of a parameter by studying the variance in a set of measurements, but that is not the situation here, where the variance in the results bears no relation whatever to any uncertainty, in any credible statistical sense. The only reason for saying this is that in the familiar case of a nonnal distribution of sample measurements, one can estimate the uncertainty from the variance, and thereby estimate the probability that the mean of the measurements differs from the true value by any ratio. One can also estimate "confidence levels," but that is another saga.

None of that is true here, and this is again not sophistry. In particu-lar, the draft Regulatory Guide supporting the proposed rule has state-ments about the "95% probability limit," "confidence level," and such things, and even states that the "use of two standard deviations for evaluating the 95% probability level is acceptable." None of this is possible within the framework described, and simply reflects confusion on the part of the -Staff about fundamental statistical concepts.

I still support the letter and the program, since it is a major step forward, but repeat a recomendation I have made many times: the NRC would benefit greatly by hiring a few gond statisticians. One cannot do competent safety analysis in the presence of uncertainty (popular use of the word) without doing the statistics carefully.

References:

1. U.S. Nuclear Regulatory Comission, Proposed Rule, "Emergency Core Cooling Systems, Revisions to Acceptance Criteria," February 26, 1987.

2. U.S, Nuclear Regulatory Comission, "Request for Coments on Draft Regulatory Guide, 'Best Estimate Calculations of Emergency Core Cooling System Performance,'" March 1987. 1

3. U.S. Nuclear Regulatory Comission, NUREG-1230, "Compendium of ECCS Research for Realistic LOCA Analysis," April 1987.

,- --, - , . , ,y - , _ , - - - - - - - . _ - - - . . , , . , - , -

M Mtw f

f 400 g UNITED STATES y  ; NUCLEAR REGULATORY COMMISSION

'; 4  : ADVISORY COMMITTEE ON REACTOR SAFEGUAROe WASHINGTON, D, C. 20066 May 10, 1988 The Honorable Lando W. Zech, Jr.

Chairman .

U.S. Nuclear Regulatory Commission Washington, D.C. 20555

SUBJECT:

PROPOSED REVISIch 0F THE ECCS RULE CONTAINED IN 10 CFR 50.46 AND APPEf! DIX X During the 337th meeting of the Advisory Comittee on Reactor Safe-guards, May 5-7, 1988, we met with members of the NRC Staff and reviewed the final version of the proposed revision to the emergency core cooling system (ECCS) rule contained in 10 CFR 50.46 and Appendix X. Our Subcomittee en Thermal Hydraulic Phenomena met on April 20, 1988 to discuss this matter. We also had the benefit of discussions with the NRC Staff and of the documents referenced. The ACRS previously coment-ad on the proposal to issue this rule for public coment in a letter dated September 16, 1986 The proposed revision to the ECCS rule will eliminate the requirement to use the mocels specified in Appendix K and allow use of realistic models combined with an uncertainty analysis of the overall calculation.

Certain criteria in 10 CFR 50.46, such as 2200*F peak cladding tempera-ture e.nd U". cladding oxidation, would be maintained. The regulatory guide wnich will acccmpany the revised rule describes features of a realistic evalu6 tion model acceptable to the NRC Etaff and contains guidance en parforming the necessary aisnciated uncertainty evaluatiot No charges have been proposed to the final r';1e version as a result of the public cotments received. The regulatory guide has been modified somewhat .o clarify tha NRC Staff's intent in certain areas.

The ACRS has long advocated use of best estimate or realistic evalua-tions for safety analysis. We believe the proposed rule is a major step '

forward in this effort, and we support its adoption, We wish to note the following points:

• Nork to demonstrate the Code Scaling, Apolicability, and Uncer-tainty (CSAU) method for the peak cladding temperature calculated to occur in the reflood phase of a large break LOCA has not been completad. This will be needed to establish guidelines for Staff review of future licensee submittals under the new rule., While the CSAU mthod has been reasonably demonstrated for the so-called

NRCMCRS Wr5HINGTCN DC P.03 .

The Honorable Lando W. Zech, Jr. May 10, 1988 blowdown peak, application to the reflood demonstration will ~ be more difficult. We do not object to plans to proceed with promulgation of the rule change, but we would like to be kept informed about the development of and allowance for uncertainty in the reflood peak tenperature.

• We note that the draft Federal Register notice provided to support the rule enange has eliminated reference to any claimed safety advantages for the rule. We believe the safety advantages are substantial.

Additional coments by ACRS Member Harold W. Lewis are presented below.

Sincerely, U. Kerr Chai nran Additional Comnents by ACR5 Member Harold W. Lewis I have no quarrel with the Committee's letter, but want to seize the opportunity to reinforce a point that has been made before. It is stimulated by unsatisfactory answers to questions at the presentation to the Committee.

The CSAU "methodology" purports to be a systematic procedure for esti-mating the uncertainty in code calcu1stions. That is a laudable objec-tive, and its achievement would be even more laudable. It would be helpful if, in so doing, there were less confusion between the concepts of uncertainty and a probability distribution, and less misuse of the term "confidence limits." These objectives will not be reached unless

• n this case, it is of some professional statisticians become involved.

more than usual importance, since the uncertainty is directly related to the acceptable level of conservatism which must be added to the realis-tic calculations. .

References:

1. u.s. Nuclear Regulatory Commission. Draft SECY paper for the Comissioners from V. Stallo, EDO, "Revision to the ECC.i Rule Contained in Apoendix K and Section 50.46 of 10 CFR Part 50,"

provided to the ACRS, April 20, 1980, V.S. Nuclear Regulatory Comission, Draft NUREG 1230 "Compendium 2.

of ECCS Research for Realistic LOCA Analysis," Office of Nuclear Regulatory Research, dated April 1987.

r. .

.' ...:-. .... .. . ...~.....n....n.....-........ . ..

a 89 Windsor Road Waban, MA 02168 617/253-5296 February 24,1988 Mr. Eric S. Beckjord Director Office of Nuclear Regulatory Research United States Nuclear Regulatory Commission Washington, DC 20555

Dear Mr. Beckjord:

This letter presents the report of the peer review oup convened by your office to Evaluation Methodology evaluate the Code Scaling, Applicability and Uncertainty uber. CSAU)This Methodology and a being developed under tlie technical direction of Dr. N.

Demonstration calculation were presented to the peer review group (attachment A lists membership) on January 12 and 13,1988 by the team which is developing this Methodology. The peer review group was asked to answer the following questions: .

1. Is the Methodology systematic, logical and practical?

2. Has the NRC demonstrated a Methodology which shows that best-estimate methods can be used to meet the requirements of the modified ECCS rule? (i.e. a proposed rule which would allow the utilization of a fully best--estimate approach for calculating ECCS performance when accompanied by an estimate of the uncertainty)

In addition, and separately, detailed technical comments on the application of the Methodology were solicited.

The task ordertaken by the NRC in preparing and demonstrating this Methodology is large and complex. Additionally, the work is still in its early stages of development.

Nevertheless, e, reasonable coincidence of views er.ists among the peer group members regardin the answers to these overall questions. However, there are significant viewpoints and insi hts presented in the individual member's letters which should be recognized.

Examin ion of thae letters, which comprise attachments B-K, should be an integral part of your review of the conclusions of this peer review Stoup.

The Methodolo5y has been developed to establish the uncertainty associated with a  :

realistic calculatia of peak clad temperature (PCT) in a LOCA. The realistic value is to  ;

be calcuhted ennt a numerical code suitable to this task. All suitable codes embody a very large number of parameters of order 100 to 200. The task is to devise a means to identify that subset of these parameters which, together with their uncertainties and 'a procedure for combining their effects, will allow the desired PCT uncertainty to be evaluated. The Methodology had the Demonstration of this Methodology can be summarized as follows. This summary is presented to provide the context of these items upon which we have subsequent comments.

W e . :, :.7.G.TE: . .

w.u:.?.;i...:.L c 3;2 ..

.:.2. . .:.. .: ....:. .. - v - - - -

~

Mr. Eric S. Beck)$rd February 24,1988 page two -

The Methodology centers first on using expert opinion to limit the phenomena which are to be treated'in detail to those considered key to the calculation of the goal parameter, the PCT. Next, by examining the selected code, the supporting data base consideration of these and cpecifically, the potential effects of nodalization and scaling,d number of specific key pher.omena is focused into the consideration of a limite parameters. Finally, the uncertainties introduced by each specific parameter are combined to evaluate the total uncertainty in the goal parameter.

The Demonstration applies this Methodology to calculate the PCT for the blowdown phase of a large break LOCA in a 4-loop Westinghouse PWR. The TRAC-PF1/-

MOD 1/ Version 14.3 code, which incorporates about 175 parameters to describe plant parameters and models/ correlations, was utilized. The key blowdown phenomena .

niluencing the PCT were identified by a structured process for creating phenomena identification and ranking tables (PIRT) as fuel pin stored energy, 2-phase pump performance and critical flow out the primary system break location. Assessment of, the uncertainties associated with the applicabilty and utilization of the code, particularly with respect to code models and scaling and nodilization effects, identified 7 parameters by which to quantify the influence of these 3 key phenomena. Of these 7 parameters,3 related to fuel pin stored energy (peaking factor, gap conductance, fuel I related to 2-phase pump pertormance (a flow degradation parameter),

' conductivity)

I related to break critical flow (a discharge parameter) and 2 related to the heat transfer calculation procedure for PCT (the convective heat transfer coefficient and the minimum temperature for film boiling which se?arates the transition and film o

boiling heat transfer regimes and is also used to establish the transition b'iling heat flux). The estimated uncertainties of each of these 7 parameters were combined by construction of a response surface from only 140 data points obtained from 7 TRAC runs by utilizing 20 sunplemental rods. The overall uncertainty in blowdown PCT, establis'ced by this ca)culational method, was com nred vrith two independent estimates using experimental data (LOFT L2-d ter.t anc other tcrts in LO!'T sad four other scaled tr.cil,tles) and found consistut. The reported overall taleulational uncertaint for the Blowdown Peak Clad Temperature is bounded, with 93%

probabillt , by 13790F.

This Metholodogy can be viewed as comprised of three main components: Problem Definition, Code Selection and Accuracy Assessment, and Execution and Results Interpretation. It is convenient to first present comments specific to these components and then present our response to the two more ;;eneral questions posed.

1. Problem Definition (includes steps 1,2,3 and plant input data of step 11 of the CSAU flow chart).

The identification of key phenomena, although accomplished through a disciplined procedure, essentially rests on subjective individual assessments. For well-studied problems, the results probably are satisfactory, it is important, however, not to apply I

this procedure, or more importantly, to trust its results for insufficiently studied 1

i

.: ..g.

y;e-. ~ . r a. .,.., ..,.s .....

& .u . .. .;...

. .. ... % . . . a . :h; ': ... , . . .: . . : .

• .t 4

Eric S. Beckjord February 3,1988 .

page three ,

problems. Nevertheless, the Methodology offers no criteria for establishing when expert opinion is so immature that the proposed PIRT approach is inapplicable.

A more subtle difficulty,,apparently not well addressed,is how to recognize significant influences from low ranking phenomena arising because their effects are known with only low accuracy. Apparently the procedure does not include this effect of considering the interaction of such low ranking phenomena with higher ranking phenomena.

2. Code Selection and Accuracy Assessment (items 4 through 10).

Topics of relevance here involve data assessment matrix, nodalization and scaling.

The data assessment matrix contains a limited number of tests in reactor-like geometries at various scales. Additionally, many other data have been used to qualif the from models universities and correlations of TRAC.

exists Extensive which havesin A not beenproblem fundamental utilitized.gle exists . effect regarding used in code tuning.

use of data in later stages of code assessment which mig methodology development to emphasize the need to locate and utilize test data which has not been already used in code model assessment. Further, contact with peer group members who emphasized this point could lead the NRC team to unexplored sources of useful data.

Nodalization has been long recognized as a source of variability in numerical calculations. Distortions in cal:ulated results from nodalization is controlled Methodology by (1) utilizing the extensive experience available to establish a codalization scheme on a component by component basis and (2) utilizing the same nodaliznion for calculsting the separate sad in+.qtalWe effects observe,testshowever, and thethat selected nuclear power plant or rationalizing the differences.

specific studies have not been done as part of the development of thisThe Methodology to examine the sensitivity of the selected nodiog scheme to further refinements. not strictly selected nodalization scheme is therefore a matter of judgement,de alterations.

mathematically defensible, which might still be sensitive to modest no An applicant desiring an alternative arrangemenc will need to make significant demonstratlon calculations in its support. However,it will be difficult for the NRC to establish the amount of demonstration calculations needed since the sensitivity of

- results to nodall:ation changes is not quantified.

no scale effect exists for a blowdown PCT The Demonstration uncertainty of contends that3610F for data3610F frombands.

five test facili Demonstration calculation subsumes scale effects if any, within the

• Therefore, in practice the claim that no scale effects must be considered is correct.

However,it is likely that these data physically reflect a scale effect which has not yet been detected. This distinction between the practical Demonstration calculation and the potential physical reality should not be submerged but reflected in the ,

. .:.. i ::2 :. .':.:l. ; .. _ .. :u ......,a...

h' ? ".~ 18.~.%.$%lbi.O :"* . .: .n .

s Mr. Eric S. Beckjord February 24,1988 page four the description of the Demonstration, Further,it should be recognized that only the blowdown PCT has been explored in detail to date.

The above discussion deals with scal: effect in the data. An additional question which does not appear to have been explored is whether the code itself shows a scale effect.

It is possible that the code will show an effect of scale even when the data does not.

For example, the code may tend to over-predict the peak clad temperature in large scale experim,nts while it does a good job of predicting peak clad temperatures in small scale tests even though the tests themselves indicate no scale effect. This particular example is a safe scenario because the code indicates a positive bias.

However, a code with a negative scale bias is a cause for concern because it can be The scale expected to underpredict the peak clad temperature for a full scale plant.

bias of the code can be determined by performing data comparisons with tests at two or three different scales. If there is a negative bias, it should be combined with the other uncertainties to obtain the total code uncertainty.

3. Execution and Results Interpretation (steps 11(part),12 and 13).

The Demonstration developed a response surface from 140 points of which 60Fo were double cross products (i.e. two of the 'l variables perturbed from their nominal values) and 257o were triple cross products. While it is recognized that j gement must be y recognized that applied to bound the number of TRAC calculations, it must be equ the fundamental foundation for the response surface created in t e Demonstration no double cross suffers from the truncated number of runs. We point out that (1)kflow and pump products exist for non-core by hydrodynamic variables (i.e. brea degradation, and (2) no run was done to assess whether in the period before DNB the rod uncertainties coupled to the hydrodynamics would give rise to different peak clad temperatures).

Finally, we observe the,t the Demonstration calculation presented focused solely on the calculation of PCT While this is the relevant parameter for the modified ECCS rule, the development of confidence in the Demonstration requires that realistic results are obtalced for other dependent variable and for the whole temperature response. While the NRC team stated that attention had been given to checking this full range of variables during the developnwnt of the Methodology General and theacceptance Demonstration, of such little such evidence was presented to the peer review group.

LOCA Demonstration calculations will require some degree of demonstration of

> consistency of prediction versus data for physically significant variables other than the peak clad temperature.

Our response to the two questions posed is as follows:

C ~~. n; ; 7 .:..T. T ~ c, a '.u . . ;- - -

1.~.... . -.-..e 1

Mr. Eric S. Beckjord February 24,1988 page five (1) Is the Methodology systematic, logical and practical? The Methodology is judged to be systematic and logical. Its practicality is not uniformly accepted because of differing views of the breadth of demonstration calculations needed to reach a judgement. When judged only in terms of the LBLOCA blowdown calculation completed to date, the Methodology appears practical.

However, an equally relevant question which should be addressed here is that of the completeness of the Methodology. We observe that the Methodology does not include quantitative criteria to establish the completeness of many of its steps. We cite as unaddressed examples the required breadth of the phenomena screening process, the required evidence of code maturity, the scope of the assessment matrix, and the amount of data necessary to construct the response surface. Questions of this type which were raised throughout the peer review group meeting were answered by the citation of engineering judgement. At this stage in the development of the Methodoloir,y and oarticularly when the experience and skill of the NRC team is considerect, such a response is appropriate. Nevertheless, the objective of establishing quantitative criteria of-completeness for relevent steps of the Methodology should be maintained.

Additionally, the representation of the Methodology as an objective, near quantiative method for establishing uncertainty should give way to its description as a formal method for combining quantitative analysis and expert opinion in a controlled way to minimize the subjectivity of the experts in arriving at computed values of uncertainty.

(2) Has the Methodology been demonstrated?

At the minimum, a promising start has been made in demonstrat!ng the Methodology for the blowdown PCT in a PWR, While a minority holds thM the Demonstration is not yet conclusiv 3. no unresolvable difficulties are identified which would prevent achievement of this goal as work progresses.10 the oractical sense then. it ran be assumed that an acceptable Demonstrt,clon of the Methodology to the blowdown PCT is or soon could be in hand.

however, this does not assure that demonstrat. ions of the S!gnificantly, Methodology for the Lill.OCA renood phase and small break LOCA's are achievable The work to date has demonstrated the importance of prior experience of the experts with the phenomena and of physical simplicity of the ,

ahenomena in developing and demonstrating the Methodology. Since Lf3LOCA alowdov>n PCT is the casiN parame'.er to evaluate, success :a entending the Methodology to these othrr I.OCA cases cannot be assured.

CLOSING OBSERVATIONS The peer group did not have snincient time to review, in. detail, the TRAC-PF1/-

MOD 1 Models and Correlations lleport nor Section 4.4 of Nt' REG-1230 which describes

L . .: - . . w . e.: ., .......=.......w.... . , . . . . . . .

Mr. Eric S. Beckjord February 24,1988 page six the CSAU hiethodology. Revision of Section 4.4 to produce an updated coherent description of the Methodology is desirable because of the broad impact this Methodology will have on LOCA and probably on other accident parameter acceptance criteria. Other means for comprehensive review of the TRAC Report are recommended because of the central role that this TRAC code has in the Demonstration of the blowdown PCT calculation and will presumably have in other NRC LOCA demonstration calculations.

The central role which software quality assurance (SQA) will play in application of this Methodolo y should be recognized now and steps taken to incorporate and appropriate formal SQ plans into the Methodology. Evidence of the need for formal consideration of SQA was t e persistent reference in our meeting to the TRAC Report as a QA document.

While the report is a good starting description of the models and correlations utilized, it is far short of a QA document.

and The peer group strongly encourages continued work on the development demonstration of this Methodology. The pursuit of such a program of comprehensive utilization of research results, not only has potential benefits for operating reactors, but The offers a focus for planners and researchers in the conduct of future investigations.

benefits also include the comprehensive documentation of computational codes and data-bases.

The NRC team is to be complimented for their competence and dedication in developing this Methodology thus far. It is not clear, however, that a significantly less skilled team can evaluate industrv .<ubmitted Methodologies whic< differ much from the -

Methodology propsed. This may likely pose a difficulty in eventually managing the <

process of evaluatang applicant respon.<cs to the proposed modified ECCS rule.

Sincerely,

• udd.d Neil E. Todress Chairman NET:ewp

' " ~ ~

l ";

... ...; b . .w . i . . - . a - h ' --

c .

' ' NUMERICAL APPUCATIONS,INC.

Engineering Sottscre Developtnent and Applcotions GEM ButtCe4G. SUITE A 825 GC)ETHALS D%1. RCHW40.WM9352

• TELEPH

Subject:

Comments on Peer Review Group Meeting on CSAG Methodology Date: January 29, 1988 ,

W, l s T li ..$ , ,;',.. d '

t .. C V .

b ". :  ;

From: Marvin J. Thurgood

. .~.

u .- -

To:

Dr. Neil E. Todreas, Chairman

> ~~ZT -

T.

Dept. of Nuclear Engineering Massachusetts Institute of Technology Room 24-102 Cambridge, Massachusetts 02139 The NRC has proposed a revised ECCS rule to allow licensees to utilize a fully real-istic approach for calculating ECCS performance. This approach requires l an esti-taate of the uncertainty in the calculations. The combination of the calculation p us the uncertainty would be compared with an acceptance criteria (i.e. , PCT < 2200 F). The cornmMion directed the NRC staff to Aevelop a methodology The to determine calculation uncertainty and to subject this methodology to peer review.

Inetnbers of the Peer Review Group met in Bethesda, Maryland on January 12 and 13, 1988 where members er the NRC staff and various consultants and national laboratorj personnel presented to the group the Code Scaling, Applicability and Uncertainty (CSAU) methodotetv that has been developed by the NRC staff to meet these requirtment . Membe rs of the Peer Review Group were requested to evaluate the methodology and to answer the following questions:

1. Is the methodology systematic, logical and practical?

er.imate 2, Has the NRC demonstrated a methodo'ogy that shows that best methods can be used to meet the requirements of the modified ECCS rule?

In addition, and siparrtdy the CSAU Technical Program Group has invited the PRG members to provide any detailed technical comments on the application of the methodology. It is the purpose of this letter to document my response tc, these questions as a member of the PRG.

First,I will summarize my review of the methodology by giving my response to the It is my best technical two questions that the PRC has been asked to a.nswer.

judgement that the CSAU methodology dces provide a systematic, logical and prac.

tical means for evaluating the calculation uncertainty for Large Break LOCA tran.

sients. I reach this conclusion, in part, because the methodology is defined in fairly general terms that allow a great deal of flexibility in the actual application of the

h u m m.. ~-,a. mn. _ 3,. s. f . g,,,c,, . .,.gn .

, -e -

l l

I l

l methodology to any particular scenario. ib the The mf assessm'ent manual, and a model and correlations QA manual din that descr ej correlations, numerical methods and interpolation and flow hregime schemel the code. These are very important steps and th ficult to predict the effect that minor code l change statements from licensees that code changes do not effect calculated resu  ;

an adequate number of data comparisons have been repeated dd with the fr sion of the code to clearly demonstrate that the effect of the code change minimal. Much of this methodclogy, and indeed I believe that presenting the results of data comparisons that were application.

done with a large number of different code versions greatly d If the clouds the vision of the code's actual predictive capability and should be discourage ll .

code has retained its predictive capability through all of its revisions it is gen not prohibitively expensive to repeat the calc changing or tunning correlations or correlation selection logic.

The methodology requires that the code be evaluated as tollits applicabi given transient and reactor system geometry. This is appropriate d requires that the code has been assessed against applicable expe has the required geometrical and operational m b'1 d

quately determine the eode's abilityh :to 6ta predic to acceroplish this, the method requires that the same r.rdiw be used for .

comparisons as will be used for the plant calculation. This f approa provided that experimental data is available in facilitia9 f data that appr the plant and little et no neding sensitivity is indicated NRCin should code predictio taken from facilitie nt different scales using the same noding. The require that noding sercitiv!ty studies be performed to assure tha 1 not overly sensitive f to noding in cases a were w no inates uncertaintics sssociated with noding by puttir.g this uncertainty is overall vncettainty of the code prediction of the experimental data. Howeve my judgement that this requires that the code be assessed again variety of facilities at different scales.

The methodology then requires that code uncertaintybelieve and bias be that the effects of scale be determined. These steps are essential and require that the code uncertainty be band on the assessment P operat- of th variety of facilities at different scale. The method then requires that NP ing parameter ranges and uncertainties be de

. . .. a . .. W ..u.. s.  ; ;- - ;, yc:; =~m . o . . . - , .. , , .___

h.. -_ .; y ,

l 4

an appropriate way. The methodology specifically states that the resultant uncer-tainty is applicable only to a specific scer!ario in a specific plant using a specific

• code. I agree with these steps.

For me, the entire methodology rests on the assumption that there is a fa:rly large base of experimental data at r, vuiety of different scales for the particular transient to which the methodology is to be applied and that the particulu frozen code used in the analysis has been assessed against the majority of this data and has done an acceptable job in predicting the major phenomena in the data base in both a qual tative and quantitative way. Because the methodology is defined in rather general terms, the application of the methodoiogy will require that the NRR staff responsi-ble for evaluating licensee applications under the revised rule demand that the licen-see demonstrate that an adequate data base has been used to determine the code uncertainty and that the important parameters have been correctly identified and that sensitivity studies have been performed such that the maximum variation in the calculated variables (i.e. PCT) has been obtained. I believe that this requires that the double and even triple cross products of both fluid and thermal variational parameters must be calculated or that the use of bounding values of the most important parameters will result in the prediction of the maximum value of the acceptance criteria variable (PCT).

This leads to my response to the second cuestion, 'Has the NRC demonstrated a methodology that shows that best estimate methods can be used to meet the requirements of the modified ECCS rule?" I congratulate the NRC staff on the sig-nificant progress that they have made towards demonstrating the methodology that shows that best eitimate methods can be used to license nuclear power plants. I believe that the work done thus fu provides a fair indication that this can be done.

However, I do not believe that the job is complete. The work done 'thus fu indi-cates that the methodology will work for determining the blowdown peak. It remains to be demonstrated how weil it works for the reflood peak. It is also unclew how applicable this method may be to other types of transients, such as small breaks, since this has not been demonstrated at all. Therefore, I cannot say that the methodology has been demonstrated for these cases.

I believe that the particulars of the approach will be quite different for these scenarios because of lack of data and the variety of possible scenarios. The methodology will have to be l

demor.strated for each new scenario.

I have the following observations on the application of the methodology:

1. The demonstration of the methodology relied heavily on a TRAC calculation of one LOFT test for determining the code uncertainty. I believe that this is inadequate. Comparing the the code against a variety of tests of different scales gives a better feeling for code uncertainty and the ability of the code itself to scale up. I think tha; the NRC should insist that licensees include simulations of several different facilities in their uncertainty calculation. Hav.

Ing compared the code to different tests facilities during developmental assess-ment does not incorporate the effect of these data comparisons into the code uncertainty and leaves one feeling uncertain about the codes ability to predict

- +

5 . . '... ;.~a;rF- .. .

.....o...:.,... ......a.. ..,

h

p. $. :

' .: l '.l. :.,.dl. S:

...n.~...

- . . z. ~  ?. . <..c.- . 1.; . a..a .::....... ..

data in different scale facilities.

2. It was not clear to me from the presentation that all of the primary variables were tested in performing the uncertainty analysis. For example, post chf heat transfer was indicated to be an important parameter in the ranking for blowdown heat transfer yet I saw little to persuade me that the uncertainty in this parameter was adequately investigated and cross correlated with the other parameters. The development of the response surface should include all of the parameters that were identified to be important and it should include a sufficient number of double and triple cross products to assure th. t the l

response surface has been correctly determined.

l

3. The uncertainty for various parameters was determined by looking at specific correlations. For example, the uncertainty in the minimum film boiling tem-perature was determined. This is necessary but not sufficient. It has been my experience that the greatest uncertainty is not in the values given by a given correlation but in the ability of the code to select the correct correlation at the correct time. In other words, does the code predict inverted annular film boiling when the data indicates dispersed phase film boiling or slug flow

v. hen it should be' film / mist f1c.w. Errors in flow and heat transfer regime selection logic result in large magnitude uncertainties in the calculation. I think the best way to demonstrate that the code can correctly select the correct regimes is to base the uncertainty analysis on data comparisons with dsta from a variety of different facilities. If economics becomes a limiting fac-tor then it is better to compare with' a single test from five facilities rather than five tests from one facility.

4. Some of the data comparisons from older code versions seemed to be accepted based on a statement from the code developer that the code changes made subsequent to that version wouli not effect the result. I have found this diffi-cult to do as it is easy to overlook subtle interactions between various thermal hydraulic models. The NRC should require a reasonable number of the calcu.

lations to be repeated on the frozen version of the code.

5. The issue of scal'.ng is difficult to address. I think the argument presented at the meeting that the data indicated no scale effect when plotted against linear heat rate and then showing that the TRAC calculation had a certain devia-tion from the mean of this data and assigning this deviation as the uncertain-ing due to noding, code scaling, the effect of tunning and 'home made correla-tions* ect. is good supportive information but is not adequate by itself. This method of correlating the data may only say that the power / volume scaling used to design the test facilities is a good scaling method when linear heat rate is the main correlating variable. It says nothing about the codes ability to scale to larger geometries and may not even say much about scale effect of the facility itself on phenornena that may occur at full scale. Here tain,I feel it is important to compare the code against facilities of different scale, cal-culate the code uncertainty at each scale and then see if the code uncertainty has a scale effect. Does the calculated code uncertainty get larger or smaller

.~ ,...

MR:aw

%.... : . .. -:.~

. e . .n =. ,. , . .. .r.

. . . .. c n u .  :.. .x;, .: : : w ... w ... u u. . ..-.i.

as scale increases. This data can then be used to extrapolate the code uncer-tainty at full scale. Also it is possible to evaluate if the code has a bias to ove predict or unpredict the data as scale increases. I don't see that the codes ability to scale up has been demonstrated at all. This issue becomes even more important for the reflood period were I feel that scale effects are indeed important. Code comparisons and uncertainty for tests from the FLECHT, NRU, CCTF, PKL, LOFT and other reflood facilities should be used rather than several tests from a single facility, CCTF. Dolng this allows the code to be tested against a variety of test scales, power ranges and fuel types (nuclear versus non-nuclear). TRAC, in particular, has heavily emphasised LOFT and CCTF in its developmental assessment. Comparisons with other tests would give a good indication if the code has been overly tuned to a restricted set of data.

6. I agree with Westinghouse that a hot bundle should be included in the analysis.

It is my conclusion that the work done thus far demonstrater that the methodology

- can be applied to the revised appendix k analysis but is is my hope that the NRC will require more of licensees than has been demonstrated thus far.

4 l

l l

g.ygo. 3.:g; g.$.7._, .;.,p,,,;,s,. ..,

,3.; g . g .

.o 1

Dr. Neil Todreas, Chairman Department of Nuclear Engineering Massachusetts Institute of Technology Room 24-102 -

i ,-

Cambridge, Massachusetts 02139 .

January 20, 1988.

Re.: Peer Review Group Meeting, CSAU Evaluation . -.. __....

Dear Neil:

This letter is in response to your request instructing the Peer Review Group members to provide an individual written assessment of the results of the group meeting that was held in Washington on January 11 & 12, 1988 with the NRC consultants and staff membe.s.

I will begin with a general statement. The proposed revision to the ECCS Rule is justilled. The experience accumulated over several decades (experience with full scale reactor operations in the field, the availability of a broad experimental d'ata base dealing with the critical phenomena and covering a wide range of equipment scales, the maturation of computational tools, the availability of very well balned and competent professional maraower), at one end, and the economic penalty associated with con ,1uation of the present rule's methodology (which is based on the t se of ' conservative' estimates),

' at the other - provides a strong rational for the effort directed toward the rule change. The procedure behind the revised licensing rule (formulated around the concept of 'best estimate"; is based on the wealth of knowledge in the NPP related engineering sciences, and the state of the art computational methodologies, it is for these reasons that i strongly support the proposed changes.

Concerning the specifics, we are asked to evaluate Gode Scaling, Applicability and U.ncertainty (CSAU) methodology. The methodology was presented to us through a particular example

- ,- - - - - - , -- - , ,,n,- , - - , , - - . ,,w - - --- -

m.. _2,. . _ .... _ _. m .._. m ..

g. . .g _~ , . . . .e.. ......._._..

2 involving a single reactor (PWR) and a single type o f acciden t (LBLOCA). The methodology has three main components: Problem Statement; Code Selection, and Execution and Results Interpretation.

This simplified classification was adopted here so that I could group my comments by addressing a block of common issues together.

Briefly, the breakdown of the revised classification and their, relation to the NRC original organization are shown below.

A. The Problem Statement is comprised of items 1,2,3 (from the CSAU flow chart), plus specification of the plant input data (including all uncertainty in operational parameters and material properties (these items are contained in part in box 11 of the master chart).

B. The Code Selection: items 4 to 10.

C. The Execution and Result Interpretation: part of item 11, and items 12 and 13.

The methodology associated with block A (Problem Statement) is weII formulated and justified. I do not see any deficiencies in the muthodology as it pertains to the specific example presented at the meeting, por do I see any inherent difficulties in extending the methodology to the other types of reactors or different accident scenarios.

Block B (Code Selection) involves elaborate step by step verification that includes; check of the code ability to simulate the main phenomena for a chosen scenario, validation of the closure relations and an overall verification of the code accuracy (by reouiring an acceptable agreement between the code prediction and the available experimental data). For a code that uses ' tuningto force agreement with experimental data obtained from subscaled systems (either through raodalization selection or adjustments in the closure relations), a procedure for the code up scaling should be provided. The final product is comprised of a working code (with full documentations), specified bounds of uncertainty for the closure relatieris (which should include effects of the scatter in

F:,9 eM=.m .v,a..,m.j7.g5.m::. .c. - 3, ..

9 3

experimental information used in their development), and the uncertainty in the predicted value of the critical parameter (s) due to the code up scaling.

The proposed methodology applied to the main Items in this section, as shown through the specific example, is quite satisfactory as far as it goes; the methodology concerning the up-for scaling uncertainty has not been presented in necessary details us to judge its validity; for the case demonstrated at the meeting, (LGLOCA), no scaling effect us detected and therefore there was no need to deal with this issue within the riarrow confines of the example. However, this absence of the scale effect might not, and probably would not, carry on to all accident scenarios, and consequently there is a need for a procedure development that would be able to handle tho' code up scaling uncertainties Until this is done the methodology for the code acceptance could not be considered entirely complete.

Block C (Execution and Result Interpretation) requires eva iation of the 'best estimate' and the total uncertainty in the l calculated value of the controlling parameter (here PCT). The input l

parameters necessary for the execution include the plant input data (with uncertainty in operational parameters and material properties, from Block A), the uncertainty in the closure relations, and the uncertainty due to up scaling (the last two from Block B). Due to the nonlinear nature of the problem, the . total uncertainty in the PCT could not be obtained by addition of the individual effects. At the

[

meeting a combined effect due to simultaneous changes in two parameters was demonstrated. Perhaps in an actual licensing fiituation more then two combined effects need to be considered.

Nevertheless, the expansion to multiparametric analysis for the combined uncertainty is constrained only by the lack of additional computer time (rather than by the absence of a correct methodology). Of some concern however is the fact that the shown example did not indicate how the code vo scalina uncertainjr gluLt?

4 oresent. would be included in the total uncertaintv. The methodology for this situation remains to be demonstrated. '

In conclusion, the entire program appears to be very well c;irected; the overall methodology is structured well but it is still not quite finished. I do not see any reason why the continuing effort would not yield shortly a solid and complete methodology which could be confidently employed as a ' blue print' for the 'best estimate' procedure.

Sincerely, h V ( A k v. J ' c /

Bora Mikic, -

Professor of Mechanical Engineering, Massachusetts Institute of Technology P.S. A final comment. I was very impressrsd with the presentation of the NRC staff and consultants. Most of them displayed a strong affection and respect for the work they were doing. I believe that this, together with the demonstrated competences in their respective areas, are the best guaranties that they will continue to work aggressively and successfully until completion of the project. ,

t> A\.

i

~

g; ch--+y.. 'd.v.MnW55 7G.i.VW. dNi. .m .- . ,

ARCONNE NATIONAL LABORATORY .

9msacasswwmm9 January 22, 1988 AU T *. ; * ' .

• • i Professor Neil E. Todreas Chairman, Peer Review 1roup '

Department of Nuclear Engineering Massachusetts, Institute of Technology v ., , , ' " .

'

• v. ~ ~

Room 24-102

• Cambridge, MA 02139

• C J." .', ,,-1 ,, . . . . .

Subject:

Coments on the Code Scaling, Applicability and Uncertainty (CSAU)

Evaluation Methodology Developed by RES, NRC During the Peer Review Group Meeting on the CSAU methodology, January 12-13, 1988, the Technical Program Team presented the basic methodology and its demonstration. Accordingly, we were asked to coment on the following two questions.

I) Is the methodology systematic, logical and practical?

II) Has the NRC demonstrated the methodology which shows that the best estimate code approach can be used to meet the requirement for the preposed revision of the ECCS rule?

Below, please find my comments concerning these two questions.

I) On CSAU Methodology (1) ThemostdifficultstepintheCSAhmethodologyistheextrapolation of a predictive tool (code) based on scaled down experiments to untested prototype conditions. Reliable extrapolation of the models or evaluation of uncertainty of a code significantly depends on the type of accidents, available data base and physical understanding of scale effects. Although this methodology is very systematic, the practicality of the methodology may not be determined independently of these three items.

(2) In a step to evaluate the total uncertainty, the CSAU methodology appears not to make a clear logical distinction between the varia-tions in code calcu'ations over possible ranges of parameters and statistical uncertairdy of the code prediction for a NPP. In order to bridge these two, it may be necessary to assume that the thermal-hydraulic models are basically correct and all important physical phenomena encountered in a NPP can be modeled by the code. This is a significant step which should be taken carefully with sufficient supporting evidence.

(3) Several steps, for example, 1, 3, 6-8 and 10 in the CSAU methodology require considerable expertise and good understanding of the acci-dent phenomenology. In other words, the reliability of this method-ology d(pends critically on the technical level and experiences of a USDgwmen d Ew Tk hny d O*xp

s. , . . ~ . , _ _ . . . m_._, . u .. . . m. v, m . . ., . . m _. v... -

_.....u. ..

1 January 22, 1988 Professor N311 E. Todreas user. Therefore, in order to evaluate its proper use for licensing purposes, it is necessary for the NRC to sustain a high level of

• l technical. capability in the thermal-hydraulic area.

l

'!!) On NRC Demonstration of the Methodology (1) I-think that, in general, the NRC has clearly demonstrated the applicability of the CSAU methodology to estimate uncertainty associated with the use of "best estimate" computer codes. This demonstration has been carried out by the Technical Program Team with a high level of technical competence and considerable past experience in this area. I consider that the ability of the team has been one of the key elements in the successful application of the CSAU methodology.

(2) In Step 3, process screening, the engineering judgment by experts has been used to identify the important processes and to rank them according to their relative importance. For the particular case of the blowdown PCT, this may be quite sufficient because of the extensive experimental data and good physical understanding of the phenomena. However, for broader applications of this methodology, systematic considerations of the following two points may be beneficial . ,

1)

An evaluation ed the level of the understrnding of the scale effect for each phenomena.

11) An evaluation of each phenomena from the degree of non-linear t

effects involved. In particular, bifrucation phenomena should be carefully evaluated in ranking the phenomene as well as in calculating the code uncertainty. This is because a bifruca-tion phenomenon can lead to considerably different results starting from only slightly different initial conditions. This implies that the uncertainty may not be handled as a linear additive quantity.

(3) In the NRC demonstration of the CSAU methodology, cross products of uncertainty associated with several thermal-hydraulic parameters are not examined or not considered to be important. However, it is necessary to demonstrate that this is a varied approach by making a few sample calculations. If these sample calculations produce '

surprise effects (nighly non-linear nature of uncertainty), some modifications of t' methodology may be needed.

Sincerely,

'fh ouwh[ '

M. Ishii Reactor Analysis and Safety Division Argonne Nstional Laboratory MI:kj l

i i

i ._ - . _ _ _ . _

n. ,y MSI $<d '.~ ,i e t, - - , ,

. c e, - , . ?. ., .;'.o.f .}. {. , ,: .,. .. .. . 3.........s...

s . - -

,  :. 'r . .. : :v...... ..

0 82 ,

'80 81/25 16:21

,1 CIENTECH INC.

me c a-camr unun p @ 60s 1404 CANO P ALA.1 @AMO fleel 25 January 1988 ,

m mwr ,

Dr. Neil E. Todreas, Chairman Department of Nuclear Engineering Massachusetts Institute of Technology Room 24102 Cambridge, Massachusetts 02139

Subject:

YBR 517 88: Transmittal of CSAU Review Meeting Report

Dear Dr. Todreas:

Attachment 1 to this letter is the repod which you requested after our 12,13 January meeting on the CSAU methodology. If you have any questions on this report or if I can assist you in any other manner in the preparation of the final report, please contact me.

I believe that you did an excellent job, as chairrnan, managing the diverse contributions and viewpoints of the committee members.

If there are further evaluations of the CSAU method,I will be pleased to panicipate.

Best Regards, L. J. Yba rondo President LJY:cbc Attachment File: DN.9 f

_ ______ W h B8.GO O]

" i; , . . M .i; >.% t W&.' ;e. . -; . L i:,. . . . ,. . l. i., . . :w.. .. s; .'*T f: ~ ~ ' ...::.,+. . ..'.

'. *  %. -- - --' -- -- ~ - ~ ~ ~ " "'-

f, .

0 es ,

.. 'es ev2s 16: 21

^

s ATTACHMENT 1 TO YBR 617 88.CSAU AEVIEW GROUP REPORT PAGEi l

l REPORT ON CSAU METHODOLOGY l

by L. J. Ybarrondo, SCIENTECH, Inc.

)

1.0 INTRODUCTION

On 12,13 January the author participated as a member of a peer review group appointed by the Office of Nuclear Regulatory Research. The function of the p6er review group was related by NRC personnel as follows:

" The NRC has proposed a revised ECCS rule to allow licensees to utilize a fully best estimate approach for calculating ECCS performance. This approach also requires an estimate of the uncertainty in the calculations. The combination of calculation plus uncertainty would then be compared with l I

- acceptance criteria (i.e., PCT < 2200 'F). The Commission directed the NRC staff to develop a methodology to determine calculation uncertainty and to subject this methodology to peer review. The peer review should answer the following qucations:

1. Is the methodology systematic, logical, and practical?

2. Has the NRC demonstrated a methodology which shows that Best Estimate Methods can be used to meet the requirements of the modified ECCS rule?

in addition, and separately, the CSAU Technical Program Group (TPG) would be interested in any detailed technical commsnts on the application of the methodology."

l 2.0

GENERAL COMMENT

S A tremendous amount of work has been done,in the last few months, by the NRC and NRC Contractor personnel who made the presentations over two days. The material was organized well. The presentations were excellent.

The information presented represents over two decades of experimental and i, analytical research into the Loss of Coolant Accident costing over a billion dollars. The piece by place, systematic approach being followed is excellent. l l believe the results of this research can and should be used in the NRC licensing process to further improve reactor safety.

,. -- - - ,-.-. - --,-, -,,.,..y,,---,m,,,,,--<--w. ,..m,-m,,,,,-, ---,n--,..--,.-.___-.~,,.,n..wm - - - .. , - - - - , . ,

M *- - -s~-~"-- - - ' ' -' ~ ' ' ~ '

D to

• .. *g4 ele 25 16822 ~

~ "

PAGE2 A1TACHMENT 1 TO YBR 517 88. CSAU REVIEW GROUP REPORT  :

2.1 RESPONSES TO QUESTICNS POSED BY NRC a Each of the questions above are repeated below for convenience, j 2

1, is the me'hodology systematic, logical, and practical?

i RESPONSE. I believe that the the methodology presented necears to be -

systematic and logical for the example demonstrated. Based upon the example presented, I cannot say whether the methodology is prsctical because the  ;

example used to demonstrate the methodology is too limited in scope to allow a a

definitive yes or no answer. The example used was confined to the Large g Break Loss of-Coolant Accident (LBLOCA) blowdown. The Refill and Reflood =

portion of the LBLOCA will be a more complex and demanding test of the practicality of the CSAU method for a LBLOCA. An encouraging first start has [ -

been made in a partial portion of one type of accident, to one type of p! ant (A =

four loop Westinghouse Pressurized Water Reactor), to a specific code (TRAC) -

to which the CSAU method is Intended to be applied. I recommend that it should be coritinued. I do not believe that the demonstration offared is g complete enough in scope to provide a definitive test of the NQ desired "

characteristics of systematic, logical, and practical at this tirr',. In my opinion, a complete demonstration of the proposed methodology should encompass the full range of expected applicability of the methodology which includes at least j a

the complete LBLOCA and SBLOCA. Also, the issue of applicability of the J

CSAU to other codes, PWR's, and the Boiling Water Reactor (BWR) was not addressed, explicitly, in the presentations. If it is judged not practical or necessary to demonstrate the CSAU method for each proposed application, at -

least such situations should be specifically addresseo in order to answer the question posed properly.

2. Has the NRC demonstrated a methodology which shows that Best Estimate i Methods can be Used to moot the requirements of the modified ECCS rule?

2 RESPONSE. An initial, limited scope approach which shows excellent -

promise has been demonstrated. A methodology which shows that Best  ;

Estimate Methods can be used to meet the requirements of the modified ECCS rule has not been demonstrated because the example used is too limited in scope compared to the wide range over which such a proposed methodology -

must be applied.

R IM M m um m

. - . . _ m , , ,

x-,w ,, .,. .~

.. e5 D

'80 Ste25'16123 PAGE3 ATTACHMENT 1 TO Y8R 517 88. CSAU REVIEW GROUP REPORT 2,2 SPECIPIC CCMMENTS .

This section contains a discussion of Individual technical and management areas which came up during the two days of presentations.

2.2.1 Definitions A set of definitions for key words and words which have relative meanings needs to be developed in order to avoid unnecessary debate add to make the ,

method explicable. For exemple:

• The use of the the word "scaling" caused substantial nonproductive discussion. Scaling is a centralissus for CSAU (see Section 2.2,4), it must be property defined and used if the CSAU method is to be explicable to technical peers.

• Important parameter, frozen code, uncertainty, capture important phenomena, peak clad temperature, maximum temperature, code uncertainty, response surface (see Section 2.2.5), minimum / maximum variations, etc. are other words ,

which created some confusion.

2.2.2 Quality of Analysis A limited but good analysis was conducted, in some areas it was elegant in the manner in which the basic physical parameters responsible for observed '

behavior were Isolatsd and evaluated.

i 2.2.3 Nodalization .

The effects of nodalizaton on computer code calculations (analytically and experimentally) are well known. The CSAU method overemphasizes the attention which needs to be devoted to the area of nodalization, it is suggested l l that the issue be addressed in a more balanced manner. Currently, nodalization represents almost one third of the diagram which presents the CSAU method. ,

2.2.4 Scaling I

This is a central issue for CSAU. Several times statements which I belisve to be l incorrect were made with respect to scaling in general and in particular for the J experimental blowdown data. For example,

• No effect of scale could be observed in peak clad temperature measurements from several scaled experimental facilities." Of course there are scale effects they have been subsumed within the several hundred degree temperature (*360F) band 4

selected. Those of us who helped design and conduct some of the referenced tests know there are scale effects:lt is a matter of degree not existence.

Sweeping statements such as the quoted one need to be more carefully stated l In order to avoid unnecessary technical argument which detracts from the general credibility of the CSAU method.

l t

, _ _ _ . . _ _ _ . - _ _ . _ - ,,_.____.-..__._.___.___._._,_.,__._.________f___. __ _ . - - , _ _ . . _ _ _ _ _ . . . _ _ . _ - ,-

= _ . . . . - _ - _ . . . - - - . ._. _- --_ - _ - _ _

D 06 **

'et 01/25 16:23

• I ATTACHMENT 1 TO YBR 617 84. CSAU HEVIEW GROUP REPORT PAGE4 2.2.5 Response surface The response surface approach described by the NRC and their contractors is completely different from the response surface approach described by the Westinghouse representative. For example, the independent variables used in i the NRC approach were physical parameters. But the approach described by Westinghouse used plant actions such as pumps on/off as the independent variables. I strongly suggest that the effect of such differences be evaluated to determine whether the CSAU method can be applied effectively and consistently with such differences.

2.2.6 Overview of Methodology

The description of the CSAU method as such is excellent. But the relation of the method to the "big picture"is missing. A MORT type chart is needed that

explicitly presents the overall relationship of the method to the objective ECCS RULE REVISION. Then, the applicability to other plants, codes, transients (e.g.

are operational transients and ATWS excluded?), assumptions, constraints, regulatory guides, need for software quality assurance, etc would be clear.

Without such an overview, it is not practical to answer the two questions posed i

in Section 1.0.

2.2.7 UserInterface Adaption of the CSAU method will require much more attention to the code user i interface. The user interf ace becomes significantly more important because the ,

extreme bounding approach mandated by Appendix K will be replaced by a realistic best estimate approach with sufficient but less margin. Therefore more NRC control is needed in the area M software quality assurance than currently exists.

2.2.8 Software Quality Assurance (SOA) l Section 2.2.7 Indicates why the author believes this area is significantly more l

l Important for the CSAU method. In addition, there seems to be some confusion .

! over what constitutes SOA. I recommend it be resolved. A "TRAC PF1IMOD1 '

Models and Correlations" report was continually referred to as a OA document l vertally and in slides (e.g.,see step 5 in the CSAU diagram);it is not a OA l document. It is a good reference document that describes and provides a basis for the models and correlaticris used. It is an Impressive document in size and with respect to the amount of information presented (see comments in Section l

j 3.0),

I believe that it is Imperative that a formal SOA plan be required of all f organizations that will be associated with the CSAU method. The eleven elements of a SOAP as defined by ANSI /IEEE Std. 7301984 include requirements in the areas of:

I i l

..:....d. . " *h W 5 R A'"'*  % ~- ~% * *T~

Q M y '3 s" -

.. , of D ,

'et et/25 16824 PAGE5 i ATTACHMENT 1 TO Y8R 517 46. CSAU REVIEW GROUP REPORT )

Management; Documentation; Standards: Practices and Conventions; j Reviews and Audits; Software Configuration Management: Problem Reporting and Corrective Action; Tools, Techniques, and Methodologies;,

Code Control; Media Control; Supplier Control; Records Collection, Maintenance, and Retention.

A methodology without these type of defined controls will not be systematic, logical, or practical for the proposed CSAU method.  :

3.0 Comments on Additional Documents in addition to the comments on the CSAU method, I reviewed the "TRAC.

PF1/ MOD 1 Models and Correlations' dratt report Comments follow.

The short time avaltable for review of the TRAC document allowed me only a glimpse through the report. Therefore, my revlow cannot be aimed at review of the quality of the presented models and their rationale or even on the quality of their description and discussion but on rather superficial judgement of overall -

completeness and quality of the report.

1. The reportis impressive with regard to its size and the amount of information presented.

2. It seems that the presented materialis basically sufficient to judge the implemented models and provides a good basis for understanding and a good reference for further model development and assessment.

3. The executive summary needs improvement. Lots of the material presented in the Introduction should be included in the Summary. Short forms of the conclusions of each section should be included in the executive summary.

Regarding the extent of the document and its subject, the executive summary can be several pages long.

4. Logic diagrams are used in this report only in the sec implementation of many other models particularly for interfacial heat and momentum transfer.

5. Section 3 on Flow Regimes is missing the subsection on Scaling Considerations and Summary and Conclusions. Scalability of flow regimes is most important with regard to evaluation of code quality for NPP calculations; only small diameter pipes are discussed.

6. Sometimes, some additional explanation of implementation would be of advantage. For example,in the Flow Regime Section calcu!ations (adjoining cell averages vs, cell centers).

g ..

. s a.:. s -,.;. . ..

Nt ,'b N 5 . ';.1'..+,s....r.a x. rw. . . . r. i- a. . . ;.:M ;;; .' a; u . . . _,

^

0 88 p, ,

'90 Ste25 16125 , ,

PAGE 6 ATTACHMENT 1 TO YBR 517 66. CSAU REVIEW GROUP REPORT

7. The role of the liquid velocity in the stratification modelis not explained sufficiently.

8. In the Section on Critical Flow (the conclusions) nothing is mentioned about ,

the known problem of superheated, single phase steam flow (accurate estimation of'the entropy).

9. Scaling is not addressed in the pump section.

10. TRAC does not calculate the pump energy source term. In the pump section, this problem is not discussed.

11. Section on Special Component Models should includs also a discussion of the accumulator model, it is known that TRAC sometimes calculates unphysical behavior of the accumulator flow.

g ;.7 m,em ,. v .t.~ . . . . . . . . . . . . .. .

UNIVERSITY OF CALIFORNIA. SANTA BARBARA SANTA B ARS AM

• S ANT 4 CRVE eg!sstsy

• 04tts

• isvtNE

• Los ANCELES

• RrvERlfDE

• SAN DIECO

• S AN FM.tsCISCO S ANT A B APS ARA. CAtJFORNIA 33104 DEPART > tent of CHEMICAL AND NVCLEAR ENCI'iEERINC .

'i ' ' 3, f,.\T. January *22. 1988 Prof. N. Todreas, Chairman Department of Nuclear Engineering "

Massachusetts Institute of Technology 77 Massachusetts Avenue Bldg. 24102 ~~~~~

Cambridge, MA 02139

Dear Neil,

I am writing to you to document my views regarding the CSAU methodology The developed by the NRC in response to the requirements of the modified ECCS rule ,

questions that were posed to us are:

1.

Is the methodology systematic, logical and practical?

best estimate

2. Has the NRC demonstrated a methodology which shows that methods can be used to meet the requirements of the modified ECCS rule?

as if we were In giving you my views on these questions I will deal with the first asked to consider the overall CSAU method its applicability without reference to different to any particular transients, application, i.e. the comments will be about With regard to the second question.

and its strengths and weaknesses in this context. Clearly the two questions are I will focus on the LBLOCA material presented to us.

somewhat interrelated and this division is arbitrary.

Comments Related to Ouettion 1 (and its weak point) the fact that

1. The CS AU methodology has as its strong point is used to reduce a very large computational problem to engineering judgementThis also allows many different transients to be considered and practical dimensions. in this process something is lost in that makes the method quite general. uncertainties handwaving arguments have to be resorted to in order toconditions, capture variations of arising out of numerous input parameters and boundary which are not studied.

leads to a practical problem is Step 3, i.e. phenomena The step in the process that screening.

I have reservations about this step because In generale theirit judgenent is only asisgood basedas the and ;orresponding judgement only on what of the experts who are responsible.has gone phenomena that rr ay only arise before - perhaps sm calculations .. and they are unlikely to identify newTo give a concrete exampl on a large scale system.

common wisdom that emergency venting of chemical reactors underping sets o f runaway reactions would lead to all vapor discharge. This was confirmed by extensive

~ " = - " ~ ~ ~ - " - -

fy g M h ; 2.-.a.. - .-..... w . ~ " ~~ ,

However, accidents contint.ed to occur experiments done on reduced. scale reactors.on full. scale systems where reac It has only been recognized recently that this was.

to the vent sizes being too small.

due to two. phase discharge which occurred in the large system but not in the power-In retrospect, this is because level swell does not to. volume sesled small systems.I can think of several other examples where important scale linearly with height. expert groups.

phenomena have, been missed in deliberations by a

I feel more stringent safeguards are required in the methodology to ensure that consensus is ns.L required of the screening group in order for a phenomenon to be ranked highly. This may result in a somewhat enlarged group of parameters for which sensitivity analysis must be done, but this will be in the interests of credibility. Finally, the effects of phenomena in SBLOCA are much less understood than for LBLOCA, so there is some question as to whether screening and ranking can be done with a high degree of confidence.

2. I have reservations about Step 8 inThe themethod methodologyproposedwhich is not relates defensible to on defining the critical nodalization for NPP calculations. used for the calculations have scientific grounds unless the computer codes being The present generation of been explicitly developed to handle lumped components. is essential that computer codes difference the governing equations and, as such, it convergence with regard to node size and time step be demonstrate be done. I believe the results.

becomes a set of input parameters for which uncertainty analysis must have seen no evidence that this uncertainty is properly accounted for (see also Comment 6 in the next section).

3. The methodology does not require, that the results of codes As be compared with discussed in the experimental next section, the data over a rarige of dependent variable responses. app!!

entirely on the comparison of peak clad temperature during blowdown with experiment.

If flows, pressures and the whole temperature transient had been compared, much greater confidence with regard to the- veracity of the code wo have been developed. I feel it is a weakness of the methodology that it does dependent not try to develop quantitative measures for uncertainty in several of the important variables.

4. The methodology is focused on use of response surfaces for which itThe is well known that a very large number of calculations have to be done to even for this situation there can be large errors in the results.

methodology into selecting only a small number As Iofmentioned parametersduring for sensitivity the analysis in generating the response surface.

presentat'ons, there are alternatives which are much more economical in terms of One such methodology (see references 1, 2 and 3) is based on adjoint computer time. Once an adjoint code has been developed, only one run is required per calculations. variable. (In addition, one run is required for the nominal response per indeoendent The development of the adjoint code does require r response by the main code.)

significant effort (e.g., for the general circulation Howevermodelonce of the theadjoint atmosphere, code was it took four man years to develop the adjoint code).

developed, sensitivities could be Note obtained to several that tha general circulation thousand modelparameters took and could be donc very economically. Documentation of tais can be found in l

several hundred hours of CRAY time to run. \

j i

2 i

t l

.. n p .

reference 4. I am simply offering this to demonstrate that it is not a practical necessity to use a methodology with a high degree of parameter screening to do large problems. Alternatives are available and as the work on the atmospheric model demonstrated, the sensitivity results can go completely counter to the intuition of experts.

In summary, my answer to question 1 is that the method proposed by the NRC is prtetical - maybe too practical. I wouM say that it is also systematic and logical, but I would have to qualify this by indicating reservations as to the phenomena screening step and the nodalization step. I also have reservations with regard to Steps 12 and 13, though these are less fundamental and arise from the demonstration for the LBLOCA case. More about this in the next section.

Comments R einf ed in Ouettinn 2

1. Step 3, which is the phenomena screening step, was conducted very thoroughly by two independent grnups. Nonetheless it appears to have missed the imponance of nitrogen in effecting the fluid mechanics of refill and reflood --This an effect is significant which calculations done in the U.K. Indicate is of primary imponance.

because a great deal is known about LBLOCAs as compared to other transients that would be considered by the methodology. Funhermore, the effect of DNB delay time This is not borne out by the was considered unimponant by the ;xpert groups.

calculations done by Wulff (see Figures 8 and 9 in his handout) where a DNB delay time of I second made 100 K tempcrature difference with regard to the peak clad temperature. This is significant in comparison to some of the other uncertainties considered.

2. As pointed out in the previous section, the nodalization appearswere No arguments to bepresented tuned to experiments, particularly with regard to LOFT.

justifying the so called "lumped parameter" aspects Muchofcould the closure be made relationships, of this problem so this approach is at present without foundation.

by intervenors as the procedure flies in the face of established mathematical practice.

No results

3. Far too much emphasis was put in all the presentations on PCT alone.

were presented to indicate that the code was giving realistic results with regard to the other dependent variables and for whole temperature response..

4 With regard to scaling, the peak cladThe temperature was plotted against the linear peak clad temperature should be plotted heat rating. This in itself is incorrect.

against the stored energy.

At first sight, this is not a serious deficiency, because However, for nuclear rods, the linear heat rating is directly related to the stored energy. ,

a considerable number of non nuclear experiments in the plots presented, there are In these experiments the rods were indirectly heated and e.g., Semiscale and THTF.

for a given linear he at rating the amounts of stored energy are quite different.

Therefore, if the plot is redone with stored energy as the abscissa, one finds that THTF and Semiscale have considerably lower stored energy than would be indicated by the We then begin to see a distinct effect of position on the linear heat rating abscissa. rod length, e.g., Semiscale and data. Note that the Semiscale and THTF results would be shifted to the left compared to the present plot. Note also that LOBI should be deleted from this set because can be it pre:.upposes a knowledge of the heat flux transient and there is no way that it put (with any credibility) on a stored energy abscissa.

3

--~ -

W 2 M : . 7.J.srium.w~. + . - w "~ ~ - yy ~~~ ~ ~

i i

there are no scale effects associated with rod in any case, I am not convinced thatThe abscissa should be changed to stored energy length from the presentation made.

which, after all, is the correct quantity based on the arguments presented. (See also reference 5.) '

5. I feel that, too few runs were done varying the hydrodynamic parameters to  !

quantify the uncertainties associated with cross products between hydrodynamic  !

effects such as uncertainties in the pump -model and in the discharge rate model. l Furthermore, in order to reduce computer usage,This the runs _ varyir.3 fuel in itself is only a valid rod parameters were decoupled from the hydrodynamics.a single run was done to see whethe '

procedure after DNB has occurredI feel and notperiod before DNB with the ro that the codes should have been give run, at rise to different least for the peak clad temperatures. initial period, with coupling between the fuel thermalhydraulics.

6. The lumpin's of all uncertain les by comparison between Thethe calculationcurve, experimental and the as so called experimental curve for PCT was unconvincing.

discussed previously under point 4, is heavily dependent on the correlation for peak I

clad temperature which was incorrectly plotted against the linear heat rating.

have no confidence in the physical basis for such a curve and therefore no confidence in the claimed uncertainties provided by the comparison between the In 9 e.g. In Lellouche's presentation. ,

NPP calculations and the linear regression lineto extrapolate experimental data from general it is incorrect  !

full scales, and use this to provide a datum against which NPP calculations It wouldare be  !

compared . leading to quantification of the remaining uncertainties. i.e.

much better to do this by honest calculations against new experiments, has not been tuned against and on the scale that experiments which the code The key here is to do blind predictions to keep things ,

experiments were done at.

honest and quantify the total uncertainty.

7. There are a number of miscellaneous points which need to be mentioned, First. though the they may not be of the same significance as those discussed above.

Second, the models and correlations assessment matrix should probably contain PKL.

document produced for TRAC should be Itgiven a thorough sense a examination byatexperts not -

involved in the development of TRAC.

is in no QA document present.

some LBLOCA experiments which have not Third, blind runs should be done against i been used to develop TRAC with nodalization as proposed for the NPP calculations.  :

at developing a  !

In summary, I feel that while the NRC has ofmade a promising the modified ECCSstartrule, it has not ,

methodology to meet the requirements even for the relatively simple case of the conclusively demonstrated its application ,

blowdown peak in an LBLOCA. This is nottotominimize say thatcomputer with more workthe usage, thework methodology >

could not be used. However, in the quest i presented perhaps took too many shortcuts and I, for one, would like to see a moreAlso I wo  !

deliberate and thorough approach to the demonstration. problem than the blowdown phase  !

methodology applied to a somewhat more complex l of LBLOCA before giving a definitive opinion. The reflood peak for an LBLOCA may j be such a problem or a particular scenario for a SBLOCA may be even more i convincing.

[

! 4 l  ;

I i

---~~~

g ,;,, ,, .... . _........ .~ ~ - -- --

To conclude, the CSAU methodology None of the points hasarethe potential difficulties that tocannot meet be NRC's require the qualifications noted above. The only one that .

resolved with funher work involving more computer usage.

"unresolvable" category relates to phenomena screening and perhaps falls in, the is thrown sufficiently wide and the expen group is made to even there, if the neta larger cross.section of the community, then there is a good represent As an aside I would like it noted that nothing really important will be omitted.

alternatives to response surface methodologies Such(e.g. adjoint methods methods) are mucy exist th screening to reduce computer usage.

require phenomena more "foolproof."

My impression in reviewing the work was that it was done quickly and u of whatever was available to shore up the case.

nodalization studies, etc. were often done with earlier versions of no there are TRAC.PFI or e It is hard to feel reassured by the glib statement that with TRAC.PD2.

significant differences in the calculated responses between all these vers Nonetheless, the presentations were sufficient for me to give qualified posi answers to both questions.

On another subject, I have not had the time to review the TRAC QA docume depth.

I strongly recommend that a separate review be conducted.

me if you need funher information.

Please contact Yours si :erely,

)

S. Banerjee Professor and Chairman Enclosures SB:bef References The

1. D. G. Cacuci, "Deterministic Sensitivity Analysis of Nonlinear Systems:AnnivrLi Forward and Adjoint Methods," in Uncenuntv Press Boca Raton, Florida (1988).

2.

D. G. Cacuci, Sensitivity Theory for Nonlinear Systems Parts I and Ehn.,22, pp. 2794 2802 and pp. 2803 2812 (1981).

"An Exact Sensitivity Analysis of a Simplified Transient

3. E. Wacholder et al. Ene,12,, pp.133 (1985),

Two. Phase Flow Problem," Nuc. Sei S

v. . . . . . ,

4. M.C.G. Hall. "Estimating the Reliability of Climate Model Pro Towards a Solution"Ca: bon Dioxide" M. C. MacCracken & F. M. Luther, eds. U

1. DOE /ER 0237 (1985) 5.

E. E. Lewis "A Transient Heat Conduction Model for Reactor Fuel Elemen Ene Des.11, pp. 233 40 (1971).

e 6

.m c.. w ..t.; .a z .o,:

$Y.Q,.?U:'%:W. i. L: .@ p .@. . . . . i. . . . ... -. : . ~ : . . .~.: ..

CM Conge Ooouwn or UniverMy cf of Ergreerrg Cremes Em Neustmt, Y esas 004 ,

713749 4407 January 20,1988  !

)

Dr. Neil E. Todress j Department of Nuclear Engineering l Massachussetts Institute of Technoingy '

Room 24-102  !

Cambridge, Mass 02139  !

Dear Nell:

I 1 have now reviewed my notes and the vast array of documents we were given at the NRC peer review group meeting on CSAU methodology l i

and can give you my comments. in your report I suggest you commend Zuber and his group for the superior organization, planning and documentation of '

the two cay meeting. I have been a consultant to NRC and ACRS for over ten l years on and off and have never before participated inwhich a review wasmeeting presentedin  :

which all the information needed was available and l in so logical a manner. This took a major job of organization and planning.

A.

GENERAL COMMENT

S:  !

1. The claim that the methodology quantifies the uncertainty in safety of a nuclear power plant is still to be fu!Iy confirmed for reasons  !

discussed below. Even so the CSAU methodology, if adopted, can made a major contribution to safety by (a) documentation to reveal code errors, forcl and a systematic review of that limitations and innacuracies and (b) forcing a systematic review of all uncertainty and ,

the integral effects data to evaluate their validity to the full scale nuclear power plant, if ,it did nothing else applicability this would be a historic accomplishmenu. 1

2. The CSAU methodology is represented as an objective near quantitative method for measuring uncertainty in performance of EC6S. As a result of this description it is immediately subject to criticism by those well informed in quantitative analysis and statistics. That happened repeatedly during our meeting and is reflected in the ACRS letter on the  ;

subject. Instead, it should be described as a formal r9ethod ed wav fortocombinine minimize cuantitative s nalvsis and exoert coinion in a control subjectivity 0 the eroerts in arrivino at comouted values of uncertaintv.

i 3. The methodology was demonstrated by application to the peak clad 4

temperature during the blowdown of a Westinghouse 4 loop reactor after a i large break. We are told that CSAU methodology is now being applied to the reflood peak temperature for which there is a large body of integral j effects tests at very large scale. Because the use of these integraland uncertainty effects for  ;

tests are so central to the method decisions on noding for the full scale calculations (for deciding) there remains the i j

e

< t

_ . _ . _ _ , . _ _ . _ . __ _ _ __ _ _ _ . _ .. - O

Q '.Q.or y.;..y5 3 .;. m .

43,333 33 y _ ,

. y question of whether the method can really be used for all other accident scenarios and plant tyses. Before this can be decided we have to be sure there are enough of 1he right kind of integraluntil effects thetestseffortavailable. *lt is actually ,

seems to me this cannot be determinedundertaken and step 7 in the is reached for each of these other scenarios or plant types or both.

4. It is' represented that the method is independent of the code being '

evaluated and the scenario selected.

it is true that once the scenario is selected then step 3, the identification and ranking of specific phenomenais independen expected to occur in the course of the ECCS by the experts,ificantly code of the code. However many of the other steps are sign dependent and will require an expert's familiarity with the code. The impression should not be left that carrying out this process o  ;

matter. .

B. SPECIFIC COMMENTS:

Step 3: Phenomena identication and Ranking: The method makes good use of expert opinion in a controlled way whicli can be c'osschecked .

and evaluated to arrive at a ranking table, in the demonstration only '

phenomena which were ranked 9 were subsequently evaluated, in discussion it was indicated that lower ranked ones could also be evaluated. However, an important judgement call is how far down on the list are the phenomena which are to be evaluated. Just preparing the list is not enough. There should be a basis proposed for deciding how far down the list one must go before the phenomena which follow can be ignored.

- Step 6: Code Applicability: I am uncomfortable with some aspects of this step. The process looks at the code at sev ,

Second does it have equations or models for the individual phenomena listed in the ranking table? Third are there any defects?i The ll first two the aretechnical de easy to execute. The third is not andquality of the investigator with recent literature on a wide range of subjects, must know the  ;

difference between an empirical correlation and a physical model, must i understand where there are gaps in the technical understanding that simp,1y are not filled. Furthermore two phase flow, theheat range of technical transfer, exoertise drop mec1anics, ,

required boiling, flow is very large ( ing phenomena and much more). 'It seems to me patternr..wett this cannot be done with any confidenc6 until the code has been  :

independently reviewed by a group of experts for soundness in its l technical detail. This idea is addressed at the end of this letter. ,

l Step 7: The Assessment Matrix: This experimental results used in 1 this assessment matrix focus largely on data from integral effects tes(

(Loft, semi scale etc). L various aspects of code development and tuning. Of the 40+ assessment l siudies of TRAC only two were made against university single effects data  ;

as indicated in the report by Boyack.There exist extensive single effect  ;

data in the literature from universities which have not been evaluated /.

Part of this process should em data not in used in code tuning.phasize the need to locate l

..-.a.......,.

B.,, . . . . . . . .

3 The weakest point in the methodology is Sto 10: Scalin Effects:at least as applied to peak clad temperature for the dis ayed in this s , data from integral tests taken on equiEment at five bio down. The PC scales are plotted vs linear heat generation rate. F. The 95/. confidence Because there is no curves show a scatter in the data of +/

361 discernable effects of scale in such a plot (by some simple stat it is assumed there is none. It is a weak argument. linear heat generation rat tests)For any narrowMore ransie of values ofdata itself varies in P likely this comes from the fact that simaly experimental error or operating conditions were dur ng these tests certain initial systematically varied between runs. (if this was not the case one would have to question the technical judgement of the whole Loft project.) These factors were not taken into consideration in step 10. If the variation in was truly exaerimental error then the PCT in the plot of the Loft data, say, table. But i; will be necessary to argument might be partially suppor carefully study the data for any one test series to establish this fact before the conclusion is acceptable.

C.THE TWO QUESTIONS:

Here is my response to the two questions posed by Lou Shotkin.

1. 1 The methodology appears to be systematic, logical and practical.

in eart demonstrated a methodology which shows

1. 2 The NRC has that best estimate metnods can be used to meet requirements of the modified ECCS rules. Some weak spots to ainwide the variety methodofhave been scenarios identified and its general applicability and plant types remains to be fully established. .

D. CONCERNS AND RECOMMENDATICNS:

1. The very formulation and acceptance of a plan for estimating uncertaintly in loss of coolant accidents will probably suggest to theth Nuclear Communit sical phenomena underly,ng i ECCS performance. One can research on the p expect that contr ctors, utilities and NRC will find it more difficult appear to justify research. The fact that appalling untilignorance such time asand a errorr strange and unanticipated sequence of events causes an accident knowledge or near will become accident and then t1e inadeqacy of the existing apparent once again.

2. The members of this group whoreflood are available should be constituted when completed. We have a ,

once again to review the results on considerable investment in time and effort getting to our present knowledge about the methodology and can probably make a more r

expeditious and reasoned review than a new group, i

3. As mentioned in B Step 6 above, there is a difficult problem en evaluating a complex code such as TRAC. NRC has invested vast sums of money in developing the code. We finally have a document (erroneously misnamed Q &A document) which explains in great detail the modelling

C

... . .a.,... .. ......:..... s.s............

- 4 used by the code for physical processes. This document should be examined in detail by an independent group of experts including wide ranging disciplines to judge whether the code includes the most recent research results for the many technical area it covers, where it is deficient because we simply do not have the fundamental research results needed, whether its methods of approximation are reasonable and where it is in error. l suggest this be undertaken by a committee of the National Academy of Engineering under sponsorship of NRC. The the results would then be of unquestioned integrity. This is quite different in nature and importance than the many so called assessment studies of the code by national laboratories.

Neil, I trust this helps you in areparing the summary statements. I presume you will attach the indivicual responses to your summary and send each of us the entire package. I, for one, would like to read all of the comments. It was a pleasure working with you.

Cordiall ,

t

. f. A I

W . . . . .. . :. .; .

.z.........~.-.. . .

TECHNISCHE UNIVERSITAT WQNCHEN LEHRSTUHL A FOR THERMODYNAMIK .

e. PRO 9. oRaf NS W. M AYtNSER 300. M Q NC H E N 3 tweswoo.wavem a . vvw . po..s ace. .. se on . woe.cnew e ARCISSTR A.SE Si n o.., ......

si.e34se E. Todreas Dept. of Nuclear Engineering Massachusetts Inst. of Techn.

Room 24-102 Cambridge Mass.02139 USA ,

va ,.nesn.n va , o.ivm

, , , , . . . .o.

- M/GQ 25.01.1988 Peer Review Group Meeting on code scaling applicability and uncertainty

Dear Professor Todreas,

before starting with my comments on the best-estimate approach for calculating ECCS performance as proposed by NRC I have to beg your pardon for being late with this letter. Returned from my short trip to Washington I was confronted with some new du-ties in my university which I had to accomplish immediately. So, I lost a few days before I could start thinking about my com-ments on the methodology proposed by NRC for calculating ECCS performance with the best-estnaate approach.

For many years calculting ECCS performance was done by utilizing a conservative approach and by giving an upper limit for the cladding temperature in the most unfavourable position of the core in case of a Less of Coolant Accident. Not only before the TMI accident but already when starting risk studies we realized that accident management may play an important role in avoiding

~

catastrophic situations which may go far beyond the large break LOCA-conditions. Conservative rules aiming mainly and almost ex-clusively on the large break LOCA may handicap or even deterio-rate the effectiveness of accident annagement proceduren. There-fore, there is a great demand to judge the 14CA by best-estimate rules instead of conservative ones. In addition we have to rea-lize that an incident resulting from a transient is much more likely than a 14CA. Therefore, I welcome the activition of the NRC developing .2 methodology for calculating the ICCS performan-ce by using a fully best-estimate appraoch.

Before going in deteils I would like to state that the methodo-logy presented Techn chl Program and Committee documentedonbythe thePeer NRCReview and theGroup members of the Meeting in Washin' ton 59.cs to m systematic logical and practicle. It was

.. .. , . . . . ,, - -- - v n.. .u. . a.....:....

.. . m . m . .... .

- . ,... ..-. . . . '. .. . . , e ' -

.. i 2.

. + . .

l methods can be used to meet the requirements of the modified ECCS rules. This, however, in the farst step is restricted to large break LOCAS.

or based.

The CSAU-methodology is and must not lie rastricted tv on to qua a sp@eci'fic ify core cooling computer code. It proposes e.7tectiveness a general and to predict procedure h reliably  ;

claddims temperatures to a m od accuracy.

After at least 15 years experimental and theoretical worldwide research in loss of cooling accidents we should and must have in a best-gained enough experience to treat the ECC procedure 14BI, PXL, s

estimate way. Experimental facilities like SEMISCALE, SCTF, CCTF, LOTT, and UPTF cover a wide range of geometrical With scaling up to the real size cf a PWR-nuclear power plant.

' UPTT a full-scale mock-up of a PWR-plant is available to produco transient data on the thermohydraulic behaviour of the primary system of a PWR and a number of very informative experiments vo-re already performed in this facility. So, with code scalling it is not necessaary to rely fully and substantially on engineering judgement only. The CSAU methodology must systematizes the ap-plication of judging best-estimate ECC-data by using the infor-

, mations community of including all thesethe facilities large scale in the international f acilities nuclear in Japan and in the Federal i3public of Germany.

Certainly, the CSAU methodology has not yet been tested over a ,

' wide range of applications, however, it is a systematic way to l appracch the questions related to core coolability during a 14C-cA.

• r There are many thermohydrsulic phenomena connected with the I

emergency core cooling procudure. These phenomena have to be cp l refully identified and selected to meet the requirements which meet a high status of experience. This experience, however, is available in the international community after doing soInspite many years experiments and computer code development for ECC.

of this experience one may ask whether we do know all importar.t phenomena and whether we are fully aware how these phenomenaThe in-

)

teract with each other producing unexpected situations.LOCA'S l

CSAU-procedure in general is certainly applicable to all l'

but it would be wise to restrict it in a first step to LBLOCA's until we have more experience with this procedure.

Nodalisation may be a key point for succ6ss in applying this nev methodology. However, there are new and more sophisticated ap-l proaches available by adapting nodal zones to a best-estimate in comparing the results with separate effect tests and in-

fit tegral ef f ect tests performed in facilities of different scale t l

up to the full-scale UPTF-facility. Scaling can be checked against UPTF and can be improved in the methodology by doing it.

Codes like TRAC or RELAP 5 were developed end im},.oved in the re-physical models implemented in, by checking the correlated sults against experimental data gained in the SEMISCALE and LOFT facilities maiily, ene may argue that no independent checking

~ ' ^ ^ - ^ ~ ^ ' 90 mann

.[ yQ:u.p .[.y;.. ,y e. :, < .

. .. . t .. a .:,,., . a

> . n.. ,

...r. .

l

.. l 1

example, from PKL, from LOBI, and also from other facilities the best-estimate methodology. So which can be used to assess users should and can take independent data available in the in-ternational literature to check it against the code.

One also may argue that new phenomena, unknown up to now, may arise which may make it dangerously to apply this CSAU methodo-logy. According to my opinion and experience the procedureleast with LOCA is so thoroughly researched and clearly defined - at as we restrict our effort to large break LOCA'S that one must not be a'fraid to be confronted with completely new phenomena in the future. 1 During the presentations on the Peer Review Group Meeting it be- i came evident that ranking the phenomena for doing best-cstimate of low ran- I calculations is of great importance. Also phenomena if their effects ars i

king may have an unexpected influence, j known with extremely low accuracy, and if by interacting with a '

phenomenon of higher ranking order they can change the fluiddy- l namic situation. Therefore sensitivity studies should be recom- J mended to check carefully possibly unexpected effects.  !

Applying CSAU successfully and in a correct way needs a certain therefore re-competence of the user, expecially the code' user, gulatory institutions have to take care whether the user is com- <

petent enough. In general good enough experience can f.e expected with users who are active in licensing calculations on the side of the utilities and of the vendors as well as on the regulatory side. These institutions are concerned with ECC-calculations since many years. . ,

It must be free to the experienced user to apply a code which i has, according to his experience, the highest status of develop- to be ment. Applying such a code regions of uncertainty have pointed out and the applicability of the CSAU-method together with this code has to be justified and documented.

When the methodology is finally matured the NRC will have an ef- In fective and independent tool to audit industry submittals.

the presentations during the Peer Review Group Meeting good methods and first results were shown to proof at principle that the CSAU methodology will allow to elaborate reliable data for LOCA-analysis, but more work in this direction should be recom-mended.

Yours sincerely,

\-

Prof. Dr.-Ing. F. Mayinger NN t

V.. .s5%.vu m:.J:nMM

~ - ^ u. u . . , , ,- -

--m ,,

e he THAY 'R T M SCHOOL O U T H OF C OENGINEERING L L E G E D A R January 22, g gs c g seg , .,

c , .. . ., ,,

_ ...e" Dr. Neil E. Todreas, Chairman D pt. of Nuclear Engineering _ e] ,,

f ~

hlassachusetts Institute of Technology e'

W "*

Room 24102 Cambridge, MA 02139

Dear Neil,

I have reviewed my notes from the Peer Review Group meeting l in Bethes 13,19SS, as weu as documentation supplied before, during and U methodology in after the m summarizes my reastions to what I saw and heard. I wiu first discuss the CSA general, then comment on its piecesc The overau CSAU methodology is reasonable and,in principle,is a logica Best Estimate predictions and experimental data for estimating fi d uncertainty.the probability t such as peak clad temperatures, will lie within some required range, with speci e The main questions concerning CSAU involve the i l adequacy ntiaDy but and practica each step in the process. As each "box" in the diagram his considered, not ent re y seq with some iteration and cross. influence, judgments have to be made about " ow muc For example, the , odes for reactor analysis have overjustified h weakly 100 empirical or which appear (as in the TRAC Q/A document) to have been chosen on af rat er basis. They have been steadily upgraded over many dyears i "mature" by tuning the c correlations to fit data. An assessment has to be made somehow w and ready to be ' frozen."

Similarly, the assessment matrix contains a limited h number plicability of a of tests in r at various scales. The decision that this provides enough evidence to ble assess error. t e ap certain code can only be made after exhau.<tive comparisons, including criteria fo Likewise, nodalization is based on 10 to 20 years h of experience

' but it is and the it has converged to an acceptav.e level of detail to "capture the important p enomena, hard to justify this in a definite, concise way.

Uncertainty analysis seems to be most rationaDy based on arrays of pre ill has to address the a code by varying a set of chosen parameters over a crediHe range. One st difficult questions of how many parameters to use, how many runs are nee ti ly surface, how to assure that the combination of au other" uncertainties has been estimated and so on.

- m w orts sa ns m ws osmo

m m _ _

D.(..j. h 2.c The various presentations in Bethesda showed d range of that doubtlong about steps have been tak j

strating practicality. However, the PRG's questioning key issuesindicated to rest. a broa Therefore, whether the evidence presented was adequate to put the fh various U methodology, ,

consideration has to be given by NRC not only to the acceptability o t e overa h but also to whether the applicants for licenses and the NRC staff are yet in a po w is that the numerous references to the use cpproach in all its details. Pa.ticularly, my own vie ible, to well defined, '

of "judgment" in making key decisions needs to bel changed, ll d out. as much as poss quantitative criteria for which a documented and Q/A'd rationale is clear y spe e Detailed Camments In the introductory talk it was stated that the l has accuracy been logically requirements derived in should be co with the degree of risk. Presumably this means that the 95% leve some way by considering the probability of an accident and its consequences.

Step 3 involved PIRT. Eventually the selection of important parameters m hellouche/Wuld would later on a more deductive basis than "asking people." We were toldf hthat "neglected" ,

confirm the expert evaluations, but they did not make separate evaluations o t eld be a variables. Showing that the statistical spread covered the data l cou i ortant.

- major variables over too large a range, rather than being evidence that all e se was Lots of "experience" went into the choice of major variables; does this mean tha research base will be needed for new designs?

Zuber's overview of the NRC sponsored eEorts over many years was realistic.

against the "good enough?" question at many points.

l To an outsider several of the assumptions inbeTRAC seem successfully tunablestrange. They wou ticularly good for predicting simple flows in straight pipes, i but appear d makes to to it hard to reactor geometries. Having 175 parameters in a code seems excess ve butan keep track of rational comparisons. In retrospect it would h dicalhave been better l

physically based, parameters that could be checked against many phenomen l

way in Steps 4 and 5.

l the criterion is for In Steps 6 and 7, where the code is being assessed, I'm not clear i wht.t l

l "passing" and insuring that faults in the code itself do b llnot get through to in

)

I issessments. I'm not sure about gathering au sorts of "uncertainty" under the same u l The comparisons we saw with LBLOCA may be good because the flow r In the SBLOCA. more phenomena are likely to be close to homogeneous and therefore simple.

important and there is a greater need for a realistic code.

Nodalization was discussed in St_pJ. There's 5'K isanot lotreassuring:

of experience behind statement that Sandia doubled the number of nodes and changed PCT by 7 this is comparable with the uncertainties due to other causes.

b.6e .;.6 e E n. & M .d.ef r. s.. . .: m; w . . . . . . .

i BNL's work on assessment of uncertainties associated with the fuel, Step 9, was logical and methodical, the type of thing that wiU have to be required to get a good uncertainty prediction.

The presentation on scaling, Step 10, was a little odd in that a very simplistic approach wa y

taken, compared with a much more detailed approach in the other anaj ses. The range of P 750'F.for all data points, so plotting them versus the main influence of heating rate 1050'KF to reduce the uncertainty by half is not surprising. Surely many patterns in the data are hidde and there are specific known reasons, such as experimentai set points or initial conditions tha account for some of the trends. I'm nervous about taking such a grand view, treating au scatter as "uncertainty," and deducing "no clear trend with scaling."

The calettlations in Step 11 were very limited, involving only seven runs in which thermohy-draulic variables were changed. The changes were large and the effects were large, making linear quadratic fits to the results less secure. I believe we saw a data point for which the pre in PCT was much larger when three variables were perturbed than the sum of the changes w each was changed individually, indicating some "cross product effects that need moreinvestig (this did not appear in the tables supplied to us, so I may be raistaken).

Since the temperature versus time curve is double humped, with either hump being higher under different circumstances, the response surface may not be smooth.

The uncertainty analysis, Steps 12 and 13, is a key element, and one of the few new ones, in the methodology.

The overau approach is pertinent and reasonable. The various statistical comparisons are appropriate and encouraging. h!y major criticism is the attitude th product was "good enough." Surely more tests are needed by running the code to eval cross products of the thermohydraulic variables

- the impact of the neglected variables trends in data that are not due to randomness but to known differences in the experiment There are more significant variables for the SBLOCA and ways wiU have to be found t many more computer runs to cover the whole spectrum. ,

I am not familiar with all of the integrated effects tests but my impression is that TRA predicts an NPP like LOFT because that is the main basis for its tuning. Were L duplicated in the other tests?

I assume you will pass these comments on to the appropriate people at NRC.

Sincerely. .

W tL -

Graham B. Wallis Professor of Engineering GBW/dec

. 7, ; . , . . . . .-mT.w m a w a e - g~- - - ~ - - - - - -

.. 800 J1.21 17123 FPCM 4305 63111 DRAFT PEER REVIEW GROUP MEETING ON CSAU METHODOLOGY BETMENDA, WABMTNOTON DC, 12-13 (ANUARY 1988 assessnent of the material presents my The following the presented at this Review and addresses the questions put to Review Panel by Dr L Shotkin.

1. TME PROBLEM cal-The USNRC proposes the use of best estimate codes to a modified approach to as part of culate ICCS performance the requires an estimate of licensing. This approach

" uncertainty in the calculation of a small set of key parameters.

for determining this has developed a methodology The USNRC methodology to uncertainty and has subjected this calculation The members of the Peer Review Group are asked to peer review.

answer four questions which I interpret as follows:

(i) Has the USNRC demonstrated a ' proof of principle'?

l Is the methodology systematic, logical and practical l

' (ii) and adequate?

I 1

i

~' -__ _ _

@  %. l FROM 8385 63151 ..  ;

an inde-Is this methodology adequate to provide (iii) submittals, which are of industry pendent audit independent likely to use a somewhat different and methodology?

guidance Does the USNRC methodology provide useful

. (iv) and to industry in producing their own methodology  ;

I submittals?

This note sets out my own conclusions, after examination of the meeting on 12 and to us at the documents presented in these It has benefited from points made 13 January 1988.

presentations and from discussion with other PRG members.

I I have also made use of UK experience, as follows:

of over a number Use of various versions of TRAC, (1) and years in the analysis of a range of experiments of plant LB/LOCA transients.

particularly in Interaction with other TRAC users, (ii) contacts, US but also through international the including ICAP.

i code i Independent UK analysis of the problems of (iii)

I uncertainty estimation.

l >

w. .w.. .,.m-._ , _ _ _ _~

w__ __

o- raon esos assis o . , ,.

CSNI with other organisations in the (iv) Collaboration in an attempt to reach an international consensus.on code uncertainty analysis.

2 SCOPE OF THE REVIER I in this is important to recognise that at all stages It For demonstration, we need to exercise ' engineering judgement'.

only believe that a ' proof of principle' is this reason I and showing developing a specific implementation possible by l ractical and that it works i.e. that it is systematie, logica , p .

O adequate. I make three points:

9 USNRC I am impressed by the work carried out by the (i) with 'the They have genuinely got to grips team.

have produced quantitative results.

problem and '

undcubtedly of of the material presented is Most permanent value.

of the There are areas where I disagree with some (ii) numbers and assumptions.

In most cases I am satis-of a this does not affect the adequacy fled that team demonstration of ' principle;' and that the yet been wottid accept that the last word has not i

them to therefore not allowed said. I have influence my overall conclusions.

3

% . . , , .. . . _ . ~ . .w m , _ , _ , .

'se.ei.at sp 2s ..

FROM e305 63111 on the presented allows conclusions (iii) The evidence PCT.

uncertainty of calculation nf the blowdown PCT There is only limited analysis for the reflood

, It is clear and no discussion of SB/LOCA problems.

more that the analysis for the blowdown PCT is much I SB/LcCA.

straightforward than for reflood and the demonstration is conclude that, at this stage, PCT uncer-confined to the estimation of blowdown extrapolated to tainty and its success can not be cover the other cases.

3. TRAC-PF1/ MOD 1 STATUR estimate l It is central to the whole approach that the bestmaturity.

of code used must have reached a satisfactory state To no this implies four conditions.

major The code modelling is adequate to capture the the (1) which influence therani-hydraulic phenomena 1

transient.

for Enough assessment studies have been carried code out (ii) representative SETS and IETS to show that the accuracy for satisfactory quantitative achieves only if this can be claimed is it major variables, concentrate on a particular than acceptable to For exanple I would variable such as blowdown PCT.

I i

4lne.[ - E.;2.7..2;;.u 2 w..o.

se.es.2: sis 2s non esss ssti:

\

code which gives an not be satisfied to use a  ;

accurate PCT prediction but got the hydraulics.qu'ite I engineering judgement wrong. This is an issue of On the other which it is difficult to quantify.

about scaling l

hand if it is not met I have doubts transient and about robustness to small changes in definition.

O been sufficiently The SETS and IETS analysis have (iii) NPP.

aw&nne(tra en enppnet the nndino rhngen fnr the the in general, one cannot defend

' I accept that, that they schemes used in TRAC by showing noding node are, in some sense, saturated against further refinement.

In practice, at the noding levels used, between a finite are a compromise the codes approach. The and a lumped parameter difference matter choice of noding for the NPP is therefore a of judgement.

for an At the stage where the code is to be used (iv) for uncertainty assessment it must have been stable a period (frozen?) and properly written up (QA).

2 I am personally satisfied that the version of TRAC-PF1/ MOD 1 certainly to used for this exercise meets all these criteria, inde-the level needed for a demonstration and possibly for an on I rely For the first two items in my list, pendent audit.

5 1

. . ,, y.

, .. ; ., . .y,. . ,

~ -

} ' . . . .I e . 'e' g E g'g'. 21 ~ 17 ' 2'6 -

FACM 9303 63111 UX. For the evidence from our own work in the independent second two my comments are:

support their The case presented by the USNRC to

"(i) a thorough noding choice has been the subject of review and is well documented.

time (ii) The TRAC QA document has been awaited for some On first review I regard ,

and has only just arrived.

time it an an excellent production but it will take

' to examine it in detail. .

4. TUNING AND SCALING These two issues led to considerable controversy at the PRG I can only state my own position.

, , discussion.

the The central role of closure relationships implies that only This presents problems codes are always fitted to data.

and the when this is not adequately separated from evaluation the It is also only objectionable if database is too narrow. scale suspected not be fitting processing is seen or data the independent. One approach is to fit only the cerrelations; claim other is to use integral data and separate effects but to There is also the philosophical difficulty to model phenomena. the (more fundamental?)

that as codes become more advanced I

correlations they use become more numerous and more complex.

6 l

l i

d

uvuav

- - - . . . . .c.%- ; . . .c , ; . . . _ , , , , , , _ _

'80.08 88 FROM 0385 6 Sill I (now we ['l do not identify any deliberate tuning in the code and such ,

l have UPTF data) I believe the database is too wide to make l tunir.g a practical proposition. .

The presentation and discussion has considered the issue of whether scaling is is scaling. In my view the real issue If it is, then it is not of properly included in the codes. scale.

direct importance that the experimental facilities also that it requirement on the experimental database is then The MPP be vide enough to capture phenomena relevant to the should code can and that analysing this database with the transient modelled. On that these phenomena are adequately

' demonstrate a the evidence available to me I am not satisfiedNPP that here is.

satisfactory procedure available that can be used to scale uncertainty.

calculations by adding a bias or by increasing the id However for t.e caso presented, the blowdown PCT, I amnosatisf e scales correctly (or that there is that the NPp calculation j additional scale effect).

I

5. OVERALL STRATEGY OF THE METHOD important The issues raised in the previous discussion are only analysis that follows is because the quantitative I also note meaningful if these issues are dealt with properly.

i that they are almost en*irely matters of opinion (or engineer The ng

to quantify.

judgement if you 'like) and are difficult N

7

~ .. + . . . . , , - . , -- -- - ..

ng ,3,3 ,3,,, ..

five stages. I quantitative part of the exercise then comes in .

identify these as follows: f 4

a small (i) There is a NPP calculation line in which identified and  !

number of major uncertainties are '

(and proba-these are given an uncertainty range distribution). A sensitivity calculation bility seven 1 then provides a response surface (PCT against variables). By straightforward techniques this then gives a probability distribution for the PCT and tha mean value.

direct I (11) A separate analysis (W Wulff BNL) provides a back-up to part of the NPP calculation. .

by com-0 A bias and uncertainty range is obtained (iii) _

puting a number of fuel pins for LOFT L2-6.

1 (iv) A set of five facilities each claimed to give a good simulation of blowdown PCT for an LB/LOCA is used te provide a mean and an SD for this ensemble.

(v)

A final analysis is presented which attempts to pull this material together to give a bias and SD for the NPP calculation.

3 I give my comments below, against these headings:

9 8

j2 8 E

' ', $.:,,.,3 ; y ;,:., , a:.,d .

. g .,. :~, , , , _

Wl,^. ,. .e 17:2s Er.' ~' * .- -,.:

'es.st.21 Fton esos 63111

6. THE NPP chtc0LATION I believe the ..

I ' regard this as the heart of the matter.

other analyses should be regarding as supporting this main line.  ?

persuade [

I rely on the other calculations to assess bias and to t using only me that essential uncertainties are not omitted by for the ,

NPP calculation route, or to make some allowonce the i missing items.

• ~

I am impressed by the A key issue is the PIRT analysis.

I think it was valuable to care and detail that went into this. analysis ,

have had two teams to carry out the quantitative PIRT clearly they are not independent in any : statistical though but sense. I see no reason to disagree with the PIRT rankings It was accept that they may change in detail with experience.

were not entirely clear but I assume that items with ranking 9 9 were given a automatically included and those from 7 to accept that there are good grounds for detailed review. I regarding the results as code independent.

i the deter-1 The selection of the seven PIRT variable and mination of their uncertainty range is the key to the method and independent. The be claimed to be code both these can for uncertainty estimation are objective and techniques used neither could be repeated by another team but they are clearly f For some problems, e.g. fuel pin stored j

mandatory or unique.

9 4

_. ..e...---

- , , ..;- v; . .ac. c..c....= w a a -- ~--

' " ~~~

i..

in detail. For used can be defended energy, the approach of much a matter modelling, it is very e.g. pump others, done is I conclude that what has been subjective judgement. d for other evidence sensible ,and defensible but there is a nee nothing essential or arguments to provide some assurance that not been has been missed and that the uncertainty inputs have underestimated.

dis-am not worried about the choice of the probability These are enough variables I

tributions for the PIRT variables. Statistics.

be able to rely on the Central Limit Theorem of to to treat the In this case I believe it is adequate to allow us than PCT distribution as gaussian so long as we do not move more two standard deviations from the peak.

judgement I am aleo somewhat influenced by my own personal nable.

that the quoted standard deviation of 172 P is not unreaso I would expect a value between 100 F and 300*F.

BNL

7. hNALYsIS BY W WULFF, main calcu-I found this analysis very supportive of the noding line.

It provides valuable back-up on scaling, lation PCT It gives a value of 80 F for the 50 of the arguments 2tc.

due to fuel stored energy.

10

\

w , > .. . . . - a. . - r a . . . . - - . - - - ~ ~ --- ~-

e e

e ,

3. Lo7T L2-6 ANALYSIS of This is an analysis of calculations of PCT for a number had pins in L2.-6 which were at different points in the core and

-TLOTT) had a a range of linear ratings. The difference (TTRAC bias of 23 F and a standard deviation of 120 F.  !

reason for the the It is very difficult to determine this to other

~ and to relate scatter in (TTRAC ILoFT) analysis The USNRC in the NPP calculation.

uncertainties instrumentation except to argue that presents no guidance My own view is that relt tively few of uncertainties are small. in the the uncertainties demonstrated in this analysis appear that instrumentation Npp analysis. If I accept the argument conclude that (and test) errors are small then I am forced to of the uncer-these errors are an estimate of at least some tainties ' missed' in the NPp analysis.

extended, would like to see this part of the exercise I help to It would also possibly by analysing other LorT Tests. was analyse the uncertainty background and compara it with what satisfied in the NPP analysis. In essence I am not includ2d via this type of analysis, that the way to use this LorT data, I conclude however that to has been fully thought through. possit:.fstic treat it is an independent uncertainty component is even if not fully argued.

11

_ I

- - - -o-

,,gg""g3gg.333, ..

9. EYPRRIMENTAL FAcILTTY ANALYSIS heat generation The blowdown PCT is plotted against linear f 360'T is rate for five facilities and a 95% i h an SD of 180 F.

confidenceIt band o computed which I take to be consistent w t predictor of is possible to regard this ensemble of data as aThi the blowdown PCT for the NPP. Some of the uncertainties in the are well designed and scaled. d some of those for NPP analysis are due to scaling up effects test conditions. an If I the experiments are from instruments and t both this that ignore these two items there is an argument rtainty input. tha and the NPP calculations have the same unce

10. FINAL ANALYSIS I see the conclusions of the exercise as follow (i) HPP sensitivity analysist 0

T mean a 1106 F i

~

SD = 172 F to:

Augmented by L2-6 analysis the SD increases l

0 SC

= 208 7 l

12 f

)

. ~ _ _

L- s. J;.--~ z; .;g;, ,, _

5 ,*  :

'88.81 21 17:37 PROM 8385 631 1 I

Experimental results:

(ii)

Tmean = 1200'F SD

= 180*F td it is possible I conclude that, on the evidence presen e , bias of -100"F d rd to take the NPP calculation culation is optimistic) with final aas reconciliationthe best es this of 200 F. I felt that d required further thcught deviation be exercise was in a preliminary stage anI believe however that and a better presentation. .

done. .

eoMcLUsToWg -

as i f and they are I have purpopoly made my conclusions br e follows:

to for me to The material presented by the USNRC is adequa believe 1 There are places where I In assess the method used. by additional work.

the argument could be strengthened terial the final section which pulls all the ma me particular, This has not prevented together should be clarified.

from coming to conclusions.

13 i

-,,,...:..a.,.:..w...:a.. a. ,.a.>. .-.. .. .: .. u , ,,

+y......t rnoa esos 63111 ing the blowdown PCT.

A method has been presented for assass will work for th.e 2

on how well it I have reservations argur.ent is PCT and accept that this part of the reflood I am not able to form a judgement as to not yet available. i nts.

how it could be applied to SB/LOCA trans e of engineering The method implies a substantial exercise I note in 3.

most of which cannot be quantified.

judgement, particulars itable and (1) Evidence that the best estimate co'de is nature (CA and validation).

their uncertainty The selection of variables and (ii) range.

d no r.ajor Evidence that the analysis is complete an (iii) factors have been omitted.

i make these judgements I note however the techniques usedthese to approaches are quantitative when possible and believe sound and a useful model.

1 1 14 1

1

M z; _, _

.-+a.e m.._ _ . , ,

~

• t i^ #

' 8 0." 01.' "3 8 -19:33 FROM $309 63111 l

logical systematic, I

accept that the proposed method is other 4.

practical, I believe it is a useful model for and independent audit.

approaches and can be used for l J TELL Deputy Director Water Reactors prograrse UKAEA AEE Winf rith Dorchester Dorset DT2 SDH UK 21 January 1988 (ND t

l l

l t

l l

l i

l l

l l

\

i s s i 15

\

MEMO To: Peer Review Group .

Code Scaling Applic bility and Uncertainty (CSAU) Methodology From: N. Todreas Date: January 22,1988 views o' the CSAU Methodology presented to us by the This letter summarizes These views, together with those submitted by each of you, NRC on January 12-13, 1988.

will be submitted with my summary letter of our deliberations to the hRC Office of Nuclear Regulatory Research (RES).

1. Value of the Process The NRC should be strongly encouraged to continue this effort not only because inherent physical logic of a best-estimate and uncertainty approach but also beca of the attendant benefits to the structuring of its research program. The creation the Methodology and Demonstration have already reinforced the value applying consistent coce version throughout an evaluation process, documentation of a key US LOCA analysis code and stimulated a reevaluation of broad range of valuable and costly experimental data. Further, future research should be able to be identified more readily and justified -

alternative Methodology.

Inevitably, however, the successful licensing application of this Methodology probably maintain large break LOCA (LBLOCA) limits in the position of dic the real operating limits. This is not a problem if the uncertainty can be establi with sufficient confidence. If. however, this process of establishing uncertainty ev to one requiring statements with statistically Rat her. rigorous rigor into the establishment of the Methodology and the Demonstration.

while elements exist where statistical methodology has been applied, the establishm of the Methodology and the Demonstration are replete with steps based principal engineering judgement albeit of experts in the field.

2. Is the Methodology systematic, logical and practical?

Its practicality really can only be I find the Methodology systematic and logical.As discussed under 3. its Demonstratio confirmed by its Demonstration.

blowdown peak clad temperature (PCT), while not fully difficult to achieve because of the complex nature of processes.

There remains important conunents on the Methodology with regard to the assessm matrix, scaling and nodalization.

a

m .- =-

~

Peer Review Group January 22,198S ,

page two

. It would be desirable to broaden the assessment matrix to -

l separate effects tests since relevant ones appear a  :

not effectively based on the same data base.

The Methodology does not present a method for scaling supplementa presumption that the code models themselves contain sufficient ph reproduce scale effects. A key element of the Demonstration was t bv exist.

considering a sufficientiv broad However, scatter the in the dat effects of scale even if thev exist, can be accommod physical sense do not exist,is inappropriate and seemingly unnecessary.

The nodalization strategv employed is to require consideration and subse establishment of noding'for each component and then utilize the sam for the nuclear power plant calculation as the data assessments. How number of data assessments made with the NPP noding is limited so use of previous experience in investigating sensitivity to nodalization critical. The sensitivity of results to the number o the selected NPP node array.

3. Has the Methodology been demonstrated?

A distinction between the blowdown PCT with regard to demonstration calculations has been made under i

and the2reflood issues of s nce P practicality and demonstrability seem intertwined.

The key issue on the Demonstration was the adequacy of the calcu define the response surface. The number of cross-products conside particularly the nonlinear effects betweenbehydrodynamic controlled but the parameter explored. The number of required computer runs must ustification for the choice made to date is not physically satisfying. A l Demonstration presented orally and in written form to the peer group wa complete with respect to evidence in the following areas. In all c oral assurances were given but they eventually need to be documented.

The data matrix used to assess the core models and the cod broader than that presented. code resulting in satisfactory perform circumstances.

. . ... ... .. . . . . . . . . 2 ,

. . : . . . . . .: c..... .

Peer Review Group

. ' .)anuary 22,1988 page three Assessments, using different versions of the code (from the frozen one), were utilized as pieces of the Demonstration. Differences between code versions were examined and found loconsequential to the individual assessment being made.

The same nodalization was utilized in all key data assessments and the NPP calculations.

The clad temperature thermal history, as well as other hydrodynamic parameters, were examined in addition to PCT in key assessments of data.

The overall prediction trends were found reasonable.

4. Closing Comments The NRC team is excellent technically, but worked under very tight time constraints to prepare for our meeting. Responses regarding content (I do not refer to opinion where we should welcome expression of differences at this of the Methodology and the Demonstration were not always consistent stage)d and le one to question whether these are "frozen" even for the blowdown PCT.

Additionally, the description of the Methodology prepared for NUREG-1230 of April 1987 should be brought up-to-date.

t 0

4 6

3 + 8

• po s VN -a w~ A..re. . e . . . .. . . ..e r * ~

. - .. s' .. .'3,  ; .1 , .,.^. . s

~

s h

ENCLOSURE D e

e O

9 0

ee

• P,:

C0fMENT LETTERS FOR ECCS RULE 'h

. NO. OF REACTOR TYPES IETTER NO, DATE ORGANIZATION COMMENTER COMMENTS B&W g W_ E ' '

,i 1 4/20/87 Representing Self D. L. Johnson D. L. Johnson 3 If; la* 8/17/87 Representing Self 2 2 6/23/87 Westinghouse Owners Group R. A. Newton 9 3 6/25/87 Georgia Power L. T. Gucwa 4 X 4 6/22/87 Westinghouse Electric W. J. Johnson 16 X  ;

5 6/21/87 Representing Self J. D. Harris 12  :

6 6/30/87 Representing Self Anonymous 2 7 6/30/87 Advanced Nuclear Fuels G. N. Ward 9 8 6/27/87 Ohio Citizens for S. L. Hiatt 1 i

Responsible Energy 9 6/25/87 Florida Power Corp C. E. Simpson 1 X i 10 6/24/87 Arkansas Power J. T. Enos 8 X X 11 6/30/87 B&W Owners Group C. H. Turk 4. X i:

12 6/30/87 Yankee Atomic D. W. Edwards 8 X  : ,

13 7/1/87 Northeast Utilities E. J. Mroczka 5 y 14 6/29/87 Combustion Engineering Penn Power & Light A. E. Scherer H. W. Keiser 8

4 X

X

{'

= 15 6/29/87 .

- 16 7/1/87 Flordia Power & Light C. O. Woody 2 X X ,

4 17 7/1/87 Wisconsin Electric C. W. Fay 7 X 18 7/1/87 Washington Public Power G. C. Sorensen 5 X J

19 7/1/87 Carolina P & L S. R. Zimmerman 6 X

, 20 6/24/87 Duquesne Light J. D. Sieber 9 X l 21 6/30/87 Virginia Electric W. L. Stewart 7 X 22 6/1/87 Portland Gen. Electri,c D. W. Cockfield 9 X 23 7/1/87 GPU Nuclear J. R. Thorpe 5 X m

24 6/30/87 TU Electric W. G. Counsil 5 X r g 25 7/7/87 BG&E J. A. Tienan 6 X g 26 6/30/87 New York Power Ath. J. C. Brons 7 X X g 27 7/15/87 Southern Company Services L. B. Long 9 X X X g 28 7/20/87 S. California Edison M. O. Medford 7  :

o 29 7/27/87 Commonwealth Edison L. D. Butterfield 2 X X p 30 7/31/87 Ohio Citizens for S. L. Hiatt 7 Responsible Energy 31 8/8/87 Ohio Citizens for S. L. liiatt 1 Responsible Energy - -

L 32 6/5/87 General Electric R. Artigas 6 X  ;

1

• Note: Mr. Johnson transmitted additional background material to be included with his comments. i

L:.L :.s. =D . .L .L . . .. . . , ,

. .. < .. n .

PARAPHRASED

SUMMARY

OF COMMENTS ON ECCS RULE

1. Can L. Johnson, 1535 Meade Ave., San Diego, Ca., 92116, an employee of Southern California Edison Company commenting on his own behalf.

COMMENTS:

1.1 Mr. Johnson is concerned about steam generator tube integrity in general and specifically concerned about hydraulic loads during a LOCA causing concurrent tube rupture, as expressed in a number of letters to the NRC staff and in formal comments on the proposed ,

ECCS rule revision. He cites tests conducted in Semiscale and incomplete EPRI sponsored work at BNL which showed that under certain conditions (ie., rupture of a specific number of tubes),

tube ruptures during a LOCA can result in significantly increased peak cladding temperatures.

1.2 The NRC should resolve the steam generator integrity safety issues left unresolved and review in detail the adequacy of existing Plant Technical Specifications pertaining to primary to secondary coolant system deterioration in light of the 95%

confidence level; ensuring a PCT of less than 2200"F during the hypothetical LBLOCA.

1.3 The NRC should impose the requirement of a design basis accident safety evaluation, for both LBLOCA and MSLB, as a 10CFR50.59 safety evaluation of any problem or condition resulting from normal operation or design error, which may threaten primary to secondary coolant system integrity.

Page 2 ENCLOSURE O

.-m_ . . . . .

7

2. Roger A. Newton, Chairman, Westinghouse Owners Group.

COMMENTS:

2.1 The proposed changes to 10CFR50.46 and Appendix K, which allow the use of realistic calculational models while retaining the current Appendix K rule, is a preferred approach.

2.2 The uncertainty analysis should not be prescriptive but should remain flexible.

2.3 The 95% probability level is adequate for the large break LOCA and is consistent with good engineering practice.

2.4 The current peak cladding temperature limits of 2200 F, the current oxidation limits, the core configuration for coolability requirement, and the requirements on long-term cooling should be retained.

2.5 The current conservative ECCS evaluation method of Appendix K should be permitted indefinitely for those utilities which currently have acceptable analyses and are within compliance (Q1)'.

2.6 A predetermined explicit degree of conservatism should not be applied to'the proposed evaluation model, rather, the amount of conservatism that is applied and retained will be determineo through the uncertainty analysis using the proposed mods 1 (Q2).

2.7 The proposed amendments to 10CFR50.46 and Appendix K should not be tiad to the resolution of all outstanding safety issues (Q3).

2.8 The Advisory Connittee on Reactor Safety and their consultants should be the peer review body for the proposed rule change (Q4).

2.9 The language in the proposeC 10CFR50,46(a)(1)(1) should be broadened to permit the use of a range of zirconium based alloys l for cladding materials. l page 3 ENCLOSURE D

t.w . :. . ... . ..J L:. a. :.;. s. .. :.: - . .

. -: ~. :- .

3. L. T. Gucwa, Manager, Nuclear Safety and Licensing, Georgia Power Company, 333 Piedmont Avenue, Atlanta Georgia, 30308.

C0lHENTS:

3.1 Georgia Power Company strongly supports the proposed rule.

3.2 The language of the proposed 10CFR50.46 (a)(3)(i) should be of modified to require that only changes or errors greater than 50 in the non-conservative direction be considered significant.

Changes or errors in the conservative direction should not have the same reporting requirements as significant changes or errors in the non-conservative direction.

3.3 The requirement in the proposed 10CFR50.46(a)(3)(ii) should be modified to require that changes or errors be reported in the Annual Report as opposed to a special report.

3. 4' The definition of "significant reduction" listed in Appendix K,Section I.C.5.c should be deleted since it is not necessary if comment 3.2 is adopted.

3.4 The Appendix K methods should be allowed indefinitely because they are conservative (Q1).

Page 4 ENCLOSURE O

4

4. W. J. Johnson, Manager, Nuclear Safety Department, Westinghouse Electric Corporation, Box 355, Pittsburgh, Pennsylvania, 15230-0355.

COMMENTS:'

4.1 The commenter supports the current notice and comment process for rule changes in the Federal Register. The rule promulgation process of soliciting public comment should provide sufficient public input such that a time consuming open adjudicative hearing

. process is not reauired.

4.2 The proposed rule change would allow application of safety research results in the licensing process and is supported by Westinghouse.

4.3 The uncertainty calculation at the 95% probability level is judged to be adequate for the types of calculttiers that would be performed. ,onsidering the low probability of a large break LOCA.

4.4 The uncertainty analysis should be flexible and not prescriptive so that different approaches can be s0ggested by licensees and evaluated independently by the NRC staff.

4.5 The current peak cladding temperature limit of 2200 F, the current oxidation limits, the core configuration for coolability requirement, and the requirement on long-term cooling should be retained. .

4.6 The proposed reporting procedure for code errors or changes and reanalysis requirements are reasonable.

4.7 The commenter agrees that the 200 F limit as the demarcation between a small and a significant code change effect be changed to 50 0F, page 5 ENCLOSURE D

g 4.8 The' current conservative ECCS evaluation method of Appendix K should be permitted indefinitely, for those utilities which I

currently have acceptable analysis and are within compliance (Q1).

4.9 No predetermined explicit degree of conservatism should be applied to the evaluation model, rather, the amount of conservatism that is applied and retained will be determined through the uncertainty analysis using the proposed model (Q2).

4.10 The proposed rule should not explicitly prohibit core power increases (Q3).

4. L1 .m est; ed safety issues which are directly related to core power should be reviewed to ensure that there is no additional adverse risk to the public or utility due to the proposed plant uprating (Q3). .

4.12 A rulo on LOCA methodology is not the appropriate vehicle to andress requirements for plant upratings. Independent of considerations of acceptable Appendix K models, NRC approval of a plant uprating should focus on safety issues which are impacted by the proposed increase in* power level (Q3).

4.13 The peer review of the proposed Appendix K changes is best ace.omplished by the public review and comment pr;cedure together with review comments and resolutions by the Advisory Committee on Reactor Safety (Q4).

4.14 The designation of the fuel rod cladding such as that given in the proposed 10CFR50.46(a)(1)(i) should be broadened to permit use of a range of zirconium based alloys for cladding material.

4.15 There is no technical basis for the continued use of the Dougall-Rohsenow correlation without further verification. A reasonable time limit, perhaps 3 years, should be placed on those licensees who use this correlation, to either show additional page 6 ENCLOSURE D

mz :.-.. . ... - . . .-- ~ .... . . . . . . . . , . .

justification for its use over the full range of conditions for safety analysis, or to replace this correlation with another suitably acceptable and verified correlation.

4.16 The word "immediate" should be deleted from the statement "The affected applicant or licensee shall propose immediate steps to demonstrate compliance or bring the plant design or operation into compliance with 10CFR50.46 requirements" in proposed 10CFR50.46(a)(3)(ii).

page 7 ENCLOSURE D

5. John. D. Harris, 781 Walker Springs Rd. N.W., Knoxville TN.87963.

COMMENTS:

5.1 The focus then (original ECCS hearings in 1973-1974) and now is on the short term (less than one hour) effects without sufficient and due regard for what must be done to achieve the long term recovery of the plant. Hence the proposed rule is inadequate to assure appropriate coverage of the long term plant response.

5.2 To modify the existing regulations without understanding the long term effects or without having a long term recovery plan in place is not appropriate. A short/long term integrative analysis approach is required but is not currently done except at a few Combustion Engineering designed units.

5.3 The long term decay heat removal issue (NRC USI A-45) has not yet been resolved and to modify the ECCS rules prior to resolution is not a good design practice. The "excessive" safety margins in current ECCS design criteria may be providing the protection needed to overcome some long term system degraded performance.

5.4 The statement that the results are highly conservative are only applicable to the short term LOCA impacts on the core.

' Consideration of A-45 is questioned.

5.6 A number of technical manuscripts (unpublished) are provided which nighlight a number of.long term decay heat removal concerns.

5.7 Tha current evaluation methods can be grandfathered provided they are made to inc.ude a long term decay heat removal analysis and a description of how the plant will be recovered (Q1).

.. .... .. : .ru i m ...,~. = .. v . 2 2. .. . - ... . . , , , .

.s.

5.8 An explicit degree of conservatism is appropriate; however, a different safety factor may be needed for different types of .

events (Q2).

5.9 The rule implementation should be delayed until the key unresolved safety issues have been resolved. It is very difficult to recognize if the solution to one problem has created another elsewhere (Q3).

5.10 A review by qualified people would not hurt'. If the proposed changes are viable, it would not be affected. If problems are found, it is best to find them now and not at a much later time (Q4).

5.11 Existing and proposed ECCS rules and procroires should cover decay heat removal and residual heat removal systems. The ability to go to cold shutdown and ultimately fuel removal should be demonstrated using only safety grade equipment. Hot shutdown is not an acceptable end state for a DBA (including both LBLOCA and other transients).

5,.12 The development of a specific post-accident safety philosophy is of paramount importance, pRA and cost-benefit analyses are not i

the only tools that backfit decisions should be made, Defense-in-depth, experience, judgement also should play dominant  :

roles.

Page 9 ENCLOSURE D

^ "' *

~

56 .r. ;, o ' , ~'; . sie ,. ? d. + ' lw- - . -9 ..-w.* '

6. "Alert Expert Insider", anonymous letter received June 30, 1987.

COMMENTS:

6.1 "EMERGENCY CORE COOLING SYSTEMS ARE VERY TENDER (SEE ENCLOSURE ONE FOR A DESCRIPTION OF TENDERNESS) AND TENDERIZER SHOULD NOT BF.

APPLIED AS RECOMMENDED BY THE NUCLEAR REGULATORY COMMISSION ...."

The enclosure is "Maestros of Technology," an interview with Arthur Squires by Hal Bowser, Invention & Technology, Summer, 1987, p.24. The article discusses problems in the way that bureaucracies deal with technology and cites the space shuttle Challenger, Chernobyl, the Swedish 'uattleship Vasa, and the British dirigible R101 as examples.

6.2 "PUBLIC COMMENT IS NOT A FAIR PUBLIC HEARING BECAUSE SOME EXPERT INSIDERS WOULD BE FIRE 0 IF THEY WERE PERCEIVED TO SQUEAL."

l 1

Dags 10 ENCLOSURE D

r w .w _ .s - .-- ,..n... . ,, . , - , . .

7_ Gary N. Ward, Manager, Reload Licensing, Advanced Nuclear Fuels Corp., 2101 Horn Rapids Rd., P.O. Box 130, Richland, Wa.

99352-0130.

C0lHENTS:

7.1 Current models are very conservative and the rule as currently implemented often results in substantial expenditures of resources to install and maintain complex mitigating systems which do not impact true safety and to analyze changes with no significant impact on plant safety. ANF would implement the revised rule to the extent ANF is encouraged by reactor licensees.

7.1 The revised rule should not require reporting of changes or errors that result in PCT changes of greater than 50 F since this is not a safety hazard unless the criteria (ie., 2200 F) is exceeded. .

7.3 Full LOCA/ECCS analyses should not be required for reload fuel suppliers since the hydr.aulic aspects of the LOCA are unchanged by different fuel. Only a specific anal.yses of the fuel heatup is required.

7.4 Continued use of Dougall-Rohsenow should bs allowed since Appendix K contains a larger total conservatism than the non-conservatism in this correlation.

7.5 Equal opportunity to obtain approval of models using either Appendix K or the new best estimate approach should be assured since the industry has invested great expense in Appendix (

models and modifications to these models within the limits of Appendix K would be less costly than shifting to the new method (Q1).

Page 11 ENCLOSURE D

,. .~ . . , , . .. - ~ . , . . .

.o 7.6 The method of Appendix K has proven to be conservative and therefore should be permitted indefinitely (Q1).

7.7 The rule as written along with the criteria of 10CFR50.46 provides for sufficient conservatism without additional conservatisms added (Q2).

7.8 Current regulations adequately address the safety issues with regard to power upgrades without promulgating additional regulations in this area (Q3).

7.9 The review of this proposed rule change that has occurred as a result of the research efforts over the past several years has l

provided a broad based technical community review. ,There is no obvious benefit of additional review (Q4).

Page 12 ENCLOSURE D

c.

4O ,

s

-8. Susan L. Hiatt, Ohio Citizens for Responsible Energy, Inc., 8275 Munson Rd., Mentor, Oh 44060.

COMMENTS:

8.1 An extension of the comment period is requested in order to allow review of NUREG-1230 which provides the technical basis for the ECCS rule change.

Page 13 ENCLOSURE D

~

P % .W J. . . : . . . . . . : JZ%ZZ.. : . ' - ", Z C -X . iC- .

9. E.C. Simpson, Director Nuclear Site Support, Florida Power Corp.,

P.O. Box 219, Crystal River, Fl . 32629.

COM4ENTS:

9.1 Increased power levels increase the rate of fission product formation, but not necessarily the total amount available for l release to the public in a core melt scenario. The amount for release depends on the megawatt-days, with or without the new rule. A five percent power increase would not result in five percent more fission products at the end of life (Q3).

Page 14 ENCLOSURE D

10. J. Ted Enos, Manager, Nuclear Engineering and Licensing, Arkansas Power & Light Company, P.O. Box 551, Little Rock, Arkansas 72203.

COMMENTS:

10.1 The ECCS methods of Appendix K should be permitted indefinitely

~

since the methods are conservative and would allow use of existing models if relief from operating restrictions is not needed or desired (Q1).

10.2 The rule should not contain an explicit degree of conservatism.

The conservatism consistent with the uncertainty of the calculation and guidance in the form of a regulatory guide should suffice (Q2). ,

t 10.3 The rule should not explicitly prohibit any increase in power if it is demonstrated that the requirements of 10CFR50.46 are met, I the uncertainties are adequately quantified and there is negligible effect on the environmental impact (Q3).

10.4 Review of the technical basis for the proposed rule change by an independent group is not warranted. ECCS performance information obtained by NRC, DOE, industry and foreign researchers is ,

considered an adequate basis to proceed with the rule change ,

(Q4).  ;

10.5 The commenter agrees with redefinition of a "significant change"  ;

from 20 F to 500F, but considers other proposed reporting I requirements unnecessary or even undesirable. The existing rule does not require reporting per se, rather only that a description of the evaluation model be revised. The proposed requirement to provide a 30 day report is not appropriate because a 50 F change is not directly related to a safety concern requiring immediate Page 15 ENCLOSURE O

3. g

_. y t 3-

, - - .;- 7 -,---,;_ . - - - - - ,  ;.,, ,

reporting, considering that the change coul'd be in the conservative direction or could still remain well within the safety limit of 2200 0F, An annual reporting of significant changes appears reasonable. Also, the requirements of 50.72 and 50.73 provide adequate controls for ensuring proper notification and reporting; consequently, the additional provisions in the proposed rule are unnecessary.

10.6 The reporting of non significant changes is considered unnecessary until such time that the cumulative effect of such changes is greater than 50 F. As stated before, annual reporting is sufficient in this regard.

10.7 Reporting of "applications of such model" that results in peak clad temperature changes is an unnecessary expansion of existing regulation. Changes to the facil'ty, such as fuel property changes, are already subject to 50.59 and thus are duplicative.

10.8 The backfit analysis is incomplete,since the proposed rule does require new reporting requirements. The current analysis does not address the additional burden of reporting minor changes to the model and each application of the ECCS model.

page 16 ENCLOSURE D I- __

W . r i. . i.';. . .s K > i..:. % : ~...;.,.

.? a :" n. . . . .

". s. . u '.

+.

e

11. Charles Turk, Chairman, B&WOG Analysis Committee, B&W Owners .

Group, Suite 220, 7910 Woodmont Ave., Bethesda Md. 20814.

COMMENTS:

11.1 The commenter agrees that use of best estimate (BE) codes should offer an improved basis for refining our understanding of the margins calculated to be available in nuclear power operation.

11.2 The commenter also feels strongly that'use of Appendix K should be permitted indefinitely for those utilities who do not intend to use the new BE methodology (Q1).

11.3 The commenter hopes that the regulatory review process will focus on the relationship of the BE codes to the data, rather than comparison of BE results to EM results. An affirmation of the focus of the review process would .be a desirable part of the final rule and would preclude any unnecessary and unproductive reevaluations using EM codes.

11.4 Higher power levels do not necessarily result in a higher fission product content hypothesized for release during a severe accident (Q3).

Page 17 ENCLOSURE D

_ . , . . . - . - . . - - - . . . ~ - _ - - -

12. D.W. Edwards, Director of Industry Affairs, Yankee Atomic Electric Co., 1671 Worchester Rd., Framingham, Ma. 01701.

COMMENTS:

12.1 The proposal to allow more realistic and accurate calculations presents numerous advantages including:

1. Improvements in plant operation such as refinement of reactor simulator benchmarking, operator training, and emergency response guidelines.

2. Integrated improvements in plant performance from enhanced fuel design, fuel management, and reactor operation, such as longer fuel cycles, flexible in-core fuel management, reduced vessel fluence, and realistic ECCS surveillance requirements.

3. Improved utilization of manpower and computational resources to support plant coerational safety and performance.

12.2 The proposed approach requiring licences to quantify and justify the uncertainty in model predictions, while not prescribing the uncertainty evaluation techniques to be used, is technically sound. This allows licencees to develop models consistent with the scope and type of application (generic vs. plant specific).

It also provides flexibility to incorporate future improvements in uncertainty analysis methods and results.

12.3 Research has shown that the five ECCS acceptance criteria provide a conservative bound for significant fuel damage and should be retained.

Page 18 ENCLOSURE D

. n..,.~ -

1: . ...u. .. . , . ... . .,

12.4 The commenter endorses the proposal to define a significant error as e deviation in peak clad temperature of more than 50 F. Such definition balances the safety significance with resources needed for reanalysis. The cor.menter agrees that the new reporting requirements will allow NRC to monitor changes as they occur, will not impose undue burden on licensees, and will remove ambiguities that exist with current reporting practices.

12.5 The method of Appendix K should be permitted indefinitely since research has proven that Appendix K methods provide conservative results (Q1).

12.6 An explicit degree of conservatism would counter the process discussed in the proposed rule change. The ECCS acceptance criteria contain additional conservatisms to preclude significant core damage from a LOCA (Q2).

12.7 The rule should not explicitly prohibit any increase in power. I Any proposed increase in power undergoes a thorough reanalysis and review by the licensees and NRC. All regulations must be met, thereby providing for the health and safety of the public and environment (Q3). ,

12.8 Government and industry research has been available for and subjected to technical review througnout this time frame. Public and industry comments from both a previous proposed rule and this proposed rule will be incorporated in the rule change. In addition, the ACRS and its consultants will provide review of the rule.

l 4 ,

a page 19 ENCLOSURE D i

,, . . . . ; c .u . . .a . , x . ~ w. + - - . - - - - - - - - - - '+.c. ... - -- - -

l

13. E.J. Mroczka, Senior Vice President, Northeast Utilities, P.O.

Box 270, Hartford, Connecticut 06141-0270.

C0tHENTS:

-13.1 The commenter supports the intent of the proposed revisions in that they provide for greater operational flexibility and potentially substantial economic benefits, while at the same time preserving appropriate safety margins.

13.2 The current Appendix K methods should be permitted indefinitely.

For plants with large LOCA margins, there shoule. be no need to spend resources to reevaluate the plant. The requirement for many reviam could significantly slow the NRC review process for ,

utilities needirig a timely review to resolve a particular safety issue (Q1).

13.3 The proposed rule should not prohibit increases in power until '

all severe accident and unresolved safety issues are resolved.

To include severe accident issues into the rule change is inappropriate since the capacity and cooling capability of an ECC system is not influenced by fission product release following a severe accident scenario (Q3).

13.4 The proposed rule is the culmination of the results of 20 years of research from a number of independent organizations. It is expected that an independent review of the proposed rule would .

. not be of significant value (Q4).

13.5 The proposed reporting requirements represent an improvement over the current Appendix K. However, the commenter believes 100'F is a more appropriate threshold for a significant change in PCT, -

l Changes in PCT of 50 F to 75 F frequently result from normal

! system and equipment degradation. Such changes are insignificant i in terms of viant safety provided there is sufficient margin between the calculated PCT and the 2200 F limit. -

Page 20 ENCLOSURE D l

-h-

. ,= _ _ __ - ._. __ ___

.y *+

14. A.E. Scherer, Director, Nuclea? Licensing, Combustion .

Engineering, 1000 Prospect Hill Rd., P.O. Box 500, Windsor Connecticut 06095-0500, i i

COMMENTS:  ;

i 14.1 Use of the amassed body of research data to remove excess conservatism is long overdue.

14.2 Past experience has shown that minor code changes are made and errors do occur. Most do not effect the safety of the plant.

Preparing a report to the NRC and having it reviewed would not be an inconsequential task. Having to report each change or error, even on an annual basis, might be counter-productive, j 14.3 Based upon experience with potent,ial sensitivity of LOCA models, a change of 100'F is a more appropriate screening value to employ -

as a measure of a signif4 cant change in peak clad temperature.

14.4 Omission of 10CFR21 from the reporting requirements raises a question as to whether or not NRC has made a determination that significant changes or errors, or exceeding the 10CFR50.46(b) criteria does not represent a substantial safety hazard as defined in 21.3(b).

14.5 There does not appear to be a safety related need to phase out Appendix K models. Economic incentives are most likely sufficient in the long term to bring about adoption of the more realistic model approach (Q1).

14.6 Inclusion of an additional explicit degree of conservatism would have the same effect as lowering the acceptance criteria. The acceptance criteria are already conservatively low (Q2).

L I

i Page 21 ENCLOSURE D t

e.a:;.s :,< , c. :a

. ;c;:_. Q. m -

c-,-~---- . . .. ,,. . . .

14.7 A prohibition on an increase in power until all severe accident issues and unresolved safety issues are resolved appears '

unwarranted. These power increases are the same order of magnitude of power increases already approved in the past (Q3).

14.8 Given the accumulation of data, a review of the technical basis for the proposed rule change by an independent group is not justified. Further, review by an independent group may be impractical due to difficulty finding a sufficiently technically qualified, yet independent and unbiased group (Q4).

l l

1 page 22 ENCLOSURE D

<p:.,,.s;w . c-,: u . :.m :2.2.,- , -- .-

. ;w; = . . -,; -, .

.. . . ~: .

O o

15. H.W. Keiser, Vice President-Nuclear Operations, Pennsylvania Power & Light Co., 2 North Ninth St., Allentown, Pa. 18101.

CONMENTS:

15.1 The proposed rule is a very thorough and beneficial approach to endorsing the use of realistic evaluation models.

15.2 The proposed new reporting requirements for changes and error corrections are reasonable and not burdensome to licensees and NRC staff.

15.3 The commenter agrees that grtndfathering use of Dougall-Rohsenow is appropriate on the basis that there exists adequate conservatism in Appendix K.

15.4 The commenter also agrees that changes which result in reduction of overall model conservatism deteriorate the basis for continued use of Dougall-Rohsenow. However, since all error corrections are improvements, regardless of the effects on calculated temperature, we disagree that a significant reduction in calculated peak cladding temperature should be justification for disallowing use of Dougall-Rohsenow. The conservatism in Appendix K is not changed by error correction.

Page 23 ENCLOSURE D j

16. C.0, Woody, Group Vice president, Nuclear Energy, Florida Power and Light Co.,P.O. Box 14000, Juno Beach, Fl. 33408-0420.

C0ffiENTS:

16.1 There is no need for the annual reporting of minor changes and errors.

16.2 The commenter suggests that the proposed 50 F definition of a significant error or change be moaified to be 100'F or 50% of the margin to the PCT limit, whichever is smaller.

l l

I

.=

2, Page 24 ENCLOSURE D

gy m w.s.;a q.;w. e m av.. .. . - -.c c. .

=

m . ... .. ...

17. C.W. Fay, Vice President, Nuclear Power, Wisconsin Electric Power Co., 231 W. Michigan, P.O. Box 2046, Milwaukee, Wi. 53201.

ComENTS:

17.1 The Appendix K methods should be permitted indefinitely. Current Appendix K methods are conservative and a requirement to reanalyze.would place an unfair burden on utilities with acceptable operating margin. Not all plants are inhibited by the current rule (Q1).

17.2 Conservatism is already included in the uncertainty analysis and the acceptance criteria. The objective of the rule change is to eliminate the excessive and unnecessary conservatism of certain Appendix K requirements (Q2). ,

17.3 The rule should not explicitly prohibit any increase in power until all severe accident issues and unresolved safety issues are resolved. A utility applying for a plant uprating should be required to satisfy regulatory requirements including severe accident and unresolved safety issues impacted by the application. Explicit prohibiting a power increase may prevent an increase even if health and safety of the public is not affected (Q3),

17.4 The technical basis of the proposed rule should be reviewed by an independent group as provided for by the NRC proposed rule I process. The ACRS is an independent group and should review the proposed rule change (Q4).

17.5 The comenter is currently performing a BE calculation of Point Beach and looks forward to taking advantage of the significant investment in analysis methods already developed. Please do not delay in approving the proposed rule.

3 Page 25 ENCLOSURE D 1

:7 . ararwr.e u .. ; 2; :g~ r.:. . ,; . . . n. - -

.. ., m .

17.6 The commenter is pleased that the proposed rule does not prescribe the uncertainty techniques. A realistic technique must be allowed to change as the state-of-the-art changes. Ensure that the regulatory guide is not interpreted as a requirement.

17.7 The proposed procedure for reporting code errors and changes appears to be reasonable. The commenter recognizes that the proposal is a relaxation of current practice and may eliminate costs associated with reporting insignificant changes.

l 1

1 t

I

{

l 4 page 26 ENCLOSURE D

18. G.C. Sorensen, Manager, Regulatory Programs, Washington Public Power Supply System, P.O. Box 968, 3000 George Washington Way, Richland, Washington 99352.

COMMENTS:

18.1 The commenter believes that the uso of more realistic or "Best Estimate" methods should be encouraged whenever research results in new information which replaces previous uncertainties, thus reducing the need to maintain unnecessary conservative requirements.

18.2 Since changing to the new methods is not necessary for safety reasons, it could be difficult for licensees to justify the not inconsiderable cost involved to their respective State regulatory bodies (Q1).

18.3 No change is being props .9 to the Peak Clad Temperature limit of 0

2200 F. Thus the "explicit degree of conservatism" referred to does exist (Q2).

18.4 Power increases of 5-10% will have an insignificant effect on risk, since the incremental increase in fission product

' inventories available for release during a worst case DBA would be no more than 5-10%. The 2200 F limit remains, thus there will be a small decrease in margin between the temperature at which significant fuel damage is predicted to begin and the calculated PCT. The increase in risk due to this small decrease in margin is not significant, especially when compared to the margin remaining (Q3).

18.5 The costs incurred by an independent peer review would likely not be justified by the results. The scientific basis for the rule change is well understood, and has been subjected to a very robust review process which extended over a period of many years (Q4).

Page 27 ENCLOSURE D

. . . . . .. . c,w2#.w .o... ..:c....;..y. . +. . _ o<x .. . .

19. S.R. Zimerman, Manager, Nuclear Licensing Section, Carolina Power & Light Co., 411 Fayatteville St., P.O. Box 1551, Raleigh, NC 27602.

C0tHENTS:

19.1 The revision makes appropriate use of the decade of research on ECCS performance and will provide for improvements to plant safety through better understanding of plant behavior.

19.2 The conservatisms of Appendix K as demonstrated by previous research will ensure that a plant's margin of safety is no less than that provided by the proposed revision. For these reasons, '

the requirement to implement the proposed rule without the option at plants not restricted by LOCA limits would be a substantial and unnecessary burden (Q1).

19.3 Adequate conservatism will be ensured by the continued use of the 10CFR50.46(b) limits and the requirement to account for uncertainties in the evaluation model. Additional conservatisms would limit the value of the proposed revision (02).

39.4 The changes to power level or power distribution are not expected to significantly impact severe accident scenarios or unresolved ,

safety issues (Q3).

19.5 The proposed revision is the product of 13 years of research, review, and improvement in the practical understanding of ECCS performance. We believe that it.is highly unlikely that additional review would discover issues which have not already  ;

been addressed. There is, however, reason to believe that additional delays in the issuance of a final rule would result (Q4).

t Page 28 ERCLOSURE D

.. ,: ., , . ~ . . . .. - : 3MaM.p, .~,: .- tu .:- .. . s ,

- . ..  :. . .. u. . .

. 19.6 The comenter encourages the Commission to implement the proposea rule revision in a timely manner without significant revision.

The revision will prove of benefit to utilities without a reduction in the health and safety of the public.

4 4

i- .

i i

j k

i 5

t Page 29 ENCLOSURE D

.o!

20. J.D. Sieber Vice President, Nuclear, Duquesne Light, P.O. Box 4 f Shippingport, Pa 15077-0004, i CotHENTS:

20.1 The more accurate calculations will be useful for our operator I training, accident management, and simulator response.

20.2 The current limits should be retained. These current limits preclude core temperatures which could result in significant core damage, and as such, provide additional conservatism for the design basis accident as well as economic protection for the '

reactor. The limits also act to protect the health and safety of the general public (Q2). ,

20.3 The reporting procedure for code errors or changes and reanalysis requirements appear to be reasonable and reflects a balanced consideration of the financial burden to the utility and plant safety. The proposed 50"F value, while larger than the current 20'F value is more consistent with recognition that a large break l LOCA is a very improbable event and that additional margin above 2200 F does exist.

20.4 The uncertainty analysis, which will be required for the proposed ,

rule change, should not be prescriptive but should remain flexible. The evaluation of the uncertainties at a 95%

probability level is adequate for the large break LOCA and is consistent with good engineering practice.

20.5 The current Appendix K method has been shown to be conservative, ,

and the requirement of reanalysis using the methods of the  ;

proposed rule change would place an unfair financial burden on f utilities which already have acceptable operating and ECC margins ,

to the licensing limits and would not result in any improvements in safety (Q1).

I Page 30 ENCLOSURE O ,

l

00 o.

20.6 The amount of conservatism that is applied and retained should be i determined through the uncertainty analysis using the proposed  !

mode) (Q2).

i 20.7 The 5 to 10 percent power uprating is well within the additional margin identified in the source term studies such that there is l no calculated increase in risk to the pu'olic (Q3).

20.8 It is unrealistic to believe that all unresolved safety issues will be resolved quickly. A requirement of having to resolve all unresolved safety issues before a plant could uprate would unfairly penalize those utilities and their customers who desire to uprate for economic reasons. However, those unresolved safety issues which are directly related to core power should be j reviewed to ensure that there is no additional adverse risk to the public or utility due to the proposed plant uprating. The l proposed rule should not prevent justified uprating until all  !

unresolved' safety issues are completely resolved. The merits of  ;

the change should be judged by the research results (Q3).

20.9 A peer review, if properly performed, can be useful. However, we believe that this is best accomplished by a public review and comment procedure together with review of the comments and resolution by the ACRS. The ACRS by law, has the proper advisory role, historical perspective and technical basis. While the ApS could perhaps address more philosophical questions as to the impact of the rule change or, society; this organization would have to rely on experts from the national laboratories which have been involved in the nuclear program. We believe that this is the proper role for the ACRS and another independent review is not warranted (Q4).

l i

, page 31 ENCLOSURE O

O*

21. W.L. Stewart, Vice President, Nuclear Affairs, Virginia Electric and Power Co., Richmond, Va. 23261,.

. C0t91ENTS:

21.1 The new rule should specify cladding as "zirconium base alloy" to permit fuel vendors to develop advanced cladding materials. The fuel vendor would then be required to demonstrate that the acceptance criteria for Zircaloy are conservative for the zirconium base alloy.

21.2 Proposed section 50.4S(s)(3)(1) should state "...cumulation of changes and errors such that the not sum of the respective temperature changes is greater than 50'F." For example, a case in which two separate errors caused changes in peak clad temperature of +35 F and -20 F, respectively, would be reported in the annual report, since the net change is only 15 F. As wricten, the proposed rule would classify this case as a significant change, requiring a report to be submitted within 30 days.

21.3 The requirement for annual reporting of non significant changes and errors) explicitly include a threshold temperature value above which this part of the rule is applical,le. Use of a threshold value would preclude reporting of trivial changes which clearly do not impact safety.

21.4 Many nuclear units have existing analyses which do not impose operational restrictions, and therefore may prefer to have the option of continued use of Appendix K for future reanalyses. If Appendix K requirements are to be phased out, consideration should be given to the impact this may have on existing or future small break LOCA analyses. Since small break LOCA analysis results usually do not impose restrictions on plant operations, retention of existing analysis methods, versus imposing the alternative analysis methods in the proposed rule, would be preferable for the analysis of these events (Q1).

Page 32 ENCLOSURE D

s. t 21.5 Any degree of conservatism required to meet the acceptance criteria should be specified in the regulatory guids rather than in the rule. This will allow some amount of discretion on the

• part of the NRC and licensees in evaluating the high level of ,

probability. Such flexibility is preferable when using best ,

estimate calculational techniques since multiple approaches i incorporating differing o grees of evaluation model conservatism could provide comparable levels of probability (Q2). ,

21.6 The proposed rule need not prohibit approval of such applications until severe accident issues and unresolved safety issues are {

resolved. Including this restriction in the proposed rule provides no additional regulatory velen, since existing' regulations require the NRC to determine whether any such issues l must be resolved prior to approval of a specific license application (Q3),

f 21.7 inis effort involved obtaining new data, developing improved ,

analysis models, and revising analysis methodologies over a 13 year period. The technical basis for this proposed rule should i not be reviewed by any independent group, primarily because such i a review at this late stage would add no new understanding on ,

which to base the rule. ,

i I

l i

i i

page 33 ENCL 05URE D

..)

22. David W. Cockfield, Vice President, Nuclear, Portland General Electric Co.,121 S.W. Salmon St. , Portland Oregon 97204.

COMMENTS:

22.1 PGE sup,norts the praposed rule change. The more accurate calculations will also be useful for our operating training, accident management, and simulator ,panse.

22.2 PGE agrees that the current limits should be retained. These current limits preclude core temperatures which could result in significant core damage, and as such, provide additional conservatis:a for the design basis accident as well as econc.nic protection for the reactor (Q2).

22.3 The reporting procedure for code errors or changes and reanalysis requirements using a 500 F limit between small and larger changes

~

is reasonable and doe = reflect a consideration of the financial burden to the utility that small changes in the evaluation model Can Cause.

22.4 The uncertainty analysis, which will be required fcr the proposed rule change, should not be prescriptive but should remain flexible. The evaluation of the uncertainties at a 95%

probability level is adequate for the large break LOCA and is consistent with good engineering practice.

22.5 The current Appendix K method has been shown to be conservative, and the requirement of reanalysis using the methods of the pro;.ned rule change would place an unfair financial burden on utilities which already have acceptable operating and ECC margins to the licensing limits and would not result in any improvements in safety (Q1).

22.6 The a.nount of conservatism that is applied and retained should be determined through tne uncertainty analysis using the proposed model (Q2).

Page 34 ,

ENCLOSURE O

22.7 The 5 to 10 percent power uprating is well within the additional margin identified in the source te m studies such that there is no calculated increase in risk to the public (Q3).

22.8 PGE believes that a proposal for a rule change should not be tied to the resolution of all outstanding safety issues. The basis for the rule change is the significant amount of research and developme it which has been performed on ECCS performance and merits that the change should be judged by the research results (Q3).

22.9 PGE recommends that the ACRS and their consultants be the peer review body for the proposed rule change.

t 1

i i

1 l

Page 35 ENCLOSURE D

1 7-

- . o

23. J.R. Thorpe, Director, Licensing & Regulatory Affairs, GPU Nulcear Corp., 100 Interpace Parkway, Parsippany, NJ 07054-1149.

C0iHENTS:

23.1 This proposed rule should be adopted without delay.

23.2 The non-routine reporting criteria should be included in 10CFR50.72 and 50.73 and not in Appendix K.

23.3 All ECCS evaluation models (EM) which were reviewed and cpproved by the NRC since the issuance of the final acceptance criteria in 1974, and including those found to be inconsistent with SECY-83472 by the NRC, should be considered acceptable for satisfying the requirements of the new paragraph 50.46(a).

Although calculational results using previously approved EM may

' provide very conservative and inappropriate results for operation and engineering purposes, there should not be new requirements to recalculate in compliance with this rule change. The recalculation should be left to the individual owners of the facilities (Q1).

23.4 Power level increase, if found to be justifiable, should not be l

prohibited. This 5-10% power increase is not expected to cause

! difficulty in meeting the existing environmental limits (Q3).

l 23.5 Through GPU's experience with GE's SAFER-CRECOOL, no additional review by an independent group such as the APS is necessary. The proposed alternative method was reviewed by several different l

I engineering groups within GE and knowledgeable and experienced

~

( utility engineers from sponsoring utilities before topical reports were submitted to the NRC. Consultants were brought in by the NRC staff for numerous occasions during the review period.

The ACRS staff also reviewed the material. Therefore, additional

( independent group review is not necessary.

Page 36 ENCLOSURE D

• • l l

i

24. W.G. Counsil, TV Electric, 400 North Olive St., LB 81, Dallas, Texas 75201.

CONNENTS:

24.1 The commenter encor ages the Commission to implement the proposed revisions without alteration.

24.2 Because the current Appendix K conservatisms ensure margins equal to or greater than those provided by the proposed rule, no safety

~

issue would be created if those plants are allowed to employ the current methodology, thereby avoiding substantial and unwarranted financial burdens involved with reanalysis (Q1).

24.3 Sufficient conservatism is assured under the proposed rule by (1) ,

continued use of conservative 10CFR50.46 limits, and consideration of uncertainties in the PCT calculated by realistic models. Any additional degree of conservatism would be an arbitrary and. unwarranted burden which we41 minimize the incentives for a utility to use the proposed rule (Q2).

24.4 The proposed rule continues to assure adequate conservatism in ECCS models. An arbitrary prohibition of power increase based on yet unquantified and, perhaps unquantifiable impacts, if any, from severe accident and unresolved safety issues would not be based on a cognizable safety concern. This results in economic burdens without a demonstrated and commensurate increase in safety (Q3).

24.5 These results have been reviewed by numerous groups, including government agencies, the ACRS, industry organizations, and independent professional societies. Additional review is unlikely to result in any new developments. Additional reviews could, however, result in further delay in issuing the already long overdue proposed amendments.

Page 37 ENCLOSURE D

. . . , ... . . . ~ .. .. . - _ . . _ ~ . . . - _ _

7.....,.n.-...~. . . . . . .

1 l

l l

25. J.A. Tiernan, Vice President, Nuclear Energy, Baltimore Gas and  ;

Electric, Charles Center, P.O. Box 1475, Baltimore, MD. 21203.  !

COMMENTS:

25.1 We continue to support the Commission's efforts to_ improve existing regulations through use of improved analysis methodology.

and recent technical developments. The proposed rule is an acceptable approach to ensure the safety of the public without unnecessarily restricting applicants and licensees.

25.2 The proposed changes to the reporting and reanalysis requirements are realistic and require immediate reporting only if the change or error is significant. However, the proposed rule does provide for a new annual reporting requirement for all changes or errors that affect peak fuel cladding temperature no matter how minor.

The licensee is required by 10CFR50.59 and 50.71 to annually submit summaries of completed 10CFR50.59 safety evaluations.

Therefore, since most changes or errors are already reported to the staff under existing requirements, the need for a new annual reporting requirement for changes and errors which are not significant is not necessary and this requirement should be deleted from the proposed rule.

25.3 Those utilities which are not LOCA limited 'd be allowed to continue to use their present ECCS evaluat. .. inethodology and thus be provided the option of expending the resources to update to the proposed methodology. The Regulatory and Backfit Analysis did not assess a mandatory requirement to shift to the proposed methodology and any mandatory requirements should be assessed in terms of the value (e.g., public benefits such as safety) and impact (e.g., consequences and costs)(Q1).

Page 38 ENCLOSURE D

g.

. 1,-

25.4 The proposed amendment would allow the use of more realistic calculations of ECCS performance along with reasonable estimates of uncertainty. -This, combined with the fact that 10CFR50.46 limits of 2200'F and'17% cladding oxidation are believed to already be appropriate and conservative limits below which substantial core damage would occur, make this application of additional conservatism unnecessary (Q2). .

25.5 Current regulations adequately account for any changes in the licensed power. A number of plants have uprated power level and addressed all regulations including increased radioactive source terms. Since the precedent of increasing power has been repeatedly established, prohibiting such an action based on a realistic LOCA analysis is unwarranted (Q3).

25.6 The technical basis for the proposed rule change is supported by over 13 years of LOCA/ECCS tests and experiments, and computer code development research in laboratories around the world.

NUREG-1230 presents a detailed and comprehensive discussion of the increase in knowledge acquired by extensive research. Every aspect of the research has already been independently reviewed and the international nuciear technical community has reached a ,

consensus that sufficient information now exists to justify more realistic LOCA calculations.

i

~

l l

1 Page 39 ENCLOSURE D

u ... . n . w. .- . . .- . ,. _ . ~~ . - . - . .

26. J.C. Brons, Executive Vice President, Nuclear Generation, New York Power Authority, 123 Main St., White Plains, NY 10601.

COWiENTS:

26.1 The commenter strongly support the promulgation of these proposed amendments. NYPA's Fitzpatrick facility has already benefited from the implementation of best-estimate ECCS calculations.

Further improvements at Fitzpatrick and equivalent improvements at Indian point are expected if these amendments are adopted.

26.2 Proposed section I.C.5.c of Appendix X should be clarified.

Suggested language could be, "For this purpose, a significant reduction in the overall conservatism in the evaluation model is defined as a reduction in the calculated peak fuel cladding temperature of at least 50 F from that which would have been calculated on (effective date of rule) due either to individual changes or error corrections or the net effect of an accumulation of changes or error corrections."

26.3 Utilities should not be required to change analysis methodology unless it is economically favorable to do so (Q1).

26.4 The purpose of performing best estimate calculations, taking into account uncertainties, is to avoid the application of an arbitrary degree of conservatism. There is no reason that an f explicit degree of conservatism should be applied to the models

! (Q2).

I l 26.5 The NRC will require an evaluation.to demonstrate that a plant will continue. to operate in accordance with all applicable regulations prior to an increase in licensed power. There is no l reason the a prohibition of such power increase should be f

3 icluded is this regulation. If the rules are currently sufficicat to protect the health and safety of the public, they will continue to be sufficient (Q3).

Page 40 ENCLOSURE D l

t

26.6 Power gained from an increase in licensed power may displace power which would otherw;ae be generated using a new power plant .

using nuclear fission as a heat source, so that the overall available source term would be the same, although the public would suffer from inefficient use of capital to build a new plant (Q3).

26.7 This rule has been prepared by the NRC staff based on research which clearly demonstrates that current evaluation models are extremely conservative. An opportunity for industry and public comment has been provided by this rulemaking proceeding (Q4).

Page 41 ENCLOSURE O

i

27. L.B. Long, General Manager, Nuclear Safety and Fuel, Southern Company Services, Inc., p.0. Box 2625, Birmingham, Alabama j 35202.

ColHENTS:

27.1 The commenter supports the proposed rule change which ... will also be useful for operator training, accident management, and simulator response.

27.2 The current peak cladding tamperature limit of 2200 F (etc.) ...

should be retained. These current limits ... provide additional conservatism for the design basis accident as well as economic 7

protection for the reactor.

27.3 The revised reporting procedure ... is reasonable and does reflect consideration of the financial burden to the utility that small changes in the evaluation model can cause.

27.4 The uncertainty analysis ... should not be prescriptive, but should remain flexible. ... a 95% probability level is adequate for the large break LOCA and is consistent with good engineering practice.

27.5 The current Appenoix K method has been shown to be conservative, and the requirement of reanalysis using the methods of the proposed rule would place an unfair financial burden on those utilities which already have acceptable operating and ECC margins

... and would not result in any improvement in safety (Q1).

27.6 The amount of conservatism that is applied and retained should be determined through the uncertainty analysis using the proposed model (Q2),

page 42 ENCLOSURE O

27.7 The research on. fission product source term indicates that significantly lower amounts of fission products would be released than currently used. The 5 to 10 percent power uprating mentioned in the proposed rule change is well within the additional margin identified in the source term studies such that there is no calculated increase in risk to the public (Q3).

27.8 The basis for tne rule change is the significant amoun,t of research ... and the merit's of the change should be, judged by the research results (Q3).

27.9 The ACRS and their consultants should be the peer review body for the proposed rule change (Q4). -

i l

Page 43 ENCLOSURE O I

u

~

28. M.0. Medford, Manager of Nuclear Engineering and Licensing, Southern California Edison Co., P.O. Box 800, 2244 Walnut Grove Ave., Rosemead, California 91770.

COMMENTS:

28.1 The proposed r.ule recognizes the degree of conservatism in the present requirements ... (and) the knowledge acquired over the past 13 years ....

28.2 Specific potential advantages ... are an increase in load maneuverability, less restrictive fuel assembly power peaking limits and greater margin when performing ECCS equipment surveillances ... reduced neutron fluence at the vessel wall and improved fuel economy.

28.3 Utiliting this more realistic method ... is a major undertaking

.... However, based on the potential for these advantages, the Edison Company strongly supports the proposed rule.

28.4 The results of a cost benefit analysis may or may not favor a change in policy in the distant future (Q1).

4 28.5 The explicit degree of conservatism is adequately addressed in the proposed rule and environmental impact analysis (Q2).

28.6 Increases in approved power levels are adequately addressed in the proposed rule and environmental impact analysis (Q3).

28.7 The proposed rule is currently receiving adequate review by all interested parties (Q4).

page 44 ENCLOSURE O

29. L.O. Butterfield, Nuclear Licensing Manager, Commonwealth Edison, P.O. Box 767, One First National Plaza, Chicago, Illinois 60690.

COMMENTS:

29,1 The proposed rule would remove needless conservatisms from the regulations while not requiring reanalysis where the analysis supporting the current licensing basis is adequate.

29.2 Proposed section 50.46(a)(2) ... appears to eliminate the need for the Director, NRR to issue an Order to impose restrictions on reactor operation. Thus, this provision expands the Director's discretion by allowing him to impose immediate effective restrictions on reactor operation without a finding that such innediately effective action is required to protect the public health, safety, or interest.

page 15 ENCLOSURE O

30. Susan L. Hiatt, Ohio Citizens for Responsible Energy, Inc., 8275 Munson Rd., Mentor, Oh 44060.

C0ft1ENTS:

30.1 OCRE believes the proposed revision to the ECCS rule is unwarranted, primarily because it is based on illegal consideration of cost saving to licensees (several court cases and statements by the Director of NRR are cited) and has questionable safety benefits.

30.2 It is complained that the present Appendix K medels are unduly conservative. What's wrong with being conservative? When the public health and safety are at stake, it is certainly better to err on the side of caution.

30.3 The commenter quotes from the 1975 GE Nuclear Reactor Study (Reed Report) and highlights statements concerning lack of understanding of phenomena and small safety margins. She also notes that the rule would decrease margins rather than increase them as recommended by GE.

30.4 The fission product inventories will likely increase .... This is an increase in risk, which MAY be offset by safety benefits for SOME plants. As this is plant-specific, generic rulemaking does not seem to be the appropriate avenue. It would make more sense to include taking credit for ECCS margins to gain risk reduction from PTS and diesel failures as an option in the IPE to be conducted as part of the Severe Accident Policy Statement implementation.

30.5 ECC bypass data for a full-scale PWR vessel will be made available from forthcoming experiments in the 20/30 program. It thus does not appear that sufficient experimental basis exists for rule revision at this. time.

Page 46 ENCLOSURE D

-30.6 The technical basis for the rule revision shouid be reviewed by an independent group such as the APS. Increases in power limits should'not be permitted until all severe accident issues and unresolved safety issues are resolved.

30.7 OCRE may submit further comments after reviewing NUREG-1230.

page 47 ENCLOSURE D

- -- - - - , - - - - - - . _ , - . - . . . - ~ . - - . - - . . . . . , _ - - - _ - . - -

.4 ~

...h.?iilch6U:.cx-: :-Oh a .. %',. .L . 2.c:NEd .'.. a.

^

. . u. . . . l. . '. L .cQv: ^ ~

e 0'

31. Susan L. Hiatt, Ohio Citizens for Responsible Energy, 8275 Munson Road, Mentor, OH 44060 .

COMENTS:

31.1 The primary motive for the ECCS rule change is cost savings to licensees and this is illegal.

b Page 48 ENCLOSURE D

. ..- _ - - _ . . . . . . _ _ _ . _ - - _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ . _ . _ , _ . _ _ _ _ _ . _ _ _ . . . _ - . . _ _ _ _ _ _ _ _ _ _ . _ , _ . _ . . - _ _ _ _ . - . _ , ~.

s, c. m.we :v.. .w

. . ., . ,u...,.. .,; . . . . . , . -.. .-. ~.. /

.. a ,,

+

32. (Redesignated as Comenter la) Dan L. Johnson,1535 Meade Ave.,

San Diego, Ca., 92116, an employee of Southern California Edison Company comenting on his own behalf.

C0t91ENTS:

Commenter provid'ed no new coments but rather supplied further background documentation supporting his original coment letter.

Page 49 ENCLOSURE D  !

1 1

Y- Ou L.,'

d....: . . A.  :,. Nr.: R.4.h z us :su . , '

. ." .. .3 ..

1:

L

33. R. Artigas, Manager of Licensing and Consulting Services, General Electric Company, 175 Curtner Ave, San Jose, California.

C0l91ENTS:

33.1 General Electric Company agrees that the proposed amendment is

~

needed to take advantage of research performed since the current rule was written.

33.2 The question of permitting conservative ECCS evaluation methods of Appendix K indefinitely is premature. It should be considered when significant experience has been gained with implementation of the new features of the proposed rule.

33.3 T,he research cited in the Federal Register Notice (Vol. 52, No. 41 Pg. 6337) demonstrates that significant fuel damage will not occur below 26000 F This supports an existing 400 F conservatism 0

relative to the 2200 F LOCA limit.

Additional conservatism is introduced when the uncertainty is accounted for such that there is a high level of probability (as defined in SECY 83-472) that the LOCA limit is not exceeded.

Further conservatism is not justified.

33.4 Increases in approved power level have been justified under the existing rule and with existing ECCS models. While tne propos'ed rule change may facilitate such increases in the future it does not, by itself, allow high power operation. Such operation must be justified by consideration of many subjects beyond ECCS performance. Furthermore, the small increase in fission product inventory resulting from a 5-10% power increase will have a negligible impact on the absolute magnitude of overall risk.

Severe accident issues should be addressed in their own forum and not be used to confound other activities which could offer the opportunity for improved plant performance and reduced power generation costs. Because of this, the suggested prohibition in power level increase is totally inappropriate.

Page 50 ENCLOSURE D

33.5 The technical basis for the proposed rule has been scrutinized for years by NRC, ACRS, National Laboratories, Vendors, et.al..

Further review by yet another party will serve no useful purpose.

33.6 GE has obtained approval of an improved LOCA model and application methodology for the BWR which fully implements the requirements established in SECY 83-472. This fact is acknowledged in the backfit analysis p.ovided in the Federal Register notice (Vol . 52, NO.41 Pg. 6337). It was clearly our intention and the intention of the NRC reviewers, that the approved model and methodology not be impacted by the proposed new LOCA Rule. In order to avoid any unforeseen complications, clarifying language should be added to the amendment of 10CFR50.46 which indicates that prior approved ECCS evaluation models which are consistent with SECY 83-472 are acceptable for satisfying the requirements of paragraph 50.46(a).

Page 51 ENCLOSURE D

, - . - . - - - - , - - , - - . , - - - - , - - - - c-,

9

• 6 ENCLOSURE E 1

s I

l l

l l

l

y _

=

-e,,

.-. [7590-01]'

NUCLEAR REGULATORY C0ffilSSION 10 CFR Part 50 Emergency Core Cooling Systems; Revisions to Acceptance Criteria AGENCY: Nuclear Regulatory Commission.

ACTION: Final rule.

SuffiARY: The Nuclear Regulatory Commission (NRC) is amending its regulations to allow the use of alternative methods to demonstrate that the emergency core cooling. system (ECCS) would protect the nuclear reactor core during a postulated design basis loss-of-coolant accident (LOCA). The Commission is taking this action because research, performed since the current rule was written, has shown that calculations performed using current methods and in accordance with the current requirements result in estimates of cooling system performance that are significantly more conservative than estimates based on the improved knowledge gained from this research. While the

. existing methods are conservative, they do not result in accurate calculation of what would actually occur in a nuclear power plant during a LOCA and may result in less than optimal ECCS design and operating procedures. In addition, the operation of some nuclear reactors is being unnecessarily restricted by the rule, resulting in increased costs of electricity generation. This rule, whila continuing to allow the use of current methods and requirements, also allows the use of more recent information and knowledge to demonstrate that the ECCS would protect the reactor during a LOCA. This amendment, which applies to all applicants for and holders of construction permits or operating licenses for light water reactors, also relaxes requirements for certain reporting and reanalyses which do not contribute to safety.

Page 1 ENCLOSURE E

) ..

EFFECTIVE DATE: (

FOR FURTHER INFORMATION CONTACT:

L. M. Shotkin, Office of Nuclear Regulatory Research, U.S. Nuclear Regulatory Commission, Washington, DC 20555, telephone (301) 492-3530.

SUPPLEMENTARY INFORMATION:

BACKGROUND On March 3,1987, the Nuclear Regulatory Commission published in the Federal Register proposed amendments (52 FR 6334) to 10 CFR Part 50 and Appendix K. These proposed amendments were motivated by the fact that since the promulgation of Section 50.46 of 10 CFR Part 50, "Acceptance Criteria for Emergency Core Cooling Systems (ECCS) in Light Water Power Reactors," and che acceptable and required features and models specified in Appendix K to 10 CFR Part 50, considerable research has been performed that has greatly increased the understanding of ECCS performance during a LOCA. It is now confirmed that the methods specified in Appendix K, combined with other analysis methods currently in use, are highly conservative and that the actual cladding temperatures which would occur during a LOCA would be much

' lower than those calculated using Appendix K methods. In soliciting

• the public's comments on the proposed rule, the NRC specifically requested its views on questions posed by Commissioner Asselstine and the Advisory Committee on Reactor Safeguards (ACRS). The ACRS requested that the Commission solicit the public's comments on whetner the existing rule should be "grandfathered" indefinitely. That is:

1. Should the conservative ECCS evaluation method of Appendix K be permitted indefinitely or should this aspect of ths ECCS rule be phased out after some period of time?

Commissioner Asselstine requested the public's comments on the following:

Page 2 ENCLOSURE E i

2. Stould this rule change include an explicit degree of conservatism that must be applied to the evaluation models?

3. This rule change would allow a 5 to 10 percent increase in the fission product inventory that could be released from

any core meltdown scenario. Should this rule change explicitly prohibit any increase in approved power levels until all severe accident issues and unresolved safety issues are resolved?

'4. Should the technical basis for this proposed rule change be reviewed by an independent group such as the American Physical Society?

SUMMARY

OF PUBLIC COMMENTS The comment period for the proposed rule revision and the draft regulatory guide (52' FR 11385) expired on July 1,1987. Twenty-seven letters addressing the proposed rule were received by the expiration l 9

date, as well as nine responses to the request for comments on questions in the regulatory guide. A number of late comments were also received. These were also considered to the extent that new and substantial comments were provided.

l The public comment on the proposed rule revisions have been divided into thirteen categories and are summarized in the following i paragraphs. Catagories one through four represent the responses to ,

the specific questions posed by the ACRS and Commissioner Asselstine.

In general, consideration of the public comments resulted in no substantive revision to the proposed rule.

1. Grandfathering of Conservative ECCS Methods of Appendix K ,

, (Question 1).

1 e Twenty-one of the commenters specifically addressed the ARCS question concerning the grandfathering of the current Appendix X Page 3 ENCLOSURE E

approach. Seventeen of these commenters recommended indenfinite c .

grandfathering of the existing Appendix K evaluation models.

Most cited the known conservatism as the basis of their recommendation. In addition, several commenters stated that in light of the known conservatisms not allowing continued use of existing Appendix K evaluation models would be unfairly burdensome to licensees who determine that they would not derive an economic benefit by performing realistic analysis of ECCS performance. The position of an additional commenter is unclear concerning grandfathering. The remaining commenter was not opposed to grandfathering but thought the question is premature.

This commenter believes that indefinite use of existing ECCS evaluation methods should be considered when significant experience has been gained with the implementation of the new features of the rule but makes no recommendation as to what policy the Commission should pursue in the meantime.

The Commission agrees with the majority of the commenters that existing Appendix K evaluation models should be permitted indefinitely. The Commission also believes that the decision to permit continued use of such models can and should be made at this time because it believes that both methods provide adequate protection of the public health and safety. As described in the regulatory analysis, the probability of a large break is so low, that the choice of best estimate versus Appendix K has little effect on public risk. The TMI action plan calls for industry to improve their small break LOCA evaluation models to be more realistic when evaluating the more probable small break accident scenario. This has been done within the context of f 50.46 and Appendix K compliance and was entirely appropriate since small breaks are not limiting in design basis performance and a better understanding of small break behavior is a desirable safety goal~

from a risk perspective. Therefore, the grandfathering provision has been retained in the final rule.

Page 4 ENCLOSURE E

• • Specification of Explicit Dearee of Conservatism (Question 2).

2.

The majority of the responses to this question indicated that the proposed rule already contains conservatism in the required uncertainty evaluation.

The use of additional conservatism would be inconsistent with the objective of the rule which is to provide a realistic evaluation of plant response during a LOCA. The NRC has not included an additional explicit degree of conservatism in this rule.

3. Resolution of all Safety Issues Prior to Allowinc Power Level increases (Question 3). Some commenters pointed out that fission product inventory is not a direct function of total power, but rather it is the rate of fission product formation that is a direct function of power. Fission product inventory available for release during a core meltdown would be a function of burnup, not total power.

Actually, the inventory of fission products is a complex function of both time and power and not as simple as described by the commenters. Short lived isotopes, such as xenon and iodine, quickly reach an equilibrium inventory and total steady state inventory of these fission products is a direct function of power. Inventories of long-lived isotopes, such as strontium and cesium, are functions of total fuel burnup, as described by the commenters. Intermediate-lived isotopic inventories are complex functions of time, power, and integrated power. In an independent study, documented in chapter XII of NUREG 1230, the staff determined that the change in risk due to a 5% power increase is negligible. The arguments above do not alter the Commission's position that the increase in fission products available for release during a core meltdown caused by a 5% power increase is negligible compared to the uncertainty in fission product release. The Commission has decided not to delay the Page 5 ENCLOSURE E

proposed rule revision pending resolution of all unresolved safety issues or severe ar.cident issues and therefore will proceed with this final rulemaking, as planned.

4. Independent review of Technical Basis (Question 4). Several commenters indicated that the technical basis for the proposed rule has had adequate review as the research was being performed.

A number of commenters stated that it was the role of the ACRS to perform any review of the proposed rule revision because it is uniquely qualified due to its familiarity with the research.

The Commission agrees that the technical basis has had acequate review, except for the uncertainty methodology which is new and untried except for the General Electric Company's use of an uncertainty evaluation of their SAFER code. As a proof of principle and demonstration of feasibility, the ACRS and a second independent peer group has reviewed the uncertainty methodology developed by the NRC for use in quantifying the uncertainty of NRC developed thermal hydraulic transient codes. Both the ACRS and the peer group made generally favorable comments concerning' the methodology; however, both groups recognized that a complete demonstration (i.e., application to small break LOCA and the reflood portion of large break LOCA) has not yet been accomplished and certain reviewers questioned whether such a demonstration could be performed successfully. The only objectives of the NRC methodology demonstration are to demonstrate feasibility, to develop an audit tool, and to prcvide the necessary experience to audit licensee submittals. The staff does not believe that NRC demonstration of the methodology is a prerequisite to this rulemaking. Licensees wishing to adopt the best-estimate approach permitted as a result of this rule are neither required to use this methodology nor to model their own methodologies on it. This methodology will play an important part in the best-estimate modal review process. The NRC has determined through twenty years of experience that independent analysis with indepundent methodologies is the most effective way Page 6 ENCLOSURE E

7

• to intelligently review new vendor or licensee methodologies. It is therefore appropriate that this new methodology be subjected to stringent technical scrutiny, as directed by the Commission.

The NRC staff is committed to completing this demonstration by the time that it will be needed to review licensee submittals and is confident that such a demonstration will be successful. Based on the paucity of negative response concerning the technical basis for the proposed rule revision and generally favorable review of the NRC uncertainty methodology, the Commission plans no further review of the technical basis.

5. General Comments on Proposed Rule. Twenty-one commenters made comments of this nature. The majority of the comments came from the nuclear industry of which 19 expressed support of the proposed rule. The industry also strongly supports the snecific ECCS rule approach proposed by the NRC. One commenter neither supported nor opposed the proposed approach. One negative comment was received from an anonymous individual within the nuclear industry who implied, without specifics, that the ECCS rule is not sound and that public comment is not a fai.' hearing because expert insiders would be afraid to comment.

Based on the absence of any supporting justification for the negative response and the unprecedented amount of research supporting the rule revision, the NRC does not consider this comment to be valid and has proceeded with this rulemaking with no major revisions.

One commenter suggested that fuel reload suppliers should not be required to complete full LOCA/ECCS analyses because the hydraulics are not changed by a fuel change.

Although this point is valid, the Commission believes that it is an unworkable situation to allow fuel suppliers to make use of previous analyses performed by others. It is believed that serious questions of accountability would arise in cases where Page 7 ENCLOSURE E

--e- .- w -, e r - . -.. ----,y -

r- . , . - , - - = .---.-+- m-- ..---m.--,- ,, ,,, -

c - -------,- - , - - --

errors are discovered in evaluation models, requests are made to revise plant technical specifications, or some other questions regarding the analyses are raised. The NRC believes that shared responsibility for evaluation models would not be in the best interest of the public health and safety and therefore has not implemented the suggestion of this comenter.

The NRC received two requests for an extension of the comment period to allow time for review of NUREG-1230, which describes the research supporting the proposed rule revision.

The NRC believes the coment period was sufficient since most of the research is not new and has been extensively reviewed in the past. Both commenters were contacted and told that coments received after the coment period would be considered if time permitted. Coments from both parties were received late and were indeed considered by the NRC.

6. Reporting Requirenents. Some comenters viewed the proposed reporting procedures as new requirements needing consideration in the backfit analysis while others stated that they are a major relaxation and clarification of existing reporting requirements.

The NRC position is that the reporting requirements are new in the sense that they will now appear in the Code of Federal Regulations and therefore have been considered in the backfit analysis. In practice, these reporting requirements are indeed a clarification and relaxation over the current interpretation of the existing requirements and therefore the net effect of these requirements will be to reduce the frequency for reporting and reanalysis.

A number of comenters requested that only significent errors or changes in the non-conservative direction or only thoss that result i.1 exceeding the 2200 F limit be required to be reported.

In addition, a number of comenters suggested that the NRC require only annual reporting of significant errors or changes.

Page 8 ENCLOSURE E

The NRC considers a major error or change in any direction a cause for concern because it raises potential questions about the adequacy of the evaluation model as a whole. Therefore, the NRC requires the reporting of significant errors or changes, in either direction, on a timely basis so that the Commission may make a determination of the safety significance. Thus, the final rule contains no change in this recuirement.

One commenter recommended that the word "immediate" be deleted from the requirement to propose steps to be taken to demonstrate compliance in the event that the criteria in 9 50.46(b) are exceeded.

The Commission considers this a very serious condition in which the plant is not in compliance with the regulations and may be operating in an unsafe manner. The word "immediate" reflects this seriousness and is further defined by reference in other sections of Part 50.

Several commenters questioned the need to report minor or inconsequential errors or changes, even on an annual basis, as required in the proposed rule.

While errors or changes which result in changes in calculated 0

d peak clad temperatures of less than 50 F are not considered to be of immediete concern, the NRC requires cognizance of such changes or corre:t ons since they constitute a deviation from what

.previously has been reviewed and accepted. The proposed annual reporting is believed to be a fair compromise between the burden cf reporting and the Commission's need to be aware of changes and error corrections being made to evaluation models. Therefore, the annual reporting of minor errors remain; in the final aule.

One commenter interpreted the use of the words "or in the i application of such a model" as requiring reporting when facility page 9 ENCLOSURE E

changes (already reportable under 5 50.59), resulting in model input' changes, occur.

The regulatory language ~ferred to is intended to ensure that applications of models to oceas not contemplated during initial review of the model do not result in errors by extending a model beyond the range that it was intended. The Commission does not believe that further clarification of this requirement is necessary and has not done se in the final rule.

Several commenters requested a further relaxation of the reporting requirement by changing the definition of significant 0

code errors from 50 F to 100 F.

While justification for the 50 F criteria is largely judgmental, the NRC believes that it is sufficiently large to screen the code error corrections and changes which have little safety significance while providing a mechanism for timely reporting of more serious errors and changes. Since 50 F is a threshold for reporting and no further action is required pending NRC determination of safety significance, the Commission has retained this criteria in the final rule.

One c.ommenter requested consideration for allowing that the cumulative effect of several errors and corrections be applied towarcs the 50 F threshold.

Tte requirement, which states that the 50*F criteria applies to the sum of the absolute magnitudes of temperatur e changes from numerous error corrections or model changes was formulated specifically because the Commission requires knowledge of serious deficiencies in evaluation models in use by licensees.

Allowing errors or corrections which offset one another to relieve a licensee of the thirty-day reporting requirement, would be counter to this objective. If this recommendation were accepted, two errors or changes, having a large impact on the t

Page 10 ENCLOSURE E

calculated peak cladding temperature but in the opposite direction, would not be reportable if the net magnitude of their difference was less than 50 F. For this reason, and the fact that no further action (beyond reporting within thirty days) is required, the Commission retained this requirement in the final rule. .

7. Continued Use of Dougall-Rohsenow. Five comments that addressed this aspect of the proposed rule were received. One commentar believed that this correlation should not be permitted without further verification and should be phased out. Other commenters supported continued use of the correlation subject to the provisions of the proposed rule.

The NRC position is that no safety concern is created by continued use of the correlation, as long as the evaluation model is overall conservative. Therefore, the Commission can not justify the burden of requiring licensees to modify their evaluation models and to perform reanalysis. As discussed in SECY 83-472, current evaluation models contain more conservatisms than just those required by Appendix K. However, error corrections or changes could alter the conservatism of the model.

Therefore, the Commission believes that it is necessary to ensure continued overall conservatism in the evaluation models as a basis for continued use of the correlation. Therefore, the final

' rule does not modify this requirement except for the correction of a typographical error identified by one commenter.

8. Uncertainty Evaluation. The comments received on the uncertainty evaluation support the proposed rule, particularly the flexibility provided by a non-prescriptive requirement. .

Therefore, the Commission is publishing the final rule without modification of this requirement.

f i

Page 11 J'4 CLOSURE E

O

9. Acceptance Criteria. The three comments received on this topic *

• were all supportive of the existing criteria, as contained in 9 50.46(b), and thus the Commission did not give consideration to altering them in the final rule.

10. Cladding Materials. Three commenters requested that the Commission consider broadening the language of the rule to allow the use of a range of zirconium based alloys for cladding material.

The Commission believes that this modification is beyond the scope of the current rule revision and should be considered in a separate rulemaking action in which it would receive appropriate public review and comment prior to implementation. In cddition, zircaloy cladding material is specified in other portions of the Code of i9deral Regulation, such as 5 50.44. Making a change of this type is more suitable in a broader regulatory context.

Therefore, the Commission is not u.vadening the definition of cladding materials within this rulemoking.

11. Other Suggested Expansions to Rule Secpe. One commenter believes that hydraulic loads occurring during a LOCA could cause steam generator tubes to rupture and that the NRC should resolve steam generator tube integrity safety issues prior to publish'ag this rule.

Steam generator tubes are designed to withstand LOCA loads at allowed thinning, and there is no evidence to contradict this.

If anything, the problem would be with inspection techniques to detect the actual tube thinning and whether there is an unacceptably high probabl?ity that a tube rupture during a LOCA due to tube thinning is in excess of the design basis. 'However, the risk from LOCA with concurrent tube rupture will not be greatly affected by the proposed rule change. As a result of the commenter's concerns, this issue has been assigned as a generic Page 12 ENCLOSURE E

• issue (GI-141) to be prioritized by the NRC staff. The results of the prioritization process will determine if further action is required.

A second commenter believes that the ECCS rule does not adequately address a plant's long term decay heat removal capability, and recommends a "short/long term integrative analysis approach." Both the existing requirements and the proposed rule contain the requirement to provide for long term cooling subsequent to a LOCA. Small increases in power that may result from the proposed rule should not greatly change decay heat removal requirements following a LOCA or any other accident or transient. Thus, the issue of decay heat removal is not materially impacted by this rulemaking. Moreover, any proposed increase in power resulting from this rule promulgation would be approved only after the licensee demonstrates that decay hast removal capacities remain adequate. The Commission is planr.ing no further action with regard to this issue.

12. Acceptability of Models Approved Under SECY-83-472. One commenter requests that the rule language be modified to state explicitly that ECCS evaluation models that have been previously approved under SECY-83-472 continue to be acceptable under this rule.

SECY 83-472 provides an alternative, acceptable method for developing ECCS evaluation models. Licensees were still required, however, to demonstrate that evaluation models developed using the SECY-83-472 approach complied with the requirements of Appendix K to Part 50. This final rule explicitly finds that ECCS evaluation models, which have been previously approved as satisfying the requirements of Appendix K, remain acceptable. Therefore, the Commission sees no need for further clarification of this issue.

4 Page 13 ENCLOSURE E

13. Comments Received After Comment Period. Six letters commenting .*

on the proposed rule were received subsequent to the end of the comment period. The Commission considered these comments to the extent that the comments provided substantive information not previously considered.

One commenter believes that the proposed 5 50.46(a)(2) expands the discretion of the Director of the Office of Nuclear Reactor Regulation (NRR) by allowing imposition of immediate effective restrictions on reactor operation without a prior determination that such action is required to protect the public health, safety, or interes';. NRC's intent is not to alter the responsibilities of the Director of NRR but to simply retain the description of the scope of'the authority that is currently found in 5 50.46(a)(1)(v). Furthermore, the provisions of 5 50.46(a)(2) do not specify the procedure to be followed by the Director of NRR. These procedures are set out in Part 2 and remain unchanged by this rulemaking.

One commenter believes that the rule is illegal because it is based solely on cost savings considerations and that there is nothing wrong with large conservatisms.

The Commission disagrees with this assessment. Safety factors are required to protect the health and safety of the public when uncertainties in plant response exist. As these uncertainties are reduced, it is appropriate to modify these safety factors to provide more realistic evaluation of actual plant response. The large conservatisms of Appendix K served the public well in 1974 when there was great uncertainty in ECCS performance. However, these conservatisms are now known to be very large, and there is no need to "over regulate" by maintaining this unnecessary margin. This type of activity can often result in the expenditure of resources that would be better spent improving safety in other areas. The benefits to safety, while difficult to quantify, are believed to be substantial. While cost savings may have been one factor resulting in the rul,e change, the Page 14 ENCLOSURE E

Commission' believes that the conservatisms contained in the acceptance criteria themselves, as well as those required in the l uncertainty evaluation required in this rule, are adequate to protect the health and safety of the public. .

This commenter also cites portions of the 3975 General Electric Company's Nuclear Reactor Study (Reed Report), which claims that

- there is a lack of understanding of phenomena and small safety margins.

Many of the conclusions of the "Reed Report" were valid in 1975 when it was written and due to this fact it was difficult to show that sufficient safety margins existed. Most of the research discussed in NUREG-1230 has been conducted since the "Reed Report" was written and has resulted in significant improvement in understanding LOCA phenomena. We now know that significant margin to the ECCS acceptance criteria exists, particularly for .

the BWR/6 which was of concern in the "Reed Report." The contents of this report have been reviewed by the Commission on several occasions, most recently in NUREG-1285, and the finding has been made that no new significant safety issues are identified. For these reasons, the NRC is proceeding with this rulemaking, as proposed.

The same commenter also recommends that credit for ECCS margins be taken in the Individual Plant Examinations (IPE) and not through generic rulemaking.

The Commission agrees that plant specific differences may justify the application of different margins and that these may be addressed through Individual Plant Examinations. However, the requirement for licensees to evaluate ECCS performance and meet the acceptance criteria specified in 10 CFR 50.46(b) is generic.

The Commission believes that margins that may be reduced due to a better understanding of a reactor's response to a LOCA should be I applied through a generic rulemaking action because it allows a broad range of technical review of the issues, enhances public Page 15 ENCLOSURE E

participation in the process, and provides a complete public record. Therefore, the Commission has decided to proceed with the rulemaking as planned. ,

Finally, this commenter questions the experimental basis for this rule because full-scale ECCS bypass data is not yet available.

The 20/3D tests which will provide this important data represent a small portion of the total research upon which this rule relies. Significant research on ECCS bypass has already been completed in small scale vessels and the full-scale work is required only tc confirm the smaller scale results and quantify any uncertainty due to scale effects. One full-scale ECCS bypass test has already been completed under the 20/3D program which showed that more margin exists than expected from the small scale tests. Completion of the full-scale tests only affects the uncertainties in the calculations, and reduces them.

Uncertainties must be addressed by licensees in any analysis und6r the revised rule whether 2D/3D results are available or not. The Commission concludes that there is no need to delay the final rule, while awaiting these data.

SUMMARY

OF RULE CHANGES Section 50.46 Acceptance Criteria for Emergency Core Cooling Systems for Light Water Reactors: ,

See. tion 50.46(e -) is amended and redesignated 5 50.46(a)(1)(i) to delete the requirwnt that the features of Section I of Appendix K I to Part 50 be used to develop the evaluation model. This section now requires that an acceptable evaluation model have sufficient supporting justific'ation to show that the analytical technique realistically describes the behavior of the reactor system during a LOCA. The NRC expects that the analytical technique will, to the extent practicable, utilize realistic methods and be based upon applicable experimental data. The amended rule also requires that the Page 16 ENCLOSURE E

i. uncertainty of the calculation be estimated and accounted for when comparing the results of the calculation to the temperature limits and other criteria of f 50.46(b) so that there is a high probability that the criteria would not be exceeded. The Commission expects the realistic evaluation model to retain a degree of conservatism consistent with the uncertainty of the calculation. The final rule does not specifically prescribe the analytical methods or uncertainty evaluation techniques to be used. However, guidance has been provided

~

in the form of a Regulatory Guidel . As discussed in SECY-83-472, the NRC has, in the past, found acceptable a method for estimating the uncertainty that was judged to be at least at the 95% probability level. This probability level of 95% is considered adequate to meet the high level of probability required by the rule. It is also recognized that the probability cannot be determined using totally rigorous mathematical methods due to the complexity of the calculations. However, the NRC requires that any simplifying assumptions be stated so that the Commission may evaluate them to ensure that they are reasonable. The NRC has independently developed and exercised a methodology to estimate the uncertainty associated with its own thermal-hydraulic safety codes. This methodology is described in the "Compendium of ECCS Research."2 This document also provides reference to the large body of relevant thermal-hydraulic research, documents NRC studies on the effects of reactor power increases on risk, and provides background information on the ECCS rule. While this nethod has not been reviewed for acceptability from the standpoint of safety licensing, it may provide additional gulcance on how the uncertainty may be quantified. In addition to providing guidance to industry, this work was undertaken to provide a proof of principle and a tool to independently audit industry submittals.

Appendix K Section II, "Required Documentation," remains generally applicable, with only minor revisions made to be consistent with the amended rule.

1 Regulatory Guide, "Best Estimate Calculations of Emergency Core Cooling Systems Performance," was issued to all licensees.

2"Compendium of ECCS Research for Realistic LOCA Analysis,"

NUREG-1230, TBp. ,

Page 17 ENCLOSURE E

~

A new paragraph (ii) has been added to f 50.46(a)(1) to allow the features of Section I of Appendix K to be used in evaluation models as an alternative to performing the uncertainty evaluation specified in the amended 6 50.46(a)(1)(i). This method would remain acceptable because Appendix K is conservative with respect to the realistic method proposed in the amended 6 50.46(a)(1)(1). This would allow both current and future applicants and licensees to use existing evaluation models if they did not need or desire relief from current operating restrictions.

In 5 50.46, paragraphs (a)(2) and (3) have been revised to ,

eliminate portions of those paragraphs concerned with historical implementation of the current rule, These provisions have been replaceo as described in the following paragraphs:

Section 50.46(a)(2) has been revised to indicate that restrictions on reactor operation may be imposed by the Director of Nuclear Reactor Regulation, if the ECCS cooling performance evaluations are not consistent with the requirements of 55 50.46(a)(1)(i) and (ii). This section has been added to retain similar requirements that have been deleted from i 50.46(a)(1)(i) by this rule revision. This section does not specify the procedures to be followed by the Director. These procedures are found in Part 2 and are unchanged by this rulemaking.

The currer.t rule contains no explicit requirements concerning re-porting and reanalysis when errors in evaluation models are discovered or changes are made to evaluation models. However, current practice has required reporting of errors and changes and reanalyses with the revised evaluation models. This final rule explicitly sets forth requirements to be followed in the event of errors cr changes. The definition of a significant change is currently taken from Appendix K,Section II.1.b which defines a significant change as one which changes 0

calculated claddi.ng temperature by more than 20 F.

Page 18 ENCLOSURE E

o.

'o The revised 9 50.46(a)(3) states specific requirements for reporting and reanalyses when errors in evaluation models are discovered or changes are made to evaluation models. It requires that all changes or errors in approved evaluation models be ceported at least annually and does not require any further action by the licensee until the error is reported. Thereafter, although reanalysis is not required solely because of such minor error, any subsequent calculated evaluation of ECCS performance requires use of a model with such error, and any prior errors, corrected. The NRC needs to be apprised of even minor errors or changes in order to ensure that they agree with the applicant's or licensee's assessment of the significance of the error or change and to maintain cognizance of modifications made subsequent to NRC review of the evaluation model. Past experience has shown that many errors or changes to evaluation models are very minor and the burden of immediate reporting cannot be justified for these minor errors because they do not affect the immediate safety or operation of the plant. The NRC therefore requires periodic reporting to satisfy NRC's need to be apprised of changes or errors without imposing an unnecessary burden on the applicant or licensee. This report is to be filed within one year of discovery of the error and must be reported each year thereafter until a revised evaluation model or a revised evaluation correcting minor errors is approved by the NRC staff.

Significant errors require more timely attention since they may be important to the safe operation of the plant and raise questions as to the adequacy of the overall evaluation model. This final rule defines a significant error or change as one which results in a 0

calculated peak fuel ciadding temperature different by more than 50 F, or an accumulation of errors and changes such that the sum of the absolute magnitude of the temperature changes is greater than 50 F.

More timely reporting (30 days) is required for significant errors or changes. This definition of a significant change is based on NRC's judgement concerning the importance of errors and changes typically reported to the NRC in the past. This final rule revision also allows the NRC to determine the schedule for reanalysis based on the Page 19 ENCLOSURE E

1 importance to safety relativo to other applicant or licensee requirements. Errors or changes that result in the calculated plant performance exceeding any of the criteria of i 50.46(b) mean that the plant is not operating within the requirements of the regulations and require immediate reporting as required by 6 50.55(e), 5 50.72 and S 50.73 and immediate steps to bring the plant into compliance with 6 50.46.

Appendix K ECCS Evaluation Models:

Amendments have been made to Appendix K,Section I.C.5.b, to modify the post-CHF heat transfer correlations listed as acceptable.

The "McDonough" reference has been replaced with a more recent paper by the same authors entitled "An Experimental Study of Partial Film Boiling Region With Water at Elevated Pressures in a Round Vertical Tubed which is more generally available and which includes additional data.

The heat transfer correlation of Dougall and Rohsenow, listed as an acceptable heat transfer correlation in Appendix K, paragraph I.C.S.b, has been removed, since research performed since Appendix K was written has shown that this correlation overpredicts heat transfer coefficients under certain conditions and therefore can produce nonconservative results. A number of applicants and licensees currently use the Dougall-Rohsenow correlation in approved evaluation models. The NRC has concluded that the continued use of this correlation can be allowed. This is appropriate (even though parts of the approved evaluation model, Dougall-Rohsenow, are known to be nonconservative) because the existing evaluation models are known to contain a large degree of overall conservatism even while using the Dougall-Rohsenow correlation. This large overall conservatism has been demonstrated through compariso . between evaluation model calculations and calculations using NRC's best estimate computer codes. Thus requiring that the applicants and licensees remove the Dougall-Rohsenow correlation from their current evaluation models cannot be justified as necessary to maintain safety. The stipulation Page 20 ENCLOSURE E

that the Dougall-Rohsenow correlation will cease to be acceptable for previously approved evaluation models applies only when changes to the model are made which reduce the calculated peak clad temperature by 50"F or more. However, the requirement to report any changes or culmir.ation of changes, such that the sum of the absolute magnitudes of the respective temperature changes is greater than 50'F, still applies.

A new Section I.C.5.c has been added to Appendix K to state the Commission's requirements regarding continued use of the Dougall-Rohsenow correlation in existing evaluation models.

Evaluation models which make use of the Dougall-Rohsenow correlation and have been approved prior to the effective date of this rule may continue to use this correlation as long as no changes are made to the evaluation model which significantly reduce the current overall conse vatism of the evaluation model. If the applicant or licensee submits proposed changes to an approved evaluation model, or submits corrections to er ors in the evaluation model which significantly reduce the existing overall conservatism of the model, continued use of the Dougall-Rohsenow correlation under conditions where nonconservative heat transfer coefficients result would no longer be acceptable. For this purpose, a significant r. duction in overall conservatism has been defined as a "net" reduction in calculated peak clad temperature of at least 50 F from that which would have been calculated using existing evaluation models. A reduction in calculated peak clad temperature could potentially result in an increase in the actual allowed peak power in the plant. An increase in allowed plant peak power with a known nonconservatism in the analysis would be unacceptable. This definition of a significant reduction in overall conservatism is based on a judgement regarding the size of the existing overall conservatism in evaluation model calculations relative to the conservatism required to account for overall uncertainties in the calculations.

Appendix K,Section II.1.b, has been removed since this requirement has been clarified in the amended i 50.46(a)(3).

Page 21 ENCLOSURE E

.o Likewise, Appendix K,Section II.5, has been amended to account for the fact that not all evaluation models will be required to use the features of Appendix K,Section I. These minor changes to Appendix K do not affect any existing approved evaluation models since the changes are either "housekeeping" in nature or are changes to "acceptable features," not "required features."

AVAILABILITY OF DOCUMENTS

1. Copies of NUREGs 1230 and 1285 may be purchased from the Superintendent of Documents, U.S. Government. Printing Office, P.O. Box 37082, Washington, D.C. 20013-7082. Copies are also available from the National Technical Information Service 5285 Pcrt Royal Rcad, Springfield, VA 22161. A copy is also available for public inspection and/or copying at the NRC Public Document Room, 1717 H Street NW., Washington, DC 20555.

2. Copies of SECY-83-472, an information report entitled "Emergency Core Cooling Systems Analysis Methods," dated November 17, 1983, is available for inspection and copying at the NRC Public Documents Room 1717 H Street NW., Washington, DC 20555. Single copies of this report may be obtained by writing (give name and address of the contact).

3. Regulatory Guide, "Best Estimate Calculations of Emergency Core Cooling Systems Performance" may be obtained by writing to the Division of Technical Information and Document Control, U.S.

Nuclear Regulatory Commission, Washington, DC 20555.

4. The Paraphrased Summary of Public Comments on ECCS Rule is available for public inspection at the NRC Public Documents Room, 1717 H Street NW., Washington, DC 20555.

FINDING OF NO SIGNIFICANT ENVIRONMENTAL IMPACT: AVAILABILITY The Commission has determined under the National Environmental Policy Act of 1969, as amended, and the Commission's regulations in Page 22 ENCLOSURE E .

Subpart A of 10 CFR Part 51, that this rule is not a major Federal action significantly affecting the quality of the human environment and therefore an environmental impact statement is not required. The primary effect of the rule is to allow an increase in the peak local power in the react 5r. This could be used either to tailor the power shape within the reactor or to increase the total power. Changing the power shape without changing the total power has a negligible effect on the environmental impact. The total power could also be increased, but is expected to be increased by no more than about 5% due to hardware limitations in existing plants. This 5% power increase is not expected to cause difficulty in meeting the existing environmental limits. The only change in non-radiological waste will be an increase in waste heat rejection commensurate with any increase in power. For

, stations operating with an open (once through) cooling system, this additional heat will be directed to a surface water body. Discharge of this heat is regulated under the Clean Water Act administered by the U.S. Environmental Protection Agency (EPA) or designated state agencies. It is not intended that NRC approval of increased power level affects in any way the responsibility of the . icensee l to comply with the requirements of the Clean Water Act. The environmental assessment and finding of no significant impact on which this ,

determination is based are available for inspection at the NRC Public Document Room, 1717 H Street NW, Washington, DC. Single _ copies of the environmental assessment and the finding of no significant impact are available from L. M. Shotkin, Office of Nuclear Regulatory Research, U.S. Nuclear Regulatory Commission, Washington DC. 20555, telephone (301) 492-3530.

PAPERWORK REDUCTION ACT STATEMENT This final rule amends information collection requirements that are subject to the Paperwork Reduction Act of 1980 (44 U.S.C. 3501 et seq.). These reporting requirements were approved by the Office of l Management and Budget (Approval Number 3150-0011).

i i

page 23 ENCLOSURE E j

i

.o REGULATORY ANALYSIS The Commission has prepared a regulatory analysis for this final regulation which exauines the costs and benefits of the alternatives considered and is available for inspection and copying at the NRC Public Document Roam, 1717 H Street NW, Washington, DC. Single copies of the analysis may be obtained from L. M. Shotkin, Office of Nuclear Regula-tory Research, Washington, DC. 20555, telephone (301) 492-3530.

REGULATORY FLEXIBILITY CERTIFICATION As required by the Regulatory Flexibility Act of 1980, 5 U.S.C.

605(b), the Commission certifies that this rule will not have a significant economic impact upon a substantial number of small entities.

This rule affects only the licensing and operation of nuclear power plants. The companies that own these plants do not fall within the scope of the definition of "small entities" set forth in the Regulatory Flexibility Act or the Small Business Size Standards set out in regu-lations issued by the Small Business Administration in 13 CFR Part 121.

Sinca these companies are dominant in their service areas, this rule does not fall within the purview of the Act.

BACKFIT ANALYSIS Except for the reporting requirements, a backfit analysis is not required by 10 CFR 50.109 because the rule does not require applicants or licensees to make a change but only offers additional options.

Nonetheless, the factors in 10 CFR 50.109(c) have been analyzed for the entire rule.

1. Statement of the specific objectives that the backfit is designed to achieve.

The objective of the rule is to modify 10 CFR 50.46 and Appendix K to permit the use of realistic ECCS evaluation models.

More realistic estimates of ECCS performance, based on the improved knowledge gained from recent research on ECCS Page 24 ENCLOSURE E

O ,

performance, may remove unnecessary operating restrictions. Also experience with the previous version of 9 50.46 has demonstrated that a clearer definition of reporting requirements for changes ,

and errors is very desirable.

2. General description of the activity that would be reouired by the licensee or applicant in order to complete the backfit.

The amendment allows alternative methods to be used to demonstrate that the ECCS would protect the nuclear reactor core during a postulated design basis loss-of-coolant accident (LOCA).

While continuing to allow the use of current Appendix K methods and requirements, the rule also allows the use of more recent information and knowledge currently available to demonstrate that the ECCS would perform its safety function during a LOCA. If an applicant or licensee elects to use a new realistic model they will be required to provide sufficient supporting justification to validate the model and include comparisons to experimental data and estimates of uncertainty, In accounting foa the uncertainty, the analysis would have to show, with a high level of probability, that the ECCS performance criteria are not exceeded. Whether or not a licensee or applicant chooses to use realistic analysis, complete with an uncertainty analysis, each licensee must comply with the requirement to report changes to their evaluation models (i.e., less than 50 F change in calculated peak cladding temperature) annually to the NRC. In addition, significant changes (those which have a greater than 500 F change in calculated peak cladding temperature) have to be reported within 30 days.

3. potential change in risk to the public from the accidental offsite release of radioactive materials.

The rule could result in increased local power within the reactor core and possibly increases in total power. Power

! increases on the order of 5 will have an insignificant effect on j Page 25 ENCLOSURE E

risk. One effect of increased power could be to increase the fission product inventory. A five percent power increase would result in a less than five percent increase in fission produ::ts.

Thus, less than five percent more fission products might be released during core melt scenarios and potentially released to the environment during severe accidents.

The rule still requirer that fuel rod peak cladding temperature (PCT) remain below 2200 F. Reactors choosing to increase power by about five percent will be operating with less margin between the PCT and the 2200 F limit than previously. The increased risk represented by this decrease in margin and increase in fission product inventory is negligible and falls In addition, within the uncertainties of PRA risk estimates.

other safety limits, such as departure from nucleate boiling (DNB), and operational limits, such as turbine design, will limit the amount of margin reduction permitted under the rule. The rule could also potentially reduce the risk from pressurized thermal shock by allowing the reactor to be operated in a manner which reduces the neutron fluence to the vessel.

4. potential impact on radiological exposure to facility employees.

Since the primary effect of the rule involves the calculational methods to be used in determining the ECCS cooling performance, it is expected that there will be an insignificant impact on the radiological exposure to facility employees, i

Because of the reduced LOCA restrictions resulting from the new calculations it is possible for the plant to achieve more efficient operation and improved fuel utilization with improved j

maneuvering capabilities. As a result, it is conceivable that i there could be a reduction in radiological exposure if the fuel reloads can be reduced. This effect is not expected to be very significant.

1 Page 26 ENCLOSURE E l

e

5. Installation and continuing costs associated with the backfit, including the cost of facility down times or the cost of construction delay.

LOCA considerations resulting from the present rule are re-stricting the optimum production of nuclear electric power in some plants. These restrictions can be placed into the following three categories:

(1) Maximum plant operating power, (2) Operational flexibility and operational efficiency of the plant, and (3) Availability of manpower to work on other activities.

The effect of the rule will vary from plant to plant. Some plants may realize savings of several million dollars per year in fuel and operating costs. Significantly greater economic benefit would be realized by plants able to increase total power as a j result of this final rule. The regulatory analysis cited above l l

indicates that the total present value of the energy replacement l cost savings for a five percent power upgrade would vary between  !

18 and 127 million dollars depending on the plant. Additional information concerning these potential cost savings are included in the regulatory analysis.

The costs associated with the new reporting requirements are deemed to be minimal. Although the existing Appendix K has no official reporting requirements, paragraph II.1.b was interpreted by the staff to require a reanalysis and report to NRC when significant changes are made which change the peak cladding temperature by more then 20'F. Therefore, this rule change, by  :

changing the definition of significant changes to 50'F, is f actually a relaxation of current practices. The annual reporting of changes that are not significant is not viewed by the NRC as a major burden since no other action is required. I t

i Page 27 ENCLOSURE E

.o

6. The potential safety impact of changes in plant or operational . .

complexity including the effect on other proposed and existing regulatory requirements.

There are safety benefits derivable from alternative fuel management schemes that could be utilized. The higher power peaking factors that would be allowed with the final rule provide greater flexibility for fuel designers when attempting to reduce neutron flux at the vessel wall. This can result in a corresponding reduction in risk from pressurized thermal shock.

The reduced cladding temperatures that would be calculated under the revised rule offers the possibility of other design and operational changes that could result from the lower calculated temperatures. ECCS equipment numbers, sizes or surveillance requirements might be reduced and still meet the ECCS design criteria (if not required to meet other licensing requirements).

Another option may be to increase the diesel / generator start time duration.

In summary, the effect of this rule on safety would have both potential positive and negative aspects. The potential for reduction of ECC system capability in existing or new plants is present. However, several positive aspects may also be realized j

under the final rule. The net effect on safety would be plant l

l specific. However, the probability of a large break LOCA is so low that the choice of best estimate versus Appendix K would have little effect on public risk.

7. The estimated resource burden on the NRC associated with the proposed backfit; and the availability of such resources.

The major staff resources required under the final rule are to review the realistic models and uncertainty analysis required by the revised ECCS Rule. Based on previous experience with the Page 28 ENCLOSURE E

~_ - - - _ _ _

General Electric Company's SAFER model and the learning tha% has resulted from these efforts, it is estimated that approximately one staff year would be required to review each generic model submitted. There are four major reactor vendors (GE already has a revised evaluation model approved under the existing Appendix K for both jet pump and non-jet pump plants and may update their .

methodology under this new rule) and several fuel iiuppliers and utilities which perform their own analyses and potentially might ,

submit generic models for review. However, it is expected that [

only 3 or 4 generic models would be submi'cted since not all plants would ben 6 fit from this rule. Thus, about 3-4 staff years would be required to review the expected generic models. Once a generic model is approved, the plant specific rsview is very short. In addition, several vendors are currently planning to

- submit realistic models in conjunction with the use of SECY-83-472. Therefore, staff resources would be expended to review these models in any event. Since these models would not change as a result of the revised ECCS rule, there should he no net increase in resources required over that already planned to be expended. In summary, while it is difficult to estimate accurately, it is expected that the rule change will have a small overall impact on NRC resources.

8. The potential impact of differences in facility type, design or age on the relevancy and practicality of the backfit.

The degree to which the rule would affect a particular plant depends on how limited the plant is by the LOCA restrictions.

General Electric Company (GE) plants do tend to be limited in operation by LOCA restrictions and would benefit from relief from LOCA restrictions. However, this relief is already available for most GE plants through the recently approved SAFER evaluation model. Any additional relief due to a rule change would be of little further benefit. Westinghouse (W) plants would appur to ,

directly benefit from relaxation of LOCA limits. W plants represent the largest number of plants, with 47 plants operating page 29 ENCLOSURE E

,O!

and 10 additional plants being constructed. )! indicates that ,

most of these plants are limited by LOCA considerations. The l potential benefit for plants of B&W and CE design is uncertain at ,

this time.  !

9. Whether the proposed backf,it is interim or final and if interim, l the justification for imposing the proposed backfit on an interim t basis.

The rule, when made effective, will be in final form and not interim form. It will continue to permit the performance of ECCS cooling calculations using either realistic models or models in 4

accord with Appendix K.

LIST OF SUBJECTS IN 10 CFR PART 50 Antitrust, Classified information, Fire prevention, Incorporation by reference, Intergovernmental relations, N'Jelear power plants ard roactors, Penalty, Radiation protection, Reactor siting criteria, Reporting and Recordkeeping requirements.  ;

For the reasons set out in the preamble and under the authority of the Atomic Energy Act of 1954, as amended, the Energy Reorganization Act of 1974, as amended, and 5 U.S.C 552 and 553, the

)

NRC is adopting the following amendments to 10 CFR Part 50.

PART 50-DOMESTIC LICENSING OF PRODUCTION AND UTILIZATION FACILITIES

1. The authority citation for Part 50 continues to read as i follows:

i AUTHORITY: Secs. 102, 103, 104, 105, 161, 182, 183, 186, 189, 68 Stat. 936, 937, 938, 948, 953, 954, 955, 956, as amended, sec. 234, 83 j i

j Stat. 1244, as amended (42 U.S.C. 2132, 2133, 2134, 2135, 2201, 2232, 2233, 2236, 2239, 2282), secs 201, as amended, 202, 206, 88 Stat.

j '

1242, as amended, 1244, 1246, (42 U.S.C. 5841, 5842, 5846).

l I

Page 30 ENCLOSURE E I

i Section 50.7 also issued under Pub. L.95-601, sec. 10, 92 Stat. 1 2951 (42 U.S.C. 5851). Section 50.10 also issued under secs. 101, l 185, 68 Stat. 936, 955, as amended (42 U.S.C. 2131, 2235); sec. 102, Pub. L.91-190, 83 Stat. 853 (42 U.S.C. 4332). Sections 50.23, 50.35, l t

50.55, 50.56 also issued under sec.185, 68 Stat. 955 (42 U.S.C. ,

2235). Sections 50.33a, 50.55a, and Appendix Q also issued under sec.

102, Pub. L.91-190, 83 Stat. 853 (42 U.S.C. 4332). Sections 50.34 and 50.54 also issued under sec. 204, 88 Stat 1245 (42 U.S.C. 5844).

Sections 50.58, 50.91, and 50.92 also issued under Pub. L.97-415, 96 l

Stat. 2073, (42 U.S.C. 2239). Section 50.78 also issued under sec.

122, 68 Stat. 939 (42 U.S.C. 2152). Sections 50.80-50.81 also issued under sec. 184, 68 Stat. 954, as amended (42 U.S.C. 2234). Section 50.103 also issued under sec.108, 68 Stat. 939, as amended (42 U.S.C.

2138). Appendix F also issued under sec. 187, 68 Stat. 955 (42 U.S.C.

2237).

For the purposes of sec. 223, 68 Stat. 958, as amended (42 U.S.C.

2273), $$ 50.10(a), (b), and (c), 50.44, 50.46, 50.48, 50.54, and 50.80(a) are issued under sec. 161b, 68 Stat. 948, as amended (42 U.S.C. 2201(b)); il 50.10(b) and (c) and 50.54 are issued under sec. 1611, 68 Stat. 949, as amended (42 U.S.C. 2201(i)); and il 50.9, 4 50.55(e), 50.59(b), 50.70, 50.71, 50.72, 50.73, and 50.78 are issued under sec. 161o, 68 Stat. 950, as amended (42 U.S.C. 2201(o)).

2. In i 50.46, paragraph (a) is revised to read as follows:

9 50.46 Acceptance criteria for emergency core cooling systems for light-water nuclear power reactors.

(a)(1)H), [Exceptlis-provided-in-paragraph-faff2)-and-f3)-ef this-sectien-] Each boiling and pressurized light-water n'.' clear power reactor fueled with uranium oxide pellets within cylindrical Zircaloy cladding must be provided with an emergency core cooling system (ECCS) that must be designed such that its calculated cooling performance following postulated loss-of-coolant accidents conforms to the criteria set forth in paragraph (b) of this section. ECCS cooling  ;

i l

i Page 31 ENCLOSURE E

performance must be calculated in accordance with an acceptable evaluation model and must be calculated for a number of postulated loss-of-coolant accidents of different sizes, locations, and other properties sufficient to provide assurance that the most severe postulated loss-of-coolant accidents are calculated. [ Appendix-Mi EGGS-Evaluation-MedelsT-sets-ferth-certain-required-and-seceptable features-of-evaluation-models-] Except as provided in partaraph (a)(1)(ii) of this section, the evaluation model must include sufficient supporting justification to show that the analytical technique realistically describes the behavior of the reactor system during a loss-of-coolant accident. Comparisons to applicable experimental data must be made and uncertainties in the analysis method and inputs must be identified and assessed so that the uncertainty in the calculated results can be estimated. This uncertainty must be accounted for, so that, when the calculated ECCS cooling performance is compared to the criteria set forth in paragraph (b) of this section, there is a high level of probability that the criteria would not be exceeded. Appendix K, part II, Required Documentation, sets forth the documentation requirements for each evaluation model. Lunfermance-with-the-criteria-set-ferth-in paragraph-fb)-ef-this-section-with-EGGS-ceeling-performance-cateu4ated in-accordance-nith-an-acceptable-evaluatien-modeli-may-require-that restrictions-be-imposed-en-reseter-eperation:]

(ii) Alternatively, an ECCS evaluation model may be developed in conformance with the required and acceptable features of Appendix K ECCS Evaluation Models.

(2) The Director of Nuclear Reactor Regulation may impose l restrictions on reactor operation if it is found that the evaluations of ECCS cooling performance submitted are not consistent with paragraphs (a)(1)(i) and (ii) of this section.

(3)(i) Each applicant for or holder of an operating license or construction permit shall estimate the effect of any change to or error in an acceptable evaluation model or in the application of such Page 32 ENCLOSURE E

-- - .-- , . p_m

l s ., .

[* a model to determine if the change or error is significant. For this purpose, a significant change or error is one which results in a 0

calculated Deak fuel cladding temperature different by more than 50 F from the temperature calculated for the limiting transient using the last acceptable model, or is a cumulation of changes and errors such that the sum of the absolute magnitudes of the respective temperature changes is greater thar. 500F.

(ii) For each change to or error discovered in an acceptable evaluation model or in the application of such a model that affects the temperature calculation, the applicant or licensee shall report the nature of the change or error and its estimated effect on the limiting ECCS analysis to the Commission at least annually as specified in S 50.4. If the change or error is significant, the applicant or licensee shall provide this report within 30 days and include with the report a proposed schedule for providing a reanalysis or taking other action ar_may be needed to show compliance with 6 50.46 requirements. This schedule may be developed using an integrated scheduling system previously approved for the facility by the NRC. For those facilities not using an NRC approved integrated scheduling system, a schedule will be established by the NRC staff within 60 days of receipt of the proposed schedule. Any change or error correction that results in a calculated ECCS performance that does not conform to the criteria set forth in paragraph (b) of this section is a reportable event as described in 66 50.55(e), 50.72 and  !

6 50.73. The affected applicant or licensee shall propose immediate steps to demonstrate compliance or bring plant design or operation into compliance with 6 50.46 requirements.

• 2 * *

• I i

3. In 10 CFR Part 50 Appendix K, paragraph II.1.b is deleted, paragraph II.1.c is redesignated II.1.b, the text of paragraph I.C.5.b and paragraphs II.1.b and II.5 are revised, and a new section I.C.S.c is added to read as follows: .

1 1

Page 33 ENCLOSURE E

APPENDIX K - ECCS EVALUATION MODELS I. REQUIR60 AND ACCEPTABLE FEATURES OF THE EVALUATION MODELS***

C. Blowdown Phenomena ***

5. Post-CHF Heat Transfer Correlations.***

b. The Groeneveld flow film boiling correlation (equation 5.7 of D. C. Groeneveld, "An Investigation of Heat Transfer in the Liquid Deficient Regime," AECL-3281, revised December 1969)[T-the-Bengall-Rehsenew-fiew-film-be44+ng-eerrelatien-(R:-S;-Beugall-and-W:

M:-Rohsenew;.uFilm-Beiling-en-the-inside-of-Vertical-Tabes-with-Wpward Flen-of-Fluid-at-Eew-Qualities 7 .MIT-Repert-Number-9079-267-Eambridge-u Massachusetts;-September-1963);] and the Westinghouse correlation of steady-state transition boiling ("Proprietary Redirect / Rebuttal Testimony of Westinghouse Electric Corporation," USNRC Docket RM-50-1, page 25-1, October 26, 1972) are acceptable for use in the post-CHF boiling regimes. In addition, the transition boiling correlation of McDonough, Milich, and King (J. B. McDonough, W. Milich, E. C. King,

[uPartial-F+1m-86414ng-with-Water-at-2000-psig-in-a-Reund-Vertical Tube;u.MSA-Research-Gers:7-Technical-Repert-62-(NP-6976);-fi958)] "An Experimental Study of Partial Film Boiling Region with Water at Elevated Pressures in a Round Vertical Tube," Chemical Engineering Progress Symposium Series, Vol. 57. No. 32, pages 197-208, (1961) is suitable for use between nucleate and film boiling. Use of all these correlations is [shall-be] restricted as follows:

c. Evaluation models approved after ( ) which make use of the Dougall-Rohsenow flow film boiling correlation (R. S.

Dougall and W. M. Rohsenow, "Film Boiling on the Inside of Vertical Tubes with Upward Flow of Fluid at Low Qualities, MIT Report Number 9079 26, Cambridge, Massachusetts, September 1963) may not use this correlation under conditions where nonconservative predictions of heat transfer result. Evaluation models that make use of the Cougall-Rohsenow correlation and were approved prior to ( )

continue to be acceptable until a change is made to, or an error is Page 34 ENCLOSURE E

~

,' corrected in, the evaluation model that results in a significant reduction in the overall consurvatism in the evaluation model. At that time continued use of the Dougall-Rohsenow correlation under conditions where nonconservative predictions of heat transfer result will no longer be acceptable. For this purpose, a significant reduction in the overall conservatism in the evaluation model would be a reduction in the eticulated peak fuel cladding temperature of at least 50 0F from that which would have been calculated on

( ) due either to individual changes or error corrections or the net effect of an accumulation of changes or error corrections.

II. REQUIRED DOCUMENTATION 1.a. ***

b. A complete listing of each computer program, in the same form as used in the evaluation model, must be furnished to the Nuclear Regulatory Commission upon reauest.

5. General Standards for Acceptability - Elements of evaluation models reviewed will include technical adequacy of the calculational methods, including : for models covered by 6 50.46(a)(1)(ii),

compliance with required features of Section I of this Appendix K [and previsien-of-a-level-ef-safety-and-margin-of-censervatism-comparable te-ether-acceptable-evaluation-medelsi-taking-inte-aceeant-signiffeant differenees-in-the-reseters-to-which-they-apply-] ; and, for models covered by 6 50.46(a)(1)(i), assurance of a high level of probability that the performance criteria of 6 50.46(b) would not be exceeded.

(

Dated at Rockville, MD this day of , 1988.

For the Nuclear Regulatory Commission.

Samuel J. Chilk, Secretary of the Commission.

Page 35 ENCLOSURE E

0 it

- 9 ENCLOSURE F 1 l

1 i

l i

i 1

l l

l W_._____.__.-_-_____._-_.-_____.-_.-_______._____________._ _ _ _ _ _ _ _ _ _ _ . ____ - _ - - - _ _ - _ _ _

l=.

'o J

REGULATORi GUIDE BEST ESTIMATE CALCULATIONS OF EMERGENCY CORE COOLING SYSTEMS PERFORMANCE t

\

'l BEST ESTIMATE CALCULATIONS CF EMERGENCY CORE COOLING SYSTEMS PERFORMANCE A. INTRODUCTION Paragraph 50.46(a)(1) of 10 CFR Part 50, "Acceptance Criteria for Emergency Core Cooling Systems for Light Water Nuclear Power Reactors,"

requires that light water nuclear reactors fueled with uranium oxide pellets within cylindrict.1 zircaloy cladding be provided with emergency core cooling systemt, (ECCS) that are designed in such a way that their calculated core cooling performance after a postulated loss-of-coolant accident (LOCA) conforms to certain criteria specified in Paragraph (b) 9 50.46. Paragraph (b)(1) requires that the calculated maximum temperature of fuel element cladding not be greater than 2200 F. In addition, Paragraphs (b)(2) through (b)(5) of f 50.46, which contain required limits for calculated maximum cladding oxidation and maximum hydrogen generation, require that calculated changes in core geometry remain amenable to cooling and that long term decay heat removal be provided.

In 1987, the NRC staff proposed to amend the requirements of 9 50.46 and Appendix K, "ECCS Evaluation Models," so that the regulations would reflect the improved understanding of ECCS performance during reactor transients that had been obtained through the extensive research performed since the promulgation of the original recairements in January of 1974. By this revision, paragraph 6 50.46(a)(1) now permits licensees or applicants to use either Appendix K features or a 1 2 realistic evaluation model. These realistic evaluation models must I

For the purpose of this guide, the terms "best-estimate" and "realistic" have the same meaning. Both terms are used to indicate that the techniques attempt to predict realistic reactor system thermal-hydraulic response. Best estimate is not used in a statistical sense in this guide.

2The term "evaluation model" re rs to a nuclear plant ,ystem computer code or any other analysis too. 'esigned to predict the aggregate behavior of a loss-of coolant accident. It can be either best-estimate or conservative and may contain many correlations or models.

Page 1 ENCLOSURE F

include sufficient supporting jus,tification to demonstrate that the analytic techniques realistically describe the' behavior of the reactor

' system during a postulated loss-of-coolant accident. 10 CFR Part 50.46(a)(1) also requires that the uncertainty in the realistic evaluation model be quantified and considered when comparing the results of the calculations with the applicable limits in 9 50.46(b) so that there is a high probability that the criteria will not be exceeded.

This regulatory guide describes models,3 correlations,4 data, model evaluation procedures, and methods that are acceptable to the NRC staff for meeting the requirements for a realistic or best-estimate calculation of ECCS performance during a loss-of-coolant accident and for estimating the uncertainty in that calculation. Methods for including the uncertainty in the comparisons of the calculational results to the criteria of paragraph 50.46(b), in order to meet the requirement that there be a high probability that the criteria would not be exceeded, are also described in this regulatory guide. Paragraph (a) of 5 50.46 also permits licensees to use evaluation models developed in conformance with Appendix K.

Other models, data, model evaluation proce'dures, and methods will be considered if they are supported by appropriate experimental data and technical justification. Any models, data, model evaluation procedures, and methods listed as acceptable in this regulatory guide are acceptable in a generic sense only and would still have to be justified to the NRC staff as being appropriately applied and applicable for particular plant applications.

The Appendix to this regulatory guide lists models, correlations, data, and model evaluation procedures that the NRC staff considers 3

The term "model" refses to a set of equations derived from fundamental physical laws and designed to predict the details of a specific phene;menon.

4 The term "correlation" refers to an equation having empirically determined constants such that it can predict some details of a specific phenomena for a limited range of conditions.

Page 2 ENCLOSURE F

., . . - - _ _ -. .. ._ - - - = - - . . . -

acceptable-for; realistic calculations of ECCS performance. It also ,e provides.a description of the acceptable features of best-estimate

' ~

c computer codes and acceptable methods for determining the uncertainty in the calculations.

This regulatory guide contains no new information collection requirements and, therefore, is not subject to the Paperwork Reduction Act of 1980 (44 U.S.C. 3501 et seq.).

Any information collection activities discussed in this draft regulatory guide are contained as requirements in 10 CFR Part 50, which -

provides the regulatory basis for this guide. The information collection requirements in 10 CFR Part 50 have been cleared under OMB Clearance No. 3150-0011.

B. DISCLSSION The criteria set forth in 10 CFR 50.46, "Acceptance Criteria for Emergency Core Cooling Systems in Light Water Nuclear Power Reactors,"

and the calculational methods specified in Appendix K were promulgated in January 1974 af ter extensive rulemaking hsarings ar v were based on the understanding of ECCS performance available at that time. In the years following the prornulgation of those rules, the NRC, the nuclear industry, and several foreign institutions have conducted an extensive program of research that has greatly improved the understanding of ECCS performance during a F tulated LOCA. The methods specified in Appendix K were found to be highly conservative; that is, the fuel cladding temperatures expected during a less-of-coolant accident would be much less than the temperatures calculated using Appendix K methods.

In addition to showing that Appenoix K is conservative, the ECCS research provided information that allows for quantification of that conservatism. The results of experiments, computer cod 9 development, Page 3 ENCLOSURE F

~

,--,-,.,...,-..-.,.,,-.,,.,-,,--,,.,,-._.--n. - . . - . - , . . -

O o.

'+

and code assessment allow more accurate calculations of ECCS performance, along with reasonable estimates of uncertainty, during a postulated loss-of-coolant accident than is possible using the Appendix K procedures.

It was also found that some plants are being restricted in operating flexibility by limits resulting from conservative Appendix K requirements; thus, preventing optimal oparation. Based on the research performed, it was determined that these restrictions could be relaxed through the use of more realistic calculations without adversely affecting safety. The Appendix K requirements tended to divert both NRC and industry resources from matters more relevant to reactor safety to analyses with known nonphysical assumptions.

In recognition of the known conservatisms in Appendix K, the NRC 5

adopted an interim approach, described in SECY-83-472 , to accommodate industry requests for improved evaluation models for the purpose of reducing reactor operating restrictions.. This interim approach is a step in the direction of basing licensing decisions on realistic calculations of plant behavior. Although the approach permits many "best estimate" methods and models to be used for licensee submittals, it retains those features of Appendix K that are legal requirements.

The current revision of 10 CFR 50.46 now permits dCCS evaluation models to be fully "best estimate" and removes the arbitrary conservatisms contained in the required features of Appendix K for those licensees wishing to use these improved methods. The NRC staff believes that safety is best served when decisions concerning the limits within which nuclear reactors are permitted te operate are based upon Information Report from William J. Dircks to the Commissioners, dated November 17, 1983, "Emergency Core Cooling System Analysis Methods,"

SECY-83-472.

Page 4 ENCLOSURE F

realistic calculations. This approach is currently being used in the ,,

resolution of almost all reactor safety issues (e.g., anticipated '

transients without scram, pressurized thermal shock, ard operator guidelines) and is now available for one of the last remaining major issues still treated in a prescriptive manner, the loss-of-coolant accident.

The NRC staff amended 5 50.46 of 10 CFR Part 50 to allow realistic methods to be used for the ECCS performance calculations in place of the evaluation models that use the required Appendix K features. This rule change also requires analysis of the uncertainty of the best-estimate calculation and requires that this uncertainty be considered when comparing the results of the calculations to the limits of 9 50.46(b) so that there is a high probability that the criteria will not be exceeded.

In this manner, more realistic calculations are available for regulatory decisions, yet appropriate conservatism would be maintained consistent with the accuracy of the calculation.

Many of the methods and models needed for a best-estimate calculation are the same as those used previously for evaluation model analyses. Although licensees and applicants are well acquainted with them, explicit guidance on acceptable methods and models (based on NRC experience with its own best-estimate advanced codes such as TRAC-PWR, TRAC-BWR, RELAPS, COBRA and FRAP) would be useful. Further, the NRC has not made acceptable methods for uncertainty analyses widely available.

Therefore, the NRC staff decided that guidance in the form of a regulatory guide would be useful in order to document procedures acceptable to the regulatory staff.

C. REGULATORY POSITION The features presented in the Appendix to this regulatory guide, "Description of Best Estimate Calculations of Emergency Core Cooling Page 5 ENCLOSURE F

Systems Performance and Estimate of the Calculational Uncertainty," are

• acceptable to the NRC staff for demonstrating compliance with Paragraph (a)(1)(i) of 10 CFR 50.46.

D. IMPLEMENTATION The purpose of this section is to provide information to applicants and licensees regarding the NRC staff's plans for using this regulatory guide.

Licensees and applicants may propose means other than those specified by the provisions of Section C of this guide for meeting applicable regulations. The guidance provided in Section C has been approved for use by the staff in the evaluation of submittals in the following categories and as an acceptable means of complying with the Commission's regulations described in Section A:

1. Construction Permit applicants that choose to make use of the provisions of the rule that allows the use of realistic models as an alternative to the features of Appendix K of 10 CFR Part 50.

2. Operating License applicants that choose to make use of the provisions of the rule that allows the use of realistic models as an alternative to the features of Appendix K of 10 CFR Part 50.

3. Operating Reactor Licensees will not be evaluated against the provisions of this guide except for new submittals which make use of the provisions of the rule that allow the use of realistic models as an alternative to the features of Appendix K of 10 CFR Part 50.

Page 6 ENCLOSURE F

E. VALUE/ IMPACT ANALYSIS A value/ impact analysis has not been prepared in support of this regulatory guide. Such a study was performed as part of the regulatory analysis which supports the rulemaking effort. This study is available through the NRC Public Document Room.

i l

l l

Page 7 ENCLOSURE F l

l i

APPENDIX DESCRIPTION OF BEST ESTIMATE CALCULATIONS OF EMERGENCY CORE COOLING SYSTEMS PERFORMANCd AND ESTIMATION OF THE CALCULATIONAL UNCERTAINTY

1. BEST ESTIMATE CALCULATIONS 1.1 General A best-estimate calculation uses modeling that attempts to realistically describe the physical processes in a nuclear reactor.

There is no unique approach to the extremely complex modeling of the processes occurring during a loss-of-coolant accident (LOCA). The NRC has developed and assessed several best-estimate, advanced thermal-hydraulic transient codes. These include TRAC-PWR, TRAC-BWR, RELAP5, COBRA, and the FRAP series of codes (References 1, 2, 3, 4, 5, and 6). These codes predict the major phenomena observed for a broad range of thermal-hydraulic and fuel tests reasonably well. Licensees and/or applicants may use, but are not limited to, these codes and/or specific models within them to perform best-estimate calculations of emergency core cooling system (ECCS) performance. However, since the NRC staff has not performed the plant-specific uncertainty analysis required by the proposed revision to 10 CFR 50. 6, the licensee must demonstrate that the code and/or models are acceptable and applicable to the specific facility over the intended operating range and must quantify the uncertainty in the specific application. General features expected in a best-estimate calculation are described in this section, and specific examples of features that are considered acceptable best-estimate models are given in section 1.2. Other models and/or correlations will be considered acceptable if their technical basis is demonstrated with appropriate data and analysis.

Page 8 ENCLOSURE F

A best-estimate model should provide a realistic calculation of a ,

particular phenomenon to the degree practi. cal with the currently available data and knowledge of the phenomenon. The model should be compared with applicable experimental data and should predict the mean of the data, rather than providing a bound to the data. The effects of all important variables should be considered. If it is not possible or practical to consider a particular phenomenon, the neglect of this phenomenon should not normally be treated by including a bias in the analysis directly, but should be included as part of the model uncertainty. The importance of neglecting a particular phenomenon should be considered within the overall calculational uncertaintv.

Careful consideration should be given to the range of applicability of the mods! when used in a best-estimate code. When comparing the model to data, judgments of the applicability of the data to the situation that would actually occur in a reactor should be made.

Correlations generally should not be extrapolated beyond the range over which they were developed or assessed. If the model is to be extrapolated beyond the conditions where valid data comparisons have been made, judgments should be made as to the effect of this extrapolation and should be considered in the uncertainty evaluation.

The use of fundamental laws of physics, well established data bases (e.g., steam tables), and sensitivity studies should be used to assist in the estimation of uncertainty that results from extrapolation.

A best-estimate code contains all the models necessary to predict the important phenomena that might occur during a loss-of-coolant accident. Best-estimate code calculations should be compared with applicable experimental data (e.g., separate effects tests and integral simulations of loss-of-coolant accidents) to determine the overall uncertainty and biases of the calculation. In addition to providing input to the uncertainty evaluation, integral simulation data comparisons should be used to ensure that important pheno"ena that are 4

Page 9 ENCLOSURE F

sxpected to occur during a loss-of-coolant accident are adequately calculated. The following paragraphs list some of the primary features that should be included in best-estimate thermal-hydraulic transient codes. In general, these features will have uncertainties associated with their use for predicting reactor system response. These uncertainties should be considered in the overall uncertainty analysis described in Section 2 of this Appendix.

The above discussion is an idealized definition of best-estimate.

In practice, best-estimate codes may contain certain models which are simplified and/or contain conservatism to some degree. This conservatism may be introduced for the following reasons:

1. The model simplification or conservatism has little effect on the result and, therefore, the development of a better model is not justified.

2. The uncertainty of a particular model is difficult to determine, and only an upper bound can be determined.

3. The particular appl.ication does not require a totally best-estimato calculation, so a bias in the calculation is acceptable.

The introduction of conservative bias or simplification in otherwise best-estimate codes should not, however, result in calculations which are nonphysical, unrealistic, do not include important phenomena, or contain bias and uncertainty that cannot be bounded. Therefore, any calculational procedure determined to be a best-estimate code in the context of this guide or for use under 10 CFR 50.46(a)(i) should be compared with applicable experimental data to ensure that realistic calculations of important phenomena result.

Page 10 ENCLOSURE F

1.2. Features of Best Estimate Codes **

Some features that are acceptable for use in best-estimate codes are described in the following paragraphs. flodels which address these features may be used with the basic requireert that a specific model is acceptable if it has been compared with applicable experimental data and shown to provide reasonable predictions. Reference 7, "Compendium of ECCS Research for Realistic LOCA Analysis," NUREG 1230, provides a summary of the large experimental data base available, upon which best-estimate models may be based. While inclusion in Reference 7 does not guarantee that the data or model will be acceptable, the report represents a large body of data generally applicable to best-estimate models. NUREG 1230 also provides documentation of NRC studies of the effect of reactor power increase on risk, background information on the ECCS rule, and a description of the NRC developed methodology for estimating thermal-hydraulic transient code uncertainty.

For any models or correlations used in a best-estimate code, sufficient justification must be provided to substantiate that the code performs adequately for the classes of transients to which it is applied. In general these features have uncertainties associated with their use for predicting reactor system response. These uncertainties should be considered as part of the overall uncertainty analysis described in Section 2 of th'is Appendix.

1.2.1 Basic Structure of Code 1.2.1.1 Numerical Methods A best-estimate code consists of a numerical scheme for solving the equations used to represent the various models. The numerical scheme is, in itself, a complex process which can play an important role in the overall calculation. Careful numerical modeling, sensitivity studies, and evaluations of numerical error should be performed to ensure that the results of the calculations are representative of the models used in the code. Numerical simulations of complex problems, such as those considered here, treat the geometry of the reactor in an approximate page 11 ENCLOSURE F

. . manner, making use of discrete volumes or nodes to represent the system.

Sensitivity studies and evaluations of the uncertainty introduced by noding should be performed. Numerical methods treat tiLa in a discrete manner, and the effect of time step size should also be investigated.

1.2.1.2 Computational Models A best-estimate code typically contains equations for conservation of mass, energy, and momentum of the reactor coolant and noncondensible gases, if important (e.g., air, nitrogen, etc). Energy equations are also used to calculate the temparature distribution in reactor system structures and in the fuel rods. The required complexity of these equations will vary depending on the phenomena that are to be calculated and the required accuracy of the calculation. NRC staff experience with its own best-estimate computer codes has indicated that separate flow fields for different fluid phases, or types, and calculation of nonequilibrium between phases may be required to calculate some important phenomena (e.g., countercurrent flow, reflood heat transfer) to an acceptable accuracy. NRC staff has also determined that certain phenomena require that the equations be solved in multiple dimensions.

However, one-dimensional approximations to three-dimensional phenomena will be considered acceptable if those approximations are properly justified. Other basic co:le features include equations of state and other material properties. Sensitivity studies and comparisons to data should be performed to determine the importance of the simplifications used.

1.2.2 Initial and Boundary Conditions and Equipment Availability The heat generated by the fuel during a loss-of-coolant accident depends on the power level of the reactor at the time of the loss-of-coolant accident and on the history of operation. The most limiting initial conditions expected over the life of the plant should be based upon sensitivity studies. It is not necessary to assume initial conditions which are impossible to occur in combination. For example, beginning-of-life peaking factors and end-of-life decay heat Page 12 ENCLOSURE F l-

may be unrealistic conditions and, therefore, would not require **

consideration. Given the assumed initia? conditions, relevant factors such as the actual total power, actual peaking factors, and actual fuel conditions should be calculated in a best-estimate manner.

Calculations should be performed that are representative of the spectrum of possible break sizes from the full double-ended break of the largest pipe down to a size small enough that it can be shown that smaller breaks are of less consequence than those already considered.

The analyses should also include the effects of longitudinal splits in the largest pipes, with the split area equal.to twice the cross-sectional area of the pipe. The detail of break sizes consid-ered should be sufficient so that the system response as a function of break size is defined well enough to confidently interpolate betwer calculations, without unexpected behavior between the break sizes considered.

Other boundary and initial conditions and equipment availability should be based upon plant technical specification limits. These other conditions include, but may not be limited to, availability and performance of equipment, automatic controls, and operator actions.

Appendix A to 10 CFR 50 requires that a single failure be considered when analyzing safety system performance and that the analysis consider the effect of using only onsite power and only offsite power.

1.2.3 Sources of Heat During a Loss-of-Coolant Accident Models should account for the sources of heat discussed below and the distribution of heat production.

1.2.3.1 Initial Stored Energy of the Fuel The steady state temperature distribution and stored energy in the fuel before the postulated accident should be calculated in a best-1 i

estimate manner for the assumed initial conditions, fuel conditions, and l

l l

\ ~

page 13 ENCLOSURE F l

l i

l.

operating history. Toaccomklishthis,thethermalcon.ductivityofthe fuel pellets and the thermal conductance of the gap between the fuel pellet and the cladding should be evaluated. Thermal conductivity of fuel is a function of temperature and is degraded by the presence of gases in crack voids between fuel fragments. An acceptable model for thermal conductivity should be developed from the in-pile test results for fuel centerline and off-center temperatures, taking into account the conductivity of gases in crack voids.

Thermal conductance of the fuel-cladding gap is a strong function of hot gap size and of the composition and pressure of the gases in the fuel rod. The calculation of hot gap size should take into account U02 or mixed oxide fuel swelling, densification, creep, thermal expansion and relocation, and cladding creep. Fuel swelling is.a function of temperature and burn-up. Fuel densification is a function of burn-up, temperature, and initial density. Densification can result from hydrostatic stresses imposed on fuel during pellet-cladding mechanical interaction and should be considered. Fuel creep is a function of time, temperature, grain size, density, fission rate, oxygen-to-metal ratio and external stress. Fuel thermal expansion represents dimensional changes in unirradiated' fuel pellets caused by changes in temperature.

An acceptable model for fuel swelling should be based on in-pile and out-of-pile test data. Cladding creep introduces compressive creep strain in cladding during steady-state operation, reducing the gap between the fuel pellet and cladding. Cladding creep is a function of fast neutron flux (>1 Mev), cladding temperature, hoop stress, and material. Cladding materials may be cold-worked and stress-relieved or fully recrystallized, and there is a significant difference in the magnitude of creepdown between these materials. During pellet cladding mechanical interaction, cladding experiences deformation due to tensile creep, which is significantly different from that due to compressive creep. An acceptable model for cladding tensile creep should be based on in-reactor tensile creep data.

Best-estimate fuel models will be considered acceptable provided that the models include essential phenomena identified above and their technical basis is demonstrated with appropriate data and analysis.

page 14 ENCLOSURE F

The'model evaluation procedure and data presented in this section ,

would be considered acceptable for assessing a model used to calculate stored energy and heat transfer in fuel rods.

1.2.3.1.1 Model Evaluation Procedure f 1

A model to be used in ECCS evaluations to calculate internal fuel l rod heat transfer should:

a. be checked against several sets of relevant data

b. recognize the effects of fuel burnup, fuel pellet cracking, fuel pellet relocation, cladding creep, and gas mixture conductivity.

The model described by Lanning (Ref. 8) compared well to inpile fuel temperature data. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analysis.

1.2.3.1.2 Cxperimental Oata The correlations and data of Reference 9 would be considered acceptable for calculating the initial stored energy of the fuel and subsequent heat transfer.

1.2.3.2 Fission Heat Fission heat should be included in the calculation and should be calculated using best-estimate reactivity and reactor kinetics calculations. Shutdown reactivities resulting from temperatures and The point voids should also be calculated in a best-estimate manner.

kinetics formulation is considered an acceptable best-estimate method for determining fission heat in loss-of-coolant accident safety calculations. Other best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses. Control rod assembly insertion may be assumed if it is expected to occur.

page 15 ENCLOSURE F

1.2.3.3 Decay of Actinides The heat from radioactive decay of actinides, including neptunium and plutonium generated during operation, as well as isotopes of uranium, should be calculated in accordance with fuel cycle calculations and known radioactive properties. The actinide decay heat chosen should be that appropriate for the operating history. Best-estimate models will be concidered acceptab.e provided their technical basis is demonstrated with appropriate data and analysis, 1.2.3.4 Fission Product Decay Heat The heat generation rates from radioactive decay of fission products, including the effects of neutron capture, should be included in the calculation and should be calculated in a best-estimate manner.

The energy release per fission (Q value) should also be calculated in a best-estimate manner. Best-estimate methods will be considered acceptable provided that their technical basis is demonstrated with appropriate data and analysis. The model of Reference 10 would be considered acceptable for calculating fission product decay heat.

1.2.3.4.1 Model Evaluation Procedure The values of mean energy per fission (Q) and the models for actinide decay heat should be checked against a set of relevant data.

1.2.3.5 Metal-Water Reaction Rate The rate of energy release, hydrogen generation, and cladding oxidation from the reaction of the zircaloy cladding with steam should be included in the calculation in a best-estimate manner. Best-estimate models will be considered acceptable provided that their technical basis is demonstrated with appropriate data and analysis. For rods for which cladding is calculated to rupture during the loss-of-coolant accident, the oxidation of the inside of the cladding should be included in the Page 16 ENCLOSURE F l

calculation in a best-estimate manner. The following model evaluation procedures and data would be considered acceptable for assessing a model used to calculate hydrogen generation and cladding oxidation.

1.2.3.5.1 Model Evaluation precedure Correlations to be used to calculate metal-water reaction rates for the temperature range which are less than or equal to 1900 F temperature range should:

a. be checked against a set of relevant data

b. recognize the effects of steam pressure, preoxidation of the cladding, deformation during oxidation, and internal oxidation from both steam and b0 2 I"'I' The data of Reference 11 would be considered acceptable for-calculating the rates of energy release, hydrogen generation, and cladding oxidation for cladding temperatures greater than 1900 F.

1.2.3.6 Reactor Internals Heat Transfer Heat transfer from piping, vessel walls, and internal hardware should be included in the calculation and should be calculated in a best-estimate manner. Heat transfer to channel boxes, control rods, guide tubes, and other in-core hardware should also be considered.

Models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analysis.

1.2.3.7 Primary To Secondary Heat Transfer (not applicable to boiling water reactors)

Heat transferred between the primary and secondary systems through the steam generators should be considered in the calculation and should be calculated in a best-estimate manner. Models will be considered page 17 ENCLOSURE F

f. .

e

.3 acceptable provided that their technical basis is demonstrated with' appropriate data and analysis.

1.2.4' Reactor Core Thermal / Physical parameters 1.2.4.1 Swellina and Rupture of the Claddina and Fuel Rod Thermal Parameters A calculation.of the swelling and rupture of the cladding resulting from the temperature distribution in the cladding and from the pressure difference between the inside and outside of the cladding, both as a function of time, should be included in the analysis and should be performed in.a best-estimate manner. The degree of swelling and-rupture should be taken into account in calculations of gap conductance, cladding oxidation and embrittlement, and hydrogen generation and in calculating heat transfer and fluid flow outside of the cladding. The calculations of fuel and cladding temperatures as a function of time should use values of gap conductance and other thermal parameters as functions of temperature and time. Acceptable beet-estimate methods to calculate the swelling of the cladding should take into account spatially varying cladding temperatures, heating rates, anisotropic material properties, asymmetric deformation of cladding, and fuel rod thermal and mechanical parameters. Best-estimate methods will be considered acceptable provided that their technical basis is demonstrated with appropriate data and analysis.

1.2.4.2 Other Core Thermal parameters As necessary and appropriate, physical / chemical changes in in-core materials (e.g. , eutectic formation, phase change, or other phenomena caused by material interaction) should be accounted for in the reactor core thermal analysis. Best-estimato models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analysis.

Page 13 ENCLOSURE F

1.2.5 Blowdown Phenomena .

1.2.5.1 Break Characteristics and Flow In analyses of hypothetical loss-of-coolant accidents, a spectrum of possible break sizes should be considered, as indicated in Paragraph 1.2.2 above. The discharge flow rate should be calculated with a critical flow rate model that considers the fluid conditions at the break location, upstream and downstream pressures, and break geometry. The critical flow model should be justified by comparison to applicable experimental data over a range of conditions for which the model is applied. The model should be a best-estimate calculation, with uncertainty in the critical flow rate included as part of the uncertainty evaluation. Best-estimate models will be considered acceptable provided that their technical basis is demonstrated with appropriate data and analyses.

The model evaluation procedure and data discussed below would be considered acceptable for assessing a model or correlation used to calculate the discharge flow rate during a loss-of-coolant accident.

1.2.5.1.1. Model Evaluation Procedure Critical flow models to be employed in ECCS eval'uations should:

a. be checked against an acceptable set of relevant data

b. recognize thermal nonequilibrium conditions when the fluid is subcooled

c. provide a means of transition from nonequilibrium to equilibrium conditions.

The uncertainties and bias of a correlation or model used to calculate critical flow should be stated, as well as their range of l

applicability.

l l

Page 19 ENCLOSURE F l

L

e

^

The mechanistic thermal nonequilibrium and slip model of Richter (Ref. 23) compares well to small- and large-scale test data (Ref. 24).

1.2.5.1.2 Exrerimental Data An acceptable set of relevant critical flow data should cover the fluid conditions, geometries, and types of breaks pertinent to light water reactor loss-of-coolant accidents. The following tests should be considered in establishing an acceptable set of relevant data:

o Marviken tests (Ref. 12) o Moby Dick experiments (Ref. 13) o BNL critical flashing flows in nozzles (Ref. 14) o Sozzi-Sutherland tests (Ref.15) o Edwards experiments (Ref. 16) o Super Moby Dick experiments (Ref. 17 and 18)

For critical flow from small breaks under stratified conditions, currently acceptable test data for assessing models and codes include those reported by:

o Anderson and Owca (Ref. 19) o Reimann and Khan (Ref. 20) o Schrock et al (Ref. 21 and 22) 1.2.5.2 ECCS Bypass The best-estimate code should contain a calculation of the amount of injected cooling water that bypasses the vessel during the blowdown phase of the loss-of-coolant accident. The calculation of ECCS bypass shou be a best-estimate calculation using analyses and comparisons wi ' ?plicable experimental data. Although it is clear that the dominant processes governing ECC bypass are multidimensional, single dimensional approximations justified through sufficient analysis and i

data may be acceptable. Best-estimate methods will be considered 4

-l Page 20 ENCLOSURE F

acceptable provided that their technical. basis is demonstrated with appropriate data and analysis. Cooling water that is not expelled, but remains in piping or is stored in parts of the vessel, should be calculated in a best-estimate manner based on applicable experimental data.

The model evaluation procedure and data discussed below would be considered acceptable for assessing a model or correlation used to calculate ECC bypass during the blowdown phase of a loss-of-coolant accident.

1.2.5.2.1 Model Evaluation Procedure A correlation or model to be used in ECCS evaluations to calculate ECC bypass should:

a. be checked against an acceptable set of relevant data

b. recognize the effects of pressure, liquid subcooling, fluid conditions, hot walls, and system geometry.

Uncertainties and bias in the correlations or models used to calculate ECC bypass should be stated, as well as the range of their applicability.

For scaled-down PWR downcomers, correlations by Beckner and Reyes (Ref. 30) compared well to the bypass data of References 25 and 26.

Correlations of Sun (Ref. 31) and Jones (Ref. 32) compare well to counter-current flow limiting (CCFL) test data of interest to BRWs.

1.2.5.2.2 Experimental Data The following test should be considered in establishing a set of data for scaled-down PWR downcomers:

BCL test (Ref. 25)

CREARE test (Refs. 25 and 26)

Page 21 ENCLOSURE F

For a full-scale PWR vessel, ECC bypass data will become available from the forthcoming upper plenum test facility (UPTF) experiments performed as part of the 20/3D program sponsored by the Federal Republic of Germany, Japan, and the United States.

For BWRs, the following test should be considered in establishing an acceptable set of relevant data:

SSTF test data (Refs. 27 through 29) 1.2.6 Noding Near the Break and ECCS Injection Point The break location and ECCS injection point are areas of high fluid velocity and complex fluid flow and contain phenomena that are often difficult to calculate. The results of these calculations are often highly dependent on the noding. Sufficient sensitivity studies should be performed on the noding and other important parameters to ensure that the calculations provide realistic results.

1.2.7 Frictional pressure Drop The frictional losses in pipes and other components should be calculated using models that include variation of friction factor with Reynolds number and account for two-phase flow effects on friction.

Best-estimate models will be considered acceptable provided their '

technical basis is demonstrated with appropriate data and analysis.

1.2.7.1 Model Evaluation Procedure A model for frictional pressure drop to be used in ECCS evaluation should:

a. be checked against a set of relevant data Page 22 ENCLOSURE F

b. be consistent with models used for calculating gravitational and acceleration pressure drops. If void fraction models or correlations used to calculate the three components of the total pressure drop differ one from another, a quantitative justification must be provided.

Uncertainties and bias of a correlation or model should be stated as well as the range of applicability.

1.2.7.2 Experimental Data An acceptable set of relevant data should cover, as far as possible, the ranges of parameters (mass flux, quality, pressure, fluid physical properties, roughness, and geometries) that are found in actual plant applications. The following tests should be considered in establishing an acceptable set of rele.vant data:

Vertical tubes Cambridge tests (Ref. 33)

CISE test (Ref. 34 and 35)

HTFS data bank (Ref. 36)

Horizontal tubes Ca'mbridge test (Ref. 33)

HTFS data bank (Ref. 36)

GE tests (Ref. 37 ard 38) l Rod bundles l

l HTFS data bank (Ref. 36)

GE tests (Ref. 39) 1.2.8 Momentum Ecuation The following effects should be taken into account in the two-phase conservation of momentum equation: (1) temporal change in momentum, (2) momentum convection, (3) area change momentum flux, (4) momentum change page 23 ENCLOSURE F

due to compressibility, (5) pressure loss resulting from wall' friction, (6) pressure loss resulting from area change, and (7) gravitational acceleration. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

1.2.9 C,ritical Heat Flux Best-estimate models developed from appropriate steady-state or transient experimental data should be used in calculating critical heat flux (CHF) during loss-of-coolant accidents. The codes in which these models are used should contain suitable checks to assure that the range of conditions over which these correlations are used are within those intended. Research has shown that CHF is highly dependent on the fuel rod geometry, local heat flux, and fluid conditions. After CHF is predicted at an axial fLd1 rod location, the calcul6 tion may use nucleate boiling heat transfer correlations, if the calculated local fluid and surface conditions iustify the reestablishment of nucleate boiling. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analyses.

1.2.10 Post-CHF Blowdown Heat Transfer Models of heat transfer from the fuel to the uurrounding fluid in the post-CHF regimes of transition and film boiling thould be best-estimate models based on comparison to applicable steady-state or transient data. Any model should be evaluated to demonstrate that it provides acceptable results over the applicable ranges. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analysis.

The model evaluation procedure and data discussed below would be considered acceptable for assessing a model or correlation used to calculate the heat transfer from the fuel to the surrounding fluid while in the post-CHF regime.

Page 24 ENCLOSURE F

%r, ,

• l 1.2.10.1 Model Evaluation Procedure l A model to be used in ECCS evaluation to calculate post-CHF heat transfer from rod bundles should:

a. be checked against an acceptable set of relevant data

b. recognize effects of liquid entrainment, thermal radiation, thermal nonequilibrium, low and high mass flow rates, low and high power densities, and saturated and subcooled inlet r conditions.

The uncertainties and bias of models or correlations used to calculate post-CHF heat transfer should be stated as well as the range of their applicability.

1.2.10.2 Experimental Data The acceptable set of relevant data should cover power densities, mass flow rates, fluid conditions, and rod bundle geometries pertinent to light-water reactor designs and applications. The following tests should be considered in establishing an acceptable set of relevant data:

o ORNL tests (Ref. 40 and 41) o INEL tests (Ref. 42) o ORNL data bank (Ref. 43) 1.2.10.3 Post-CHF Heat Transfer from Uncovered Bundles During some time periods of small break LOCAs and during portions of large breaks prior to reflood, partial or complete core uncovering may be calculated to occur. During these time periods, very little liquid and low flow usually exist in the uncovered regions. Special considerations are therefore appropriate.

Page 25 ENCLOSURE F a

1.2.10.3.1 Model Evaluation procedures A correlation to be used in ECCS evaluations to calculate heat transfer from uncovered rod bundles should:

a. ~be checked against against an acceptable set of relevant data

b. recognize the effects of radiation and laminar, transition, and turbulent flows.

Uncertainties and bias in the models and/or correlations used to calculate post-CHF heat transfer should be stated as should the range of their applicability.

The correlation derived should include a stated procedure for correcting for radiative heat transfer and for estimating the vapor temperatures. The Hottel procedure, cited in Reference 9 is a satisfactory example.

The turbulent correlation may be of the general form:

Nu = A Re* Pr" for higher Reynolds numbers, where the coefficients A, m and n are modifications from the basic Dittus-Boelter form and may be functions of other variables. The physical properties may be defined as wall, film, or vapor values.

A distinction from, and transition to, laminar convection (i.e.,

Re< 2000) should be made, with a value of the laminar heat transfer for rod bundles that is appropriate for the applicable bundle geometry and flow conditions.

l Page 26 ENCLOSURE F

Other forms and values, depending on the bundle geometry and flow conditions are also appropriate.

1.2.10.3.2 Experimental Data An acceptable set of relevant data for post-CHF heat transfer from uncovered rod bundles should cover power densities, fluid conditions, and rod bundle geometries pertinent to light-water reactor design and application. The following tests should be considered in establishing an acceptable set of relevant data:

o ORNL-THTF tests (Ref. 44 and 45) o ORNL Data Base (Ref. 46) 1.2.11 Pump Modeling The characteristics of rotating primary system pumps should be derived from a best-estimate dynamic model that includes momentum transfer between the fluid and the rotating member, with variable pump speed as a function of time. The pump model resistance and othcr empirical terms should be justified through compari:ons with appliccble data. The pump model for the two-phase region should be verified by comparison to apolicable two-phase performance data. Pump coastdown following loss of power should be treated in a best-estimate manner. A locked rotor following a large break loss-of-coolant accident neet, not be assumed unless it is calculated to occur. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analysis.

1.2.12 Core Flow Distribution During Blowdown The core flow through the hottest region of the core during the For the purpose of blowdown should be calculated as a function of time.

these calculations, the hottest region of the core should not be greater than the size of one fuel assembly. Calculations of the flow in the hot i

Page 27 ENCLOSURE F

- w ,- .,n.- , , . . , . , , - . , . , . , , - . - . . _ , -

region should take into account any crossflow between regions and any flow blockage calculated to occur during the blowdown as a result of cladding swelling or rupture. The numerical scheme should ensure that unrealistic oscillations of the calculated flow do not result.

Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analysis.

1.2.13 Post-Blowdown Phenomena 1.2.13.1 Containment pressure The containment pressure used for evaluating cooling effectiveness during the post-blowdown phase of a loss-of-coolant accident should be calculated in a best-estimate manner and should include the effects of containment heat sinks. The calculation should include the effects of operation of all pressure reducing equipment assumed to be available, as discussed in Paragraph 1.2.2 of this Appendix. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analysis.

1.2.13.2 Calculation of Post-Blowdown Thermal-Hydraulics for Pressurized Water Reactors The refilling of the reactor vessel and the ultimate reflooding of the core should be calenlated by a best-estima*e model that takes into consideration the ther.nal and hydraulic characteristics of the core, the emergency core cooling systems, and the primary and secondary reactor system. The models should be capable of calculating the two-phase level in the reactor during the postulated transit Best-estimate models will be considered acceptabla provided their technical basis is demonstrated with appropriate data and analysis.

1.2.13.2.1 Model Evaluation procedures A correlation or model to be used in ECCS evaluation to calculate level swell should be checked against an acceptable set of relevant Page 28 ENCLOSURE F ,

' .o S- data, and shsuld recognize the effects of depressurization, boil-off, power icvel, fluid conditions and system geometry.

The correlation proposed by Chexal, Horowitz, and Le11outhe (Ref. 52) provides acceptable results when compared to experimental data reported in References 44, 48, 49, and 51.

Uncertainties and bias of a correD non or model used to calculate level swell should be stated, as should the range of applicability.

The primary coolant pumps should be assumed to be operating in the expected manner, based on the assumptions of Paragrap5 1.2.11, when calculating the resistance offered by he pumps to fluid flow. Models will be considered acceptable provided that their technical basis is demonstrated through comparison with appropriate data and analysis.

The total fluid flow leaving the core exit (carryover) should be calculated using a best-estimate model which includes the effect of crossflow on carryover and core fluid distribution. Thermal-hydraulic phenomena associated with unique emergency core cooling systems, such as upper plenum injectior, nnd upper head injection, should be accounted for. The effects of the compressed gas in the accumulator, which is discharged, following accumulator water discharge should be included in the calculation. Any model or code used for this calculation should be assessed against applicable experimental data. Reference 7 describes a large body of refill /reflood thermal hydraulic dt:a obtained from the 20/30 program which is appron-iate for consideration 1.2.13.2.2 Experimental Data The following test should be considered when establishing an acceptable set of relevant data:

o GE tests (Ref. 47 and 48) o ORNL tests (Ref. 44 and 49)-

o FLECHT-SEAFET test (Ref. 50) o THETIS tests (Ref. 51)

' Page 29 ENCLOSURE F

t*

1.2.13.3 Steam Interaction With Emergency Core Cooling Water in Pressurized Water Reactors The thermal-hydraulic interaction between the steam or two-phase fluid and the emergency core cooling water should be taken into account in calculating the core thermal-hydraulics and the steam flow through the reactor coolant pipes dur,ing the time that the accumulators are  ;

discharging water. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analysis.

1.2.13.4 post-Blowdown Heat Transfer for pressurized Water Reactors During refilling of the reactor vessel and ultimat. aflooding of the core, the heat transfer should be based on a best-es' aate calculation of the fluid flow through the core, accounting for unique emergency core cooling systems. The calculations should also include the effects of any fi w blockage calculated to occur as a result of cladding swelling or rupture. Heat transfer calculations, which account for two-phase conditions in the core during refilling of the reactor vessel, should be justified through comparisons with experimental data. ,

Best-estimate models will be considered acceptable provided that their technical basis is demonstratwd through comrarison witn appropriate data and analysis.

The FLECHT-SEASET tests (Ref. 53, 54, 55) should be considered when establishing an acceptable set of relevant data. Reference 7 contains extensive information regarding a large amount of experimental reflood heat transfer data. This information should also be considered when developing and assessing models. In particular, the results from the .

2D/30 are particularly relevant.

1.2.14 Convective Heat Transfer Coefficients for Boiling Water Reactor Rods Under Spray Cooling Models will be conside*ed provided their technical basis can be justified with appropriate data and analysis.

Page 30 ENCLOSURE F

• l Following the blowdown period, convective heat transfer coefficients should be determined based on the calculated fluid conditions and heat transfer modes within the bundle and on the calculated rod temperatures.

During the period following the flashing of the lower plenum fluid, but prior to ECCS initiation, heat transfer models should include cooling by steam flow or by a two-phase mixture, if calculated to occur.

Following initiation of ECCS flow, but prior to refiroding, heat transfer should be based on the actual calculated bundle fluid conditions and best-estimate heat transfer models which take into account rod-to-rod variations in heat transfer.

After the two-phase reflood level reaches the level under consideration, a'best-estimate heat transfer model should be used. This model should include the effects of any flow blockage calculated to occur as a result of cladding swelling or rupture.

Thermal-hydraulic models which do not calculate multiple channel effects should be ccmpared with applicable experimental data or more detailed calculations to ensure that all important phenomena are adequately calculated.

1.2.15 The Boiling Water Reactor Channel Box Under Spray Cooling Following the 'olowdown period, heat transfer from the channel box and wetting of the channel box should be based on the calculated fluid conditions on both sides of the channel box and should make use of best-estimate heat transfer and rewetting models that have been compared with applicable experimental data. Best-estimate models will be considered acceptable provided their technical basis is demonstrated with appropriate data and analysis.

.Page 31 ENCLOSURE F

!+

1.2.16 Special Considerations for a small-Break Loss-of-Coolant Accident in pressurized Water Reactors The slower small-break loss-of-cor' ant accident leads to fluid conditions characterized by separation of the fiuid phases versus the more homogeneous fluid conditions that would result from rapid large-break loss-of-coolant accident transients. Phenomena that would occur in a PWR during a small-break loss-of-coolant accident would, therefore, be significantly different from those phenomena that would occur during a large-break locs-of-coolant accident. The distribution of liquid throughout the reactor system, in addition to the total liquid inventory, is of increased importance for the small-break loss-of-coolant accident. A number of special factors must be given increased consideration in small-break loss-of-coolant accident calculations to correctly predict phenomena influenced by the liquid inventory distribution.

Break flow may be greatly influenced by the location and specific geometry of the break. For a break in a horizontal pipe containing stratified flow, the quality of the break flow will be a strong function of the assumed location of the break on the pipe (e.g., top or bottom).

Small-break loss-of-coolant accident calculations should, therefore, include various assumed break locations in the spectrum of breaks analyzed. The assumed operating state of the reactor coolant pump will also influence the distribution of liquid throughout the system and the amount of liquid lost through the break.

The pump operation assumptions used in the calculations should be the most likely, based on operating procedures, with appropriate consideration of the uncertainty of the pump operation during an actual event. Level depression in the core regien and subsequent core heatup may be influenced by liquid holdup in the steam generator tubes, manometric effects of liquid in the piping and loop seal region, and Page 32 ENCLOSUP.E F

l liquid levels relative to vent paths for steam through upper plenum bypass flow paths and vent values. Steam generator heat transfer under i

"reflux" or "boiler-condensor" modes of operation may also strongly influence core inventory through level depression and the effect on total system pressure and, thus on ECCS flow. These phenomena should be carefully considered in the calculation, as should sensitivity studies of the importance of these effects for use in the uncertainty evaluation.

Heat transfer from an uncovered core under high pressure i

conditions typical during a small-break loss-of-coolant accident may include contributions from both convective and radiation heat transfer ,

to the steam. Models will be considered acceptable provided that their technical basis is demonstrated through comparison with appropriate data and analysis. Specific guidsnce regarding uncovered bundle heat transfer is given in Section 1.2.10.3.

1.2.17 Other Features of Best Estimate Codes No list of best-estimate code features could be all-inclusive, because the iraurtant features of a best-estimate code may vary depending on the transient to be calculated and the required accuracy of the calculation. Because of this, no attempt has been made to construct an exhaustive list of best-estimate code features. Rather, features which were identified as important for inclusion in Appendix K were used as a basis for the above list. These features are not necessarily any more or loss important than other code features, but were highlighted because it is necessary to give specific examples of how current best-estimate models may vary from methoas used traditionally in evaluation model codes using the various. Appendix K conservatisms. In addition, models have not been included in areas where it was felt that the best '

model would be highly dependent on the specific plant design or the specific transient under consideration.

Page 33 ENCLOSURE F F

The NRC staff believes that good examples of best-estimate thermal-hydraulic transient codes are those developed by the NRC (e.g., TRAC-PWR,-TRAC-BWR, RELAP5, COBRA and FRAP). Although these codes are subject to further improvement, based on their ongoing use and assessment, they currently provide reasonable best-estimate calculations of the loss-of-coolant accident in a full-scale light-water reactor. This is substantiated through the code development and assessment literature generated by the NRC and its contractors over the past several years.

It is possible. however, to generally describe how other features of best-estimate codes should be constructed. Two basic criteria should be applied:

1.2.17.1 Completeness -

Best-estimate codes should contain models in sufficient detail to predict phenomena that are important to the desired result of the calculation (e.g., peak cladding temperature). S1mplifications are acceptable as long'as code uncertainties or bias do not become so large that they cast doubt on the actual behavior that would occur or on the true effect of assumed initial and boundary conditions (e.g., equipment sizing, safety system settings, etc.). Coroparisons of the overall calculations to integral experiments should be performed to ensure that important phenomena can be predicted and to help in making judgments on the effect of code ;i.mplifications. Consideration should also be given to the uncertainty and validity of the experiment to ensure that meaningful comparisons are being made.

1.2.17.2 Data Comparisons Individual best-estimate models should be compared to applicable experimental data to ensure that realistic results are predicted and

\

Page 34 ENCLOSURE F

that relevant experimental variables are included. Uncertainty analyses are required to ensure that a major bias does not exist in the models and that the model uncertainty is small enough to provide a realistic estimate of the effect of important experimental variables. Uncertainty analyses should also consider experimental uncertainty to ensure that meaningful comparisons are being made.

2. Estimation of Overall Calculatio7 Uncertainty 2.1 General The term "uncertainty," when applied to best-estimate thermal-hydraulic transient codes, is used at two levels. At the lower, or more detailed level, the term refers to the degree to which an individual model, correlation, or method used within the code represents the physical phenomenon which it addresses. These individual uncertainties, when taken together, comprise the "code uncertainty."

The combined uncertainty associated with individual models (i.e.,

code uncertainty) within the best-estimate codes does not account for all of the uncertainty associated with the model's use. In addition to the code uncertainty, various other sources of uncertainty are introduced when attempting to use best-estimate codes to predict full-scale plant thermal hydraulic response. These include uncertainty associated with the experimental data used in the code assessment process (including applicability of the data to full-scale reactors),

the input boundary and initial conditions, and the fuel behavior.

Additional sources of uncertainty stem from the use of simplifying assumptions and approximations. A careful statement of these assumptions and approximations should be made, and the uncertainty associated with them should be taken into account. Therefore, the "overall calculational uncertainty" is defined as the uncerteinty arrived at when all the contributions from the sources identified above, including the code uncertainty, are taken into account.

Page 35 ENCLOSURE F

~ '

'm-' -- , _,___ ___. , _ _

A 95% probability level is considered acceptable to the NRC staff for comparison of best-estinate predictions to the applicable limits of 9 50.46(b) to meet the requirement of 9 50.46(a)(1)(i) to show that there is a high probability that the criteria will not be exceeded.

The basis for selecting the 95% probability level is primarily for consistency with standard engineering practice in regulatory matters involving thermal-hydraulics. Many parameters, most notably the departure from nucleate boiling ratio (DNBR), have been found acceptable at the 95% probability level by the staff in the past.

This 95% probability level would also be applied to small-break loss-of-coolant accidents, which have a higher probability than large breaks. The dominant factors influencing risk from small-break loss-of-coolaat accidents include equipment availability and operator actions. Calculational uncertainties are much less important than factors such as operator recognition of the event, the availability of equipment, and the correct use of this equipment. The use of a best-estimate calculation with reasonable and quantifiable uncertainty is expected to provide a reduction in the overall risk from a small-break loss-of-coolant accident by providing more realistic cal.ulations with which to evaluate operator guidelines and determine the true effect of equipment availability.

(

This section provides a description of the features that should be included in the overall code uncertainty evaluation that is called for in 10 CFR 50.46(a)(1). This uncertainty evaluation should make use of probabalistic and statistical methods to determine the code uncertainty bias (if any). For a calculation of this complexity, a

- completely rigorous mathematical treatment is neither practical ner required. In many cases, approximations and assumpt' ns may be made to make the overall calculational uncertainty evalua, tion possible. A careful statement of these assumptions and approximations should be made so that the NRC staff may make a judgement as to the validity of Page 36 ENCLOSURE F

the uncertainty evaluation. The purpose of the uncertainty evaluation is to provide assurance that for postulated loss-of-coolant accidents a given plant will not, with a probability of 95% or more, exceed the applicable limits specified in 10 CFR 50.46(b).

2.2 Code Uncertainty This regulatory guide makes a distinction betwoen the terms "code uncertainty" and "overall calculational uncertainty." The latter term is defined in Section 2.1 of this Appendix and includes the contributions to the uncertainty described in Sections 2.2 and 2.3. The features of the code uncertainty (i.e., the contribution to the overall uncertainty due to the models and numerical methods used) are described in this section.

Code uncertainty should be evaluated through direct data comparison with relevant integral systems and separate effects experiments at different scales. In this manner, an estimate of the uncertainty attributable to the combined effect of the models and correlations within the code can be obtained for all scales and for different phenomena. Comparison to a sufficient number of integrai systems experiments, from different test facilities and different scales, should be made to assure that a reasonable estimate of code uncertainty and bias has been obtained. Where necessary, separate effects experiments should be used to establish code uncertainty for specific phenomena (e.g., comparisons to Cylindrical Core Test Facility data to ascertain code uncertainty in modeling upper plenum injection performance). Code comparisons shou:d account for limitations of the measurements and calibration errors.

These comparisons should be performed for important key parameters to demonstrate the overall best-estimate capability of the code. For large-break loss-of-coolant accidents, the most important key parameter page 37 ENCLOSURE F

5-is. peak cladding temperature, which is addressed by one of the criteria of 10 CFR 50.46(b) and has a direct influence on the other criteria. In addition, a code uncertainty evaluation should be performed for other important parameter (s) for the transient of interest to evaluate compensating errors. For small-break loss-of-coolant accidents, the cladding temperature response is the most important parameter; however, the ability of the codes to predict overall system mass and reactor vessel inventory distribution should also be statistically examined.

In evaluating the code uncertainty, it will be necessary to evaluate the code's predictive ability over several time intervals, since different processes and phenomena occur at different intervals.

For example, in large-break loss-of-coolant accident evaluations, separate code uncertainties may be required for the peak cladding temperature during the blowdown and post-blowdown phases. Justification for treating these uncertainties individually or methods for combining them should be provided.

The experimental information used to determine code uncertainty will usually be obtained from facilities that are much smaller than nuclear power reactors. Applicability of these results should be justified for larger scales. The effects of scale can be assessed through comparisons to available large-scale separate effects tests and through comparison to integral tests from various sized facilities. If there are scaling problems, particularly if predictions are nonconservative, the code should be improved for large scale plants (i.e., nuclear plants). Codes not having scaling capability will not be acceptable if their predictions are nonconservative.

2.3 Other Sources of Uncertainty When a best-estimate methodology is used to predict reactor transients, sources of uncertainty other than the limitations in the Page 38 ENCLOSURE F l

l

individual models and numerical methods (i.e., code uncertainty) are introduced. The foliowing contributors to the overall calculation:1 uncertainty should also be considered in the uncertainty analysis.

2.3.1 Initial and Boundary conditions and Equipment Availability When a plant input model is prepared, certain relationships describing the plant boundary and initial conditions t.nd the availability and performance of equipment are defined. These include factors such as initial power level, pump performance, valve activation times, and control systems functioning. Uncertainties associated with the boundary and initial conditions and the characterization and performance of equipment should be accounted for in the uncertainty evalua+. ion. It is also acceptable to limit the variables to be considerad by setting their values to conservative bounds.

2.3.2 Fuel Behavior Variability of the results of plant transient calculations can result from uncertainties associated with fuel behavior that is not included in the comparisons of code results with integral experiments, since most integral tests use electrical heater rods. This uncertainty includes many effects such as fuel conductivity, gap width, gap condue.tivity, and peaking factors. These uncertainties should be quantified and used in the determination of the overall calculational uncertainty.

2.3.3 Other Variables There may be individual models within the best-estimate code whose effect may not have been evaluated by the comparison to the integral systems data. For example, since most integral systems experiments use electrically heated rods, uncertainties associated with the prediction Page 39 ENCLOSURE F

of core decay heat and cladding metal-water reaction have not been evaluated. In addition, to demonstrate the overall adequacy of the predictive ability of the best-estimate code, it may be necessary to use empirical break discharge coefficients to obtain a reasonable break flow. The uncertainty in the indiviaual models, that have not been evaluated by comparison to integral systems data should be quantified and used in the determination of overall code uncertainty.

2.4. Statistical Treatment of Overall Calculational Uncertainty 1

The methodology used to obtain an estimate of the overall calculational uncertainty at the 95% probability limit should be provided and justified. If linear independence is assumed, suitable justification should be provided. The influence of the individual parameters on code uncertainty should be examined by making '

comp:risons to relevant experimental data. Justification should be provided for the assumed distribution of the parameter and the range considered.

In reality, the true statistical distribution for the key parameters (e.g., peak cladding temperature) is unknown. The choice of a statistical distribution should be verified using applicable engineering data and information. The statistical parameters appropriate for that distribution should be estimated using available data and results of engineering analyses. Supporting documentation should be provided for this selection process. These estimated values are assumed to be the true values of the statistical parameters of the distribution. With these assumptions, an upper one-sided probability limit can be calculated at the 95% level. As the probability limit approaches 2200 F, more care must be taken in the selection and justification of the statistical distribution and in the estimation of its statistical parameters. If a normal distribution is selected and justified, the probability limit can be conservatively calculated using two standard deviations. The added conservatism of the two standard deviations compared to the 95 percentile is used to account Page 40 ENCLOSURE F

for uncertainty in the probability distribution. Other techniques which account for the uncertainty in a more detailed manner may be used. These techniques may require the use of confidence levels which ,

are not required by the above approach.

The evaluation of the peak cladding temperature at the 95%

probability level need only be performed for the worst case break identified by the break spectrum analysis in order to demonstrate conformance with 10 CFR 50.46(b). However, in order to use this approach, justification which demoastrates that the overall calculational uncertainty for the worst case bounds the uncertainty for other breaks within the spectrum must be provided. It may be necessary to perform separate uncertainty evaluations for large- and small-break loss-of-coolant accidents due to the substantial difference in system thermal hydraulic behavior.

The revised S 50.46 (a)(1)(i) requires that it be shown with a high probability that all the criteria of f 50.46 (b) will not be exceeded, ,

not just the peak cladding temperature criterion. However, since the other criteria are strongly dependent on peak cladding temperature, explicit consideration of the probability of exceeding the other criteria may not be required if it can be demonstrated that meeting the temperature criterion at the 95% probability level ensures with a similar or greater probability that the other criteria will not be exceeded.

2.5 NRC Approach to LOCA Uncertainty Evaluation Chapter 4 of the "Compendium of ECCS Research for Realistic LOCA Analysis" (Ref. 7) presents a methodology,that is being considered for evaluating a best-estimate LOCA analysis code and determining the overall calculational uncertainty in peak cladding temperature r predictions.

I Page 41 ENCLOSURE F

ta References

1. Les Alamos National Laboratory "TRAC-PF1/M001: An Advanced Best Estimate Computer Program for Pressurized Water Reactor Thermal-Hydraulic Analysis," NUREG/CR-3858, July 1986.

2. Idaho National Engineering Laboratory, "TRAC-BD1/ MOD 1: An Advanced Best Estimate Computer Program for Boiling Water Reactor Transient Analysis," NUREG/CR-3633, April 1984.

3. Idaho National Engineering Laboratory, "RELAP5/ MOD 2 Code Manual,"

Vol. 1 & 2, NUREG/CR-4312, August 1985.

4. Pacific Northwest Laboratory, "COBRA / TRAC - A Thermal-Hydraulics Code for Transient Analysis of Nuclear Reactor Vessels and Primary Coolant Systems," NUREG/CR-3046, Vol. 1-5, March 1983.

5. L. J. Siefken et al., "FRAP-T6: A Computer Code for the Transient Analysis of Oxide Fuel Rods," NUREG/CR-2148, May 1981.

6. G. A. Berna et al., "FRAPCON-2: A Computer Code for the Calculation of Steady State Thermal-Mechanical Behavior of Oxide Fuel Rods," NUREG/CR-1845, December 1980.

7. "Compendium of ECCS Research for Realistic LOCA Analysis,"

NUREG-1230, (to be published).

8. Idaho National Engineering Laboratory, "MATPRO Version 11 (Revision 2): A Handbook of Materials Properties for Use in the Analysis of Light-Water Reactor Fuel Rod Sehavior," NUREG/CR-0497, Rev. 2, August 1981.

1 .

Page 42 ENCLOSURE F

2

9. D. Lanning and M. Cunningham, "Trends in Thermal Calculations for Light Water Reactor Fuel (1971-1981)," in Ninth Water Reactor Safety Research Information Meeting, USNRC, NUREG/CP-0024, March 1982,

10. American Nuclear Society, "American National Standard for Decay Heat Power in Light Water Reactors," ANSI /ANS 5.1-1979, (ANS 555 North Kensington Avenue, LaGrange Park, Illinois, 60525),

August 1979.

11. J. V. Cathcart and R. E. Pawel, et al., "Zirconium Metal-Water 0xidation Kinetics: IV Reaction Rate Studies," Oak Ridge l

National Laboratory, ORNL/NUREG-17, August 1977.

' 12. USNRC, "Marviken Full Scale Critical Flow Tests, Summary Report,"

(Joint Reactor Safety Experiments in the Marviken Power Station, Sweden),NUREG/CR-2671,1982.

13. M. Reocreux, "Contribution a l' etude des debits critiques en ecoulement diphasique eau-vareur," Ph.D. Thesis, L'Universite Scientifique Medicale de Grenoble, 1974. NRC Translation:

NUREG/TR-0002.

14. N. Abuaf, G. A. Zimmer, B. J. C. Wu, "A Study of Nonequilibrium Flashing of Water in a Converging-Diverging Nozzle," Brookhaven National Laboratory, NUREG/CR-1864, l'ols.1-2, March 1982.

15. G. L. Sozzi, and W. A. Sutherland, "Critical Flow of Saturated and Subcooled Water at High Pressure," General Electric Company, GE Report NE00-13418, 1975.

I s

Page 43 ENCLOSURE F

to

16. R. A. Edwards and T. P. O'Brien, "Studies of phenomena Connected with the Depressurization of Water Reactors," Journal of the British Nuclear Energy Society, Vol. 9, No. 2, April 1970.

17. C. Jeandey, et al., "Auto vaporization d'ecoulements eau /vapeur" Rapport TT, No 163, Centre d' Etudes Nucleaires de Grenobic, Grenoble, France, Juillet 1981. NRC translation: NUREG/IA 00010.

18. C. Jeandey and L. Gross d'Aillon, "Debit Critique en tuyere courte sur Super Moby Dick," Rapport TT/SETRE/71, Centre d' Etudes Nucleaires de Grenoble, Grenoble, France, September 1983. NRC Translation: NUREG/IA 00011.

I

19. J. L. Anderson and W. A. Owca, "Data Report for the TPFL Tee / Critical Flow Experiments," Idaho National Engineering Laboratory, NUREG/CR-4164, EGG-2377, November 1985.

20. J. Reimann and M. Khan, "Flow Through a Small Break at the Top of a Large Pipe with Stratified Flow," Nuclear Science and Engineering, Vol. 88, pp. 297-310, 1984.

21. V. E. Schrock, et al . , "Steam-Water Critical Flow Through Small Pipes From Stratified Upstream Regions," Proc. 8th International Heat Transfer Conference, San Francisco, August 18-22, 1986, i 22. V. E. Schrock, et al., "Small Break Critical Discharge-The Roles of Vapor and Liquid Entrainment in a Stratified Two-Phase Region Upstream of the Break," Lawrence Berkeley Lacoratory, NUREG/CR-4761, LBL-22024, October 1986.

23. H. J. Richter, "Separated Flow Model: Application to Critical Flow," EPRI Report NP-1800, April 1981.

Page 44 ENCLOSURE F

. - _ _ _ _ ~

~

24. D. Abdo11ahian, et al., "Critical Flow Data Review and Analysis,"

5. Levy Incorporated, Campbell, California, EPRI Report NP-2192 January 1982.

25. W. D. Beckner, J. N. Reyes, R. Anderson, "Analysis of ECC Bypass Data," U.S. Nuclear Regulatory Commission, NUREG-0573, July 1979,

26. C. J. Crowley, et al., "1/5-Scale Countercurrent Flow Data Presentation and Discussion," Creare Incorporated, New Hampshire, NUREG/CR-2106, April 1981.

27. J. A. Findlay, "BWR Refill-Reflood Program Task 4.4 - CCFL/ Refill System Effects Tests (30 Sector). Evaluation of Parallel Channel Phenomena," General Electric Company, NUREG/CR-2566, EPRI NP-2373, GEAP-22044, November 1982.

28. D. G. Schumacher, et al., "BWR Refill-Reflood Program Task 4.4 -

CCFL/ Refill System Effects Tests (30 Sector). SSTF Systems Response Test Results," General Electric Company, NUREG/CR-2568, EPRI NP-2374, GEAP-22046, April 1983.

29. J. A. Findlay, "BWR Refill-Reficod Program Task 4.4 - CCFL/ Refill System Effects Tests (30 Sector). Evaluation of ECCS Mixing Phenomena," General Electric Company, NUREG/CR-2786, EPRI NP-2542, GEAP-22150, May 1983,

30. W. D. Beckner and J. N. Reyes, Research Information Letter No. 128, "PWR Lower Plenum Refill Research Results," USNRC

( Available in the NRC Document Room,1717 H Street, N.W. ,

Washington, DC 20555), Decenber 8, 1981. -

Page 45 ENCLOSURE F

'.+.

31. K. H. Sun, "Flooding Correlations for BWR Bundle Upper Tieplate ,

and Bottom Side-Entry Orifices," in proceedings of the Second Multi-phase Flow and Heat Transfer Symposium Workshop, T. N. Veziroglu, Editor, School of Engineering and Architecture, University of Miami, Coral Gables, Florida, April 1979.

32. B. D. Jones, "Subcooled Counter-Current Flow Limiting Characteristics of the Upper Region of a BWR Fuel Bundle," -

General Electric Company, NEDG-NUREG-23549, July 1977. .

33. R. W. Haywood, et al., "An Experimental Study of the Flow Conditions and Pressure Drop of Steam-Water Mixtures at High Pressures in Heated and Unheated Tubes," in Proceedings of the Institute of Mechanical Engineers, Vol.175, pp. 669-708,1961.

34. G. P. Gaspari, et al., "Pressure Dropa in Steam Watr.' Mixtures.

Round Tubes Vertical Upflow," CISE-R83, 1964.

, r A. Alessadrini, et al., "Large Scale Experiments on Heat Transfer 35.

and Hydrodynamics with Steam-Water Mixtures. Critical Heat Flux and Pressure Drop Measurements in Round Tubes at the Pressure of 51kg/cm2 abs," CISE-R86, 1963.

36. Heat Transfer Fluid Flow Services (HTFS) Data Bank, UKAEA, Harwell, UK.*-

37. E. Janssen, "Two-Phase pressure Drop in Straight Pipes and Channels: Water-steam Mixtures at 600 to 1400 psia," AEC R&D Report GEAP-4622, 1964. ,

• This data is already available to subscribers. The NRC'is working with Maxwell Laboratory to make these data sets available to the public. Contact Harry S. Tovmassian, Office of Nuclear Regulatory Research, Washington, D.C. 2C555.  ;

Page 46 ENCLOSURE F

s

38. E. Janssen, Two-Phase Pressure Drop in Straight Pipes and Channels: Water-Steam Mixtures at 600 to 1400 psia," AEC R&D Report GEAP-4616, 1964. l l

39. E. Janssen, et al., "Two-Phase Flow in Multirod Geometries," AEC (

R&D Report GEAP-1024, 1970.  ;

40. G. L. Yoder, et al., "Dispersed Flow Film Boiling in Rod Bundle l Geometry - Steady State Heat Transfer Data and Correlation Comparisons," Oak Ridge National Laboratory, NUREG/CR-2435, ORNL-5822, March 1982.

41. D. G. Morris, et al., "Dir,persed Flow Film Boiling of High 1 Pressure Water in a Rod Bundle," Oak Ridge National Laboratory, NUREG/CR-2183, ORNL/TM-7864, August 1982.

42. R. Gottula, et al., "Forced Convective, Nonequilibrium Post-CHF Heat Transfer Experiment Data and Correlation Comparison Report,"

EG&G Idaho, NUREG/CR-3193, EGG-2245, March 1985.

43. G. L. Yoder, "Rod Bundle Film Boiling and Steam Cooling Data Base and Correlation Evaluation," Oak Ridge National Laboratory, r NUREG/CR-4394, ORNL/TM-9628, August 1986.

44. T. M. Anklam, et. al., "Experimental Investigations of Uncovered-Bundle Heat Transfer and Two-Phase Mixture-Level Swell Under High-Pressure Low Heat-Flux Conditions," Oak Ridge National Laboratory, NUREG/CR-2456, ORNL-5848, March 1982.

45. G. L. Yoder, et. al., "High Dryout Quality Film Boiling and Steam Cooling Heat Transfer Data from a Rod Bundle," Oak Ridge National Laboratory, NUREG/CR-3052, ORNL/TM-8794, December 1983.

t Page 47 ENCLOSURE F

)

i

\*

46. G. L. Yoder, "Rod Bundle Film Boiling and Steam Cooling Data Base and Correlction Evaluation," Oak Ridge National Laboratory, NUREG/CR-4394, ORNL/TM-9628, August 1986.

47. J. A. Findlay, "BWR Refill - Reflood Pregram Task 4.8 - Model Quantification Task Plan," General Electric Company, NUREG/CR-1899, EPRI NP-1527, GEAP-24898, August 1981.

P

48. D. Seedy, et al., "BWR Low Flow Bundle Uncovery Tests and Analysis," General Electric Company, NUREG/CR-2231, EPRI NP-1781, GEAP-24964, June 1981.

i-

49. T. M. Anklam, "0RNL Small-Break LOCA Heat Transfer Series 1:

Two-Phase Mixtures Level Swell Results," Oak Ridge National Laboratory, NUREG/CR-2115, ORNL/NUREG/TM-447, August 1981.

50. S. Wong and L. Hochreiter, "Analysis of the FLECHT-SEASET Unblocked Bundle Steara Cooling and Boil-Off Tests," Westinghouse 4 i Electric Corporation, EPRI-NP-1460, 1981.  :

i

51. D. Jowitt, "A New Void Correlation for Level Swell Cond'itions," j Winfrith UK, AEEW-R-1488, December 1981, t

52. B. J. Chexal, J. Horowitz, G. Lellouche, "An Assessment of Eight l Void Fraction Models for Vertical Flow," Electric Power Research Institute, NSAC-107, September 1986, e

53. S. Wong, and L. E. Hochreiter, "Analysis of the FLECHT-SEASET Unblocked Bundle Steam Cooling and Boiloff Tests," Westinghouse  ;

' Electric Corporation, NUREG/CR-1533, EPRI NP-1460, WCAP-9729, ,

i January 1981, i

Page 48 ENCLOSURE F l

54. L. E. Loftus, e2. al., "PWR FLECHT SEASET 21-R00 Bundle Flow Blockage Test Data an Analysis Report," Westinghouse Electric Corporation, NUREG/CR-2444, EPRI NP-2014, WCAP-9992, Vols. 1-2, September 1982.

55. N. Lee, et al., "PWR FLECHT SEASET Unblocked Bundle, Forced and Gravity Refload Task Data Evaluation and Analysis Report,"

Westinghouse Electric Corporation, NUREG/CR-2256, EPRI NP-2013, WCAP-9891, November 1981.

Page 49 ENCLOSURE F

7___

9 L'

hO a

ENCLOSURE G

I .

INCLOSURE G SUPEARY OF PUBLIC COMMENTS On March 3, 1987, the Nuclear Regulatory Commission published in the Federal Register (52FR6334) prnposed amendments to 10 CFR 50 and Appendix K. These proposed amendments were motivated by the fact that since the promulgation of Section 50.46 of 10 CFR 50, "Acceptance Criteria for Emergency Core Cooling Systems (ECCS) in Light-Water power Reactors," and the acceptable and required features and models specified in Appendix K to 10 CFA 50, considerable research has been performed that has greatly increased the understanding of ECCS performance during a loss-of-coolant accident (LOCA). It is now known that the methods specified in Appendix K, combined with other analysis methods currently in use, are highly conservative and that the actual cladding temperatures which would occur during a LOCA would be much lower than those calculated using Appendix K methods.

In soliciting the public's ccmments on the regulatory guide, the I

NRC requested general comments and responses to the following j questions: {

l

1. Should the regulatory guide list models, data, and model evaluation procedures that the NRC considers to be acceptable for realistic l calculations of ECCS performance? l l

2. Are the models, data, and mods 1 evaluation procedures listed in Appendix A to the draft regulatory guide appropriate?

3. Does Appendix B of the draft regulatory guide appropriately discuss the features of the best-estimate codes and the estimation of overall calculational uncertainty?

page 1 ENCLOSURE G

. . - - - _ _ _ .=. .- _ _ _ __- - .

The concent period for the proposed rule revision and the draft  !

regulatory guide (52FR11385) expired on July 1, 1987. The NRC l received nine responses to the request for comments on the regulatory guide questions. The names and organizations of the nine people who  !

responded to the request for comments are provided in Section A-1.

Section A-2 contains the paraphrased summaries of the comments submitted by the nine people listed in Section A-1. The NRC responses ,

to the draft regulatory guide comments are presented in Section A-3.

The comments are numbered to facilitate cross-referencing with the appropriate response category. The response categories cover general comments and responses to the three questions listed above.

The NRC staff incorporated Appendix B of the draft regulatory guide into Appendix A after the original draft had been submitted for review. The numbering sequence in the original Appendix B is nearly identical to the sequence in the newly merged Aprendix in the current guide. Comments will be referenced to the sections in the new Appendix. If the numbers are different between the draft and the current guide, s oss-referencing to the sections in the draft guide will be indicated in square brackets, (e.g., . . . Section 1.2.5.2

[2.1) . . . ).

i i

page 2 ENCLOSURE G J

, , _ , - , , . _ _ . , , _ ,---y- . , . , , . - - , - - . , -

,t. SECTION A-1 LIST OF PUBLIC COMMENTERS This section lists the names of the people and organizations that responded to the NRC staff's request for public comment on the draft regulator,y guide.

1. G. S. Lellouche, S. Levy Incorporated, 3425 S. Bascom Avenue, Campbell, California 95008-7006.

2. Gary N. Ward, Manager, Reload Licensing, Advanced Nuclear Fuels Corporation, 2101 Horn Rapids Road, P.O. Box 130, Richland, Washington 99352-0130.

3. Susan L. Hiatt, OCRE Representative, Ohio Citizens for Responsible Energy, 8275 Munson Road, Mentor, Ohio 44060.

4. W. L. Stewart, Vice President, Nuclear Operations, Virginia 4

Electric and Power Company, Richmond Virginia 23261.

5. C. O. Woody, Group Vice President, Nuclear Energy, Florida Power and Light Company, P.O. Box 14000, Juno Beach, Florida 33408-0420.

6. C. W. Fay, Vice President, Nuclear Power, Wisconsin Electric Power Company, 231 W. Michigan, P.O. Box 2046, Milwaukee, Wisconsin 53201,

7. G. C. Sorensen, Manager, Regulatory Programs, Washington Public Power and Supply System, P.O. Box 968, 3000 George Washington Way, Richland, Washington 99352,

8. A. E. Scherer, Director, Nuclear Licensing, Power Systems, Combustion Engineering, Incorporated, 1000 Prospect Hill Road, P.O. Box 500, Windsor, Connecticut 06095-0500.

Page 3 ENCLOSURE G

• /

9. J. R. Thorpe, Director, Licensing and Regulatory Affairs, GPU Nuclear Corporation, 100 Interpace Parkway, Parsippany, New Jersey 07054-1149.

Page 4 ENCLOSURE G

SECTION A-2  !

PUBLIC C0tHENTS ON THE DPAFT REGULATORY GUIDE -

1. G. S. Le11ouche, S. Levy incorporated,;342h.S. Bascom Avenue, Campbell, California 950Q8-}006._ ,

C0tHENT:

1 1.1 Mr. Le11ouche comments that the extant literature does not j support a constant Nt.4selt number (Nu = 8) for laminar flow, independent of oundle geometry. He recommends using the method employed by Weisman (NSE 1959) to take into account the effect of differer,t pitch to diameter ratios on the Nusselt number.

2, Gary N. Ward, Manager s Reload Licensing Advanced Nuclear Fuels Corporation 2101 Horn Rapids RoGd, P.O. Box 130, Richland, Washington 99352-0130.

C0tHENTS:

2.1 Mr. Ward indicates that th6 fraft guide is overly prescriptive and that i NRC would be limited in its receptiveness to models other than those identified as "acceptable" by the draft guide.

l 2.2 The regulatory guide iCfslies that c?:nplex "best-estimate *I LOCA models developed by NRC, or th61r equivalent, are a necessery and I For the purpose of the regulatory guide supporting the LCCS rule, the terms "best-estimate" ard "realistic" have the same meaning. Both terms are used to indicate that the techniques attempt to predict realistit reactor system tha. mal-hydraulic response.

Page 5 ENCLOSURE G

integral part of quantifying uncertainties of simpler models used '  !

in commercial licensing calculations. Development of a best estimate model, comparison of this model to experiments, and quantificat' ion of the unce.'tainties are burdensome for a single reactor vendor, fuel vendor, or utility. ,

4 t

2.3 It would seem appropriate for NRC to provide the industry with  ;

best estimate model LOCA reference plant results and quantification of uncertainties by conducting a spectrum of LOCA  !

calculations for each generic plant type. Simpler comme <cial licensing ms 'is could then be compar6d to the NRC generic reference calculations.

2.4 Appendix A is a reasonably complete listing of models, dita and model evaluation procedures.

2.5 Appendix B [Section A.2) is not sufficiently dev6ioped for a completa statistical appt-oach with which to calculate j uncertainties. It does not add to the information on how to I

interpret the regulation.

i

3. Susan L. Hiatt, OCRE Representative, Ohio Citizens for Responsible Energy, 8275 Munson Road, Mentor, Ohio 44060.

i COMMENT:

L 3.1 Ms. Hiatt requests that NRC extend the comment period for the draft regulatory guide to September 1,1987, because of irisuf ficient f I, time to review NUREG-1230.

4. W. L. Stewart, Vice president, Nuclear Operations, X1roinia Electric and power Company, Richmond, Viratn_i+ 23261.

L i I l r Page 6 ENCLOSURE G j I

s.

COMMENTS:

4.1 The draft s . ' quid .<oecify cladding as zirconium base alloy rather l then Zircaloy, .c not to impede the development of future  !

improvements i- .dding. ,

l 4.2 In evaluating Section 1.2.2 of Appendix B, it is suggested that conservative combinations of input parameters ba allowed in best estimate calculations, provided their impact upon the calculated result can be quantified and the total calculational uncertainty adjusted to take this impact into account.

4.3 Is it intendad that the confidence level used in a specific uncertainty evaluation be chosen and justified as part of the licensing application?

5. C. O. Woody, Group Vice President, Nuclear Energy, Florida Power and Light Company, P.O. Box 14000, Juno Beach, Flurida

{ 33408-0420.

COMMENTS:

5.1 FPL states that the changes proposed are, for the most part, a very positive modification. They are currently using the guidelines of the proposed regulatory guide to perform a small.

break LOCA analysis in a statistical mar.ner.

5.2 The guide should note that the goal of the uncertainty analysis is to ensure that the calculated PCT is at least at the 95% probability level, and not to exactly quantify the amount of uncertainty.

Page 7 ENCLOSURE G

6. C. W. Fay, Vice Preaident, Nuclear Power, Wisconsin Electric Power Company, 231 W. Michigan, P.O. Box 2046. Milwaukee, Wisconsin 53201.

COMMENTS:

6.I' The guide should state clearly that the models, data, and model evaluation procedures listed are not the only set acceptable to the NRC.-

6.2 The models for calculating the initial stored energy in the fuel should recognize, in addition to the variables listed, the effect of thermal and elastic strain of the fuel and cladding material.

6.3 Correlations for metal-water reaction rate calculations should recognize, in addition to the variables listed, the effect of cladding temperature.

6.4 The regulatory guide should recommend use of the Cathcart cor' relation for all temperatures of concern (1500*F to 2200*F).

6.5 Section 2.5 should be rewritten to rcflect the reality of liquid in the flow field when the bundle is uncovered.

6.6 With respect to Section 1.2.3.1, the fuel model need not consider changes in the fuel properties due to burnup during the course of the LOCA, since they do not change significantly during the course of a LOCA.

6.7 The requirement in Section 1.2.4 that cladding swelling r udels take into account asymmetric deformation of the cladding is unclear. The direction of asymmetry uld be clarified. The guidance would be unreasonable if tt- maatry of the deformation is in the azimuthal diret ^.. e fuel rod.

6.8 There does not appear to be any basis for providing noding studies in areas of large thermal non-equilibrium (Section 1.2.6). -

Page 8 ENCLOSURE G t - - - - - - , . .

s..

-6.9 Research shows that the effect of flow blockage 'on post-blowdown heat transfer is' minimal; therefore, there is no reason to require calcul'a tion of that effect (Section 1.2.13.4).

6.10 The best estimate codes cited in Section 1.2.17 should be identified more clearly. The code version should be. identified and the list expanded to cover edditional aspects of the analysis, such as fuel performance and containment analysis.

6.11 To avoid double accounting of uncertainties, the experimental data should be treated as absolute. No uncertainty should be associated with experimental data.

7. G. C. Sorensen, Manager, Regulatory programs, Washincton public Power and Supply System, p.0. Box 968, 3000 George Washington Way, Richland, Washington 99352.

COMMENTS:

7.1 The regulatory guide represents a step forward toward a more reasonable approach to regulation.

7.2 Issuance of the regulatory guide should be delayed until NUREG-1230, "Compendium of ECCS Research for Realistic LOCA Analysis," is available and reviewed.

7.3 The sectis- ,n Metal-Water Reaction Rate should be ccrrected to read "greater than 1500 F and less than 19000F."

7.4 Finite break opening time should be allowed for ECCS performance as it is for other portions of LOCA analyses. Appropriate evaluation of uncertainties in break opening time would then be included in i the overall model uncertainty evaluations, l

i l

l Page 9 ENr,LOSURE G l

)

, 8. '.. E. Scherer, Director, Nuclear Licensing, power Systems, Combustion Engineering, Incorporated, 1000 Prospect Hill Road, l p.0. Box 500, Windsor, Connecticut 06095-0500.  !

COMMENTS:

8.1 Use of a regulatory guide is an appropriate vehicle'for relating those features of a best estimate evaluation model which the NRC staff has found to be acceptable.

8.2 CE believes that the regulatory guide must indicate that alternate models, data, and model evaluation procedures would be equally acceptable to the NRC if they can be shown to be appropriate for tne intended purpose.

8 ', CE has reviewed Appendix A and can accept the models, data, and model evaluation procedures listed.

8.4 The regulatory guide should be updated periodically as new data are obtained.

8.5 The sections on ECC Bypass (1.2.5.2), Core Flow Distribution During Blowdown (1.2.12) and Containment P ure Calculations (1.2.13) should be annotated to indicate that they are either not necessary or do not require the detail specified for small oreak LOC / evaluation models.

8.6 A statement should be incorporated into Section 2.0 [ App A, Section B] indicating that alternate approaches for dealing with the various components of overall calculational uncertainty would

! be acceptable if it can be shown that these approaches provide a degree of conservatism equivalent to'or greater than that provided by strictly adhering to the 95% probability approach ,

l described.

l Page 10 ENCLOSURE G l

l

1 9.. J. R. Thorpe, Director, Licensing and Regulatory Affairs, GpU Nuclear Corporation, 100 Interpace Parkway, Parsippany, New Jersey 07054-1149.

COMMENTS:

9.1 The guide fails to define what qualifications should actually certify a best-estima~ca code. The burden of proof that the codes identified in the regulatory guide are really best-estimate is unfairly handed to the licensee. The licensee can only be responsible for identifying cources of uncertainty darived from specific applications, not the code uncerteinty. Code uncertainty analysis should be performed by the NRC with a well-planned and committed schedule.

9.2 The NRC should establish an unequivocal standard for qualifying best-estimate codes. With respect to acceptance criteria for a best-estimate code, the followirg questions need to be clearly addressed in the guide:

a. Does the guide imply that a code can be certified as best-estimate if the code uncertainty analysis results in a relatively small uncertainty?

b. Does the guide impsy that industry developed codes cannot be considered as best-estimate because of their relatively larger uncertainty?

c. What is the borderline between best- and next-to-best estimate codes?

d. Should there be application categories so that certain codes are acceptable for a limited category of application?

9.3 Sections 1.2.2, 1.2.13.1, 2.1, and 2.3.1 are unclear regarding what is an acceptable or preferred method of factoring equipment availability into the best estimate calculations and uncertainty analysis.

Page 11 ENCLOSURE G

~.

SECTION A-3

SUMMARY

OF COMMENTS 3N DRAFT REGULATORY GUIDE This section consists of a synthesis of the comments from the reviewers of the draft regulatory guide and NRC staff's proposed response to those comments. The comment numbers cross-reference the commenter with the comment number. For example, Comment 9.3 references the ninth commenter's third comment.

GENEPAL COMMENTS ON THE DRAFT REGULATORY GUIDE (Refs. 3.1 and 7.2):

Two reviewers requested that the NRC extend the comment period for the draft regulatory guide review because of insufficient time to review NUREG-1230, "Compendium of ECCS Research for Realistic LOCA Analysis," which describes the research supporting the proposed ECCS rule revision.

NRC RESPONSE:

The NRC staff believes the comment period was sufficient, since most of the research is not new and has been extensively reviewed in the past. Both commenters were contacted and told that comments received after the comment period would oe considered if time permitted. Comments from both parties were received late and were, indeed, considered by the NPC.

COMMENT (Ref 5.1, 7.1 and 8.1):

Three reviewers commented favorably on the draft regulatory guide. One of the reviewers is using the guidelines of the proposed regulatory guide to perform a smull break LOCA analysis in a statistical manner.

NRC RESPONSE:

No response required.

Page 12 ENCLOSURE G

l i

0UESTION 1:

Should the regulatory guide list models, data, and model evaluation l procedures that the NRC considers to be acceptable for realistic calculations of ECCS performance?

COMMENT (Ref 2.1):

One of the reviewers indicated that the draft guide is overly prescriptive and that NRC sould be limited in its receptiveness to models other than those identified as "acceptable" by the draft guide.

NRC RESPONSE:

The regulatory guide, by virtue of references to qualified data, gives the user the option of creating new models. The only requirement is that the new model must be able to reproduce the referenced experimental data acceptably well.

COMMENT (Ref 2.2):

One reviewer stated that development of a best-estimate mocel, comparison of this model to experiments, and quantification of the uncertainties are burdensome for a single reactor vendor, fuel vendor, or utility.

NRC RESPONSE:

This regulatory guide does not imply that NRC models must be used to benchmark simpler commercial models. Paragraph 2.2.2 clearly states that code uncertainty should be evaluated through direct data comparisons with relevant integral systems and separate effects experiments at different scales. In any event, this regulatory guide is not a requirement. Each licensee must evaluate whether any additional burden in complying with this rule is sufficiently offset by the operating benefits that would result.

Page 13 ENCLOSURE G

. ., l COMMENT (Ref 2.3):

1 I

The sLme reviewer suggested that the NRC provide the industry with best-estimate model LOCA reference plant results and quantification of uncertairities by conducting a spectrum of LOCA calculations for each generic plant type. Simpler commercial licensing models could then be compared to the NRC generic reference calculations.

NRC RESPONSE:

The NRC makes available, to any licensee who requests it, any of its thermal-hydraulic safety codes and any non-proprietary assessment calculations. In addition, the NRC has developed a Code Scaling, Applicability, anj Uncertainty (CSAU) Methodology, which is generally described in NUREG-1230. Detailed applications of this methodology will be documented in the near future. Although licensees may avail themselves of any of this work, the objective is not to perform uncertainty analyses for licensees, but to provide an independent audit tool for the NRC. lnerefore, the NRC plans no further-work analyzing "reference plants" as part of the CSAU demonstration.

COMMENT (Ref 6.1, 8.2):

Two reviewers stated that the regulatory guide must indicate that alternate models, data, and model evaluation procedures would be equally acceptable if they can be shown to be appropriate for the intended purpose.

NRC RESP 0NSE:

The NRC staff agrees with this comment. The introduction to Section 2 of Appendix A clearly states that the basic requirement for acceptability of a specific model is an accurate comparison with applicable data.

Page 14 ENCLOSURE G

4 s.

QUESTION 2:

Are the models, data, and model evaluation procedures listed in Appendix A to the draft regulatory guide appropriate?

COMMENT (Ref 1.1)i .

The reviewer commented that the laminar Nusselt number depends on rod pitch to diameter ratio for triangular rod bundle lattices, so that the implied constant laminar Nu number of 8 is not supported by the extant literature.

NRC RESPONSE:

The NRC staff encourages the use of models that accurately calculate the phoneomena. The acceptance of the model will be based on comparison of applicable experimental data. The correlation of Nu = 8 for laminar flow was intended as an example of a distinct transition to the laminar flow region. If such a model were used for evaluations, the model would have to be compared to appropriate experimental data, accounting for the particular geometry and flow condition.

In order to avoid confusion over the applicability of this illustrative example, the laminar flow correlation has been deleted in the final regulatory guide.

COMMENT (Ref 2.4):

One reviewer agrees that Appendix A is a reasonably complete listing of models, data, and model evaluation procedures.

NRC RESPONSE:

No response required.

Page 15 ENCLOSURE G

e COMMENT (Ref 8.3 and 8.4):

A reviewer accepted the models, data, and model evaluation procedures and suggested that the regulatory guide should be updated periodically as new data are obtained.

NRC RESP 0NSE:

The staff agrees that the regulatory guid9 should-be updated periodically to incorporate the results of new research.

COMMENT (Ref 4.1):

One reviewer stated that the regulatory guide should specify cladding as zirconium base alloy rather than zircaloy so as not to impede the development of future improvements in cladding.

NRC RESPONSE.

The staff believes that such a modification is beyond the scope of the current regulatory guide and should be considered in a separate rulemaking action in which it would receive appropriate public review and comment prior to implementation. Additionally, zircaloy cladding material is specified in other portions of the Code of Federal Regulations, such as 10 CFR 50.44, thereby making a change of this type more suitable to a broader regulatory context. Therefore, the staff is not broadening the definition of cladding materials in 10 CFR 50.46, as requested by this commenter.

t COMMENT (Ref 6.2, 6.3):

One reviewer commented that the models for calculating the initial stored energy in the fuel should consider the effect of strain of the fuel and cladding.

l l

' Page 16 ENCLOSURE G l

l l

u

NRC RESP 0NSE:

The models for calculating initial stored energy and gap heat transfer during the LOCA should indeed consider thermal and other factors which cause elastic and plastic strain (deformation) of the fuel and cladding.

C0fHENT (Ref 6.3, 6 , 7.3):

The reviewers commented on the temperature range over which the metal-water reaction rate correlations are applicable.

NRC RESPONSE:

The language of the draft regulatory guide ould have been better 0

structured. For cladding temperatures above 1900 F, metal-water reaction data are well correlated by the Cathcart-Pawel rate equation 0

(Ref. A-11). For temperatures below 1900 F but above 1500 F, no specific correlation was recommended. The Cathcart-Pawel model is not best-estimate in this range. For temperatures below 1500 F, the Biederman model has been deleted. Therefore, all calculations below 19000 F should be conducted with a model validated by data comparisons in the appropriate temperature range.

QUESTION 3:

Does Appendix B of the draft regulatory guide appropriately discuss the features of the best-estimate codes and the estimation of overall calculational uncertainty?

COMMENT (Ref 2.5):

One reviewer commented that Appendix B is insufficiently developed for a complete statistical approach with which to calculate uncertainties. Furthermore, it does not add to the information on how to interpret the regulation.

Page 17 ENCLOSURE G

NRC RESP 0NSE:

The regulatory guide is supported by the information in draft NUREG-1230. Th's information will become formally available when NUREG-1230 is published in final form. The staff desires to keep the description of acceptable uncertainty methodologies general so that the licensee and vendor may develop methods appropriate to their codes. The staff is aware that many licensees and vendors have develcped their own uncertainty methodologies.

COMMENT (Ref 4.2):

A reviewer suggested that conservative combinations of input parameters be allowed in best-estimate calculations provided their impact upon the calculated result can be quantified and the tctal calculational uncertainty can be adjusted to take this impact into account.

NRC Resp 0NSE:

This approach is acceptable to the staff provided the licensee or vendor can quantitatively show that the conservatisms do not function to compensate errors and that the important phenomena relevant to each transient analyzed are predicted acceptably well. Note that the guide does not prohibit this approach (Sect'on 1.2.2).

COMMENT (Ref 4.3):

One reviewer pointed out that the regulatory guide proposed the use of 95% as the probability of not violating the limits of 10 CFR 50.46(b) for specific applications of best-estimate ECCS analyses; however, the guide did not discuss the confidence level associated with this probability. The reviewer asked whose responsibility it would be to assign the confidence interval and would that interval have to be justified as part of each licensing application.

Page 18 ENCLOSURE G

'- NRC RESPONSE:

1 The NRC agrees that the draft regulatory guide is silent in its treatment of confidence limits. Therefore, the final guide has been modified to describe an acceptable statistical technique which does not require the utilization of confidence limits. Other techniques which account for uncertainty in a more detailed manner may also be used. These techniques may require the use of confidence levels which are not required by the NRC suggested approach.

COMMENT (Ref 5.2 and 8.6):

Two reviewers commented that the guide should note that the goal of the uncertainty analysis is to ensure that the calculated PCT is at least at the 95% probability level, rather than to quantify exactly the degree of uncertainty.

NRC RESPONSE:

The purpose of the uncertainty analysis is as stated by these reviewers and was stated in Section 2.1 of the guide. Nevertheless, the analyst must perform a thorough uncertainty analysis before concluding that the calculated PCT is within the desired limits. The staff believes that the licensee and vendor should not be required to conform to a rigorous set of probability values; rather, the licensee should be all',wed to work within a set of constraints that assures the safety of tne public.

COMMENT (Ref 6.5):

Section 2.5 (Uncovered Bundle Heat Transfer) should be rewritten to reflect the reality of liquid in the flow field when the bundle is uncovered.

Page 19 ENCLOSURE G

.-, , - 4 - , - - , , - - , _ - - -

1 NRC RESPONSE:

The Uncovered Bundle Heat Transfer section in the draft guide addresses steam cooling only. The guide addresses liquid entrainment in Section 1.2.9, Post-CHF Heat Transfer.

COMMENT (Ref 6.6):

One reviewer commented that the fuel model need not consider changes in the fuel properties due to burnup during the course of the LOCA, since they do not change significantly during the course of a LOCA.

NRC RESPONSE:

The NRC staff agrees that the effect of fuel burnup during a LOCA is negligible compared to the effects of burnup prior to the postulated accident. One of the purposes of the fuel models is to calculate the steady-state temperature distribution and stored energy in the fuel before the postulated LOCA.

COMMENT (Ref 6.7):

The reviewer states that Section 1.2.4 is unclear regarding the requ'rement that cladding swelling models take into account asymmetric deformation. The reviewer requests that the direction of asymmetry be clarified. Additionally, the reviewer states that the guidance woL1d be unreasonable if the asymmetry of the deformation is in the azimuthal direction of one fuel rod.

NRC RESPONSE:

The staff believes thot the inclusion of asymmetric deformation capabilities in a cladding swell model will result in the closest approach to a true best-estimate calculation, However, the staff also Page 20 ENCLOSURE G

q recognizes the difficulties encountered in developing computer codes that can calculate azimuthal, asymmetric cladding deformation. Since the purpose of this guide is to provide guidelines for performing best estimate analyses of ECCS performance, the staff concedes the use of codes that can calculate symmetric deformation only. As more sophisticated computer codes become available, the staff will decide if the added sophistication needs to be considered in design basis analysis. In the interim, the licensee and vendor should attempt to quantify the uncertainty associated with only calculating symmetric deformation of the cladding only.

COMMENT (Ref 6.8):

The reviewer states that noding studies are most often performed for areas of high fluid velocity or complex fluid flow. Therefore, there is no basis for providing noding studies in areas of large thermal non-equilibrium.

NRC RESPONSE:

The break flow in a large break LOCA begins with high fluid velocities, and ECCS injections generally result in the mixing of highly subcooled fluid with a two-phase, liquid-vapor mixture (which can result in a metastable state). Many studies of break nodalization sensitivities have been performed that show nodalization significantly affects these types of flow con' figurations. To clarify the intent, the wording of the regulatory guide has been changed to reflect the reviewer's comment.

COMMENT (Ref 6.9):

The reviewer states that detailed modeling of flow blockage significantly reduces or eliminates the heat transfer penalty traditionally associated with flow blockage. Since iasearch shows that the effect is minimal, there is no reason to require calculation of the effect of flow blockage on heat transfer.

Page 21 ENCLOSURE G

NRC RESP 0NSE:

The staff requires that the effect of flow blockage on total heat transfer be quantified for a best-estimate analysis, because the

-analysis is no longer bounded by the conservatisms found in Appendix K of 10CFR50. Detailed calculations of flow blockage heat transfer may not be necessary if the licensee or vendor can show that the results of the research are applicable to the specific fuel bundle design.

COMMENT (Ref 6.10):

The reviewer states that the regulatory guide should list the names of specific codes instead of code families. Additionally, the reviewer questions the inclusion of the FRAP family of codes in the list of best estimato codes because the FRAP codes are not considered to be thermal-hydraulic analysis tools. The reviewer also suggests that containment' codes should be included in the list.

NRC RESP 0NSE:

The intent of listing families of codes was to keep the regulatory guide as general as possible to allow the greatest degree of flexibility in the selection of the analysis tools. The fuel behavior codes (e.g., FRAP codes) are included ia this list because the licensee and vendor must be able to analyze the effect of periods of reduced core cooling on fuel rod cladding integrity and to estimate fuel initial conditions at the start of the postulated accident. The staff agrees that a list of suggested containment codes might be useful but has decided not to include them because they do not substanially contribute to the calculation of ECCS performance or the uncertainty, COMMENT (Ref 6.11):

Sc cion 2.2 states that code comparisons should account for limitations of the measurements and calibration errors. The reviewer states that including these items in code uncertainty may result in Page 22 ENCLOSURE G

s double accounting of uncertainties when the uncertainty of the same variables is included in other sources of uncertainty.

NRC RESPONSE:

The limitations of measurement and calibration errors should be considered in the data qualification phase prior to acceptance of the data for comparison with code results. Any uncertainties in data should be considered when comparing code results, especially when those uncertainties could result in an approach to the upper PCT

. limit. Where it can te shown that double accounting for uncertainties

. occurs, proper adjustments may be made to correct the estimate of the uncertainty.

COMMENT (Ref 7.4):

The reviewer states that a finite break opening ti;ne should be allowed for ECCS performance as it is for other portions of LOCA analyses. Evaluation of uncertainties in break opening time would then be included in the overall model uncertainty evaluations.

NRC RESPONSE:

The commenters approach implies the utilization of the "leak before break" concept. The recent revision to GDC-4 specifically stated that ECCS design would not be affected by the use of "leak before break" as now allowed in other areas by the new version of GDC-4. However, the commission, in a separate rulemaking action, is soliciting comments on expanding this technology to ECCS design application. Until this separate rulemaking is complete, it would be premature to allow this technology to be applied to best estimate ECCS analysis.

COMMENT (Ref 8.5): ,

The reviewer states that the sections on ECC Bypass (1.2.5.2),

Core Flow Distribution During Blowdown (1.2.12), and Containment Page 23 ENCLOSURE G

Pres 9are Calculations (1.2.13) should be annotated to indicate that ,

-they are either not.necessary or do not require the detail specified for small break LOCA evaluation models.

NRC Resp 0NSE:

The staff does not agree that these models should be given blanket exclusion for small break LOCAs. These models may be required depending on the transient scenario. They shculd not be ignored until they are determined to be of no consequence for a specific postulated accident.

COMMENT (Ref 9.1):

A reviewer stated that the regulatory guius forces the licensee to assume too much of the burden of proving that the codes in the regulatory guide are really best estimate.

NRC RESPONSE:

In formulating the rule, the staff recognized that this might be a concern for some licensees. It is for this reason that the rule permits the licensee to choose whether it will perform a best-estimate ECCS evaluation or continue to use the existing Appendix K methodology. However, most of the commenters already use best-estimate codes for various analyses, and some are participating in programs to develop uncertainty methodologies for their codes.

Many licensees have already assumed some of the burden of proof voluntarily because they recognize the necessity of understanding the limitations of their codes before they commit substantial analysis resources to a project.

COMMENT (Ref 9.2):

l The reviewer states that Appendix B does not appropriately discuss the features of a best-estimate code and overall calculational i

Page 24 ENCLOSURE G l

i

' uncertainty. Specifically, the guide fails to define what qualifications should actually certify a best-estimate code.

The reviewer believes that the following' questions need to be clearly addressed in'the regulatory guide:

a. Does the guide imply that a code can be certified as best estimate if the code uncertainty analysis results in a relatively small uncertainty?

b. Does the guide imply that industry-developed codes cannot be considered as best-estimate because of their relatively larger uncertainty?

c. What is the borderline between best-and next-to-best estimate ,

codes?

d. Should there be application categories so that certain codes are acceptable for a limited category of application?

NRC Resp 0NSE:

The staff intended that the regulatory guide should not prescribe specific measures to be taken in certifying a code as best-estimate.

This approach is taken because often regulatory guides are interp-eted as de facto requirements. The staff desires to keep the description of acceptable uncertainty methodologies general so that the licensee may develop methods appropriate to their codes. The staff is also aware that many licensees and vendors have already developed their own uncertainty methodologies. A suggested methodology for quantifying the uncertainty of several NRC-sponsored codes is presented in NUREG-1230.

COMMENT (Ref 9.3):

The reviewer states that Appendix B Sections 1.2.2, 1.2.13.1, 2.1, and 2.3.1 are unclear regarding what is an acceptable or Page 25 ENCLOSURE G

preferred method of factoring equipment availability into the best-estimate calculations and uncertainty analysis.

NRC RESPONSE:

Most best-estimate calculations include a series of sensitivity studies to determine the effect of equipment unavailability and degraded / enhanced equipment performance. These sensitivity studies are then used to bound the best-estimate calculation. Often, the sensitivity studies do not require the extensive use of a best-estimate code; rather, an engineering analysis.is used to estimate the bounds. The staff believes that the licensees and vendors possess the engineering resources to perform these types of analyses.

Page 26 ENCLOSURE G t

i h

t ENCLOSURE H 9

j

)

i

)

q

~ ~ ' *" "'v~w----, __,_ _

-NUCLEAR REGULATORY COMMISSION

[10 CFR PART 50]

Acceptance criteria for Emergency Core Cooling Systems; Environmental Assessment and Finding of No Significant Impact The U.S. Nuclear Regulatory Commission (The Commissipn) is revising 9 50.46 and Appendix K of 10 CFR Part 50 which specify requirements of emergency core cooling systems (ECCS) for light water reactors.

ENVIRONMENTAL ASSESSMENT IDENTIFICATION OF ACTION TO BE TAKEN:

Section 50.46(a)(1) has been revised to eliminate the reauirement to use the features of Appendix K when calculating ECCS performance during a loss-of coolant accident (LOCA). Section 50.46(a)(1)(i) of the final rule allows use of realistic analytical techniques and requires that the uncertainty of the calculation be evaluated and considered when comparing the results of the calculation with the temperature limits and other criteria of 6 50.46(b). Section 50.46(a)(1)(ii) has been added to allow continued use of the features of Appendix K as an alternative to the uncertainty evaluation required by 5 50.46(a)(1)(i). Sections 50.46(a)(2) and 50.46(a)(3) have been revised to eliminate historical implementation sections and to specify requirements for reanalyses and reporting which are excluded from consideration in this environmental assessment per i 51.22 of 10 CFR Part 51. Appendix K of 10 CFR Part 50 has been revised to make minor technical changes to the acceptable features of the calculations.

NEED FOR ACTION:

.The revisions of 10 CFR Part 50 and Appendix K are required in order to permit new knowledge of E'CCS performance gained through research to be used in the calculations of ECCS performance. The Page 1 ENCLOSURE ll

improved calculations allow relaxation of restrictions which are i preventing optimal operation of some reactors and are not necessary to adequately protect the health and safety of the public.

ENVIRONMENTAL IMPACTS OF THIS ACTION:

If power levels remain unchanged, the revisions will reduce the cladding te:nperatures that are calculated during a LOCA. If peak calculated cladding temperatures are allowed to remain at or below regulatory limits, then the peak local power of the reactor can be '

increased. An increase in the allowed peak local power can be used in either or both of the following manners:

1. The total maximum allowed power of the reactor remains unchanged, but plant efficiency can be improved by increased flexibility in the allowed power shape. More efficient fuel utilization, more flexibility in changing power shape, and reduced derating of plants due to fuel limits are possible.

2. The total maximum allowed power of the reactor can be increased. The expected maximum increase in total power for existing and currently planned reactors is approximately 5% based on practical limits of plant hardware.

Either of these actions require an amendment to the olant license to change the technical specification limits and, therefore, result in an environmental assessment specific to that particular plant and tha specific amendment being considered. This environmental assessmer.'. is a generic evaluation considering the typical impact of the rule revision.

. A change in the allowed peak local power, without an increase in total pcwer, produces no significant environmental impact. The total fission product inventory, routine releases of radioactive materials i

and thermal releases to the environment are essentially unchanged.

l l Fuel cycle changes are in the direction of improved use of fuel and

(

should not significantly change the environmental impact of the fuel l

l Page 2 ENCLOSURE H l

[

. cycle unless major new fuel cycle methods (e.g., plutonium recycle) are adopted. Such changes are beyond the sccpe of this rule revision and are not considered.

For the case of a small (i.e., 5%) increase in total power, there is a correspondingly small increase in fission product inventory, routine releases of radioactivity and fuel use. However, maximum allowed releases of radioactivity during both accident situations and routine operation are specified by technical specifications and other sections of 10 CFR Part 50 which are un: hanged by this rule. It is not expected that the small increase in total power that could result from this revision would result in difficulty in meeting the existing ,

release limits. An increase in total power increases the thermal discharge to the environment by an amount approximately proportional to the increase in power. The discharge of heat to surface waters is regulated under the Clean Water Act by the U.S. EPA or designated state agencies. NRC defers to procedures under that Act to establish the acceptability of any increase in waste heat discharge. It is not intendeo that NRC approval of increased power level affect in any way the responsibility of the licensee to comply with requirements of the Clean Water Act. These being the only potential environmental considerations, the NRC staff believes that site specific environmental impact assessments will not be of help to the decision process.

ALTERNATIVES TO THIS ACTION:

The staff has considered a number of alternatives to revise 5 50.46 and Appendix K. However, all the alternatives considered would allow similar increases in local or total power, with the exception of the alternative of making no changes. Since the environmental impact of the proposed revision is considered to be not significant and the revision is required to reduce unwarranted restriction on the operation of some reactors, the Commission has proceeded with the final rule revision.

Page 3 ENCLOSURE H

\ '

AGENCIES AND PERSONS CONSULTE0: ,

The NRC staff consulted U.S. manufacturers of nuclear power plants to determine the maximum increase in local or total power that might result from application of_.the proposed rule revisions. The staff did not consult other agencies or persons.

FINDING OF NO SIGNIFICANT IMPACT The Commission has decided not to prepare an Environmental Insact Statement for this action. The foregoing environmental assessment of this action has concluded that the proposed action does not significantly effect the quality of the human environment.

Dated at Rockville, Maryland this 1st day of June , 1988.

For The Nuclear Regulatory Commission Eric S. Beckjord, Director Office of Nuclear Regulatory Research I

t l

Page 4 ENCLOSURE H

~ _ _ _ _ _ _ _

3~

ENCLOSoCE I l

l l

l I

J I

/

~~

s REGULATOP.Y ANALYSIS FOR REVISION OF THE ECCS RULE AND SUPPORTING REGULATORY GUIDE l

1. STATEMENT OF PROBLEM 1.1 Background I 1.2 Discussion of Proposed Rulemaking

2. OBJECTIVES

3. ALTERNATIVES

4. CONSEQUENCES 4.1 Safety and Risk Effects 4.2 Costs and Benefits l

5. DECISION RATIONALE

! 6. IMPLEMENTATION l

0

~ ,

REGULATORY ANALYSIS FOR REVISION OF THE ECCS RULE AND SUPPORTING REGULATORY GUIDE h

~

This regulatory analysis provides the value/ impact analysis eor both the ECCS Rule and the supporting regulatory guide. The calculations performed in support of this analysis utilized information in the regulatory guide as well at the final rule. m

1. STATEMENT OF THE PROBLEM F

1.1 Background

Section 50.46 of 10 CFR Part 50 requires that calculations be l; performed to show that the emergency core cooling systems (ECCS) will ('

I adequately cool the reector in the event of a loss-of-coolant accident (LOCA). Appendix K sets forth certain required and acceptable _

features that the evaluation models, used to perform these E

calculations, must contain. The results of these calculations are used to determine the acceptability of the ECCS performance. In 'many -

instances, these calculations result in technical specification limits on reactor operation (e.g., peak local power) in order to comply with .g the 2200 F cladding temperature limit and other limits of 50.46.

These limits restrict the total power output and optimal operation of some reactors in terms of efficient fuel utilization, maneuvering h capability and surveillance requirements.

The NRC, DOE (including AEC and ERDA), U.S. nuclear industry and foreign research on ECCS performsnce since the present ECCS rule was issued provides a technical understanding which shows that the existing ECCS rule restrictions are more stringent than required co protect the health and safety of the public.

Page 1 ENCLOSURE I

9 1.2 Discussion of the Rulemaking This final rale and supporting regulatory guide allow alternative methods to be used to demonstrate that the ECCS would protect the nuclear reactor core during a postulated design basis loss-c.f-coolant accident (LOCA). While continuing to all w the use of current Appendix K methods and requirements, the rule and regulatory guide  :

also allow the use of more recent information and knowledge currently available to demonstrate that the ECCS would perform its safety function during a LOCA. Procedural changes have also been made to relax requirements for certain reanalyses which do not contribute to safety. This regulation applies to all applicants for and holders of construction permits or operating licenses for light water reactors.

The effect of this rule change will be to reduce the peak cl' adding temperatures that are calculated to occur during a LOCA.

This will permit some phnts to increase allowed peak local power by increasing allcwed peaking factors and/or increasing total power.

This regulatory analysis discusses the effect of the rule enange in terms of "value" (e.g., public benefits, such as safety) and "impact" (e.g., consequences, such as costs). The intent of the rule and regulatory guide is to reduce the prescriptivenest, which unnecessarily restricts applicants and licensees (negative impact),

while continuing to ensure the health and safety of the public.

This final rule and regulatory guide will result in a s .gnificant flexibility and cost benefit for some plants. However, application of the rule will 5e optional, so that each applicant or licensee may pe' ~.m their own 'malysis to determine whether or not they should avali themselves of the greater flexibility permitted. The value of this rule may have some negative aspects, since an increase in plant power could increase risk to the public. This increar' in risk will be shown to be negligible.

Information used in this regulatory analysis was obtained from a number of sources:

Page 2 ENCLOSURE I

(1) Previous studies sponsored by DOE.(1,2)

(2) Formal responses frrm the major reactor vendors to a request by NRC for input.(3,4,5)

(3) Informal discussions with reacier vendor, utility, national laboratory, and NRC staff.

In summary, this regulatory analysis wili not be the same as a conventional value-impact analysis, because this rule reduces the prescriptiveness of former ECCS regulations in order to remove unnecessary operating restrictions. It is expected that a significant effect will be a cost savings to industry. While there may be a slight negative effect on risk, this effect may be off set for some plants. This rule will continue to provide a more than sufficient safety margin to ensure the health and safety of the public.

2. OBJECTIVES The objec61ve of this final rule and regulatory guide is to incorporate into the regulations the improved knowledge gained from recent research on ECCS performance, so as to remove unnecesrary operating restrictions.

3. ALTERNATIVES A number of alternative approaches have been considered by the staff, and each approach has been evaluates in terms of safety, impact l on the industry, NRC and industry resources required, and the

possibility of challenge both during tha rulemaking process and during application of the rule. The alternatives that were considered are

l A. Retain the existing ru;e with its present conservatism (no

! change).

I l

Page 3 ENCLOSURE I l

i B. Modify the ECCS rule, as stated in the advance notice of proposed rulemaking published in the Federal Register on December 6; 1978.

C. Modify only certain models contained in Appendix K for which research investigations have been completad and a well-documented data base exists. These changes have been l selected in areas for which new experimental data have shown j that the existing models contain a larger degree of  !

conservatism than justified by current data uncertainties or are obviously unrealistic, i

D. Eliminate the requirement to use Appendix K models and allow realistic models to be used. Reduce the 2200 0F and 17%

oxidation limits of 5 50.46 appropriately to ensure that sufficient conservatism exists to cover uncertainties in the realistic calculation.

E. Eliminate the requirement to use Appendix K models and allow realistic models with an evaluation of the uncertainty in~

the overall calculation added to the results, similar to that discussed in SECY-83-472,(8) to be used. The55b.46 limits of 2200 F and 17% oxidation would be unchanged.

These alternatives are discussed in Section 5.

4. CONSEQUENCES Since 'icensees can, at their discretion, continue under the existing requirements, one possibility is that a licensee's ccsts and benefits would be unaffected by this proposed action. It is only under those circumstances where a licensee's own analysis demonstrates a net benefit that the relaxation of these requirements would be adopted. Therefore, this regulatory analysis of the. relaxed requirements is not as specific in the identification and auantification of attributes as most regulatory analyses.

Page 4 ENCLOSURE I

o

- 4.1 Safety and Risk Effects The value of this rule must be evaluated in terms of the effect on safety. The rule will probably result in increased local power within the reactor core and possibly in increases in total power.

Power increases on the order of 5% will have an insignificant effect on risk. One effect of increased power will be to increase the fission product inventory in the fuel rods. A five percent power increase would result in a no more than a five percent increase in fission products. Therefore, the upper bound on fission products which would be released during core melt scenarios and potentially released to the environment during severe accidents would increase by no more than five percent.

This final rule still requires that the calculated fuel rod peak cladding temperature (PCT) remains below 2200 F. However, those reactors wit

• power increased by five percent could be operating with less margin be. tween the PCT and the 2200 F limit than previously. The increased risk represented by this decrease in margin and increase in fission product inventory is negligible and falls within the uncertainties of probabilistic risk assessment (PRA) risk estimates.(9) In addition, other safety limits, such as departure from nucleate boiling (DNB), and operational hardware limits, such as turbine design, will limit the amount of margin reduction permitted 4

under the revised rule.

There are also safety benefits derivable from alternative fuel management schemes that could be utilized when this rule is impl er, .nted. An important safety benefit mill be realized due to less stringent restrictions on core power peaking. High neutron leakages at the core outer beundary are used in PWRs to flatten the radial core power profile. This inefficient fuel management procedure is needed to meintain peaking within tight limits. In addition, the resultant high neutron fluence leakage can enhance vessel embrittlement resulting in pressurized thermal shock (PTS) concerns. The higher Page 5 ENCLOSURE I

/

• power peaking facto'rs'thr.t will be allowed will provide greater fuel management flexibility when attempting to reduce neutron flux at the

. vessel. This can result in a corresponding reduct ien in risk from PTS .

The reduced cladding temperatt:res that would be calculated under this rule offer the possibility of other design and operational changes that may result from the lower calculated temperatures. ECCS equipment numbers, sizes, or surveillance requirements might be reduced and still meet the ECCS design criteria (if not required to I meet other licensing requirements). The diesel / generator start time duration could be increased from the current technical specification limit of 10 seconds, typical for BWRs, to up to 70 seconds.III These

~

two potential modifications could result in offsetting effects on risk. On one hand, elimination of an ECCS component would tend to

' increase risk due to loss of redundancy (single failure criteria still required); however, increasing the time for diesel / generator start-up may result in improved operational flexibility and diesel reliability with concomitant decrease in risk. The improved diesel reliability could result from decreased stress, wear, and possibly test frequency during survcillance testing, as well as the effect of the relaxed de-sign requirement (70- versus 10-second start-up). In either case, any proposed changes in technical specifications of the facility woul'd be f at the option of the licensees and would include the consideration of the financial burden associated with technical specification changes.

i

)

In summary, the effect of the rule on safety will have both

positive and negative aspects. The potential for redt.. tion of ECC system requirements in existing or new plants is present, but the

[ likolihood of reduction of safety systems in operating plants is e).tremely low. However, several safety benefits will also be realized under the final rule. The net effect on risk is believed to be plant specific. The probability of a large break is so low, however, that the choice of best estimate versus Appendix K has little effect on 4

public risk as shown in a limited generic analysis of the affect on safety.I9) 4 i Page 6 ENCLOSURE I

• - .-,,e.,,,.--..,_,-_--m,, -

, - , , _ , _ . , , . ,,-,-,,,,,e ,.,,e,,----,.-e-ww,,3

o 4.2 Costs and Benefits LOCA ennsiderations resulting from the current requirements are restricting the optimum production of nuclear electric power in numerous ways. These restrictions can be placed into the following three categories:

(1) Maximum plant operating power.

J(2) Operational flexibility and crerational efficiency of the plant.

(3) Availability of manpower to work on other activities.

Maximum plant operating power at some nuclear facilities is limited by the present Appendix X licensing results. However, it can be very difficult to clearly separate these LOCA rule restrictions from other licensing issues and limitations. There are numerous limits that can restrict total plant power, as well as the ability to maneuver the power over a wide range. Typically, this limit is .

asscciated with either the PCT calculated to occur during LOCA transients or as a result of DNB restrictions. Additionally, there are limits to plant power because of NRC guidance on total allowable thermal power of 3800 MW and because of physical hardware limits in existing reactors on the balance of the plant (turbine, condenser, pumps, and steam generators).

Regarding the second category of operational flexibility, some plants have very little LOCA margin. Such a limited margin necessitates additional ' core power surveillance to prevent peaking factor violations. This may also require special supplementary nuclear er safety analyses and restrictive fuel management schemes, resulting in inefficient fuel burnup and no extended burnup cycles.

Page 7 ENCLOSURE '

_, . _. _my .-,

The third category concer u periodic reanalyses which are required by the current rule. I' an error (1 20oF change in PCT for the limiting transi(nt) is found in an accepted evaluation model, a new LOCA analysis must be performed immediately even if the error correction results in a decrease in PCT in the limiting transient.

This was a major problem several years ago, and several cases have i co occurred recently, with each reanalysis costing the licensee about $150KII) and diverting both licensee and NRC staff from other, more productive activities. One NRC staff year is normally needed to perform a reanalysis review when required because of errors discovered or because of other moderate changes to existing analyses. A current estimate of the incremental cost corresponding to one staff year of review time is $72K. New models and analysos would require about 3-4 staff years for review. Very often, this reanalysis contributes very little to plant safety. This final rule requires that a complete rea slysis be performed whenever changes to the evaluation model result in changes to the calculated peak clad temperature exceeding the original prediction by 150*F. If the new limiting transient calculation exceeds the criteria of 6 50.46(b), then immediate steps must be taken to achieve compliance. If the criteria of 9 50.46(b) are not exceeded, a reanalysis will be performed on a schedule proposed by the licensee and approved by the NRC.

The degree to which the rule will benefit a particular plant depends on how limited the plant is by these LOCA restrictions. The Babcock and Wilcox (B&W) and Combustion ingineering (CE) companies have informally indicated that they do not feel that the plants which they design are limited by LOCA restrictions and, therefore, B&W ano CE plants may not benefit from the first two categories. General Electric Co. (GE) plants do tend to be limited in operation by LOCA restrictions and will benefit from relief from LOCA restrictions.

However, this relief is already available for most GE plants through the recently approved SAFER evaluation model. Any additional relief Page 8 ENCLOSLRE I

due to a rule change will be of little further benefit.I4)

Westinghouse (W) plants are the only plants which appear to directly benefit from relaxation of LOCA limits. Forty-nine operating reactors, as well as ten under construction, are of the Westinghouse design. W indicates that most of these plants are limited by LOCA considerations.I3)

It can, therefore, be estimtted that there are at least 49 nurlear plants on line that are limited by LOCA considerations either in total power and/or in flexibility of operation. Up to ten additional plants being limited by LOCA considerations may also come 0

on line. Any rule change that produces a PCT decrease of 100 F can be translated into a total plant power increase of approximately 5%,

based on LOCA limit co.'siderations. This 5% increase in power represents a power increase which is within the capabilities of typical W plants based on existing hardware and is still well below other limits (e.g., DNB limits). Calculationr. show that this final rule change will provide a reduction in PCT of more than 100 F.(3,11,12)

The economic impact of this increase in power can be viewed in terms of energy replacement cost savings.II3) Since Westinghou'se plant,s would be most likely to implement these power upgrades, an analysis has been performed to determine the present values of the energy replacement cost savings which would be derived over the remaining life of each operating Westinghouse plant. The analysis was performed for the 49 plants currently operahie and used the following assumptions:

1. Replacement energy cost penalties are assumed to be constant in real terms over the remaining useful life of the reactor.

This means that costs are not assumed to increase faster than the rate of general inflation.

Page 9 ENCLOSURE I

8

's

2. The commercial operating life of a reactor is assumed to be 30 years. Thus, the remaining useful life of a reactor equals 30 minus the number of years in operation prior to 1987.

3. All costs will be expressed in 1985 constant dollars and discounted back to 1985 assuming a 10% real discount rate.

The final cost estimate will represent a 1985 present worth value in 1985 dollars.

4. The average cost to upgrade equipment in order to increase power is $150/kWe.(5)

Table A presents the average daily energy replacement cost for each Westinghouse plant, the yearly energy replacement cost savings resulting from a five percent power increase, and the present value (1985 dollars) of these cost savings over the remaining life of each plant. The total present value of the energy replacement cost savings for a 5 percent power increase in all 49 Westinghouse plants is estimated to be $2.8 billion dollars. Expressed in 1987 dollars, this figure converts to $3.0 billion dollars. As part of a sensitivity analysis, a 5% real discount rate applied to the same plants resulted in a total present value of energy replacement cost savings of $4.2 billion dollars in 1985 dollars. Neither of these values includes the cost to upgrade plant eouipment, which is small compared to the present value of the savings for most plants.

Page 10 ENCLOSURE I

. s o

TABLE A. PRESENT VALUE OF ENERGY REPLACEMENT COST SAVINGS DUE TO A 5% p0WER INCREASE (Millions of 1985 Dollars)

• Average Daily ** Yearly Energy Present Value Energy Replacement Replacement Cost of Cost Savings-Coat Savinas Over Plant Life Beaver Valley 1 .191 3.48 29.1 Beaver Valley 2 .208 3.80 35.8 D.C. Cook 1 .255 4.64 38.1 D.C. Cook 2 .265 4.83 44.2 Comanche Peak 1*** .736 13.43 126.6 Comanche Peak 2*** .736 13.43 126.6 Salem 1 .482 8.79 74.4 Salem 2 .494 9.02 80.9 Braidwood 1 .574 9.61 90.6 Bryon 1 .556 10.45 97.3 Bryon 2 .494 9.01 85.0 Callaway 1 .313 5.72 53.1 Kewaunee 1 .150 2.73 21.9 Point Beach 1 .145 2.64 18.'7 Point Beach 2 .145 2.64 20.1 Zion 1 .526 9.60 75.1 Zion 2 .526 9.60 75.1 Prairie Island 1 .126 2.30 18.0 Prairie Island 2 .129 2.35 18.9 Ginna 1 .271 4.95 35.2 Haddam Neck 1 .281 5.12 29.5 Indian Point 2 .466 8.50 68.2 Indian Point 3 .517 9.43 78.8 Millstone 3 .620 11.32 105.9

• Includes plant-specific capacity factors and power.(10,13)

• • Yearly savings associated with a 5% power upgrade.

• • • Based on assumed 1987 start-up using 1984 dollar estimates adjusted to 1985.

Page 11 ENCLOSURE I Lm._

2 TABLE A. ! cont.)

PRESENT VALUE OF ENERGY REPLACEMENT CSSTS DUE TO A 5% POWER INCREASE (Millions of 1984 Dollars)

• Average Daily ** Yearly Energy Present Value Enorgy Replacement Replacement Cost of Cost Savings Cost Savings Over Plant Life Catawba 1 .313 5.71 53.0 Catawba 2 418 5.73 53.6 Seabrook 1 .642 11.71 109.6 Yankee Rowe 1 .090 1.64 5.2 Farley 1 .396 7.22 61.4 Farley 2 .391 7.14 64.1 Harris 1 .340 6.20 58.5 McGuire 1 .440 8.04 72.1 McGuire 2 .440 8.04 74.1 North Anna 1 .343 6.27 52.4 North Anna 2 .344 6.28 35.6 H.B. Robinson 2 .242 4.42 32.6 Sequoyah 1 .196 3.58 32.1 Sequoyah 2 .203 3.70 33.6 Surry 1 .284 5.19 39.5 Surry 2 .282 5.15 40.3 Turkey Point 3 .350 6.39 48.6 Turkey Point 4 .3530 6.44 50.4 V.C. Summer 1 .324 5.92 54.6 Watts Bar 1 .186 3.40 28.7 Wolf Creek 1 .306 5.59 51.9 Diablo Canyon 1 .701 12.78 118.7 Diablo Canyon 2 .722 13.18 123.4 San Onofre 1 .288 5.26 34.2 Trojan 1 .380 6.94 58.0

• Includes plant-specific capacity factors and power.(10,13)

• • Yearly savings associated with a 5% power upgrade.

Page 12 ENCLOSURE I

u

• e l'

O Table B provides the present value cost savings for several different increments of power increase and does not include the average cost of upgrading plant equipment. This table' assumes a 10%

real discount rate.

TABLE B. .

CHANGES IN TOTAL PRESENT VALUE COST SAVINGS DUE TO A PERMANENT POWER INCREASE (Billiens of 1985 Dollars)

Power Total Present Value Increase Cost Savinos 1.0 % 0.56 2.0.% 1.12 3.0 % 1.68

~

4.0 % 2.24 5.0 % 2.80 6.0 % 3.36 7.0 % 3.92 8.0 % 4.48 9.0 % 5.04 10.0 % 5.60 These potential cost savings represent hypothetical maximum savings which do not include plant upgrade costs. It is difficult to estimate how many plants would take advantage of such a rule revision and upgrade power. Factors influencing the decision of an individual t

utility to upgrade a plant would vary and wculd depend upon the need for additional capacity and other means of obtaining additional capacity, other limits such as environmental factors (thermal i pollution), potential local opposition to plant modification, and plant-specific cost-benefit analyses.

1 Page 13 ENCLOSURE I

The reduced LOCA restrictions would result in further savings due to more efficient plant operation, irrespective of whether or not the total power of the plant was upgraded. These improvements include improved fuel utilization and improved maneuvering capabilities. Core management and advanced fuel management concepts are complicated subjects, and LOCA limits are only one of many factors to consider. (2) Thus, obtaining preciso estimates of potential savings is difficult. However, savings of 3 to 6 million dollars per plant per year would not'be unreasonable.(5) Even if a utility did not increase power or change fuel management, the simpler generic reload calculations possible with less restrictive LOCA limits would save

$250,000 per plant per year.II) 4.3 potential Impacts In contrast to the economic and safety conse'uences described in section 4.2, there are some detrimental impacts ti.ct might result from this rule. Some believe that a rule change might destabilize the present licensing process which is overly conservative but is well known and predictable, However, others believe that the licensing process will not be fully stable until the rule is revised. Any disruption of the licensing process is prevented by "grandfathe' ring,"

which will give utilities the option of adopting the new analysis methods or continuing with the old Appendix K procedures.

5. DECISION RATIONALE The various alternatives of Section 3 were considered and evaluated before finally settling on the recommended regulatory action. The consideratinos follow:

Alt A: Retain the existing ECCS rule with its present conservatism (no change).

PRO: a. The current well established and stable licensing process would be retained. ,

page 14 ENCLOSURE I

s

"^

b. No staff resources would be required for ,

rulemaking.

CON: a. Many plants would continue to be ,

unnecessarily restricted in operation by the current rule,

b. Many liconsees would continue to seek relief ,

from restrictions through requests for exemptions or by using the approach discussed in SECY-83-472.

Both these approaches are ir.terim measures and shou'd .-)t substitute for revising the rule to make it consistent with current knowledgi and practice. .

Alt. B: Modify the ECCS rule as stated in the advanced notice of proposed rulemaking published in the '

[

Federal Register an December 6, 1978; 43 FR 57157.

Consistent with previously stated plans and would PRO:

allow minor changes to be quickly implemented.

CON: a. Substantial changes would be delayed until a later (

phase,

b. Does not resolve comments received on the advanced notice recommending more substantial changes.

Alt. C: Modify certain models, discussed in Appendix K for which research investigations have been completed and a well-documented data base exists. These changes'would be selected in areas for wb'ich new experimental data have shown that existing models contain a targer degree of conservatism than justified by curren .ta uncertainties or are obviously unrealistic. ,

I I

Page 15 ENCLOSURE I

6 PRO: a. Plants'would no longer be limited in operation by current ECCS rule restrictions.

b. This would be in agreement with Commissior, policy (NUREG-0885) to incorporate the results of research into the licensing process.

CON: a. The revised ECCS rule might not contain sufficient, quantified conservatism to account for ,

calculational uncertainty. Additional analys2s would be required to demonstrate that sufficient conservatism remained in the calculation,

b. The ECCS rule would have to be changed in the future to make use of research results or other

.information which may become available.

Alt. D: Eliminate the requirement to use Appendix K models and allow realistic models to be used. Reduce the 2200*F and 17% oxidation limits of f 50.46 appropriately to ensure that sufficient conservatism exists to cover uncertainties in the realistic calculation.

PRO: a. Maximum use of completed research could be made in licensing to relax unnecessary operating restrictions. This would be in agreement with Commission policy (NUREG-0885) to incorporate the results of research into the licensing process.

b. Licensing models would provide more realistic calculations te allow more accurate determination of the effect of equipment changes or failures and operating procedures.

CON: a. The conservatism used to account for uncertainties would be fixed through the revised i 50.46 limits Page 16 ENCLOSURE I

. s and could not be varied to account for more accurate calculations of uncertainty which may be available in the future (i.e., little incentive for further improvement).

b. Additional ctaff resources would be required to establish fixed conservatisms applicable to all plant types,

c. The introduction of a less prescriptive rule would provide a greater opportunity to challenge licensing amendments.

Alt. E: Eliminate the requirement to use Appendix K models and allow realistic models combined with an evaluation of the uncertainty in the overall calculation, similar to that discussed in SECY-83-472. The 3 50.46 limits of 2200 F and 17% oxidation would be unchanged.

PRO: a. Maximum use of completed and future research could be made in licensing to relax unnecessary '

operating restrictions. This would be in agreement with Commission policy (NUREG-0885) to incorporate the results of research into the licensing prccess.

b. Licensing models would provide more realistic calculations to allow more accurate determination of the effect of equipment changes or failures and operating procedures.

c. The uncertainty evaluation would quantify the conservatism in the calculations which could change as the accuracy of the calculations improved.

. Page 17 EliCLOSURE I

J

d. The industry and NRC staff are already investing effort to follow this approach.

CON: The introduction of a less prescriptive rule would provide a greater opportunity to challenge licensing amendments.

In all alternatives considered, ths current Appendix K would remain available for those applicants or licensees not desiring to use a revised evaluation model.

The staff believes that Alternatives A and B, which would provide little or no change in the ECCS rule, are unacceptable. The ECCS rule should be changed because:

(1) A data base now exists that supports relaxation of the ECCS rule.

(2) A revised ECCS rule would remove unnecessary operating restrictions on plants.

(3) Almost all U.S. research on LOCAs has been completed.

The remaining portions of the MIST and 20/30 progrars are expected to provide valuable information for assessment of models, but should not affect the proposed rule.

(4) Nuclear reactor vendors are currently working on future plant designs which would be. influenced by the revised ECCS rule. .

The staff has also considered Alternative C, which would modify certain models in Appandix K for which research investigations are completed. The revisions considered under Alternative C include:

' Page 18 ENCLOSURE I

5

a. Reanalysis requirements

b. Post-critical heat flux heat transfer

c. Return tv nucleate boiling

, d. Refill and reflood transfer (steam cooling below reflood rate of 1 inch per second)

e. Fission product decay

f. Metal-water reaction

g. Discharge model

' Based en recent supporting analyses performed by vendors and national laboratories, the staff has determined that if Appendix K were to be revised according to Alternative C, it is nossible that the emaining overall conservatism in the evaluation models would be on the same order or less than the uncertainty of the calculation.I4) This would be unacceptable, since one could no longer assume that Appendix K contained sufficient conservatism to account foe the total uncertainty in the calculation. Thus, use of Alternative C without supporting uncertainty analysis was dropped from consideration.

The staff also censidered revising Appendix K in a manner similar to Alternative C, but requiring an additional uncertainty analysis to ensure that the evaluation model contained sufficient conservatism. This option would require two calculations, a '

- realistic calculation with uncertainty analysis and an evaluation model calculation. This option was also rejected, since the licensee would be required to perform two calculations, of which one (i.e., the evaluation model) would provide little benefit to safety.

Page 19 ENCLOSURi !

• Based on the current understanding of ECCS cerformance, the approach of a prescriptive /ppendix K is no longer believed to be appropriate. A realistic calculation, taking into account the overall uncertainty in the analyses, is believed to be the correct approach to ensure the safety of the public without unnecessarily restricting applicants and licensees. Thus, the ECCS rule should be revised accordingly and the requirement to use Appendix X eliminated. Therefore, Alternatives D and E were considered, both of which use realistic calculations. The difference between the alternatives is in the treatment of uncertainties. Alternative D would reduce the i 50.46 limits of 2200 F and 17% cladding oxidat'on to cover uncertainties in the calculation and uncertainties in the point at which substantial core damage would occur. Alternative E would require an uncertainty factor to be added to the best estimate calculation. Alternative D was not selected because (1) the i 50.46 limits of 2200 e and 17% cladding oxidation are believed to be appropriate and conservative limits below which substantial core damage will not occur,(14) (2) the conservatism used to account for uncertainties in the calculation would be fixed and could not be varied to account for more accurate calculations, and (3) further staff analyses would be required to support establishing these limits.

It was decided that Alternative E be adopted. This alternative will require that the licensee show that the criteria of f 50.46 are met using a realistic calculation combined with an evaluation of the uncertainty of the overall calculation. This uncertainty evaluation, combined with the additional conservatism in the 4 2200 F peak clad temperature and the 17% cladding oxidation criteria, should ensure a negligible risk to the public. This approach to licensing is consistent with the interim method discussed in SECY-83-472, except that the additional Appendix K calculation, which contributes little to safety, would not be required. As discussed in SECY-83-472, the combined conservatism in Appendix X methods resulted in calculated peak clad temperatures of nearly 22000F, Extensive research has shown .

Page 20 ENCLOSURE I

s that the Appendix K models provide excess conservatism in the calculated PCT, and t'hereby afford greater margin.than deemed necessary. Realistic calculation methods indicate that tha peak 0

cladding temperature during a LOCA ranges from 1400 F to 1700 F.

Therefore, the margin of excess conservatism in the current Appendix K aopears to be large. Thus, increases in the operating linear heat generation rate can be obtained while continuing to meet the criteria of 50.46(b). Further, Appendix K would remain available (with minor modifications) as an alternative.

Therefore, licensees and applicants who neither need nor desire relief from current operating restrictions would have no new requirements and could continue to meet existing Appendix K requirements. The burden of performing new calculations will be placed only on those applicants and licencees who elect to gain relief from LOCA restrictions.

This rule will also provide relief from the reanalysis requirements that do not contribute substantially to safety and

! will allow use of research data that has been obtained since the current rule was written. The rule will allow applicants and licensees relief from unnecessary operational restrictions

, resulting from loss-of-coolant accident (LOCA) analyses an'd still result in an adequate level of conservatism in the ECCS analyses.

The net effect will be to allow increased operational flexibility in the form of increased linear heat generation rates and more optimum fuel utilization while retaining the conservative margins set forth in 50.46(b).

6. IMPLEMENTATION i

6.1 Schedule J

No implementation problems are now anticipated. Each applicant I or licensee may, at its discretion, continue to utilize the existing Appendix K criteria for development of the ECCS evaluation model.

With regard to the reporting of significant changes to the accepted l '

i l

Page 21 ENCLOSURE I

[ ~

4 evaluation models, the rule provides for the establishment af a mutually agreed upon schedule for completing required actions.

6.2 Relationship to the Existing or Proposed Requirements In view of the fact that an integrated schedule is to be used for prescribing necessary actions, it is not expected that actions resulting from other requirements will be seriously affected.

A backfit analysis, developed primarily from this regulatory analysis has been included in the Notice of Rulemaking.

7. VALUE/ IMPACT OF REGULATORY GUIDE Having decider,to proceed with Alternative E for the promulgation of tha ECCS rule, the staff considered the question of whether or not i regulatory guide should be prepared and what form such a guide mlght take.

7.1 ALTERNATIVES CONSIDERED The staff identified three viable alternatives concerning the preparation of a regulatory guide. Each alternative has been evaluated using the same criteria as those applied to the rule.

The alternatives considered are:

A. Do not prepare a regulatory guide and rely on licensees existing expertise with thermal-hydraulic and interactions with NRC licensing staff to ascertain compliance.

B. Prepare a regulatory guide which describes technical areas to be considered, gives examples of acceptable features and data bases of best-estimate models, and generally describes uncertainty estimation methods.

l C. Prepare a prescriptive regulatory guide which identifies the best-estimate models, data, and l

Page 22 ENCLOSURE I i l

s correlations which should be used and the uncertainty methodology which should be applied.

7.2 DECISION RATIONALE The various alternativos of section 7.1 were considered and evaluated before deciding upon the approach taken. The considerations follow:

Alt A: Do r.ot prepare a regulatory guide and rely on licensees existing licensee expertise with thermal-hydraulic and interactions with NRC licensing authorities to ascertain compliance.

PRO: a. Great latitude would be afforded the licensee in complying with the rule.

b. NRC staff time to prepare such a guide would be saved.

CON: a. May be more costly to both the NRC and licensee because more potentially unacceptable or questionable licensee submittals would be expected,

b. There is limited experience in performing uncertainty analysis and licensees would be precluded from having NRC's views on how it might be done.

Alt 8: Prepare'a regulatory guide which describes technical areas to be considered, gives examples of acceptable features and data bases of best-estimate models, and generally describes ,

uncertainty estim: tion methods.

Page 23 ENCLOSURE I

PRO: a. Retains the flexibility of Alt A but allows for some features to be preapproved as acceptable should licencees want to use them.

b. Less costly for NRC review because many features will be familiar and preapproved.

CON: a. Regulatory guides are sometimes treated as "defacto" regulations by the licensee,

b. Limits NRC flexibility to determine deceptability of models on a Case by Case basis.

• Alt C: Prepare h prescriptive regulatory guide which identifies the best-estimate models, data, and correlations which should be used and the uncertainty methodology which should be applied.

PRO: a. Less costly to both NRC and licensee because prescriptiveness will result in less interaction,

b. Compliance should be straightforward because of similarity to existing Appendix K approach, CON: a. Prescriptiveness will prevent licensees from utilizing current or future safety research in their best-estimate models so that they will not be able to take optimum advantage of the rule. .

Page 24 ENCLOSURE I

5

b. Prescriptivens's could result in specification of features in excess of what is required for 1 safety without giving licensees the opportunity to justify alterr.atives.

The staff concluded that the disadvantages of Alternative C far outweighed the advantages particularly in light of the problems with the current Appendix K approach. In addition, this approach afforded licensees the least amount of flexibility in complying with the revised rule's features. Therefore, Alternative C was eliminated from consideration. Alternative A was attractive from the standpoint of allowing maxinum flexibility but was thought to be the most burdensome to both licensee and NRC because no information about acceptable ways of meeting the requirements would be available. This is seen as a serious deficiency particularly since the uncertainty quantification is relatively new to most licensees. This I alternative would be the most disruptive to the licensing process and would most likely cause delays in obtaining NRC approval.-

Therefore, the staff concluded that by adopting Alternative B optimum flexibility and guidance are provided.

Page 25 ENCL.05URE I

.J' o

REFERENCES

1. "Revision of Loss-of-Coolant Accident (LOCA) Rule Licensing Requirements - A Survey of Opinion within the n> clear Industry,"

NUS-4221, April 1983.

2. P. Wei, et al, "Boiling Water Reactor Uranium Utilization Improvement Potential," GEAP-24965, June 1980.

3. Ltr fm Hochreiter to Beckner, "Results of Westinghouse Calculations Using a Modified Version of Appendix K," April 23, 1984.

4. Ltr fm Quirk to Beckner, "Impact of Proposed Appendir K Rule Changes on General Electric SAFER /GESTR ECCS Results,"

July 26, 1984 (Proprietary).

5. Ltr Rahe to Ross, "LOCA Margin Benefits," February 8, 1985.

6. "Probabilistic Risk Assessment (PRA) Reference Document,"

NUREG-1050, September 1984.

7. "Effect of Diesel Start Time on BWR/6 Peak Cladding Temperature,"

NSAC-96, January 1986.

8. "Emergency Core Cooling System Analysis Methods," SECY-83-472, November 13, 1983.

9. "Compendium of ECCS Research for Realistic LOCA Analysis," Draft NUREG-1230, Chapter VII (to be published).

10. J. C. VanXuiken, et al, "Replacement Energy Costs for Nuclear 1

Electricity Generating Units in the United States: 1987-1991,"

NUREG/CR-4012 Vol 2, January 1987.

Page 26 ENCLOSURE I

l 4 l

11. "Assessement of Proposed Changes to Appendix K on LOCA Limits,"

B&W Report 77-1150444-00, April 1984.

12. Ltr Scherer to Beckner, "NRC Proposed Changes to 10 CFR 50, l

Appendix K," LD-84-021, May 25, 1984.

13. S.W. Heaberlin, et al., "A Handbook for Value-Impact Assessment," NVREG/CR-3568, Pacific Northwest Laboratory, December 1983.

14. Van Houten, "Fuel Rod Failure as a Consequence of Departure from Nucleate Boiling or Dryout," NUREG-0562, June 1979.

I Page 27 ENCLOSVRE I

.-c i

'I ENCLOSURE J Margin Inherent in the 2200*F Limit The Commission paper (SECY-86-318) which transmitted the notice of proposed rulemaking (Enclosure A of this package) also contained an enclosure (Enclosure C to SECY-86-318) which addressed the conservatism in Appendix K and in 50.46. That document stated that the criteria "of 17% maximum allowed cladding oxidation, combined with the 2200*F limit, is an apprcpriate limit --- , which will ensure survival of the cladding following reflood." This conclusion was based on a large body of research mandated by the Comission when 50.46 was originally promulgated. This conclusion regarding the performance limits of 2200*F and 17% oxidatier, as they relate to cladding embrittlement is well founded and equally valid today.

In SECY-86-318 the staff also addressed the question of the 2200 F limit j

as it related to the contribution of the metal-water reaction to the burden for heat removal. The staff stated at that time that additional conservatism eristed in the 2200*F peak cladding temperature (PCT) limit in Section 50.46 for ECCS analysis. It was further stated that this conservatism was abcut 400'F based on data from the Power Burst Facility l (PBF) severe accident experiments which showed that "at a temperature of 2600'F, the steam-zircaloy reaction becomes sufficiently rapid to produce an auto catalytic temperature excursion (i.e., cladding oxidation is a self-sustaining process). Thus 2600'F is a real limit, above which significant fuel damage would occur." The 400*F margin between 2200*F and 2600'F was also referred to in the rulemaking notice (Enclosure A of l this package). On this basis it was recomended that the 2200*F limit not l l

be modified. Additional experimental and analytical information from PBF ano other sources have been examined which show that some rapid temperature rises comparable to the original PBF data begin below and i

2 o

2-i some above 2600'F. Thus, the original basis for the 400*F margin proposed in SECY-86-318 was questioned. A careful examination was perfomed for cladding temperature transients in both design basis LOCA l analysis and in severe accident experiments and calculations. It was l

observed that even more rapid temperature rises occur in design basis LOCA analysis beginning at temperatures as low as 1200'F during refill, f

l but these rapid rises are easily arrested with nominal reflood heat transfer. The rapid temperature rise in design basis analyses is due to the high decay lieat which occurs early in time, and other conservative l l assumptions used regarding power peaking factors. The rapid temperature l l rises observed in severe accident experiments are not teminated because

\

i the fluid flowing past the fuel rods and hence the heat transfer is j l markedly less than that observed in reflood LOCA analysis and f experiments. This condition exists because severe accident experiments ]

and analyses are performed with no intentional core cooling in order to j achieve and study degraded core conditions. Since severe accident  !

thermal behavior is so different from LOCA behavior there is very little relevance to the issue of performance limit margins. Therefore, the staff does not believe that 2600 F should be characterized as a "real"

)

l limit above which phenomena would occur which make significant ccre I

ddmage inevitable. Other recent experiments show that control rod f ailure can occur below 2200'F in such a way that fuel cladding damage j t

i may result. Some of these experiments are discussed in the Compendium of I ECCSResearch(huREG-1230). Because control rods are not a heat source, control rod temperatures would be substantially below fuel rod l

temperatures during a large break LOCA. Only if fuel cladding temperatures were substantially above 2200*F would thermal radiation from fuel to control rods cause the control rods to reach temperatures at which they would fail. To achieve this condition would require a -

l I

,4

,3-significant degradation in reflood heat transfer or increase in rod power. Therefore, even though some severe accident research appears to show lower thresholds for temperature excursion or cladding failure than- l previously believed, when design basis heat transfer and decay heat are

- considered, some margin above 2200*F exists, but it is too uncertain to outati'y at this time. We do not believe any information suggests that the 2200*F limit, and most certainly .the 17% reaction limic are the edge of a cliff. Nevertheless, the viability of this rule does not' depend on the existence of any margin above 2200*F, only that quantifiable nargin f

exist between the calculated results and the 2200*F limit, i I

i 1

l

)

i l

l l

4  !

l l

i l

1 l

i

d%%d%%%%WNk$WW64%%Wdh%fWdnfWgW@dffffffffi g g g g , 3 TPAH5MITTAL TO: I Document Control Desk, 016 Phillips ADVANCE 0 COPY TO: The Public Document Rocm DATE: // f[

FROM:

SECY Correspondence & Records Branch I Attached are copies of a Comission meeting transcript and related meeting i document (s). They are being forwarded for entry on the Daily Accession List and

[ placement in the Public Document Room. No other distribution is requested or j required.

Meeting

Title:

h/ 0 A i el I

I Meeting Date: f/khb

//

Open N Closed Item Description *: Copies '

g Advanced DCS

• 8 ~

g to POR C3 i

1 1

1. TRANSCRIPT 1

l I

2. b #[- /6 d M /

1 f 3.

~

4.

N

5. __

6.

• POR is advanced one copy of each document, two of each SECY paper, b g C&R Branch files the original transcript, with attachments, withcut SECY h -

papers.  ; \ &

2 h

alR6 k '

hl hlklk l bb i l l lYiYlYlk lflflhlflhlfhflflflflfl