ML20236U832

Rulemaking Hearing on Acceptance Criteria for ECCS for light-water-cooled Nuclear Power Reactors

ML20236U832
Person / Time
Site:	Waterford
Issue date:	12/28/1973
From:	Bender P US ATOMIC ENERGY COMMISSION (AEC)
To:
Shared Package
ML20236U831	List: ML20236U829 ML20236U832
References
CLI-73-39, NUDOCS 9807310104
Download: ML20236U832 (66)

Text

l I

\

UNITED STATES OF AMERICA ATOMIC ENERGY COMMISSION COMMISSIONERS:

Dixy Lee Ray, Chairman Clarence E. Larson William O. Doub William E. Kriegsman William A. Anders in the Matter of RULEMAKING HEARING Docket No. RM 501 ACCEPTANCE CRITERIA FOR EMERGENCY CORE COOLING SYSTEMS FOR LIGHT-WATER-COOLED NUCLEAR POWER REACTORS December 28,1973 The following opinion of the Commission, the concurring opinion of Commissioner Anders, and the Appendix to the Commission's opinion are hereby issued this 28th day of December,1973 in Washington, D.C.

l By the Commission  !

Paul C. Bender Secretary of the Commission OPINION OF THE COMMISSION

1. INTRODUCTION The Atomic Energy Commission herewith announces its decision in the rulemaking proceeding concerning' acceptance criteria for emergency core cooling systems for light water cooled nuclear power reactors. The subject of emergency core cooling systems (ECCS) has become a focal point of attention for those concerned with the safety of nuclear po' wer plants. As the massive record developed during this rulemaking shows, a wide spectrum of knowledgeable opinion exists concerning the adequacy of our current regulation on this subject-the Interim Acceptance Criteria-and with respect to the nature and scope of regulations which should be adopted at the present time. We have carefully considered the entire record and the many points of view it encompasses in reaching the decision we announce today. We believe that our decision affords the required reasonable assurance of protection for the public health and safety with a substantial margin. In this introduction we briefly review the history of this long proceeding and explain the principal reasons underlying the key elements of the decision. The introduction concludes with a brief discussion summarizing the technical context of the issues presented and outlining the changes introduced. The remaining sections set forth these changes and discuss the reasons for them in detail.The Appendix to this decision set forth the amendments to 10 CFR Part 50 which incorporate the rule announced herein, in the format in which those changes will be submitted to the FederalRegister.

On June 29, 1971, we published an immediately effective interim statement of policy establishing interim acceptance cri eria for emergency core cooling systems for light water-cooled nuclear power reactors (36 F.R.12247). These criteria, which were adopted following a review by the Commission Regulatory Staff and the Advisory Committee on Reactor Safeguards, provided the basis for our belief of 9807310104 980729 PDR ADOCK 05000382 1085 G PDR

reasonable assurance that such systems would be effective in the highly unlikely event of a loss-of coolant accident (LOCA). The notice requested comments from interested persons and additionally stated that we would consider holding a public rulernaking heanng on this interim policy statement. Thereafter, on November 30,1971, we announced our decision to hold a legislative type rulemaking hearing for the purpose of aiding us in our determination as to whether or not the interim policy statement of June 29, 1971, should be retained as issued or criteria should be adopted in some other form. Expanded ground rules published on January 8,1972, established procedures, including broad rights of cross-examination, to guarantee development of a record that would be as complete as possible The Hearing Board consisted of Nathaniel H. Goodrich, Esq.,presidmg, Dr. Lawrence H. Quarles, and Dr. John H. Buck. Participation in the rule making hearing was extensive. The primary participants included the Commission Regulatory Staff, four reactor manut'acturers, a consolidated group of electric utility

! companies, and the Consolidated National Interveners (CN!), a group of about 60 orgamzations and i

individuals. In addition, three states, the Lloyd Harbor Study Group, and severalindividuals participated to a lesser degree. The hearings lasted a total of 125 days and generated a record of more than 22,000 pages of transcript and thousands of pages of written direct testimony and exhibits. We heard oral argument by the

, seven principal participants on October 9,1973.

Simultaneously with its participation in the hearings, the Regulatory Staff prepared both draft and final environmental impact statements in implementation of section 102(2Xc) of the National Environmental Policy Act. A separate phase of the hearings was scheduled to permit questioning of the Regulatory Staff witnesses on the Final Environmental Statement, and of other participants concerning their comments on the Draft Environmental Statement. In its order of June 13,1973, the Hearing Board established a schedule

for questioning which conditionally allotted approximately five-and-one-half days for questioning by CN1 dunng this phase, subject only to reasonable advance specification of the subject matter of such questioning. More than two additional days were allotted by the Board in its Supplemental Order of  ;

July 13,1973, for other participants to question CN1 witnesses on their environmental testimony. j Dissatisfied with these reasonable limitations, which were entirely consistent with our order of 1 December 12,1972, CN1 withdrew its additional direct testimony and declined to participate in the environmental phase of the hearings. See " Statement of CN1 With Respect to Board Orders of June 13, July 10, and July 13,1973 " docketed July 23,1973. This incident is especially disappointing to us, for it was largely at the request of constituent members of CNI that we experimented with the substantial use of adjudicatory-type procedural features in this rulemaking. Moreover because of its broad based make up, CNI might well have made a meaningful contribution to the environmental phase of the hearings. At oral argument before us, it was apparent that CN1 differed sharply from the other participants with reference to cost benefit balancing. The absence of any development of its views on the record is therefore particularly ,

disconcerting. I We followed the ongoing ECCS hearings with great interest. On numerous occasions during the course of those proceedings, questions were certified to us, and we issued several interlocutory orders dealing with, Inter alia, discovery, subpoenas, treatment of proprietary information, and the permissible scope of i questioning. We reaffirm all of these previous rulings, except insofar as we may specifically depart from them in this discussion.

During the course of these proceedings the Hearing Board issued nearly 100 written orders, often accompanied by lengthy opinions. Many more oral rulings are embodied in the transcript. CNI, in its concluding statement, referred to its " continuing exception to each adverse ruling in the proceeding,"8 though it declined to specify where, within this massive record, each of the alleged " adverse rulings" occurred. We have dealt with the major contentions presented by CNI, and by the other participants. We have not, nor do we believe we should have, scrutinized every page of the record for the possibility of minor procedural errors, which may well exist. While we might have differed with the board on occasional details, we conclude, on balance, that the Board generally exercised its discretion in an appropriate manner so as to develop a record-tested by abundant cross examination-more than adequate for the formulation of the rule we announce today.

A major item of controversy was the method and manner by which the views of the Advisory Committee on Reactor Safeguards would be solicited. The Committee's views concerning the interim

' Concluding Statement-Safety Phase, of Participant Consolidated Nationalintervenors, March 15,1973, p. 2.5.

I 1086 j

V Acceptance Critena were contained in a letter of January 7,1972, which became part of the hearing record.

In an order dated January 26,1972, we denied the request of one participant to subpoena one or more members of the ACRS in light of several considerations. Nevertheless, we did permit solicitation of an expansion of the views of the Committee as a body through the use ofinterrogatories,a procedure which was subsequently followed. We reaffirm our belief that this procedure struck an appropriate balance between the competing concerns, permitting a useful additional viewpoint to be added to the record, without unduly inhibiting the deliberative process by which the Committee fulfills its statutory purpose of advising the Commission. Following the certification of the evidentiary record to us, we again solicited the views of the Committee. These were furnished by letter of September 10,1973, which was served upon all hearing participants who were requested to comment thereon at the oral argument before us on October 9, 1973.

Another sigmficant area of controversy during the proceedings related to the permissible scope of subject matter to be included. In Chapter 111 ofits concluding statement,CNI takes strenuous objection to rulings that certain subjects-defense.in depth, causes of a loss of-coolant accident, the technical basis for applying a single failure criterion, the matter of ECCS design and design changes, and finally the question of fuel densification-were becond the scope of the hearing. For a variety of reasons we find CN1's arguments unpersuasive.

There was an obvious practical necessity to place some reasonable limitations upon the subject matter of the proceeding. Otherwise there would have been no way to compile a meaningful record and reach a decision within a reasonable time. As a practical matter, the proceeding could not have covered every conceivable technical question arguably bearing upon the subject matter. This proposition has an important corollary-the rule announced can be no broader in scope than the record supporting it. Thus the rule deals with matters developed on this record; matters outside the record form no part of this decision.Where ,

relevant, que'stions not explicitly addressed by the rule must be considered on a case-by case basis in individual licensing proceedings. For example, we specifically require that the matter of fueldensification be taken into account in analysis but do not specify the method (see discussion on calculation of the initial  ;

stored energy in the fuel, pp. I1001102 infra). Matters such as pressure vessel integrity and steam i generator tube failure are the subject of other Commission regulations which must be complied with in every case. A showing of compliance must be made on the record of each case. Moreover, our Rules of Practice contain a mechanism whereby parties may show that such circumstances with respect to a particular reactor are such that a regulation would not serve its purposes.10CFR 2.758. See also Consolidated Edison Co. (Indian Point Unit No. 2), memorandum and order dated October 26,1972, dealing with the admissibility of evidence as to pressure vessel integrity.

Three additional . matters-the causes, probability and consequences of a LOCA were properly excluded from the " safety" phase of the hearing, which started with the assumption that the highly unlikely LOCA had occurred, for whatever reason. That phase involved the question of the validity of j performance enteria for systems designed to mitigate the admittedly severe consequences of such a l postulated accident. With respect to the environmental phase of the hearings, however, these subjects assumed a far more prominent and relevant position. There the focus of att:ntion was upon the balancing of costs-including potential adverse environmental effects-and benefits, the most notable of which is an increase in the margin of safety. CNI, which now argues that these subjects should have been treated in greater depth, elected not to participate in the environmental phase, where discussion of such subjects would have been appropriate.

The rnatter of ECCS design and design changes is beyond the scope of the instant proceeding for yet another reason. Our General Design Criterion number 35 (10 CFR part 50, App. A, Criterion 35) requires provision of "a system to provide abundant emergency core cooling." That criterion embraces the concepts of performance (reasonable assurance that the system will cool the core) and reliability (reasonable assurance that the individual components will work). Acceptance criteria for such systems-both our earlier Interim Acceptance Criteria and the r.rw criteria announced today-relate only to the performance  !

aspect of this design criterion. In other words, ECCS acceptance criteria establish limits on design parameters (in terms of quantities such as time and temperature), which, if not exceeded, would provide assurance that the intended function of cooling the core will be accomplished by the specific system j -

provided in any given case. These ECCS cnteria establish a uniform set of standards by which the combination of design and operating limits can be judged acceptable or net solely from the standpoint of 4

1087 J

y 1l

}

calculated clad damage shoula an unlikely loss-of coolant accident occur.The adequacy or superiority of a specific ECCS design in all respects would draw on more generci considerations, end so is a subject peculiar

)

to the specific licensing case. As noted, the remaining aspect of criterion 35, reliability, deals with the separate question of whether particular designs will work as intended and meet the limitations imposed by

{

performance criteria. This question remains proper for individual licensing cases and not for generic rulemaking.

The rule we announce today is supported in every respect by the evidentiary record of the hearings.

This is consistent with the procedures we spelled out in the Supplemental Notice of Hearing. 37 F.R. 288.

(January 8,1972). Those procedures also included the following proviso:

if reliance is placed on information which is not in the record, notice will be given of such informaticm and an opportunity provided to comment thereon and to request an opportunity to respond thereto.

j Several requests to consider material furnished to the hearing board and to us subsequent to the conclusion f

of the safety phase of the hearings are before us. The material referred to in these requests encompasses:

l (1) Appendix B to the Concluding Statement on Behalf of Babcock and Wilcox (February 22,1973); '

(2)The information referred to in Attachment A of the February 22,1973, letter of Transmittal accompanying the Concluding Statement of Position Submitted on Behalf of Combustion Engineering. Inc.;

and (3) The October 26,1973, letter (with enclosures) from counsel for Babcock and Wilcox to each of the Commisuoners. In each instance, the submissions have been furnished to all other participants.

We have examined each of these extra-record submissions but have placed no reliance on them in reaching our decision. We agree with the Regulatory Staff that ample opportunity was afforded to all participants in the proceeding to develop their case by way of sworn testimony subject to questioning at the time of the hearings. Accordingly, we deny the requests to utilize the mechanism of Rule 2 of the Supplemental Notice of Hearing to elicit comments and responses with respect to these belated submissions. - t in adopting this course, we are not blinding ourselves to new knowledge acquired as a result of ongoing research. On the contrary,'we believe that it is important that research programs-both analytical and experimental-continue, in order that wh may increase the knowledge relevant to ECCS performance. The nuclear industry and the Commission have several such programs underway at the present time.

We are aware that some information exists that may permit a more liberal allowance on maximum f

calculated clad temperature than our present criteria provide. This information is not unambiguous, ,

however, and there has been no adequate exploration ofit even if these extra-record submissions were to be I considered along with all of the record evidence concerning this subject, We have recently directed the l Director of the Division of Reactor Safety Research to give priority attention to study to determine more exactly the temperature at which clad embrittlement ceases to be simply a function of oxidation. This is l the same subject principally discussed in the three extra record submissions referred to above. As new '

knowledge is acquired, the Commission will analyze it, and at an appropriate time consider the possibility of amending the rule we announce today. We do not, however, believe that the limited amount of comparatively recent knowledge now available, justifies delay in the issuance of a rule based upon extensive examination of this issue.

Proprietary Data The Commission memorandum of June 6,1972, issued in this proceeding, made clear that the ultimate rulemaking decision in this case would not necessarily result in permanent protection of data claimed to be proprietary. We there recognized the " strong public interest in conducting a rulemaking proceeding which is f as open as possible to full public scrutiny," and explained that open consideration of the technical issues I was a motivating factae in the experimental use of a public rulemaking hearing. In that vein, we approved j procedures whereby all counsel-including those representing competitors of the participant offering the data-were permitted to examine data claimed to be proprietary, and to attend and participate in in coment sessions involving such data. Subject to review by the board and by other participants, each counsel was allowed to submit issues involving the proprietary matter to his client for consultation or guidance.

With respect to the decisional phase of the case, we explicitly stated:

. . . we wculd underscore that our present holding is confined to treatment of proprietary information during the hearing phase of this proceeding. Should such information form part of the basis for the 1088 s .

_ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ - - Y

1 ultimate rulemaking decision, the Commission will again-and in that context-address the question of that information's public disclosure.

Subsequently, Combustion Engineermg (CE) voluntarily submitted data asserted to be proprietary with full knowledge that such data might be made public at a later time. Included within that submission is a chart appearing on page 6-5 of CE's Redirect and Rebuttal Testimony (Ex.1144). See page 1119 infra. This item consists of data experimentally contirming the adequacy of certain heat transfer correlations at low pressures. Each licensee must use one of these correlations in order to bring its reactor operations into conformity with the limitations here prescribed. The correlations are part of a rule which will be used by the entire industry in strengthening the degree of reasonable assurance of public health and safety. The data conforming the correlations are thus part of"the basis" for the rule.

Upon reexamining the matter of data claimed to be proprietary,in the context of promulgating the decision itself, we conclude-in the circumstances of this case-that this aspect of the CE data should no longer be protected and should be disclosed in the public interest. Accordingly, and unless good cause be shown, this portion of the submitted data will be placed in the public document room 30 days following the date of publication of the rule in the FederalRegister. Combustion Engineering and other participants may submit views as to the treatment of this item by filing appropriate papers with the Secretary within 15 days following the date of this decision.

For use in implementation of the rule, we have also approved two correlations claimed to be proprietary. First, we approve Westinghouse's transition boiling correlation (Exhibit 1I52, section 25).See infra, pp. 1109 1110, 1116 1118. Second, we approve General Electric's Hench Levy CHF correlation (referred to in Ex.1001, p. 4 21: Ex.132, p. C-9; and Tr.14184 er seq.). See pp. I109,11131115, infra.

Both of these items, previously approved for use under the interim policy statement (36 Fed. Reg.12247),

apply only to an individual company's evaluation models, and are more akin to matters involved in individual licensing proceedings, and thus must presently be considered to be distinct from the CE data.

Accordingly, our present inclination is to continue to protect these two items in a manner consistent with our approach in individual licensing cases (see 10 CFR 2.790)-subject, of course, to the outcome of our pending reexamination of policy and rules concerning data for which proprietary protection is requested (38 Fed. Reg. 31543). At the same time, we are prepared to receive com nents from all participants as to the public disclosure of these two items as well. Such comments may also be submitted by Gling appropriate papers with the Secretary within 15 days following the date of publication of the rule in the FederalRegister.

Our conclusion as ta the CE item is limited to the particular facts of this case.2 Existing rules (10 CFR 2.790) shall continue to apply to individual proceedings. Similarly, the conclusion as to this item is without prejudice to whatever determinations we may reach in the pending reexamination referred to above.

Implementation Schedule The matter of implementation of the rule ultimately adopted has generated strongly divergent views.

Indeed, at the oral argument on October 9,1973, the implementation schedule was the source of more controversy than any other issue. As appropriate in a rulemaking proceeding of this magnitude, the varied interests represented by the participants produced a broad range of views.

CNI, while arguing that the record supports no rule at all, apparently took the view that in any event cost benefit balancing was irrelevant to implementation. The Regulatory Staff, recognizing that certain social and economic costs would flow from implementation ofits proposal, argued that those costs should nonetheless be borne in order to fashion a timely response. The majority of the industry participants, on the other hand, argued that the various costs were so burdensome (and the benefits so minimal) as to 8 Among the factors which combine to make this case unique are the following:

(1)This is a rulemaking proceeding. involving Commission reliance on the CE data, claimed to be proprietary,in promulgating safety rules applicable to all vendors and licensees.

(2) This proceeding was conducted under its own rules and not under the rules of general applicability. See 10 CFR Part 2, Subpart G.

(3) Prior to the voluntary submission of the item, all vendors were placed on notice of the fact that proprietary data forming a part of the basis for the ultimate decision might be publicly disclosed.

l 1089

)

i warrant an implementation schedule which would accord with the normal reactor fuel cycle-which means that for some reactors a rule would not be fully effective for as long as three years. Finally, one participant argued for permanent exemption of existing plants.

i We note that the calculations that must be made to conform to the rule will be time consuming.They will require analysis well beyond that defined in the Interim Policy Statement. Additional phenomena must be taken into account in the evaluation models: clad deformation. clad bursting. expansion of the gas in the gap between fuel and cladding, variable gap conductance (including r.h of densification). clad oxidation (two-sided oxidation, heat generation by oxidation). decay heat from u.fdes, flow redistribution. The sensitivity of calculated results to variations in noding and input paramett- must be explored in repeated computer runs. Calculations will be required for a spectrum of pipe break sizes. including splits as well as l double-ended breaks. At least three values of the discharge coefficient must be used for each break '

examined, spanning the range from 0.6 to 1.0. This large number of calculations will now be needed for all of the newer power reactors, as well as for other reactors currently under licensing review. Only the reactor vendors will have the capability of performing the analyses for the plants they have sponsored. Each vendor will suddenly be inundated by needs to provide calculations for all the plants it has designed.

We reject at the outset any schedule which does not take proper account of the steps needed for well-founded implen.mtation. It is evident that the vendors cannot develop the models and produce the necessary calculations in the four-month period proposed by the Staffin its Concluding Statement. This is so even though prudent manufacturers may well have done advance calculations upon the assumption that the Staff's proposal would ultirnately be the rule, inasmuch as our rule is somewhat different from that proposed by the Staff, the calculation process will have to begin anew. It is also unlikely that the AEC evaluation model can be completed and proof tested to be ready in a four month period. We do not wish to require a schedule for implementation that makes impossible demands on either the vendors or the l

Regulatory Staff or that might lead to taking hasty steps on the basis of unconsidered analysis, perhaps

~  !

requiring retraction or revision later. Rather, we wish to ensure that the calculations be thorough as well as timely. ,

l A requirement of immediate compliance would be tantamount to an order shutting down or substantially derating all reactors until requisite calculations were complete. We would not hesitate to take that action if circumstances so warranted. But the record shows, and we find, that the Interim Acceptance

)

Criteria will provide reasonable assurance of protection for the public health and safety during the relatively I brief transitional period which will culminate in compliance with the new rule. In addition, there is not sufficient basis for presently accepting tne view, espoused primarily by the utility and industry participants, (

which would defer implementation for as long as three years.

The evidence has demonstrated that a number of improvements could be made in the models used to evaluate emergency core cooling systems, and these have been incorporated into the rule we announce today. These proceedings have reduced the number of unknowns, and thus reduced the degree to which unknowns must be bridged by conservative safety assumptions. In short, the new rule provides a more objective basis for safety analysis. Because of our confidence in the Interim Acceptance Criteria, we have permitted them to remain in effect throughout our reevaluation. We also conclude that they will continue to provide reasonable assurance for a transitional period. However, since we now have an improved rule, supported by an evidentiary record, we choose to implement it at a rate which meets the following principles:

1.The scheaule should allow for thorough development of the complex evaluation models necessary to comply with the new rule; 2.The schedule should seek to effectuate the new rule's incrementalincrease in safety at the earliest practicable time;

3. If pouible, this goal should be accomplished without unwarranted disruption in the nation's production of electric energy, Guided by these principles, we adopt the following implementation schedule with respect to all facilities for which opetating licenses have previously been issued and for which operating licenses may issue during a period of one year from the date of this decision:

1.The rule shall become effective for the purpose of computing the time within which required or permitted actions must be done 30 days following its publication in the FederalRegister.

l l 1090

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ )

. 2. Within six months following said effective date. all licensees shall submit the requisite evaluation to the Director of Regulation for review. The evaluation shall be accompanied by such proposed changes in technical specifications or license arnendrnents as may be necessary to bring reactor operation in conformity with the rule.

3. Any licensee may request an extension of the six month period for good cause. Any such request shall be sub.nitted not less than 45 days prior to the expiration of the six month period, and shall be )

accompanied by affidavits showing precisely why the evaluation is not complete and the minimum time I believed necessary to complete it. The Director of Regulation shall cause notice of such a request to be published promptly in the Federal Register;such notice shall provide for the submission of comments by 1 interested persons within a time period to be established by the Director of Regulation. If, upon

- reviewing the foregoing submissions, the Director of Regulation concludes that good cause has been s shown for an extension, he may extend the six month period for the shortest additional time which in his judgrrant will be necessary to enable the licensee to comply with paragraph 2 above. Requests for extensions of the six month period, submitted under this paragraph, shall be ruled upon by the Director of Regulation prior to the expiration of that period.

4. Upon the submission of the matters specified in paragraph 2 above,(or under paragraph 3 above,1f ]

f-  ;

the six month period is extended), the facility may continue or commence operation within the limits of both the proposed technical specifications or license amendments submitted under the above procedure, and all tec5mical specifications or license conditions previously imposed by the Commission. ,

l S. Furmer restrictions on reactor operation will be imposed by the Director of Regulation if his review of the evaluations submitted under paragraphs 2 and 3 above so warrants.

6.Exemp*tions from the operating requirements of paragraph 4 above may be granted by the )

Commi91on for good cause. Requests for such exemption shall be submitted not less than 45 days prior to the date upon which the plant would otherwise be required to operate in accordance with the procedures of paragraph 4 above. Any such request shall be filed with the Secretary, who shall cause l

notice of its receipt to be pubhshed promptly in the Federal Register; such notice shall provide for the

! submission of comments by interested persons within 14 days following FederalRegister publication. The Director of Regulation shall submit his views as to any requested exemption within five days following expiration of the comment period.

7. Any request for an exemption submitted under paragraph 5 above must show, with appropriate l

affidavits and technical submissions, that it would be in the public interest to allow the licensee a

! specified additional period of time within which to alter the operation of the facility in the manner i required by_ paragraph 4 above. The request shall also include a discussion of the alternatives available for establishing compliance with the rule.

Description of a LOCA The Commission noted, in the Interim Policy Statement that:

Protection against a highly unlikely loss-of coolant accident (LOCA) has long been an essential part of the defense-in-depth concept used by the nuclear power industry and the AEC to assure the safety of nuclear power plants. In this concept, the primary assurance of safety is accident prevention by correedy designing, constructing, and operating the reactor. Extensive and systematic quality assurance practices are required and applied at every step to achieve this primary assurance of safety, l i

NeverWieless, deviations from expected behavior are postulated to occur, and protective systems are

! Installed to take corrective action as required in such events. Notwithstanding all this, the occurrence of serious accidents is postulated, in spite of the fact that they are highly unlikely, and engineered safety

[

L features are installed to mitigste the consequences of these unlikely events. The loss-of coolant accident is such a postulated improbable accident; the emergency core cooling system (ECCS) is one of the engineered safety features installed to mitigate its consequences, j l

The following elementary description of a hypothetical LOCA is given to indicpe the points of  ;

principal attention and to provide a bsckground for a discussion of the changes in the regulations and their  !

conservatism. It should ue remembered that the calculations that are made of the effectiveness of the ECCS i center on maintaining the integrity of the zircaloy cladding, since if it remains intact we can be sure that  !

I the uranium dioxide fuel pellets will be kept separate and coolable. To keep the zircaloy intact requires controlling its maximum temperature and its oxidation.

1091 Y

I Although the ECCS is intendea to cope with a wide range of possible breaks in the primary system piping, most attention has been focused on the sudden severance of the large diameter cold leg pipe for the pressurized water reactor (PWR) and of the recirculation jet-pump inlet pipe for the boiling water reactor (BWR). Postulated breaks of these lines define the onset of the hypothetical arcident for the two kinds of {'

reactor. Before the hypothetical accident, the reactor is assumed to be operating at 102% of full power, thus irmluding a margin for such factors as instrument error. The temperature of the zircaloy cladding at this time would be near that of the adjacent water,in the neighborhood of 600*F. The average temperature of the hottest UOs pellet would be well above 2000*F with a peak temperature in the center greater than 4000*F. The excess heat content of the UOs at this average temperature, called the " stored heat." is approximately proportional to the power density and is affected by the thermal resistance of the " gap" between the UOs pellets and the cladding. The stored heat is important because it would contribute significantly to the later temperature history of the cladding. I Very early in the LOCA the prompt fission heat would stop as the density of the water moderator decreases. At the same time, the cooling of the outside surface of the zircaloy cladding would diminish sharply because of the altered hydrodynamic flow. Under these conditions, the temperature distribution across the uranium dioxide and the zircaloy would tend to even out, dropping the peak temperature in the .

center of the UOs, increasing the UOs temperature near its surface and increasing the temperature of the zircaloy. If there were no heat removal from the outer surface of the zircaloy, its temperature would quickly approach the average temperature of the UOs. An appreciable amount of the stored heat would be removed by the rushing water and steren as they escape from the reactor vessel thus limiting the initial rise in temperature of the zircaloy.

After the accident began the fuel pellets would continue to be heated by the decay of the fission products and of the actinide elements (neptunium and plutonium) that were produced during reactor operation. In addition,if the zircaloy cladding reached temperatures of about 1800*F or above,its reaction with steam to form zirconium dioxide would begin to add to the heat generation. These heat sources would cause the average temperature of the fuel rods to start to increase after the cooling effect of blowdown ceases, and the temperature of the zircaloy would now continue its increase, keeping pace with that of the fuel pellets. The temperature excursion would eventually be terminated as the ECCS begins to reflood the core. Both PWR's and BWR's have ECC systems in which water would reflood the reactor. In BWR's the reflood would be provided by accumulation of water from the low pressure injection system and the core spray system. Direct core spray is discussed below. To accomplish reflood in a reasonable time, the rate at which the emergency cooling water would encroach on the core (the reflood rate) must be high enough to provide a heat transfer rate from the core that would be sufficient to counter the heat input rate from decay heat and from zircaloy oxidation. The Commission believes that the calculated reflood rate should have a substantial margin over the rate that isjust sufficient to turn the temperature excursion aro:'d in a short time.

As the cooling water reaches the hot cort t.ich of it would be converted to steam, and it is this steam i together with entrained water droplets that would provide the initial cooling of the hotter regions of the core. For tne reflood water to continue entering the core it must displace the steam, which would have to escape from the reactor vessel and find its way into the containment atmosphere. In the pressurized water reactors the steam would have to flow through the steam generator and pump to escape through a cold leg break; the reduction of reflood rate by the relatively high resistance to flow of this path is called " steam )

binding" Steam binding would severely limit the rate of reflooding the core, reducing it from an intended 6 i to 11 inches per second to from 1.0 to 2.5 inches per second, depending on the reactor design. 'the rule we announce considers aR the evidence in the record on this important subject of steam binding and provides an acceptable overaE assurance of ECCS effectiveness. The inquiry, however, should not end there. Thus the Commission urgse the pressunrrd water reactor manufacturers to seek out design changes that would overcome steam binding. This same point of view is reflected ir. the September 10,1973, letter of the Advisory Committee on Reactor Safeguards.

Boiling water reactors would not be subject to steam binding, because 4 eir system design provides a er. ore direct path for the steam to escape, buds same requirement for rapid reflood would have to be met

!f excessive clad damage were to be avoided. Boiling water reactors do have a core spray system that would start about 30 seconds after occurrence of the break, but its cool % effect on the central rods of a fuel aur.A might be insufficient in itself to prevent exceeding the temperature limits we have set. The 1092

(

5 occurrence of reflooding within three minutes after a postulated break of the recirculation line would terminate the excursion.

To recapitulate, after a large LOCA the principal cooling of the core would occur in two stages,first by the uncontrolled escape of the pressurized water and steam during blow down, and second by the reflood of the core. The first would be effective in reducing the stored heat, albeit to an extent not yet fully settled, and for which we require conservative estimates, and the second would overcome the afterheat generation rate of the core. Both would be important in determining the temperature history of the zircaloy cladding.

Without redesign and back. fitting, the only measures available to the operator in relation to limiting the design basis accident within the given design framework are to linut the power and the power density of the reactor. The power density can be manipulated somewhat independently of the total reactor power by adjustments of fuel enrichment and control rod action to provide more uniform power generation throughout the core. The Commission notes that there has been a tendency to reduce the maximum alloweti peaking factor (ratio of the highest power density to the average throughout the core) to satisfy ECCS criteria. These lower allowed peaking factors leave less margin above the normal operating range for maneuvenng; thus greater care m reactor operation is required to ensure that these factors are not l ex.eeded. {

Principal Changes From Interim Policy Statement I

The Interim Policy Statement includes: (1) general criteria for emergency core cooling systems applicable to all light. water power reactors (the Interim Acceptance Criteria, or I AC),(2) requirements for analysis using a suitable evaluation model,(3) provisions for application to various classes of reactors by specified dates,(4) provision for variance under stated conditions, and (5) a listing of acceptable evaluation models. The new regulation has sections serving the same purpose as /l), (2), (3), and (4) above. No j complete listings of acceptable evaluation models accompany this decision. The required and acceptable  ;

features of evaluation models, however, will provide the basis for the Reralatory Staff to determine the acceptability of such models as may be furnished, t The pnncipal changes from the Interim Policy Statement are ss follows.The old criterion number one, specifying that the temperature of the zircaloy cladding shou!J not exceed 2300*F,is replaced by two criteria, lowering the allowed peak zircaloy temperature to 2200*F and providing a limit on the maximum allowed local oxidation. The other three criteria of the IAC are retained, with some modification of the wording. These three criteria limit the hydrogen generation from metal-water reactions, require maintenance of a coolable core geometry, and provide for long term cooling of the quenched core.

The most important effect of the changes in the required features of the evaluation models is that swelling and bursting of the cladding must now be taken into consideration when they are calculated to occur, and that the maximum temperature and oxidation criteria must be applied to the region of clad swelling or pursting when the maximum temperature and oxidation are calculated to occur there. Another important change is the requirernent that,in the steady state operation just before the accident, the thermal conductance of the gap between the fuel pellets and the cladding should be calculated taking into consideration any increase in gap dimensions resulting from such phenomena as fuel densification, and should also consider the effects of the presence of fission gases. When these effects are taken into consideration a higher stored energy may be calculated. Other changes in the evaluation models are mostly in the direction of replacing previous broad conservative assumptions with more detaile'd calculations where new experimentalinformation is available or where better calculational methods have been developed.

Tae wording of the definition of a loss of-coolant accidem has been modified to conform to its long. accepted usage, limiting it to breaks in pipes. Justification for the exclusion of consideratic t of pressure vessel failures from the LOCA is extensively discussed throughout Volume 39 of the transcript (April 11,1972), and we have referred to it earlier (pp. 6 8). I

The new regulations also require a more complete documentation of the evaluation models that are used.

Conservatism The Commission believes that the implementation of the new regulations will ensure an adequate margin of performance of the ECCS should a design basis LOCA ever occur. This margin is provided by r

l 1093 l

J

conservative features of thi evaluation models and by the criteria themselves. Some of the major points that contribute to the conservative nature of the evaluations and the criteria are as follows:

(1) Stored Heat. The assumption of 102% of maximum power. highest allowed peaking factor, and highest e.stimated thermal resistance between the UO2 and the cladding provides a calculated stored heat that is possible but unlikely to occur at the time of a hypothetical accident. Wlule not necessarily a margm over the extreme condition, it represents at least an assumption that an accident happens at a time which is not typical.

(2) Blowdown. The calculation of the heat transfer during blowdown is made in a very conservative manner. There is evidence that more of the stored heat would be removed than calculated, although there is not yet an accepted way of calculating the heat transfer more accurately. It is probable that this represents a conservatism of several hundred degrees F in stored energy after blowdown. most of which can reasonably be expected to carry over to a reduction in the calculated peak temperature of the zircaloy cladding.

(3) Rate of Heat Genemtion. It is assumed that the heat generation rate from the decay of fission products is 20% greater than the proposed ANS standard. This represents an upper limit to the degree of uncertainty. The assumption that the fission product levelis that resulting from operation at 102% of rated power for an infinite time represents an improbable situation, with a conservatism that is probably in the range of 5 to ISE The use of the Bakerdust equation for calculating the heat generation from the steam oxidation of zircaloy should also provide some conservatism, but the factor is uncertain. 1 (4) The Peak Tempemrure Cnterion. The limitation of the peak calculated temperature of the cladding to 2200PF and the stipulation that this criterion be applied to the hottest region of the hottest fuel rod provide a substantial degree of conservatism. They ensure that the core would suffer very little damabe in l the accident.

l Suggestions have been made during the hearing for quite different types of criteria, bearing more

]

directly on the actions aIrailable to reactor operators or to the mechanism by which the ECCS would terminate the temperature transient. For reactors that have already been constructed, the only limits in this sense available to the operator would be on the power of the reactor and on the distribution of power

, witt.in the core. This fact led to the suggestion by some of a criterion limiting the power density in some l fashion. (See, for example, Exhibits 1043 and 1044.) This proposalis rejected on the basis that its relation to the ECCS is tenuous and arbitrary, and that the imposition of such a restriction might inhibit innovation in reactor design. Similarly, since the temperature transient would eventually be terminated by core flooding, the suggestion has been made by some th.t the allowed power density be tied in a conservative l way to the reflood rate (Exhibit 1113, pp.14-18 and pp.17 4 to 17 5). Although this suggestion has considerable merit for application to the present design of ECCS for PWR's, the setting oflimits would still require calculation of the temperature history of the cladding by some evaluation model. For this reason this proposal was rejected in favor of modifying of the Interim Acceptance Criteria. In doing this, however, the Comnussion has intentionally incorporated the effects of the conservative features indicated above to provide a suitable safety factor in the relationship between the capability of the ECCS and the power density in the reactor.

In its Concluding , Statement the Consolidated National Interveners c! aim that there is an " inadequate base on which to base predictions of the course of an accident" and that " uncertainties and errors . . . are unknown", and in Chapter 5 they present a list of "information needs," for Commission guidance. The

Commission realizes that the knowledge in regard to a number of facets of the analysis of a lo

t of coolant accident is imprecise;it is partly for this reason that there is an on-going Water Reactor Research Jrogram.

The Commission is.ccafident, however, that the criteria and evaluation rnodels set forth here are more than sufficiently conservesho to compensate for remaining uncertainties in the models or in the data.

Continuing reseasch and development will provide a more extensive data base for such items as heat transfer coefficients during blowdown and during spray and reflood cooling, oxidation rates for zirconium, fission product decay heat, steam. coolant interaction, oscillatory reflood flows, fuel densification, pump modeling and flow blockage. With the additional data it n.ay become practical to assign a statistically meaningful measure of precision to the calculation. It is probable that, with a better data base, some relaxation can be made in some of the required features of the evaluation models. However, the Comnussion believes that any future relaxation of the regulations should retain a margin of safety above and beyond allowances for statistical error.

1094

___________________a

E T II. ACCEPTANCE CRITERIA FOR EMERGENCY CORE COOLING SYSTEM EFFECTIVENESS A. THE CRITERIA (1) Peak Cladding Temperature. The calculated maximum fuel element cladding temperature shall not exceed 2200*F.

(2) Maximum Cladding Oxidanon. The calculated total oxidation of the cladding shall nowhere exceed 0.17 times the total cladding thickness before oxidation. As used in this subparagraph total oxidation means the total thickness of cladding metal that would be locally converted to oxide if all the oxygen absorbed by and reacted with the cladding locally were converted to stoichiometric zirconium dioxide.lf cladding rupture is calculated to occur, the inside surfaces of the cladding shall be included in the oxidation, beginning at the calculated time of rupture. Cladding thickness before oxidation means the radial distance from inside to outside the cladding, after any calculated rupture or swelling has occurred but before significant oxidation. Where the calculated conditions of transient pressure and temperature lead to a prediction of cladding swelling, with or without cladding rupture, the unoxidized cladding thickness shall be defined as the cladding cross-sectional area, taken at a horizontal plane at the elevation of the rupture,if it occurs, or at the elevation of the highest cladding temperature if no rupture is calculated to occur, divided by the average circumference at that elevation. For ruptured cladding the circumference does not include the rupture opening.

(3) Maximum Hydrogen Generation. The calculated total amount of hydrogen generated from the chemical reaction of the cladding with water or steam shall not exceed 0.01 times the hypothetical amount that would be generated if all of the metalin the cladding cylinders surrounding the fuel, excluding the cladding surrounding the plenum volume, were to react.

(4)Coolable Geometry. Calculated changes in core geometry shall be such that the core remains amenable to cooling.

(5)Long-Term Cooling. After any calculated successfulinitial operation of the ECCS, the calculated core temperature shall be maintained at an acceptably low value and decay heat shall be removed for the extended period of time required by the long-lived radioactivity remaining in the core.

B. TECHNICAL DISCUSSION OF THE CRITERIA (1) Peak Cladding Temperature. The calculated maximum fuel element cladding temperature shall not exceed 2200* F.

(2) Maximum Cladding Oxidation. The caiculated total oxidation of the cladding shall nowhere exceed 0.17 times the total cladding thickness before oxidation. As used in this subparagraph total oxidation means the total thickness of cladding metal that would be locally converted to oxide if all the oxygen absorbed by and reacted with the cladding locally were converted to stoichiometric zirconium dioxide.lf cladding ' rupture is calculated to occur, the inside surfaces of the cladding shall be included in the oxidation, beginning at the calculated time of rupture. Cladding thickness before oxidation means the radial distance from inside to outside the cladding, after any calculated rupture or swelling has occurred but before significant oxidation. Where the calculated conditions of transient pressure and temperature lead to a prediction of cladding swelling, with or without cladding rupture, the unoxidized cladding thickness shall be dermed as the cladding cross-sectional area, taken at a horizontal plane at the elevation of the rupture,if it occurs, or at the elevation of the highest cladding temperature if no rupture is calculated to occur, divided by the average circumference at that elevation. For ruptured cladding the circumference does not include the rupture opening.

Discussion of Peak Cladding Temperature and Maximum Oxidation The purpose of these first two criteria is to ensure that the zircaloy cladding would remain sufficiently intact to retain the UOs fuel pellets in their separate fuel rods and therefore remain in an easily coolable

! array Conservative calculations indicate that dunng the postulated LOCA, the cladding of many of the fuel rods would swell and burst locally with a longitudinal split. The split cladding would remain in one piece if it were not too heavily oxidized, and would stdl restrain the UOs pellets. The possibility of destructive damage to the zircaloy cladding must be exanuned f or times late in the course of the LOCA when oxidation 1095

at high temperatures in the stem atmosphere could render the cladding brittle. The limits specified in these criteria will assure that some ductility would remain in the zircaloy cladding as it goes through the quenching process, and therefore that the core wou;d remain essentially intact,in a condition amenable to long term cooling. .

The steam oxidation process is described in Exhibit 1122, pp. 21 to 2-7 (Scatena, of General Electric).

Water molecules are absorbed on the surface and dissociate to hydrogen atoms and hydroxyl radicals.

Within the surface the hydroxyl radicals are (in several steps) converted into oxygen ions and hydrogen atoms. The hydrogen atoms, wherever formed, combine into hydrogen molecules at the surface and escape.

The oxygen ions diffuse into the surface, and are dissolved in the metal. When their concentration is high enough zirconium dioxide is formed.

The initial reaction rate is limited by the counter diffusion of gaseous water molecules to the surface and of hydrogen molecules away from the surface. This limitation lasts for only a short time, until an oxide film is formed. Thereafter, at sufficiently high temperatures, the reaction rate is controlled by solid state diffusion procenes: largely the diffusion of oxygen ions through the zirconium oxide and the metal. A gradient of oxygen concentration will exist from the surface to the center, with high concentrations on the outside surface and lesser concentrations in the interior.

At room temperature pure zirconium is in the alpha phase, with a close packed hexagonal crystal structure. On heating above !!50 F, it is trar.sformed to beta phase, with a body-centered cubic structure.

The zirconium-oxygen phase diagram (Fig. 2-4 of Exhibit 1122) is useful as a guide to the phases that will be present under oxidizing conditions. At the temperatures at which the oxidation takes place (between 1400 and 1700*K in the figure), the outer surface will likely be oxidized to stoichiometric ZrOs ;inside this will be some alpha phase zirconium stabilized with a high concentration of dissolved oxygen; then a region of mixed alpha and beta phase; and in the interior, pure beta phase. When this is quickly cooled to room temperature three regions,can be distingu;shed metallographically: the zirconium oxide, the unchanged alpha phase zirconium (called stabilized alpha phcse), and a region of alpha phase zirconium that existed as beta phase while it was. hot (the so-called prict beta phase). These three regions are shown in the photomicrograph of figure 2 5 of Exhibit 1122.

It is well known that the oxide and stabilized alpha phases are brittle, and that what ductility and resistance to shattering is exhibited by the oxidized zircaloy is associated with the prior beta phase material.

For this reason it has been the custom to correlate the strength remaining in the oxidized metal with either the sum of the thicknesses of the oxide and the stabilized alpha material (called Xi, or (), or, conversely, the fraction of the original thickness remaining as prior beta phase (called Fw). For example, Hobson and Rittenhouse (Exhibit 509) preferentially use Xi to correlate their data on ductility and hardness, while Combustion Engineering (see for example Exhibit 1144) has used Fw.

The situation is complicated by the fact that not all of the prior beta phase 3 equally strong or ductile, since these properties depend on the amount of dissolved oxygen. This fact has been suspected for some time and the basis for using Xi or Fw, aside from the case of measurement, has been the assumption that the' oxygen content in the prior beta phase is closely related to the thickness of the oxide and stabilized alpha layers. For example, figure 7 of Exhibit 509 gives a correlation between the hardness of the prior beta phase and XI. Westinghouse, in their concluding statement (page A 5), indicated a belief that a different exposure parameter, related to the amount of oxygen i's the prior beta phase,would be more appropriate.

The matter is discussed at Transcript pages 20,935-41 and 21,629-30; and in Exhibit 1133, pp.27 and 29-33.

As a result of the diffusion process, the concentration of oxygen in the pior beta phase will not be uniform within a sample, but will depend upon the depth below the surface.(See, for example, figure 2 3 ,

of Scatena, Ex)nelt 1122, or figure A 3 of the Westinghouse direct testimony. Exhibit 1078.) From the  !

l phase diagram, given by both Scatena and Westinghouse, it is obvious that it is possible for the beta phase l zirconium to take on a higher oxygen content at 2600*F than at 2000*F. Furthermore, since the diffusion  ;;j rate depends exponentially upon temperature, one might expect a greater incursion of oxygen into the beta l phase for a given thickness of oxide and stabilized alpha phase at higher temperatures. Westinghouse has calculated three cases (page A 10 of Exhibit 1078), and although the results are not unequivocal, the l I

comparison of the 2200*F and the 2000*F cases indicates that it is possible to have significantly more oxygen in the beta phase at the higher temperature for abot.t the same value of Xi.

In discussing the slow compression tests of Hobson, the Regulatory Staff in their supplementary testimony (Exhibit 1113, page 1814) stated that ". . . the 2400*F specimens seem to be more brittle than 1096

. )

their Fw values would indicate, compared to the lower temperature specimens. The basic inference Hobson feels should be drawn from this observation is that embrittlement is notjust a monotonic function of Fw, or Xi penetration, but is also related to the exposure temperature. The most likely explanation for this behavior is increased oxygen in the beta phase, due to increased oxygen solubility and to more rapid diffusion kinetics."

Others have also observed that the resistance to rupture depends upon the temperature at which oxidation occurs as well as the extent of oxidation. For example, Combustion Engineeringin Exhibit 1144, page 2 2, stated with respect to their compressive load tests, "The limiting value of s/6t of 0.012 cm (a measure of the extent of oxidation)is based upon experimental tests at 2300 to 2500"F. Oxidation tests at 2100 F demonstrated that the limiting value of 0.012 was significantly increased so that the 0.012 cm is increasingly conservative with decreasing temperature below 2300 F." Babcoci and Wilcox, although strongly disagreeing with much of the Oak Ridge interpretation of their results, apparently confirmed the major thesis that embrittlement is a function of both the extent of oxidation and the oxidizing temperature. In their concluding statement, page 239, they stated, "However, the tests confirmed, under similar experimental conditions, the implication from the Oak Ridge tests that a correlation between ductility and remaining beta fraction does not adequately characterize cladding ductility above a threshold temperature ranging from 2200 to 2400*F furnace temperature." (B&W has stated a belief, however, that the actual temperatures of the zircaloy specimens were at least 100 F iigher than the furnace temperatures.)

To recapitulate, measures of zircaloy oxidation, whether by percent, Xi, or Fw, are largely or wholly determined from the brittle layers of zirconium oxide or stabilized alpha phase, while the ductility and strength of oxidized zirconium depend upon the condition and the thickness of the prior beta phase.

According to Westinghouse's sample calculations (Exhibit 1078, page A 10) the proportion of the total oxygen content that is in the prior aeta phase is of the order of 4 to 101 Thus a criterion based solely on the extent of total oxidation is not enough, and some add!:fonal criterion is needed to assure that the prior beta phase is not too brittle. The specification of a maximum temperature of 2200*F willaccomplish this adequately. The dats cited in Exhibit 1113 would not support a choice of a less conservative limit.

There is relatively good agreement among the industrial participants as to what the limit on total oxidation should be. Babcock and Wilcox and General Electric have sug ested a limit of 17% of the zirconium being oxidized, while Westinghouse proposed 16% Although Westinghouse has preferred the i limit to be expressed as percent total equivalent oxidation, they equated their 16% to a ratio of brittle layer thickness (Xi) to original thickness of 0.47. Combustion Engineering's recommendation of a minimum Fw .

of 0.65 is also equivalent to a maximum Xi ratio of 0.47, so that all of the reactor venders are in essential I agreement. The Utility Group also recommended the 17% oxidation limit, but said that if one wants to be more conservative and avoid brittle behavior even when the clad were to be cooled to room temperature,a ,

12% lhnit would be reasonable. The Regulatory Staff in their concluding statement compared various measures of oxidation (page 90) and concluded that a 17% total oxidation limit is satisfactory,if calculated

{

j by the Baker-Just equation. The Consolidated National Interveners in their direct testimony (Exhibit )

1041) indicated satisfaction with the Rittenhouse criteria, and as argued by the Regulatory Staff,it appears that the 17% oxidation limit is within the Rittenhouse criteria. Thus a remarkable uniformity of opinion seems to exist with regard to the 17% oxidation limit.

None of the reactor manufacturers agreed with the Staff's proposed stipulation of a 2200*F maximum l calculated temperature (Concluding Statements and Responses to the Staff Concluding Statement).

Westinghouse proposed a maximum cal,:ulated temperature limit of at least 2700*F; Combustion l Engineering and the Utility Group agreed on 2500*F as the peak allowable calculated temperature on the basis that much of the data on oxidation and its effects stops at 2500*F. Babcock and Wilcox suggested a more conservative 2400*F as the peak calculated temperature to be allowed, presumably because "significant eutectic reaction and an excessive metal.to-water reaction rate would be prech.ded below l 2400*F." (Concluding Statement, p.242.) General Electric argued strongly that the limit should not be reduced to 2200 F; that 2700*F is really all right as far as embrittlement is concerned, but that the Interim i Acceptance Criterion value of 2300*F should be retained. In addition to being consistent with their expressed desire not to change any of the criteria, the GE recommendation of retaining the 2300*Flimit is intended to ensure that the core never "gets into regions where the metal-water reaction becomes a serious concern." (Initial Closing Statement, Vol. 2, p. M.49.) i r

1097 t _D

5 The bases for the higher temperature recommendations of the reactor manufacturers are: (1) cal-culations of the stresses imposed on fuel rods during the LOCA and particularly during quenchmg.

(2) measurements of the stresses and strains that oxidized zircaloy tubes can withstand, and (3) quench tests of oxidized zircaloy tubes to deternune their resistance to shattering. The calculations are important to both items (2) and (3) because they tend to show that the greatest stresses that the rods would encounter are those of the thermal shock during quench. The thermal shock stress calculations are usefulin indicating the general order of magnitude, or limit, of these stresses. Taken as a whole. however, they do not provide a confident basis for an accurate number. Combustion Engineering calculates a peak thermal stress of 24,600 lbs/sq in; General Electric-38,000 lbs/sq in.; Babcock and Wilcox--23.000 lbs/sq in.; and l Westinghouse-36,000 lbs/sq in. in Exh.1078. later corrected to 3500 lbs/sq in in Exh. I151. Some of the '

difference beta een various values may be due to different assumptions about the heat transfer coefficient during the rap' a quench. (Trans. 21,619 20)

Other stresses on the fuel rods that have been calculated indude those from circumferential temperature vanstion, internal gas pressure, rod rod interaction, assembly restraint, and cross flow.

Westinghouse's rod to rod interaction seems to be the same as GE's circumferential temperature variation, and aside from the thermal shock stresses, these seem to be the largest calculated (1000 to 2000 psi). There is some lack of certainty as to just what nature of stresses would be encountered dunng the LOCA. As an example it is stated by Westinghouse Exhibit 1151, page 16-2: "At the inception of the LOCA, a pressure wave passes through the system and imposes a dynamic loading on the fuel assembly. However, at this time, the fuel rod ductility has not yet been reduced by oxidation and hence this blowdown load is not of interest to this analysis" It seems possible that such a load might distort the ductile fuel rods and leave i them in a state that would aggravate the later rod to-rod interaction. The only point in mentioning this possibility is to draw attention to the fact that it may not be possible to anticipate and calculate all of the stresses to which fuel rods yvould be subjected in a LOCA. Although we believe the calculations of thermal )

shock stresses are worthwhile and informative, we agree with the regulatory staff that they are not j sufficiently well defined to depend on for regulatory purposes.

Since the principal stresses calculated for the fuel rods were those of thermal shock during quench,it is natural that in the testimony considerable reliance has been placed on the demonstrated ability of the oxidized clad to withstand a rapid quench, General Electric in Exi:ibit 1122, page 3 3, showed a plot of i data indicating that no zircaloy tubes shattered on quenching if they were oxidized to less than 17% orif their oxidation temperature was le'ss than 2700*F. Additional data and plots of others did not change this conclusion. However, many of the oxidation tests of zircaloy have involved quick quenching all the way from the peak temperature at which oxidation took place. Scatena pointed out (Exhibit 1122, page 2 4) j

" the time spent in the alpha + beta transition region can enhance embrittlement." Scatena attributed the j

additional embrittlement to precipitation of zirconium oxide along grain boundaries, although the j Regulatory Staff in their concluding statement (page 83) indicated that a more likely explanation is the precipitation of alpha zirconium in the grain boundaries while the bulk of the material is stillin the beta phase. This explanation is fortified by Hobson's observation of pronounced incursions of alpha phase into ,

the beta phase for specimens exposed at 2400 and 2500 F. (Exhbit 1126, page 11). One bit of experimental evidence pn the effect of slow cooling is given by the ANL experiment number 3, as discussed by Scatena, page 3 7. This sample remained at a temperature between 1500 and 1600 F for several minutes, and failed upon final quenching, although the amount of oxidation was less than that at which failure was expected. The transition temperature from beta to alpha is in the range of 1500 to 2000 F, and judging from the PWR FLECHT tests (e.g., Exhibit 150, page C 15), the cooling rate experienced through l this temperature range during a LOCA might be quite slow as compared to the fast quench typical of most laboratory tests.(General Electric stated that the quench from peak temperature in BWR would be fast, but the curves displayed in evidence do not provide quantitative data in this regard.) The uncertamty l introduced by the effect of cooling rate in the temperature region above 1500*F casts some doubt upon the i applicability of some of the quench tests that have been carried out. Nevertheless we find the quench j results encouraging in that they provide assurance that the 2200 F limit is conservative.

Our selection of the 2200 F limit results primarily from our belief that retention of ductility in the zircaloy is the best guarantee ofits remaining intact during the hypothetical LOCA. The stress calculations, the measurements of strength and flexibility of oxidized rods, and the thermal shock tests all are reassuring, but their use for licensing purposes would involve an assumption of knowledge of the detailed process taking place in the core during a LOCA that we do not believe is justified.

1098

! I 1

1

~

All of the reactor manufacturers except Combustion Engineering objected strongly to the application of oxidation and maximum temperature criteria to the hottest spot of the cladding,especially to regions that are calculated to have swollen and burst open. It was argued that this represents only an extremely small fraction of the reactor core and that even if this small amount became fractured it would do n Another contention is that the criteria are so conservative that even if these hot spots were oxidized more .

than the criteria allow they would remain intact. It is also pointed out that if any damage were to occur to

}

these hot spots it would happen as they were being quenched at about 1000 F after the great bulk of the core had already been quenched and was at a temperature in the range of 300*F. As put by flabcock and  ;

Wilcox in their concluding statement, page 255, "it is difficult to conceive of conditions in which the core l

has been substantially cooled, the transient completely arrested for the bulk of the core, and yet find a local fragmentation of a negligibly small portion of the core causing a significant thermal effect." )

These are valued arguments, supporting the thesis that damage to.a small part of the core would not lead to more extensive damage. However, they must be recognized as opinions as to what would happen in a situation that has never yet occurred. Others are not so sure that a local failure would not be propagated more widely throughout the core. (CNI, Exhibit 1041, p. 5.67.) In view of the lack of experience in this hypothetical situation, we think it prudent to apply our criteria to all of the core and not to exempt any part.

There remains the question of the extent to which oxidation on the inside of burst cladding should be calculated. The two principal pertinent experiments are the in. pile experiments FRF.1 and FRF.2. FRF.I went to only 1800 F, but the oxidation on the inside of burst cladding was approximately the same as on the outside in the region of the burst area. FRF.2 went to a higher temperature and had significantly less oxidation ol'the inside than the outside. However, this experiment was steam limited, as indicated by cessation of. steam effiuent during the time when the peak temperatures occurred. Thus there is no basis for assuming anything but equal access of steam to the inside of the burst tubes, as stated in the criterion.

(3) Maximum Hydrogen Generation. The calculated total amount of hydrogen generated from the i chemical reaction of the t.ladding with water or steam shall not exceed 0.01 times the hypothetical amount that would be generated if all of the metalin the cladding cylinders surrounding the fuel, excluding the 1 cladding surrounding the plenum volume, were to react.

l Discussion. The object of this criterion is to ensure that hydrogen would not be generated in amounts

{

that could lead to explosive concentrations. The criterion is essentially the same as Interim Acceptance l Criterion #2, but is more explicit in detailing how much of the zircaloy is to be used as the base for the one l

percent calculation. This criterion is non. controversial. However,its purpose was misunderstood in some 1 intervenor and vendor analyses. It has nothing to do with the need to retain strength of cladding.

(4)Coolable Geometry. Calculated changes in core geometry shall be such that the core remains  ;

amenable to cooling.

Discussion. If there were no emergency core cooling after a LOCA, the core would probably eventually fuse together into a large mass with insufficient external surface area to allow the fission product heat generated within it to be transferred away. Intermediate steps in arriving at such a state might be the j oxidation and melting cf the zircaloy cladding, allowing the uranium dioxide fuel pellets to fall together '

into a heap that would be difficult to cool. Other difficulties have been envisaged, some the result of an ECCS that was. not fully effective. Those most often mentioned are the bulging of the cladding to the extent of closing off the coolant passages and the possibility of zircaloy becoming embrittled by steam oxidation and shattering during quench, allowing the fuel pellets to fallinto a heap. (Exhibit 1041, sec. 7; Exh.10078).

Thinking largely of the latter difficulty, Combustion Engineering in their Concluding Statement suggested that, in view of the restrictions on cladding oxidation placed by criteria 1 and 2, a specific criterion on coolable geometry is no longer needed. Similarly, Babcock and Wilcox omitted a criterion on coolable core geometry from their proposed criteria in appendix A of their Concluding Statement.

Considering all of the required features of the evaluation models, we are inclined to agree that, for an: I situation that we have been able to anticipate, this criterion should be superfluous. However,in view of the fundamental and historical importance of maintaining core coolability, we retain this criterion as a basic objective, in a more general forra than it appeared in the interim Acceptance Criteria. It is not controversial

)

a.s a criterion, although the extent of flow blockage resulting from clad swelling is a matter of controversy.

This subject is discussed in the section on Required and Acceptable Features of the Evaluation Models.

l 1099 '

)

(S)Long Term Cooling. After any calculated successful initial operation of the ECCS, the calculated core temperature shall be maintained at an acceptably loiv value and decay heat shall be removed for the extended period of time required by the long lived radioactivity remaining in the core.

Discussion. Although most of the attention of the ECCS hearings has been focussed on the events of the first few minutes after a postulated major cooling line break, up to the time that the cladding would be cooled to a temperature of 300*F or less, the long term maintenance of cooling would be equally .

imponant. The intent of this enterion is self-evident and it is non-controversial.

l I

i Ill. REQUIRED AND ACCEPTABLE FEATURES OF THE EVALUATION MODELS An evaluation model is the calculation framework for evaluating the behavior of the reactor system l

during a postulated loss of coolant accident (LOCA). It includes one or more computer programs and all other information necessary for application of the calculation' framework to a specific LOCA, such as

' mathematical models used, assumptions included in the programs, procedures for treating the program  :

input and output information, specification of those portions of analysis not included in computer f

programs, values of parameters, and all other information necessary to specify the calculational procedure. j Loss-of coolant accidents (LOCA's) are hypothetical accidents that would result from the loss of reactor coolant, at a rate in excess of the capability of the reactor coolant makeup system, from breaks in pipes in i

the reactor coolant pressure boundary up to and including a break equivalent in size to the double-ended '

rupture of the largest pipe in the reactor coolant system.

A. SOURCES OF HEAT DURING THE LOCA For the heat sources listed in paragraphs 1 to 4 below it shall be assumed that the reactor has been j operating continuously at a power level at least 1.02 times the licensed power level (to allow for such uncertainties as instrumentation error), with the maximum peaking factor allowed by the technical specifications. A range of power distribution shapes and peaking factors representing power distributions that may occur over the core lifetime shall be studied and the one selected should be that which results in the most severe calculated consequences, for the spectrum of postulated breaks and single failures analyzed.

l 1. The InitialStored Energy in the fuel. The steady. state temperature distribution and stored energy in i

the fuel before the hypothetical accident shall be calculated for the burn up that yields the highest calculated cladding temperature (or, optionally, the highest calculated stored energy.) To accomplish this, the thermal conductivity of the UOs shall be evaluated as a function of burn up and temperature, taking into consideration differences in initial density, and the thermal conductance of the gap between the UOs

[ and the cladding shall be evaluated as a function of the burn-up, taking into consideration fuel densification i l and expansion, the composition and pressure of the gases within the fuel rod, the initial cold gap dimension with its tolerances, and cladding creep.

2. fission Heat. Fission heat shall be calculated using reactivity and reactor kinetics. Shutdown reactivities resulting from temperatures and voids shall be given their minimum plausible values, including allowance for uncertainties, for the range of power distribution shapen and peaking factors indicated to be studied above. Rod trip and insertion may be assumed if they are calculated to occur.

l- 3.Decey of Actinides. The heet from the radioactive decay of actinides, including neptunium and

! plutonium generated during operation, as well as isotopes of uranium, shall be calculated in accordance with fuel cycle calculations and known radioactive properties.The actinide decay heat chosen shall be that appropriate for the time in the fuel cycle that yields the highest calculated fuel temperature during the l LOCA.

l 4. Fission hocruct Decay The heat generation rates from radioactive decay of fission products shall be l assumed to be equal to 1.2 times the values for infinite operating time in the AN3 Standard (Proposed l American Nuclear Society Standard " Decay Energy Release Rates Following Shutdown of Uranium-Fueled Thermal Reactors", Approved by Subcommittee ANS 5, ANS Standards Committee, October 1971). The fraction of the locally generated gamma energy that is deposited in the fuel (including the cladding) may be different from 1.0; the value used shall be justified by a suitable calculation.

1100 j

1 S. Metal-Water Reaction Rare,, The rate of energy release, hydrogen generation, and cladd'ing

oxidation from the metal / water reaction shall be calculated using the Baker Just Equation (Baker, L,Just, L C.," Studies of Metal Water Reactions at High Temperatures, Ill. Experimental and Theoretical Studies of the Zirconium-Water Reaction," ANL-6548, page 7, May 1962). The reaction shall be assumed not to be steam limited. For rods whose cladding is calculated to rupture during the LOCA, the inside of the cladding shall also be assumed to react after the rupture. The calculation of the remion rate on the inside of the cladding shall also follow the Baker Just ' equation, starting at the time when the cladding is calculated to rupture, and extending around the cladding inner circumference a:.d axially no less than 1.5 inches each way from the location of the rupture, with the reaction assumed,not to be steam limited.

6. Reactor Internals Heat Transfer. Heat transfer from piping, vessel walls, and non-fuel internal hardware shall be taken into account.

7. Pressurized Water Reactor Primary-to-Secondary Heat Transfer. Heat transferred between primary and secondary systems through heat exchangers (steam generators) shall be taken into account. (Not applicable to Boiling Water Reactors.)

DISCUSSION

1. The Initial Stored Energy in the Fuel. During reactor operation just before the accident is assumed to happen the temperature of the UO 2would depend strongly on the radial position in the fuel pellet.The temperature would be highest at the center of the rod and much lower at its outer surface next to the zircaloy cladding. The heat generated anywhere within the UO2 must be conducted outward through the remaining UO2 ., and all of heat generated within the UO 2must pass through a poorly conducting region at the boundary between the UOs and the zircaloy, called the gap. The poorer the heat conductivity,either within the UOs or in the gap, the higher would be the average temperature of the UO2 .The stored energy is defm' ed to be the energy that would be released by the UOs ifits temperature were reduced to that of the zircaloy cladding. Just before the postulated accident the average of the radial temperature distribution of the UU 2in the hottest region would be near the melting temperature of zircaloy, and since the heat '

transfer rate between the zircaloy and the water would drop abruptly within a fraction of a second after the I initiation of the accident, the stored energy would be an important heat source which could cause a sharp rise in the temperature of the cladding during the blow down period. The increase in zircaloy temperature resulting from the redistribution of stored energy tends to contribute to the calculated peak temperature of the zircaloy.

Although the importance of stored energy was recognized at the time the Interim Policy Statement was issued, no gereral rule was formulated. Rather, the methods for arriving at the initial fuel temperature were specified in the descriptions of the individual evaluation models. A considerable amount of evidence regarding stored energy has been introduced in these hearings, indicating a wide diversity of approach.

(Exhibit 1113, p.10-10 & ff). Furthermore, a new factor affecting gap conductance and therefore stored energy has been brought to light (fuel densification: see transcript 15,242-3; 15,296.) We therefore believe that it is proper to require a more uniform approach to the calculation of stored energy, and to require that the evaluation of gap conductance and stored energy be made on a case by case basis.

A summary of several reviews of the thermal conductivity of uranium dioxide is given in Exhibit i113, pages 10 23 to 10 29. The views of the Regulatory Staff are given on page 10-29 of that document, and the Commission believes that these views are reasonable and can serve as a reference forjudging the adequacy of the values of thermal conductivity used by licensees.

The steady state gap conductance depends on the thickness of the gap (ranging from 0.000 to 0.012 inches) and the thermal conductivity of the gas filling the gap. If the gap is closed its thermal conductance depemis upon the pressure of the contact between the UOs and the zircaloy. Thus any change in fuel density, whether by thermal expansion, fuel densification, or swelling will change the gap conductance. The gap dimensions will also change if the external pressure causes the zircaloy to creep inward. As fission gas is released the composition of the gas in the gap changes from pure helium, which has a high thermal conductivity, to a mixture with xenon and krypton, which have low thermal conductivities. In ths past.

Babcock and Wilcox and Westinghouse have been calculating the gap conductance as a function of bum up.

Babcock &Wilcox (transcript p. 5536 and Concluding Statement, page 70) included all of the above features except fuel densification and cladding creep. This latter neglect was conservative, esulting in lower gap conductivities and yielding the highest stored energy at the end of fuellife. Westinghouse has provided i

l 1101

T similar calculations, btit including creep. (Exhibit i151, section 24). As e result their gap conductances have been much higher at the end of fuellife.

l These examples illustrate the wide divergence of these calet'la. ions in the past. Furthermore, with the

! advent of the fuel densification problem, new fuels are being u,troduced, making it necessary to evaluate stored energy on a case-by case basis.

General Electric in their response to the concluding statement of the Regulatory Staff (pp.60-70) and in their initial closing statement (Vol. II, K 1 through K 11) argued that there is no basis in the record to 2

require GE to abandon their use of the fixed value of 1000 BTU /hr ft F for gap conductance.Much of their argument depends upon the fact that the fixed value of gap conductance was included in the approved GE evaluation model, and that at the outset of the hearings the Staff in Exhibit 1001 affirmed its suitability. Their argument also depends on the fact that much of the technical discussion in the record seems to be directed toward PWR's, and that BWR's were seldom mentioned. The Commission notes thrt much was learned during the course of the heanngs, both from new experimental and calculational results

! and from further study ofinformation that may have been previously available. As a result the Commission I

places greater weight on the opinions of the Regulatory Staff as expressed in Exhibit 1i13 than on those expressed in 1101. The Commission also notes that boiling water reactors are fueled with uranium dioxide encased in zircaloy and that the uranium dioxide is subject to the fissioning process and temperature vanatiens much the same as in the pressurized water reacters. There is therefore every reason to believe that the basic physical processes are similar in the two reactor types and the Commission sees no reason to l

exempt boiling water reactors from the provisions of this section. The importance of stored energy in the LOCA has been well established, as has its dependence upon the steady state gap conductance, and there is ample evidence that the initial gap conductance of fuel rods in boiling water reactors can vary widely.

l (Exhibit i113, pages 10-11).

2. Fission Heat. This represents no change from previous practice.

3. Decay of Actinides. The decay of ectiaides (isotopes of U, Np, or Pu) produced in the fuel can contribute a significant amount of heat, as much as 10% of that of the fission products.The rule proposed by the Regulatory Staff would have required that the maximum decay heat from actinides be used in the calculation of the consequences of the hypothetical LOCA. Both Westinghouse and Babcock & Wilcox l objected to this in their responses te the Regulatory Staff's concluding statement on the grounds that it is physically impossible to have the maximum actinide decay heat simultaneously with the maximum stored l energy and the maximum peaking factor. The rule adopted by the Commission recognizes this situation, and is in accord with the principle ofimproving the realism of the calculation.

4. Fission Product Decay. The major source of continuing heat production after reactor shut down is radioactive decay of ti.e fission products, their daughters, and other radioactive isotopes produced from them by neutron absorption. The Regulatory Staff reviewed the literature, as described in Exhibit 1001, pp. 3 24 to 3-26, and concluded that the tentative ANS standard gave the best representation of the available information. The ANS standard is derived from the review and compilation by Shure* which, for (

the region of most interest to the ECCS (up to 1000 seconds after shut down) is taken largely from a I compilation by Stehn and Clancy.t in making their review, the Regulatory Staff also considered experimental work done since the Shure review was published. (Exhibit 1113, page 22 3). The ANS committee recommended uncertainty limits of +20% - 40%. The Staff concluded that there was an uncertainty factor of ISE (1113, p. 22 3). Thus, in adopting ANS + 20% for the Interim Policy i Statement they were providing a small margin above what they thought the uncertainty factor to be. l Although no new experimental work was presented during the hearings, new computer calculations from the doctoral thesis of T. R. England were brought up and emphasized by the Consolidated National i Interveners (Exhibit !!$2, pp. 2.2 2.6). England's work was essentially a computer calculation and summation of the contributions of individual nuclides, including for the first time the effect of neutron capture in the fission product chains. As originaPy presented, England's results indicated large deviations above the ANS prescription, particularly for high neutron fluxes and fuel burn-ups. (See, for example, Exhibit 1113, p.22 5). However a series of errors in both input data and the calculational program were l

'K. Shure, " Fission Product Decay Energy" in WAPD-BT-24, Dec.1961, pp.1 17.

tJ. R. Stehn & E. F. Clancy, " Fission Product Radioactivity and Heat-Generation," in Prm. the Second U.N.

International Conference on the Peaceful Uses of Atomic Energy, Genesa 1958, Vol.13, pp.49-54. United Nations, Geneva,1958.

l 1102 l

l

_ )

I, found both by England (Exh.1178, p. 7) and in the course of .s review by Shure (Exhibit 1178), which markedly reduced the deviations found by England's approach. With the corrections made, the positive dettations found by the England approach from the ANS standard ne nowhere greater than 10%,and are generally much less (1113, at 2215). In addition, there is the possib9ity that the selection of input data (fission product yields and decay energies) may not have been the best (l!13,22 8 and 22-9).

While England's approach is a valuable contribution, it is only one piece of work out of many cited in the record; furthermore it presents no new experimer tal determinations. On the basis of the record of these proceedings, however, one is led to believe that the ANS standard curve may be about 5%lowin the time region of principal interest, namely, zero to live minutes after shut down. Enghnd's revised values are well within the previously expressed limits of uncertainty, and to the extent of the credence given the new calculations, they tend to narrow those limits of uncertainty. At present it appears to us that the 20% on top of the ANS decay heat formula fairly represents the uncertainty and does not proWde any margin above that uncertainty. It is still conservative.

There is some margin provided, however, in the prescription requiring that the reactar shall have been considered to have operated continuously at 1.02 times rated power, with the maximum allowed peaking factor, for an infinite length of time. The exact amount of margin is uncertain, and it will vary with time, but it is probably in the range of 5 to 15%.(Exhibit 1137, pp.113 to 5;and Staff Concluding Statement, p.114).

Considering all of the above, the Commission believes that the prescription of ANS + 20% for the fission product decay heat is reasonable and should be continued.

5. Metal-Water Reaction Rate. A rate equation is needed for the calculation of hydrogen generation, the extent of " cladding oxidation, and the heat generation from oxidation of the zircaloy by steam. The Baker Just equation has so far been used in evaluation models for this purpose.This equation was derived by Baker and Just from their measurement of the rate of oxidation at the melting point of zirconium,in conjunction with Lemmon's and Bostrom's data at lower temperatures. The equation is a straight line representation of a plot of the logarithm of the reaction rate vs. the reciprocal of the absolute temperature.

The slope of this line is the activation energy, and depends in an important way on the single point of Baker and Just at the melting point of zirconium.

The Baker Just equation has been criticized extensively, principally on three bases. (Exh.1048, pp. '

75 78 & 137140; Exh. 258, p. 40; Exh. I122, p. 2-11.) One is that the data point at the melting point of zirconium (3365*F) is quite unrelated to the phenomenon of oxidation at the temperatures of interest, namely,1800 F to 2200*F. The second is that more recent data yield lower oxidation rates,especially at about 2300*F* and above. The third basis for criticism is the theoretical one that a single activation energy should not be expected for different crystal forms at different temperatures.

Recent proposals for rew rate equations (with the exception of Klepfer's, Exhibit 258, page 40) depend largely upon the data of Hobson and Rittenhouse, Exhibit 509. An example is the derivation by Westinghouse in Exhibit 1078, pp. 75 78 and pp.137140. However, Babcock and Wilcoy in their concluding statement (page 241) suggested that the Oak Ridge experiments may have suffered from steam starvation, and also suggested that the temperatures may have been higher than assumed from the furnace temperature measurements. These two observations, which may be justified, cast some doubt on the Oak Ridge results and possibly others of similar origin. The required measurements are difficult to make accurately and there is usually a wide spread in the results, especially from different investigators. This is evident, for example, from figure 3-4, page 137, of Exhibit 1078, where values of the oxidation thickness are plotted against a measure of the time-temperature history.

Until new data are obtained and present doubts are resolved we believe it best to continue the use of the Baker Just equation. It apparently gives about the correct results at 2000*F, and although it probably over predicts the oxidation at 2200*F, this over prediction does not exceed the range of the data available.

For example, the Baker Just equation fairly represents the oxidation data depicted in figure 1, page 3, of

, Exhibit 258, up to about 2372* F, well above our allowed maximum ternperature.

There is evident need for new and better experimental data to resolve this issue and to provide a rate equation with a more representative activation energy. If we were to establish such a new equation at this time, we would choose one that provides about the same calculated oxidation over the temperature range i of 1800 to 2200* F as the Baker Just equation, so that its continued use makes little practical difference.

6. Ret.ctor Internals Heat Transfer. A substantial amount of heat is stored in the metal parts of the primary circuit of a reactor system. This heat would be transferred rather slowly to the coolant and would 1103

_. _ _ _ _ _ _ . _a

therefore be 'nore unportant to the analysis of small breaks. Although General Electric in their response to

' the Staff's concluding statzment obj cted to its broader application, the Commission rules that this

. requirement should be applied to all evaluation rnodels in the interest of having them as realistic in the l phenomena comprised as is reasonable.

)

7,hesaurired IMater Resctor himary toSecondary Heat Transfer. This feature is recognized as important and does not represent any change from present practice.

B. SWELLING AND RUFIURE OF'IllE CLADDING AND FUEL ROD THERMAL PARAMETERS Each evaluation model shall include a provision for' predicting cladding swelling and rupture from consideration of the axial temperature distribution of the cladding and the difference in pressure between the inside and outside of the cladding, both as functions of time.To be acceptable the swelling and rupture calculations shan be based on' applicable data in such a way that the degree of swelling and incidence of rupture are not underestimated. The degree of swelling and rupture shall be taken into account in L calculations of gap conductance, cladding oxidation and embrittlement, and hydrogen generation.

The calculations of fuel and cladding temperatures as a function of time shall use values for gap conductance and other thermal parameters as functions of temperature and other applicable time-dependent variables. The gap conductance shall be varied in accordance with changes in gap dimensions and any other applicable variables.'

DfSCUSS/ON. In the postulated LOCA the reactor system pressure would drop rapidly and would soon fan below the pressure of the helium and fission gases within the fuel rod. The resulting differential pressure would exert an expansive force on the cladding. At the same time, as the cooling effectiveness dropped, the temperature of the cladding would increase rapidly, decreasing the yield strength of the claddag. At some time duping the LOCA the yield strength of the zircaloy might become less than'the tenaus stresses exerted by the differential pressure, and the cladding would then swell and perhaps burst.

For example, Babcock and Wilcox, using the evaluation model of the Interim Policy Statement, estimated that, for pressurized fuel, rupture of the cladding wocid be predicted over 70% of the core 1.3 seconds after the maximum size cold leg break. (Exhibit 1059, p. 6 4.) This corresponds to the time when the differential pressure would be about 200 psi and the cladding temperature about 1800* F.

Westinghouse, in a similar calculation, conservatively estimated that 25% of the fuel rods would burst sometime during blowdown, and that, by the end of the reflood period,70% of the rods would burst.

(Exhibit 1078, pp.D48 and D49.) Combustion Engineering calculated the degree of flow blockage resulting from rod swelling for each fuel assembly in the core for both unpresurized and pressurized fuel.

l In both cases, as judged from the blockage, they were calculated to be perforated or swollen rods in nearly

every fuel assembly. The ma}or difference between the pressurized and unpressurized fuel was that the L unpressurized fuel was estimated to tindergo less swelling and perforatiori during blowdown, as of course might be expected. (Exhibit 1144, sec. 5, using material from Exhibit 1%6, sec. 2.)

For the Boilms Water Reactor the situation seems to be somewhat different. The blowdown would p provide a longer period *of assured effective cooling of the fuel elements, and the initial calculated rise in  !

temperature of the cladding is not so great. Furthermore, the pressure within the fuel rods is said to be low,

so that bauoaning of the cladding would not be expected to occur during the blowdown. (Exhibit 1001,

p. 2 24.) General Electric offered one calculation for a 1967 product line flWR for which the peak cladding ,

temperature was 21054. (Exhibit 1148, sec.P.) Using some of the assumptions made by CNI (Exhibit i 1041, sec. 7.2), but ushg a constant internal fuel rod pressure, they calculated that 13% of the rods in the j hottest bundle would perforste. CNI, uains the probably erroneous assumption that there was no I conununinarian between the hot spot and the fission gas plenum 'at the top of the fuel rod, estimated that j 22% of au the fuel rods in the whole reactor would rupture. They said that this compares with 21%  !

estimated by General Electric for the Pilgrim reactor. (Exhibit 1041, p. 7.9.) In Exhibit 1032, page !!. '

8.2-1, reference is made to a calculation for a Boiling Water Reactor in which 60% of the fuel pins were ,

expected to rupture by the time the ECCS core sprays came on, with 75% of the pins expected to rupture  !

ultimately.  !

For the most part General Electric seems to rely on the experimental evidence for cl ddmg expansion '

and channel blocking in the BWR FLECHT test Zr 2. The conditions in this test were stated to be

{

1104-L L

_.m er

significantly more severe than the conditions reasonably expected to prevail during a postulated BWR LOCA, even for the " hot" bundle. (Exhibit 1148, p. P.15.) In Zr.2, the maximum cladding temperature was approximately 2250*F, and 39 out of the 49, or 80%, of the heater rods perforated. (Exhibit 1069, pp. 53-54.)

From the above it is obvious that, when the course of the LOCA is calculated according to the conservative prescriptions of an approved evaluation model, swelling and bursting of the cladding will be estimated to occur, in abundance. Three of the reactor manufacturers (Babcock and Wilcox, General Electric, and Westinghouse) have objected to applying the ECCS :riteria to regions of the cladding for which swelling is calculated (Concluding Statements). On the other hand, Combustion Engineering said, "The extent of oxidation thus limited"(by CE's version of the criteria)"for the worst case in the rupture i

! region assures that the remaining cladding will remain intact in a LOCA." (CE Concluding Statement,

p. 213.) The Consolidated National Interveners in their Concluding Statement were very much concerned with swelling of fuel cladding, particularly dwelling on flow blockage. They assumed that swelling should be taken into consideration, but were concerned that there may not be enough information to calculate blockage accurately or to take its effects into account properly. The ECCS Utility Group, in their Concluding Statement, recommended "that no changes be made to the current evaluation model: to

' account for effects of clad thinning and inside oxidation near the locations of fuel rod ruptures,"

commenting that "it seems most appropriate to continue to focus on the other more than 98 percent of the core."

Some of the objections to inclusion of the effects of cladding swelling and perforation will be dealt with m more deta$ below, but the Co'nmission sees no merit in ignoring these consequences of a calculation made with an officially approved evaluations model.

Much of the information on which calculations of cladding rupture are ba' sed comes from experiments in which zircaloy tubes were pressurized and heated at specified rates. Observations were made of the temperatures at which the tubes swelled and t urst, and profiles of the expanded tubes were used to obtain the degree of swelling. Many of the tests havo been made with single rods which were free to expand l radially. Other tests (multirod tests) have bee. made with the zirconium tubes arranged in arrays with spacing similar to that which they would have in a reactor. In this kind of test, interference between adjacent rods can limit their expansion.

From these tests, curves can be constructed relating the internal pressure to the temperature at which

! the tubes burst. Such curves have been published by ORNL (Exhibit 1007b, fig.1), General Electric l (Exhibit 137, fig. 9), Combustion Engineering (Exhibit 1066, fig. 2.4) and Westinghouse (Exhibit 1078, l

p. D-64. The amount of swelling that the cladding would experience before rupture appears to depend upon i the teniperature at which it would begin to deform,in dependence on the ductility of the crystal structure i

of the zircaloy at that temperature, and on the rate of heating.(Exhibit 1007b.) The amount of swelling is usually plotted as a function of the internal pressure, which is of course related to the deformation

_J temperature (e.g., flg. 2.5 of Exhibit 1066). This information is then transformed by calculation to percent flow blockage as a function of rod internal pressure (e.g., Exhibit 1007b, flg. 4; Exhibit 1078, fig. D-2).

L From this type of information, together with deuiled information on the internal pressure of the fuel rods i and their temperature history, the degree of expected clad swelling for each reactor can be calculated.

l The data from the rod burst tests show a great deal of scatter, particularly in the degree of swelling experienced. (See, nor example, flg. 2.5 of Exhibit 1066). The greatest controversy, however, has been with respect te laterpretations and predictions of the resulting blockage to coolant flow. (See, for example, concluding statements: Babcock and Wilcox, p.177; General Electric Vol.II, sec.0; Westinghouse, Appendix B; CN! Ch. 5, sec. A; ECCS Utility Group, p. 50; Combustion Engineering, Exhibit 1144, p. 5 1.)

It is expected that variations in cladding thickness. fuel pellet properties, gap thickness and eccentricity, and the texture of the zircaloy because ofits anisotropic character would go into the determination of just where along the length of a fuel rod the perforation would occur. (Transcript i1,51518).

The swollen and perforated region is expected to be about 1% to 3 inches long, and to occur at random as determined by the above variables over a length of relatively uniform temperature of from 7 inches to 27 inches. (Trans.12,701; Exhibit 1066, fig. 2.7; Exhibit i144, p. 5.2.) Thus it is not expected that a large omber of adjacent rods would have their maximum swelling in the same plane. The maximum blockage bserved to date in any multirod experiment containing 16 channels or more has been approximately 70%

on any horizontal plane. (Trans. pp. 9166 7L 1105

_ _ _ = _ -_- - _ _ _-

I All in all, the record still supports the Regulatory Staff position in Exhibit 1001, pp. 212, namely, that the core-wide flow area reduction in the plane of greatest blockage would not exceed 60% and that local flow channel reductions, over perhaps a 4 x 4 array of fuel rods, would not exceed 90%. As shown by calculations of blowdown and ECCS heat transfer considered elsewhere, these reductions of flow stea, while necessary to be considered, would not be disastrous. In other words, the Commission concludes that estimated fuel rod swelling and rupture would not render the geometry of the core to be uncoolable.

The swelling of the cladding would have an important effect on the thermal conductance between the fuel pellets and the cladding (the gap conductance), but none of the evaluation models approved under the Interim Policy Statement has the capability of calculating the change of the geometry of the cladding during the course of the postulated accident and the effects of this change. Neither do they include a thermal radiation term in the calculation of the gap conductance, a term that becomes important at high temperatures and for large gap dimensions (Concluding Statement of the Regulatory Staff). At low temperatures the effect of prior swelling would be to decrease the gap conductance drastically,especially if there were an appreciable arnount of fission product gas mixed with the original helium fill.(Trans.5536; Exhibit 1I13, pp 10-21.) A decrease in gap conductance during the blowdown would tend to insulate the cladding from the fuel pellets, reducing the amount of stored energy removed during blowdown, and keeping the temperature of the cladding low during this time. (Exhibits 1151, p.131; 1113, sec.10; 1137, p.10 3.) For pressurized water reactors during refill and especially during reflood, radiation from the fuel pellets to the cladding would become important for the rods with the swollen cladding, initially exceeding the rate of heat removal from the cladding. The greater stored energy thus transferred to the cladding would tend to cause the peak temperature of the cladding to exceed that obtained by present evaluation models. (Exhibit 1113, sec.10.) The scoping studies carried out by reactor manufacturers did not have the radiation term and so do not properly calculate the peak temperature.

For boiling water reactors, the change in gap conductance from clad swelling probably would have little effect on the heat removal, since swelling would be calculated to occur so late in the LOCA. However, the swelling of the cladding would be expected to hr,ve a direct effect on heat removal by both spray cooling and radiation. It is true that GE reported a calculation leading to their conclusion that radiative heat transfer is little affected by clad bulging (Exhibit 132, p. D 55), but that calculation may have contained the doubtful assumption of a uniform circumferential temperature for each rod, We cannot rely on the stated result of the calculation without knowing what assumptions were made. Also, the data of the Zr 2 BWR FLECHT experiment were cited as evidence for the effectiveness of spray cooling, although no temperature measurements were made at the positions of maximum buying. We believe that additional assessments need to be made of these effects.

In addition to the primary heat transfer effects of taking into consideration the swelling and rupture of the cladding, there would be important secondary effects arising from the steam oxidation of the cladding by the steam. Higher temper-tures would lead to increased oxidation, which would contribute to a further increase in temperature, and the opening in the cladding would allow oxidation on the inside, again increasing the calculated temperature.

Babcock and Wilcox, in discussing swelling in their concluding statement, stated that clad swelling would not have a significant effect on cladding temperature response and therefore need not be considered in LOCA evaluations (p.183). The Regulatory Staff challenged the assertion of insignificance in their concluding statement (p.153), in part on the basis that the B&W calculation did not account for clad thinning or for radiation from the fuel to the clzdding, and added the statement that a Sl*F increase in peak temperature can be significant.

With regard to the related subject of transient gap conductance calculations, Babcock and Wilcox gave a number of reasons why they contend that the problem of transient gap conductance "is an artifact of she calculation and not a significant problem in any physical sense." (Concluding Statement, pp. 184-193; Response to the Concluding Statement of the Regulatory Staff, pp. 48 58.) These reasons are:(1) That the initial stored energy is overestimated. (2)The peak power density that is assumed is higher than can be reasonably expected. (3)The blowdown heat transfer would be substantially better than the calculation allows. This would be especially important because, if trua, it would prevent clad swelling and rupture during blowdown. (4) B&W interpreted the Staff statement on page 1711 of Exhibit 1113 to rnean that oscillatory flow during reflood would greatly improve reflood heat transfer. In our discussion on reflood heat transfer we urge caution in making this interpretation, at least for the present.(5) According to B&W 1106 l

l l

studies the calculated peak temperature is insensitive to variations in gap conductance. However, as pointe out above, their calculation did not include radiative heat transfer between the UOs and the cladding.

(6)The requirement that no heat shall be considered lost from the fuel rods during refill until the water reaches the bottom of the core was said to be extremely conservative. Although this is not a requirement of r

this rule,it makes little difference because there would be very little heat transferred during this period to the stagnant superheated steam that would be present. (Exhibit 1001, at 3 36).

l

' The Commission agrees that items (1), (2) and (6) provide moderate degrees of conservatism, and believes it probable that item (3) provides a substantial degree of conservatism. These are welcome conservatism;it is intended that there should be a margin to allow for extremes in statistical fluctuations from the expected behavior of the systems, components, flows, cooling mechanisms, and materials. The fact that the calculation is intentionally conservative should not, in the opinion of the Commission, keep one from following the consequences of that calculation to its logical conclusion.

General Electric,in opposing any change in the Interim Policy Statement for Boiling Water Reactors, stated that the " preponderance of the evidence in the record provides no basis for changing the BWR evaluation model." With regard to clad bulging, they said ". . . the overwhelming weight of the evidence (and all of the reliable evidence) indicates the flow blockage is not a concern for the BWR."(Closing I' Statement, p. 0 4.) With regard to taking into consideration clad thinning and oxidation inside burst rods, "GE strongly opposes the . . . suggestions . . . that this highly localized wall thinning and inside oxygen penetration should form the basis of any aspect of the embrittlement assessments used for licensing (

purposes." (Closing Statement, p. M.22). On pages K.35 to K.40 of their Closing Statement they argued that the requir,ement for calculating char.ges in gap conductance during the LOCA can not apply to BWR's,

]

largely on the basis that the calculations showing the possible magnitude of the effects were made for PWR's. ,

General Electric further commented on " Clad Swelling and Rupture" in their Response to the Concluding Statement of the Staff, Vol. 2, pp. 5176. Some differences between our wording of the rule and that proposed by the Regulatory Staff are in response to some of the comments on page St.The rule proposed by the Staff was criticized by GE as vague and unworkable. Phrases such as " applicable data" and "in a conservative way" were characterized as imprecise. The Commission believes that the reactor vendors have the capability to interpret the existing data in formulating adequate evaluation models that include the effects of clad swelling and rupture. For instance, calculating the incidence of cladding rupture could be I done using the methodology described in Exhibit 1144, pp. 51 and 5 2.

As stated above, we believe that the effect of calculated clad swelling on heat transfer may be small for a BWR, as claimed by GE, although the evidence of this in the record is not very extensive. However small the effect might be, we believe that ignoring it would be nonconservative, and so it should be included to make sure the evaluation model is as comprehensive as the present state of understanding allows. With regard to taking into consideration the calculated thinning of the cladding and internal oxidation of burst rods, we h, ave already ruled in connection with our discussion of the first two criteria that the maximum l temperature limit and the oxidation limit should be applied to the section of the cladding that would reach l the highest temperature and be most oxidized. Therefore the estimated burst of cladding must be taken l into consideration.

l Westin$ouse objected to the inclusion of clad swelling and rupture on the basis that there is no reason to beheve that highly localized and limited rod cracking or severance during quench or at any other time in the transiset would impair core coolability. (Comments on the Concluding Statement of the Regulatory Staff). However, since there is ir. sufficient evidence of a substantial nature to the effect that that such localized damage would prove harmless, the Commission must make the conservative choice.We therefore

! require that the most highly damaged places calculated to be found along the length of the fuel rods shall be taken into consideration.

In the Consolidated Naticaal Interveners' Concluding Statement, Chapter V discusses some of the technical issues. Sections A on Flow Blockage and C on Gap Conductance are pertinent to this section. In Section A CNI implied that the state of knowledge of flow blockage is deficient, and they stated further:

"It would appear that swelling can and likely will yield an unacceptable degree of coolant channel closure in some LOCA circumstances." With regard to the state of knowledge, the Commission recognizes that there are differences of opinion as to how flow blockage should be calculated from the basic swelling data, and that there is a high degree of scatter in the swelling data. Neither of these situations seems unnatural;in 1107

_ -. __ _ )

fact they make it possible to understand better the degrres of conservatism in the different calculations.

Additional multirod tests will help to refine the calculations, but if proper regard is shown for the divergent views, the present situation allows sufficiently conservative estimates. As to the possibility of "an unacceptable degree of channel closure," the Commission has seen no hard evidence that this is at all likely, and believes that the risk of such a circumstance occurring is acceptably small.

C. BLOWDOWN PHENOMENA

.1. Break Characteristics and Flow

a. In analyses of hypothetical loss of-coolant accidents, a spectrum of possible pipe breaks shall be considered. This spectrum shall include instantaneous double-ended breaks ranging in cross sectional area up to and including that of the largest pipe in the primary coolant system. The analysis shall also include

. the effects of longitudinal splits in the largest pipes, with the split area equal to the cross sectional area of the pipe,

b. Discharge Nodel. For all times after the discharging fluid has been calculated to be two-phase in composition, the discharge rate shall be calculated by use of the Moody model(F.J. Moody," Maximum Flow Rate of a Single Component, Two-Phase Mixture," Journal of heat Transfer, Transactions of the American Society of Mechanical Engineers, 87, No.1, February,1965). The calculation shall be conducted with at least three values of a discharge coefficient applied to the postulated break area, these values spanning the range _ from 0.6 to 1.0. If the results indicate that the maximum clad temperature for the hypothetical accident is to be found at an even lower value of the discharge coefficient, the range of discharge coefficients shall '

be extended until the maximum clad temperature calculated by this variation has been achieved.

' c.End of Blomfown., (Applies Only to Pressurized Water Reactors.) For postulated cold leg breaks, all emergency cooling water injected into the inlet lines or the reactor vessel during the bypass period shallin .

the calculations be subtracted from the reactor vessel calculated inventory. This may be executed in the calculation during the bypass period, or as an alternative the amount of emergency core cooling water l- calculated 1o be injected during the bypass period may be subtracted later in the calculation from the water remaining in the inlet lines, downcomer, and reactor vessel lower plenum after the bypass period. This bypassing shall end in the calculation at a time designated as the "end of bypass," after which the expulsion or entrainment mechanisms responsib!c for the bypassing are calculated not to be effective. The end of-bypass definition used in the calculation shall be justified by a suitable combination of analysis and experimental data. Acceptable methods for defining "end of bypass" include, but are not limited to, the following: 1. Prediction of the blowdown calculation of downward flow in the downcomer for the remainder of the blowdown period; 2. prediction of a threshold for droplet entrainment in the upward velocity, using local fluid conditions and a conservative critical Weber number. j d.Noding Near the Break and the ECCS Injection Points. The noding in the vicinity of and including the broken or split se,ctions of pipe and the points of ECCS injection shall be chosen to permit a reliable J analysis of the thermodynamic history in these regions during blowdown. j 2.Frictisant Pressere Drops i

. The frictional lossue in pipes and other components including the reactor core shall be calculated using l, models that include sealistic variation of friction factor with Reynolds number, and realistic two-phase l

friction multipliers that have been adequately verified by comparison with experimental data, or models ,

1 l that prove at least equeuy conservative with respect to maximum clad temperature calculated during the l- hypothetical accident. T.he modified Baroczy correlation (Baroczy, C.J., "A Systematic Correlation for L Two Phase Pressure Drop," Chem. Enging. bog. Symp. Series. No,64, Vol. 62,1965) or a combination of the Thom correlation (Thom, J.R.S., " Prediction of Pressure Drop During Forced Circulation Boiling of

! . Water," Int. J. of Heat & Mass Tnsasfer, 7, 709 724, 1964) for pressures equal to or greater than 250 psia ,

f- ' and the Martinelli-Nelson correlation (Martinelli, R.C., Nelson D. B.," Prediction of Pressure Drop During l Forced Circulation Boiling of Water," Thmsections ofASME, 695 702,1948) for pressures lower than 250

_ psia is acceptable as a basis for calculating realistic two phase friction multipliers. (See Combined Discussion of Frictional Pressure Drops and Momentum Equation below.)

1108 I

e l

[. , .

3. Momentum Equation The following effects shall be taken into account in the conservation of momentum equation:

1. temporal change of momentum, 2. momentum convection, 3. area change momentum flux, 4. mo-mentum change due to. compressibility, 5. pressure loss resulting from wall friction, 6. pressure loss resulting from area change, and 7. gravitational acceleration. Any omission of one or more of these terms under stated circumstances shall bejustified by comparative analyses or by experimental data.

4. Critical Heat Flux

a. Correlations developed from appropriate steady-state and transient state experimental data are acceptable for use in predicting the critical heat flux (CHF) during LOCA transients. The computer prograrns in which these correlations are used shall contain suitable checks to assure that the physical parameters are within the range of parameters specified for use of the correlations by their respective authors,

b. Steady state CHF correlations acceptable for use in LOCA transients include, but are not limited to, the following:

(1) W 3. L. S. Tong, " Prediction of Departure from Nucleate Boiling for an Axially Non uniform Heat Flux Distribution," Journal o/ Nuclear Cnergy, Vol. 21. 241248,1967.

(2) Bd W-2. J. S. Gellerstedt, R. A. Lee, W. J. Oberjohn, R. H. Wilson, L. J. Stcnck, " Correlation of Critical Heat Flux in a Bundle Cooled by Pressurized Water," Two-Phase FlowandHeat Transferin Rod Bundles, ASME, New York,1969.

(3)Hench'-Levy. J. M. Healzer, J. E. Hench, E. Janssen, S. Levy " Design Basis for Critical Heat Flux Condition in. Boiling Water Reactors," APED-5186, GE Company Private report, July 1966.

(4)Macbeth. R.V.Macbeth,"An Appraisal of Forced Convection Burnout Data," Proceedings of the institute ofMechanicalEngineers, 1965 1966.  ;

(5) Barnett. P. G. Bamett,"A Correlation of Burnout Data foz Uniformly Heated Annuli and its Uses I for Predicting Bumout in Uniformly Heated Rod Bundles," AEEW R 463,1966.

l (6)Hughes. E. D. Hughes, "A Correlation of Rod Bundle Critica! Heat Flux for Water in the Pressure i

Range 150 to 725 psia." IN.1412, Idaho Nuclear Corporation, July 1970.

c, Correlations of appropriate transient CHF data may be accepted for use in LOCA transient analyses if comparisons between the data and the correlations are provided to demonstrate that the correlations predict values of CHF which allow for uncertainty in the experimental data throughout the range of parameters for which the correlations are to be used. Where appropriate, the comparisons shall use statistical uncertainty analysis of the data to demonstrate the conservatism of the transient correlation.

i

d. Transient CHF correlations acceptable for use in LOCA transients include, but are not limited to, the

! following: i (1) CE tmnsient CHF. B. C. Slifer, J. E. Hench, " Loss-of-Coolant Accident and Emergency Core Cooling Models for General Electric Boiling Water Reactors," NEDO.10329, General Electric Company, Equation C 32, April 1971.

e. After CHF is first predicted at an axial fuel rod location during blowdown, the calculation shall not use nucleets boiling heat transfer correlations at that location subsequently during the blowdown even if the calculated local fluid and surface conditions would apparently justify the reestablishment of nucleate boiling. Heat transfer assumptions characteristic of return to nucleate boiling (rewetting) shall be permitted when justified by the calculated local fluid and surface conditicns during the refiood portion of a LOCA.

5. Post CHF Hest Transfer Correlations

( a. Correlations of heat transfer from the fuel cladding to the surrounding fluid in the post-CHF regimes of transition and film boiling shall be compared to applicable steady-state and transient state data using statistical correlation and uncertainty analyses. Such comparison shall demonstrate that the correlations predict values of heat transfer coefficient equal to or less than the mean value of the applicable experimental heat transfer data throughout the range of parameters for which the correlations are to be used. The comparisons shall quantify the relation of the correlations to the statistical uncertainty of the applicable data.

I 1109 j

T l b. The Groeneveld flow film boiling correlation (equation 5.7 of D.C. Groeneveld,"An investigation of H at Transfer in the Liquid Deficient Regime," AECL.3281, revised December 1969), the Dougall-Rohsenow flow film boiling correlation (R. S. Dougail and W. M. Rohsenow. " Film Boiling on the inside of I

Vertical Tubes with Upwstd Flow of the Fluid at Low Qualities," MIT Report Number 9079-26, Cambridge, Massachusetts, September 1963), and the Westinghouse correlation of steady. state transition boiling (" Proprietary Redirect / Rebuttal Testimony of Westinghouse Electric Corporation," U.S.A.E.C.

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, " Partial Film Boiling with Water at 2000 psig in a Round Vertical Tube." MS A Research Corp.,

Technical Report 62 (NP-6976), (1958)) is suitable for use between nucleate and film boiling. Use of all these correlations shall be restricted as follows:

(1) the Groeneveld correlation shall not be used in the region near its low-pressure singularity, (2) the first term (nucleate) of the Westinghouse correlation and the entire McDonough.Milich.and King correlation shall not be used during the blowdown after the temperature difference between the clad and the saturated fluid first exceeds 300*F, (3) transition boiling heat transfer shall not be reapplied for the remainder of the LOCA blowdown, even if the clad superheat returns below 300 F, except for the reflood portion of the LOCA when justified by the calculated local fluid and surface conditions.

6. Pump Modeling The characteristics of rotating primary system pumps (axial flow, turbine, or centrifugal) shall be derived from a 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 used for analysis should be justified. The pump model for the two-phase region shall be verified by applicable two phase pump performance data. ,For BWR's after saturation is calculated at the pump suction, the pump head may be assumed to vary linearly with quality, going to zero for one percent quality at the pump suction, so long as the analysis shows that core flow stops before the quality at pump suction reaches one percent.

. Jore Flow Distribution During Blowdown (Applies only to pressurized water reactors.)

a.The flow rate through the hot region of the core during blowdown shall be calculated as a function of time. For the purpose of these calculations the hot region chosen shall not be greater than the size of one fuel assernbly. Calcu.8ations of average flow and flow in the hot ' region shall take into account cross flow between repons and any flow blockage calculated to occur during blowdown as a result of cladding swelling or rupture. The calculated flow shall be smoothed to climinate any calculated rapid oscillations (period less than 0.1 seconds).

b. A method shall be specified for determining the enthalpy to be used as input data to the hot channel heatup analysis from quantities calculated in the blowdown analysis, consistent with the flow distribution calculations. ,

DISCUSSION

1. Break Characteristics and Flow

a. The Interim Acceptance Criteria and the models that have been accepted for their application require cop.;idering the consequences of postulated rupture of primary system pipes. The requirements differ sorr.ewhat from one accepted model to another. In general, analysis must be made of the effects of rupture of pipes whose areas range up to that of the largest pipe in the primary system. Double-ended rupture must be considered, which means that it must be assumed that the pipe is severed instantaneously by a circumferential break, and the two parted sections undergo mutual lateral displacement so that nch can discharge primary system water without interference from the other.

For pressurized water reactor models, e requirement was also established for analysis assuming instead of double-ended breaks, longitudinal splits whose areas ranged from a factor of 0.6 to 1.0 times that of the maximum. No such requirement was placed on analysis applied to boiling water reactors.

1110

1

' For the approved AEC Model,, Babcock and Wilcox Model, and Combustion Engineering Model, a lower a

end of the break spectrum to be analyzed was defined as 0.5 ft . For the Westinghouse Model, no such lower limit was set, but exceptions to the Westinghouse Model were stated ordy for pipes exceeding 0.5 ft' in area.

The spectrum of pipe breaks to be considered was not an issue during the hearing. The original testimony presented by Aerojet ir.dicated little difference to be found between the effects of a split and a

&ble-ended break of the largest pipe for a 3-loop PWR, though Ybarrondo stated in questioning his belief on theoretical grounds that a split should lead to lower rates of fluid discharge (Transcript, p.10848).

Other information in the hearing record also confirms the essential equivalence of effects of alits and double.cnded breaks.

The rule proposed by the Staff in its Concluding Statement of Position has rectified the existence of l~ differences in assumed break sizes and types for models to be accepted, but has done so at some expensnf clarity of direction. We believe that the rule should be explicit in its requirement that analysis be made of the effect of splits as well as double-ended guillotine breaks. We agree with the Staffin eliminating specific reference to a minimum in the size of break to be analyzed. We have worded our statement of requirements to reflect these points.

b. Discharge Model. The blowdown period can be divided into two parts, from the standpoint of models to describe the discharge of primary coolant. During the first, the fluid leaving the postulated break would be liquid or almost entirely liquid. The second would begin when the fluid is definitely two-phase.

The record of the hearing considers at length an analytical method based on use of a correlation due to Moody, incorporated in all accepted calculational models for treating the rate of discharge of two-phase coolant duri6g the second interval. Moody's correlation was derived assuming the two phases in the discharge nqzzle to be in thermodynamic equilibrium. This assumption leads to a relationship for the maximum rate of fluid discharge.

The record shows that predictions based on the use of the Moooy model have been extensively tested against blowdown experiments in several facilities. These include blowdowns of the Containment Systems Experiment at Hanford, the LOFT Semiscale vessel at NRTS, and simple pipe blowdowns by Edwards in England. Some comparisons have also been made with the results of other experiments including some by vendors.

The Regulatory Staff, ANC, and the four vendors all agreed that calculated blowdown rates based on direct use of the Moody correlation overestimate the observed rate of discharge for any given break size, especially for the larger breaks. ANC testimony and answers during questioning of the Staff (Ross, i Transcript, pp. 1800 1806) and ANC personnel during the hearing attributed the difference possibly to Moody's use of equilibrium thermodynamics. The testimony by Roy (Babcock and Wilcox) expressed the same view. The lack of thermal equilibrium would be expected to have a greater effect when cold accumulator water is injected in the nozzle near the hypothetical cold leg break. Ybarrondo stated that ANC calculations can now fit discharge rates without use of a critical flow model such as Moody's and associated use of a discharge coefficient (see below). But the Staff's Supplementary Testimony attributed this to ANC's incorporation of the momentum flux into the computational model used, rather than a non-equilibrium calculation.

Because the calculation overestimates the discharge rate and underestimates the time until end of blowdown, ANC and the vendors have all tried modifications of their calculations based on use of a socallededischarge coefficient. This coefficient is applied directly to the break area for which the calculation is made; thus the calculated blowdown assuming a break area A and a discharge coefficient D would be made assuming a mythical area AD. It is found that if a fixed value of the discharge coefficient is used, it is chosen best to be below unity, especially for the larger breaks. This reflects the need to reduce the calculated discharge rate to agree with observation.

The initial Staff testimony and the Staff Supplementary Testimony reported a best value of about 0.6 for the discharge coefficient for large LOCA breaks; this value was supported by Roy's testimony (Babcock & Wilcox), and by Zane (Transcript p.10791), Cermak (Westinghouse, Transcript p.15134), and j Moore (Westinghouse, Transcript p.15504). I There was also widespread agreement that a variab!c discharge coefficient provides a better fit to the data than a constant one (Brockett, Transcript p. 7484; Moore, Transcript p.15161; Staff Supplementary

! Testimony; Babcock and Wilcox Redirect and Rebuttal Testimony; Bingham, Transcript p.21143; Roy, Transcript p. 21144).

1111

+j i

'l l There was some difference of opinion as to the effect of the value of the discharge coefficient on the calculated clid temperature. The Babcock and Wilcox Direct Testimony report a reduction in maximum l calculated clad temperature when the discharpe coefficient was taken as 0.6 instead of 1.0. Colmar's l l Testimony stated that a reduced discharge coefficient could increase the maximum clad temperature l through delaying the onset of ECCS, and that a variable discharge coefficient can lead to a higher maximum calculated clad temperature than a fixed value of unity. Ybarrondo (Transcnpt p. 6362) reported the result of an ANC calculation with a discharge coefficient initially 2.0 and later in the blowdown 0.6, where the first peak in the clad temperature exceeded by about 100 the value obtained with a fixed discharge coefficient of 1.0 Roy (Transcript p.12944) reported lowered maximum clad temperatures when a variable discharge coefficient was used by Babcock and Wilcox in place of the value 1.0. lanni(Transcript pp.1406814071) reported a reduced maximum clad temperature calculated by General Electric when a discharge coefficient less than unity was used. Moore (Transcript E.26) reported a similar result from proprietary Westinghouse analyses.

The Direct Testimony of the Consolidated National Interveners argued that the use of the Moody model leads to incorrect results because the liquid discharged would be metastable (supercooled) and so would emerge at a rate about 1.7 times that predicted by Moody. This would lead to higher traasient containment pressures and to faster core dryout which would cause maximum calculated clad temperatures to be highu, it was stated that this tendency would be found especially if the break were near the vessel.

Data from experiments by Barton and by Fauske were cited in support of CNI's view. During questioning, Dawson (Transcript pp. 18348 18349) added that CNI also has some supporting data based on cryogenic fluids. The CNI position was discussed at length during questioning reported on pp. 1829018653 of the Transcript.

The Redirect and Rebuttal Testimonies of Babcock and Wilcox and of General Electric generally addressed the points made,by CNI, and disagreed with CN1 both in detail and in general. General Electric testimony claimed that CN1 misinterpreted the Fauske data and ignored data of Barton that do not support their case. GE reported that Edwards' blowdowns, Bogarty data, and other data show metastability to have decayed in less than 1 millisecond which is less than the transport time of fluid from the nozzle entry to the hypothetical break. The Babcock and Wilcox testimony pointed out that the Fauske data were obtained with a % inch diameter orifice near the bottom of a vessel 10 ft. high, and the liquid at the bottom of the I vessel was subcooled because of the liquid head. For this and other stated reasons Babcock and Wilcox claimed that the Fauske data used to support the CN1 position were not relevant.

The Staff took the position in its concluding Statement that the Moody modelis well supported, apart from the possible need to use a discharge coefficient, and that this model should be used for quality exceeding 0.02.

We agree with the Staff position as to correctness of use of the model based on critical flow, since the length of time available during the blowdown far exceeds the amount needed for nucleation and build up of two-phase discharge. Furthermore, the evidence is strong that use of the Moody correlation does not underestimate observed experimental discharge rates, as would be the case if discharge were really metastable, but in fact it definitely overestimates the discharge rates.

We require calculat, ions with at least three values of the discharge coefficient ranging from 0.6 to 1.0.

Use of a discharge coefficient less than unity is equivalent to assuming a smaller break, and the requirement to calculate the consequences of a spectrum of breaks whose sizes range up to the area of the largest primary coolant pipe makes the calculations with discharge coefficients less than unity seem unnecessary.

However, the practical effect of the redundancy seems negligible. The need to calculate for smaller breaks takes care of any concern raised in connection with Colmar's view as to the greater severity of longer I discharge times.

We note finally the point raised in the ANC direct testimony, to the effect that fits of calculated blowdown transients to experiments were based primarily on comparisons of pressures. Measurements of fluid density on of mass discharge rate were not reliable enough to be used in tests of analytical methods.

Lack of thermal equilibrium in the as-yet-to-be discharged fluid can lead to incorrectly estimated fluid inventory even when the pressure is correctly calculated. This is not likely to be very much of a problem near the end of blowdown, however, where rates of change in pressure and flow rate have eased off in rapidity. At this point the primary system inventory is mostly vapor, and therefore the average mass discharge rate over the entire transient will turn out to have been predicted adequately if the pressure time curve has been predicted properly.

I 1112

3 c.End of Blowdown. The event that was most induential in raising questions about the adequacy of emergency core cooling was the publicity given to the loss of injected coolant during blowdown in Semi. scale test 845 at Idaho, and one of the major features of the interim Policy Statement was the requirement that all emergency core coolant injected dunng blowdown should be considered lost. Although there is still objection to companng PWR's with the Semi. scale tests (Combustion Engineenng Response to (

the Regulatory Staffs concluding Statement. p. 23). there appears to be no major objection to the rule.

The rule moderates the arbitranness of the " accumulator bypass" section of the interim Policy Statement in a conservative way, requiring that the calculation of the end of bypass be supported by suitable experimental data. Although the Regulatory Staffin their concluding statement reject for the time being the Wallis correlation proposed by Westinghouse and the " Waterfall" concept suggested by Combustion Engineering, these methods of calculating the extent of cooling water bypass can clearly be considered under the rule when adequate calculational detail and experimental verincation are available.

The primary importance of calculating the correct curve of system pressure versus time for a PWR is found in calculation of the time at wiuch blowdown ends and effective emergency coolant injectka is assumed to start. Several definitions of the end of blowdown have been proposed, and each bia ,ome difficulties. The Staff Direct Testimony stated that end of blowdown is assumed to be the first time when Guid now from the break ceases. Parks (B&W, Transcript pp. 1252612527) stated that this may not be the best definition, because zero break flow may never be reached for some breaks such as hot leg breaks and very small breaks. But he added (Transcript p.12527) that B&W calculations show a time of zero break flow for all cold leg breaks but the smallest ones. Rosztoczy (Transcript p.13230) reported that no CE calculation has shown the core recovered before zero break flow occurs,and so zero break flow is adequate to define the *end of blowdown.

The B&lVt Redirect and Rebuttal Testimony discussed calculations of accumulator water entrainment which considered the end of blowdown to be the time at which entrainment ended. Roy stated however that B&W does not at this time recommend such a definition (Transcript pp. 21164-21165).

Definition of "end of blowdown" is replaced in the Staffs Concluding Statement by "end of bypass,"

which means assumed end of accumulator water bypass. This is the time at which calculations supported by experiments indicate the ECC bypass or entrainment mechanisms to be no longer effective. We believe that it will be possible to provide calculations satisfying these requirements,and this procedure for defining the start of effective emergency coolant injection is superior to the one used formerly. The proposed procedure I on page 188 of the Regulatory Staff's Coraluding Statement is correct and precise, and we have elected to adopt it verbatim.

f J

d.Noding Near the Break and the ECCS Inlection Points. A number of witnesses in the course of the hearing pointed out that incorrect noding near tae assumed break or the point of ECCSinjection can lead i to non. conservative conclusions. Excessively large nodes near ECCS injection could lead to unrealistic calculated cooling of the water in the downcomer and/or the lower plenum, and more rapid reduction of the driving force during blowdown than would really be expected. The noding detailin the vicinity of the break as discussed on page 40 of the Staff's Concluding Statement is directed to avoiding this problem.

The advantages of correct thermodynamic description of the discharge nodes have been discussed earlier.

2 and 3. rhen of Frictional Pressure Drops and Momentum Equation (a)/destification of the Terms. The momentum equation describes the Newtonian behavior of the coolant as it is subjected to such forces as Guid pressure and gravitation. Two terms in the momentum equation received special attention during the hearing. One of these is the momentum flux. The other is the frictional pressure drop.

l These two terms have assumed extra importance because they can become more than usually significant l when the coolant is two. phase.

l (b) Momentum Flux. The momentum Rux represents the spatial convection of momentum.In regions of the system where accelerations of the coolant are occurring,and where heat transfer changes the ratio of liquid water to steam, the momentum flux differs from zero. Thus the momentum flux assumes finite i values at places where area changes in now paths occue (form losses)and where heat sources and sinks are

)

l

• l 1113 l

.dedaaE Am. .

found. It is also finite if phase changes are the result of system pressure changes. such as accompany depressurization.

The Staffs supplementary testimony pomted out that none of the vendor computational models includes the motTentum Oux as it appears in the momentum equation. The form of RELAP 3 approved for use with the mterim criteria did not incorporate the momentum Oux, either. The models used by Combustion Engineenng and by Babcock and Wilcox did include the area change part of the momentum flux. The Westinghouse model, the General Electric Model, and the ANC model contained empincal form losses or abrupt losses to account for the area change and the heightened turbulence at such locations as the l core entrance and exit. Different methods were sometimes used to modify form factors for two phase conditions, and to change magmtudes of pressure losses when flow reversal occurred.

The questioning of witnesses led to a rather uniform set of statements to the effect that the momentum Rux is not very important to analysis of the transient. Bingham said that the B&W model does not include the density change component of momentum flux,but calculations that had been done accountmg for this component of the momentum Oux had not changed the results significantly (Transenpt p.12621).

I Rosztoczy reported insensitivity of the peak clad temperature as to whether calculanons are made with and without the momentum Dux term. Cermak said that the Westinghouse model used the momentum Oux in 1967, but the term was removed because it led to computational stabilities and had only a small effect on the result (Transenpt 15592). Rockwell said that the acceleration terms in the momentum flux amount I

only to 4 percent of the bundle pressure drop during the early part of a BWR transient. and the effect decreases as the transient proceeds (Transenpt p. 21420). He also said that the momentum flux is included in GE's empirical loss coefficients (Transcnpt pp. 21421. 21542 3).

j The Staffs Supplementary Testimony expressed the view that vendors' codes should be modified to include all momentum flux terms. This view was supported by consultants at ANC and at ORNL. The Concludmg Statement of the Staff in its proposed rule took the same position, but also indicated that specific terms can be left out if the omission is justified by comparative analysis or experimental data.

The B&W Concluding. Statement took issue with the Staffs position, generally on the basis that the ,

B&W CRAFT code can already predict the course of a blowdown so as to achieve a conservative assessment of system response to a hypothetical LOCA,and also that the record does not support any need to include all terms of the momentum flus in the model. These views were also stated m B&W's Response to the Concluding Statement of the Regulatory Staff.

The Concluding Statement of GE and also GE's Responsive Closmg Statement. Responding to .

Concluding Statement of Position of the Regulatory Staff, also gave similar reasons why the GE computer model need not be modified to include the momentum flux. We note that the question of the need to include the momentum flux in the momentum equation was in iact argued during the heanng. Views both pro and con were expressed, though it appears that the arguments against the need for momentum flux were a little more quantitative than those favoring the need. .

On the other hand, the evidence in all respects was not overwhelming. Since the momentum equation does contain the momentum flux term,it seems that the burden of proof must rest on those who wish to leave this term out. The contentions in the record do not add up to such proof.

The Staffs proposed course is fair and is supported by the record. Momentum flux is seen to be a reality, and in any stated circumstance its components must all be included m the analysis or any omission l i must bejustified. We have adopted the Staffs careful wording for this reason.

The record contains only a few references directly or indirectly as to including momentum flux in calculating the effect of any core blockage that might occur during blowdown. It should be clear that the requirements as to use of the momentum flux apply to all LOCA blowdown flow calculations.

(c)Two#hase Multipliers. The practice in estimating fluid friction for a two-phase Guid is normally to calculate the friction coefficient for a single phase fluid and then to multiply this value by a factor that corrects for the quality. Tids is the two phase friction factor.

The vendors' practices differ in calculating the single phase fnction coefficients. as is seen on page 4 5 of the Staffs Supplementary Testimony. Westinghouse and GE blowdown codes do not melude smgle phase friction factors with explicit dependence on the Reynolds number. The B&W and CE models do include such a dependence, as does the GE transient CHF model. Parks said that B&W's calculation of Reynolds number dependence is not extended into the laminar flow regime, because no lammar flow is encountered during blowdown (Transcript, p.12766-7). Rosztoczy stated that the CE smgle phase friction 1114 4

factor is applicable to turbulent Gow and not laminar flow. but only turbulent now is expected. He said that the Reynolds number dependence used has only a mild effect on the transient, and it is not one of the important factors.(Transcript p. 13285 6).

The Westinghouse Redirect and Rebuttal testimony stated that friction factors and form factors are lumped together before application of the two-phase correction correlation.

The correlations used to correct for the two-phoe effects also differ. Westinghouse uses the Armand correlation (Westinghouse Redirect and Rebuttal Testimonyt GE uses the Martinelli Nelson correlation (NEDO-10329). B&W and CE use the Thom correlation at pressures above 250 psi and the Martinelli Nelson below.

Evidence was provided that the results of calculations of LOCA consequences are not very sensitive to the choice of two phase friction factor (but see our reservation below relative to the effect on now redistribution.) Rosztoczy also discussed such conclusions (Transcript pp. D-12.13). llench said that in GE calculations the entire friction in the core only accounts for about 10?c of the core pressure drop (Transcript p.14443 4). Moore said that in Westinghouse calculations the entire friction in the core accounts for as much as 45"c of the core pressure drop dunng blowdown (Transenpt p.15591).

These comments as to adequacy are only relevant to the one-dimensional calculations with the blowdown code and the heat balance code. Here the core pressure drop is only a small part of the full system pressure drop, and so large changes in the core pressure drop can be made without affecting the blowdown significantly, it may be added that changes of this kmd can be compensated by changmg the empirical discharge coefficient, so that matching experimental blowdowns with a given two phase friction factor correlation and the discharge coefficient as a free variable proves nothing.

The recor'd makes it clear that proper treatment of the two phase friction is more important for calculating Gow redistribution in the core. This point was made by the Staff in its Supplementary Testimony, and by Morgan (Transcnpt p.12678 9). It was also made in the Oak Ridge comments on the Staffs Supplementary Testimony. Because the quality will vary significantly from the hotter to the colder channels, the fluid velocity correspondmg to the common pressure drop plenum-to pnum will also vary accordingly.

The Staff in its Supplementary Testimony and in its Concluding Statement revealed an intention of requiring use of realistic two-phase friction correlations. Review of the adequacy, realism, and experimental basis for competing possibilities has led the Staff to the conclusions that either the modified Baroczy correlation should be used, or the Martinelli Nelson correlation should be used below 250 psi, and the Thom correlation above 250 psi. These choices are expected to provide realistic estimates of the two-phase friction. The Staff states that correlations that overpredict the friction (e.g. Martinelli Nelson above 250 psi) are not obviously conservative in all applications.and so realism should be the guiding pnneiple in this case. The Staff points out (Supplementary Testimony > that the Armand correlation has not been l adequately tested by comparison with the two-phase multiplier versus quality data used in other cases.

GE argues in its Concluding Statement and in its Responsive Closing Statement that the Staff has not made its case that the Martinelli Nelson correlation may not be used throughout for BWR calculations,in the GE Concluding Statement it is said that use of Martmelli Nelson provides good fits to blowdown data.

We have noted above the inadequacy of this argument. GE states that the use of channel walls to confine the flow limits possible flow redistnbution in a BWR; thus an argument for realistic correlations based on a need to calculate flow redistribution accurately would not apply to a BWR. We do not altogether agree.

Howevee, this point was not well explored dunng the hearing, and in fact the discussion of flow reversal given later does propose that flow redistnbution be considered only in connection with pressurized water reactors.

The record sustains the Staffs position relative to unacceptability of the Armand correlation at this time. The Staffs position relative to acceptability of the Martinelli Nelson correlation throughout the range for BWR's is not clearly based on the record, in arriving at a rule based on the record we must note this defect. We conclude that unacceptability of the Martinelli Nelson two-phase friction correlation for use m BWR accident analysis cannot be part of a rule based on the record of this hearing. However,it would be appropriate to have a rule that required use of reahstic two-phase multiphers or other multipliers that led to l results at least equally conservative. We have therefore rewritten the staffs proposed requirement as to two phase friction multipliers to permit GE to use the Martinelli Nelson correlation throughout if this choice proves as conservative as the realistic correlations. Any use of Martinelli Nelson this way would have 1115

l to be supported by sufficient parcilel calculations using an acceptable realistic correlation to show that the latter choice will not lead to higher clad temperature dunng the hypothetical accident.

4. Discussion of Critical Heat Flux (CHF) i The entical heat flux would be impor' tant dunng blowdown because it marks the po'nt at which a transition would be made from the more efficient heat transfer from fuel to clad produced by nucleate j boiling, to a less efficient heat transfer regime. The change of heat transfer mode is called departure from l

nucleate boiling (DNB).

l The values of the critical heat Oux used in practice have usually been determmed in expenments with j slow transients, such as a slow rate of merease of heat dux, or small step-wise increases in heat flux. The transients during a postulated loss.of coolant accident from a large pipe rupture would be much more rapid. .

For a PWR subjected to an assumed mstantaneous rupture of a cold leg pipe, Gow reversal in the core is l calculated to occur within about 0.1 seconds of the break. At approximately this time, for any pomt in the core, flow stagnation would exist. Just prior to flow reversal the conditions in the coolant would be such

(

the DNB would have taken place if the transient were slow.The computational models accepted according to the Interim Pohey Statement have assumed that DNB would take place at this time,and afterwards the heat would be transferred through stable film boiling or a transition boilmg which is intermediate between <

nucleate boihng and stable film boihng.

For a BWR subjected to an assumed instantaneous break of a recirculation pipe, conditions appropriate i to DNB would occur much later, from 5 to 10 seconds mto the transient, and would accompany reduced core flow as the jet pump nozzles are uncovered.

CHF is predicted in Intenm Acceptance Critena models through use of a number of correlations.Most are based on expenments conducted under quasi. steady state conditions and these are called steady state f

i correlations. GE has used "i correlation based on transient data, that leads to a bnef delay in onset of DNB.

The heanng explored the use of steady state heat transfer correlations for DNB under transient conditions. The evidence was uniformly that under rapid transients DNB if anything would be delayed, and so actual heat removal would be better than calculated assuming steady state correlations.

Combustion Engineenng's Redirect and Rebuttal Testimony cited numerous experiments showing this, though there is some confusion from grouping together results of experiments with different kinds of transients (flow decrease, pressure decrease). But this and other evidence considered during the hearing lead us to agree with the position stated in the Staffs Supplementary Testimony, that the steady state correlations can be used to predict the CHF realistically in slower, quasi-steady state blowdowns and to predict the CHF conservatively in fast blowdown transients.

Also in its Supplementary Testimony, the Staff suggested that the steady state CHF correlations should be modified in use to reDect statistical uncertainty. The suggestion contained a technical Daw, but the intent was clear. In its Concluding Statement, the Staff withdrew this suggestion and appears content now to accept realism in this part of the calculation m place of conservatism.

The PWR vendors have all used steady state CHF correlations. GE has used its own correlation based on transient data. GE has reported sensitivity studies that show the results to depend only weakly on the use of a transient model.

• In its concluding statement the Staff proposed to accept use of severa. steady state CHF correlations where comparisons with data show that the correlations predict values of the CHF that allow for uncertainty in the experimental data throughout the range of parameters for which the correlations are to be used. Acceptance of the GE correlation was also proposed on this basis.

We believe that the record upholds the conservatism of this approach.

Westinghouse has proposed to add the condition that DNB should be assumed to occur in reality only if conditions appropriate to DNB persist for a period greater than 50 ms. Although the record includes reports of laboratory studies showing that DNB does not occur during rapid flow reversal, we believe that this observation is not well enough substantiated under the conditions that would prevail during a reactor blowdown to be accepted in LOCA computational models. We agree with the Staffs omission of the Westinghouse assumption from its approved evaluation models.

On the same basis, we agree with the omission of the similar proposal by GE, to the effect that DNB be neglected if conditions appropriate to it persist less than 3 ms.

1116

T We have again accepted the careful wording used by the Regulatory Staff in their Concluding Statement to establish the requirements as to calculation of critical heat flux.

5. Discussior.of Post CHF Heat Transfer Correlations:

The rate at which heat is transferred from the clad to the water after departure from nucleate boiling (DNB) is vital to estimation of the course of a hypothetical loss-of coolant accident for a PWR. DNB is calculated to occur within about a tenth of a second after a postulated instantaneous double ended break of a large pipe, or a large split. The heat transfer after this time would primarily determme the temperature history of the clad during blowdown and the possib lity that clad damage would occur during this phase. It would also determine the effectiveness of removal of heat from the oxide fuelitself and thus the stored energy in the fuel at the time refill of the plenum by ECCS fluid starts.

The lack of now reversal in the design basis accident to a BWR leads to a very different situation as regards DNB and post-CHF heat transfer. DNB would not occur until about 5 to 10 seconds into the blowdown, when the jet pump intakes were uncovered by the blowdown. The heat transfer regime subsequent to this would then be a principal determmant of heat removal from the clad and the oxide fuel later in the blowdown. By this time. however, the heat generation rate from fission and fission products would have decayed further.

Intenm Policy Statement 3/odels for PWR 's: The Interim Policy Statement has required the use of heat transfer coefficients for PWR's calculated using stable film boiling correlations after DNB. Exceptions have been defined. Westinghouse has been perrrutted use of a proprietary transition boiling correlation,which is meant to bridge the region of time between nucleate boiling and stable film boiling,and the other vendors and ANC have similarly used the transition boiling correlation of McDonough, Milich, and King for the same purpose,. The transition period affected by this bridgmg typically occupies a small fraction o* a second after DNB, and its use instead of immediate use of a coefficient appropriate to stable film boiling has little effect on the maximum calculated clad temperature. The use of a transition boiling coefficient after DNB mostly serves to provide a smooth transition from nucleate to film boiling, and to prevent developing mathematical oscillations in the results of computer code calculations.

The practice as regards correlations for film boiling has varied from case to case. The Staff has noted in l its direct testimony that use of the Groeneveld correlation is the most conservative for this purpose. The Westinghouse practice has been to continue use of the transition boiling correlation during the period when the quality is above zero. The first part of this correlation is a nucleate boiling term which is unimportant after the first brief interval, and the remainder of the correlation leads heat transfer close to that predicted by the Groeneveld correlation. The THETA code used for hot channel analysis by ANC and D&W uses the Groeneveld correlation, as does the RELAP blowdown code used by ANC. The STRIK!N and CEFLASH codes used by Combustion Engmeetint incorporate the Dougall Rohsenow film boiling correlation. The CRAFT code used by B&W employs a modified Dougall Rohsenow correlation, which includes a factor (T . /Tei 4)% suggested by McEligot to correct for cases of high clad superheat.

All of these but the Westinghouse correlation were derived from steady state data. The Westinghouse correlation was derived from both steady state and transient data.

The validity of these practices followed according to the interim Policy Statement was reviewed at i length during the hearing, and a number of points bearing on the acceptability of the various correlations were considered.'Ihese are discussed below.

Interen Policy Statement Afodel for BWR's GE has been required to use the Groeneveld correlation after DNB, as long as the quahty of the fluid is between zero and unity. The extension of the requirement through the period oflower plenum flashing is discussed at greater length below.

Blowdown Heat Transfer The transition from nuc!eate boiling to stable film boihng would diminish the heat transfer from clad to fluid by about two orders of magnitude. Evidence was presented that in fact the l

heat transfer would not be reduced this much, because the existence of bulk nucleation during j depressurization would cause the coefficient to remain near the high value appropriate to nucleate boiling. j This evidence was mostly provided by B&W who have carried out an extensive research program at their Alliance. Ohio facility on blowdown heat transfer. The B&W reports indicate that heat transfer would be dominated by bulk nucleation until the channel is almost empty of liquid. These experiments were l performed with a fluid velocity several timer, that appropriate to a LOCA, and with heat fluxes several times ,

l those that would exist at appropriate times dunng a LOCA. They were also puformed with rods six feet l t

l l 1117 i

L____-___________

long, about half the length used in water reactors. Some questions were therefore raised by ANC and the Staff as to the direct appbcabihty of the results to analysis for a PWR.

Transient Heat Transfer Some criticism was expressed as to the use of steady state correlanons during the fast transients analyzed for a large LOCA. These views were stated by Lawson (Transcript, pp. 5766 7).

Ybarrondo (Transcript, pp. 6069, 6272.10282.10890,10906 7). and Brockett tpp. 7480, 7588). The tenor of the criticism was that evidence was not conclusive that steady state correlanons overpredicted the transient coefficients or predicted them accurately.

Considerable evidence was provided nonetheless to the effect that during depressurization the use of steady state correlations for stable film bonhng was a conservative course. The strongest evidence was the BG work cited above. Additional work m the U. S. was referenced by B&W as mdicating that transient heat transfer dunng blowdown will be better than predicted using the steady state correlations; this included work at Columbia Umversity and at the Bettis Atomic Power Laboratory. Other experiments referenced were performed at MAN in Germany and in the Soviet Union. In our view the eviaence is near overwhelming that the use of steady state correlations for stable film boihng after CHF will provide a conservative estimate of heat transfer dunng blowdown.

Rewetting and Hysteresis-like Effects: It was stated by the Staff that Westinghouse used its transient heat transfer correlations throughout calculations of blowdown for safety analysis of the Point Beach reactor, This extended use of a transition boiling correlation was questioned by the Staff, because it implies that rewetting occurs as soon as the Guid conditions agam become appropriate for racleate boiling after DNB has once occurred. This is equivalent to assummg instantaneous rewetting af er the superheat has fallen below the Leidenfrost value.

The issue is posed in a somewhat different way in that both GE and Westinghouse have proposed the use of switchmg critena foi Critical Heat Flux. % westinghouse has proposed that if conditions providmg DNB persist for less than 50 ms, it should be assumed that nucleate boihng continues unimpaired. This period of interest for press ~urized water reactors would be that attached to flow reversal almost instantly after the break. The Westinghouse position would then be essentially that the full transition boiling correlation with its nucleate boihng term should be used throughout most of the blowdown phase of the transient, in this form, the Westinghouse argument is that rewet does not need to be considered because there has not been enough time for dryout.

GE's proposed switching criterion would lead to continued use of nucleate boiling heat transfer coefficients if the conditions appropriate to DNB persist for less than 3 ms. The practical result would be 3

use of a heat transfer coefficient calculated from the GE nucleate boilmg correlation during essentially the entire period oflower plenum flashing of a blowdown of a BWR.

We note the inconsistency of vendor positions that would rely on hysteresis-like effects as the basis for switching criteria such as those above, as compared to other positions calkng for instantaneous rewet.We I note also the inconsistency of Staff positions to the contrary in both cases.The point remains that there is I not adequate understanding of either rewet after CHF or of hysteresis-like effects during flow reversalin a fast transient. The Staffs position has been to approve use of only stable film boiling once DNB has been calculated to occur, even when fluid and clad temperature conditions appropriate to rewet exist. This j course is conservative. No less conservative position is justified by the record of the heanng. We concur with J the Staffs proposal tha't the nucleate boiling term of the Westinghouse correlations not be used after DNB

{

is calculated to occur, and that other models incorporate the equivalent assumption of stable film boihng ]

throughout the period after DNB.

The Groeneveld Correlation: The Interim Policy Statement approved use of the Groeneveld correlation as conservative for calcuisting stsble film boiling coefficients. The criticism of this choice on the grounds of its derivation from steady state data has been discussed above. We have coricluded that the use of a steady state correlation for stable film boiling is conservative in this case, even taking into account the use of realistic rather than bounding correlations in this application.

A difficulty has ariaen as to which Groeneveld correlation should be used. The Interim Policy Statement and the models accepted in this connection simply identify "the Groeneveld correlation". This situation persists through the Staffs Testimony, the Staffs Supplementary Testimony, and in fact through most of ,

the record up till near the end. Under GE questioning, Mattson agreed that in all the foregoing the l i correlation referred to was that stated in equation 5.7 of D. C. Groeneveld, "An Investigation of Heat l l Transfer in the Liquid Deficient Regime", AECL 3281, revised December 1969 (Transcript p. 20696). l Mattson also sgreed that the analysis of the acceptability of the Groeneveld correlation that Slaughterbeck i

1118

T l

had conducted at ANC had used the expression Groeneveld had stated in Equation 5.9 of the same paper, l rather than that coeff;cient developed by Groeneveld in Equation 5.7. He also said that heat transfer l coefficients measured for rods and annuli should fall below those appropriate to rod bundles because of  !

cross-flow effects. He said that this situation has been regularly seen in correlations tested in recent years (Transcript, p.20563 5). Groeneveld's statement of the same point was quoted from his paper. This point was continued by the observation that the Groeneveld 5.7 correlation, derived using only single rod data, underestimates heat transfer in bundles that have been tested (20709).

The Staff's concluding statement proposes acceptance of the Groeneveld 5.9 correlation, but not the l Groeneveld 5.7. This position is not upheld by the record of the hearing,and we agree in general with the argument made in this connection by GE in their Responsive Statement. If there is any position maintained throughout the hearing on the acceptability of a Groeneveld correlation, it is that the Groeneveld 5.7 correlation is acceptable. The rule must be in accord with this point for both PWR's and BWR's,and any departure from this choice must be considered in later reviews whether rule-making or licensing.

Heat Transfer at Low Pressure: It was observed a number of times that no non proprietary data have been provided for heat transfer at pressures below 500 psi. Data developed at Columbia University under Combustion Engineering sponsorship were provided to the Staff and the Hearing Board by CE, on a proprietary basis. The Staff's position in its Concluding Statement was based in part on the existence of these data and their implications. GE Ms objected in their Responsive Concluding Statement to a position on GE reactors based on propnetary data provided by other vendors, as prejudicial to them. Ris point is considered from other aspects m another section of this opinion. (see pp.1088-1089 supra).

We note tnat CE data effectively fill the void in data below 500 psi used by Groeneveldin deriving his correlation.

Doagall-Rohsenow Correlations: Combustion Engineenng has used the Dougall Rohsenow correlation for calculating heat transfer coefficients in the stable film boiling regime. Babcock and Wilcox has used the modified Dougall Rohsenow correlation, which differs from the D R through the use of McEligot's overall factor of (Tsa t/Tei,d)4. In regions of interest at low pressure, McEligot's factor is significantly below unity.

Bus, the heat transfer predicted by the D-R correlation is better than that of the modified D-R by an appreciable factor. CE has provided data and analysis showing that the D.R correlation leads to heat transfer coefficients for stable film boiling that are lower bounds to the range of values observed. CE believes that their proprietary data justify use of the D R correlation after CHF.

The Staff position has been to accept use of the modified D-R correk. tion, but not the D R itself.

Presumably this has been done because the modified D-R correlation goes over in a smooth fashion to the Dittus Boelter correlation for single phase steam at the limit of unit quality (pure steam).

In their Response to the Staff's Concluding Statement, Combustion Engineering objects to elimination of the D R correlation from the proposed Staff rule. On this point we believe the Staff was wrong. The evidence supplied by Combustion Engineering supports use of the Dougall-Rohsenow correlation. Of course,if the Dougall Rohsenow is acceptable, so is the modified Dougall Rohsenow. I

6. Discussion of Pump Modeling For a postulated cold leg break in a PWR the flow through the core would be reversed during the major part of the blowdown. This reverse flow would be opposed by the pumps in the unbroken loops, so that the calculated flow magnitude and the resulting coolirg would depend upon the model used for the pumps. In some evaluation models it was assumed that the pump delivered zero head as soon as the suction pressure was reduced to saturation pressure. The limited data available indicated, however, that pump operation would continue into the two phase flow region. (Exh.132) Continued pump operation in the PWR intact loops would diminish the core flow during the early part of blowdown and,in some instances, change the calculated peak Zircaloy temperature appreciably, sometimes in one direction,and sometirnes in the other.

(Exh. I 13, p. 6 4).

During the hearings pump modeling was criticized on the basis that pumping action might continue into the two phase region when it was assumed not to (Tran:cript pp. 631112 and 7477), and for the lack of applicable two phase performance data, at least in the public domain (Transcript p. 5643). The new rule addressesitself to these criticisms.

For the BWR's, the core flow would not reverse as a result of the hypothetical accident. The recirculation line pumps would continue to drive the jet pumps and their continued operation would tend 1119

to prolong core flow and provide longer initial cooling of the core. GE's present raodel assumes that the pump head could start to decrease as soon as saturation pressure is reached at the pump suction, going linearly to zero as the quahty at the pump suction goes to one percent. In analyses reviewed by Regulatory to date, the jet pumps would become uncovered and the core now fall to zero before the one percent quality occurs at the pump suction (Exh.1113, p. 6 2). This situation is recognized in the rule.

In their responses to the Concluding Statement of the Regulatory Staff, Combustion Eng neering and Westinghouse stated or implied that applicable data do not yet exist for two-phase pump performance.We encourage obtaining such data. There is no disagreement with regard to the use of a dynamic pump model.

7. Discussion of Core Flow Distribution During Blowdown.

The analytical models used in reviewing the course of a hypotheticalloss of coolant accident under the Interim Policy Statement have all been one dimensional, with no direct treatment of How redistribution in the core. The detailed flow in the reactor core following initiation of the hypothetical loss-of coolant would be complex, and would be different dependmg on the kind of reactor and fuel and the specific time during the LOCA.The nature and degree of flow redistribution were discussed at length during the hearing.

Flow redistribution between channels is a phenomenon primarily affecting Pressurized Water Reactors, because they have no channel walls to restrict cross-flow. The principal forces affecting cross-Gow are friction, acceleration, drag in channels, drag through spacer grids and fittings, and buoyancy (Morgan, Transcript pp.12678-9). The presence of a two-phase fluid affects buoyancy and frictional drag (through the two phase multipliers). In up0ow the buoyancy effects tend to produce higher hot channel flow than average channel flow (though frictional effects act in the opposite direction). In downflow the buoyancy and the friction act together to reduce hot channel flow relative to average channel flow (Morgan.

Transcript 12679).

The Interim Acceptance Criteria models have accounted for these effects by a requirement that the average channel flow during blowdown of a PWR be multiplied by a factor 0.8 to obtain the Dow in the hot channel calculation. Westinghouse Testimony stated that calculations made using the THINC code are the basis for this choice of factor. The calculation assumed parallel channels, and zero cross-flow resistance between channels, so that the pressure was constant in every horizontal plane. These calculations did not unambiguously lead to flow redvetions bounded by the factor 0.8.

Although the vendors' discussions of the effect of flow redistribution generally tended to support choice of the factor 0.8, there was little sympathy elsewhere for it. Its conservatism was questioned by Rosen (Testimony), Lawson (Transcript 5755), and Ybarrondo (Transcript 6076. 6270,10255), and its continued use has not been proposed by the Staff.

It appears that considerations of flow redistribution prior to the hearing comprised only circumstances in which the clad is not deformed. Questioning during the hearing also dealt with effects of clad swelling, fuel deformation, and partial blockage on flow redistribution. It was less apparent that the factor 0.8 would be adequate if partial channel blockage occurred than if channels were undeformed.

The Staff's Supplementary Testimony recognized this point, and proposed that models be developed and used that explicitly calculate the effect of flow redistribution during both the upflow and downflow phases of blowdown. The view included use of models that calculate the flow redistribution resulting from floD blockage if that should be calculated to take place. i We believe this is the correct course to follow. We believe the wording in the Staff's proposed rule l

adequately expresses the position supported by the record, with one exception. There is no basis in the record for continued use ~of the ilow reduction factor of 0.8 after now redistributloa effects have been calculated for the hot channel We have not included this requirement in th : ".ile.

i D. POST-BLOWDOWN PHENOMENA; HEAT REMOVAL BY THE ECCS l 1. Single Faflure Criterion. An analysis of possible failure modes of ECCS equipment and of their l efhets on ECCS performance rnust be made in carrying out the accident evaluation the combination of ECCS subsystems assumed to be operative shall be those available after the most damaging single failure of l

l ECCS equipment has taken place.

2. Containment Pressure. The containment pressure used for evaluating cooling effectiveness during reflood and spray cooling shall not exceed a pressure calculated conservatively for this purpose. The calculation shallinclude the effects of operation of allinstalled pressure reducing systems and processes.

1120 l

3. Calculation of Repood. Rate for Pressurized Water Reactors. The refilling of the reactor vessel and the time and rate of refloodmg of the core shall be calculated by an acceptable model that takes into consideration the thermal and hydraulic characteristics of the core and of the reactor system.The primary system coolant pumps shall be assumed to have locked impellers if this assumption leads to the maximum cMeulated cladding temperature;otherwise the pump rotor shall be assumed to be running free. The ratio of the total fluid now at the core exit plane to the total liquid flow at the core inlet plane (carryover fraction) shall be used to deternune the core exit now and shall be determined in accordance with applicable experimental data (for example, "PWR FLECHT (Full length mergency Cooling Heat Transfer) Final Report," Westinghouse Report WCAP 7665, April 1971; "PWR Full length Emergency Cooling Heat Transfer (FLECHT) Group i Test Report," Westinghouse Report WCAP 7435, January 1970; "PWR FLECHT (Full 12ngth Emergency Cooling Heat Transfer) Group 11 Test Report," Westinghouse Report WC AP 7544, September 1970; "PWR . FLECHT Final Report Supplement," Westinghouse Report WCAP 7931, October 1972).

The effects on reflooding rate of the compressed gas in the accumulator which is discharged following accumulator water discharge shall also be taken into account.

4. Steam Interaction with Emergency Core Cooling Water in Pressuri:ed Water Reactors. *lh'e thermal-hydraube interaction between steam and all emergency core cooling water shall be taken into account in calculating the core refloodmg rate. During refill and reflood. the calculated steam flow in unbroken reactor coolant pipes shall be taken to be zero during the time that accumulators are discharging water into those pipes unless experimental evidence is available regarding the realistic thermalhydraulic interaction*

between the steam and the liquid. In this case, the experimental data may be used to support an alternate assumption.

5.R.efilland Repood Heat Transfer for Pressuri:ed Water Reactors. For teflood rates of one inch per second or higher, reflood heat transfer coefficients shall be based on applicable experimental data for unblocked cores including FLECHT results ("PWR FLECHT (Full length Emergency Cooling Heat Transfer) Final Report," Westinghouse Report WCAP-7665, April 1971). The use of a correlation derived from FLECHT data shall be demonstrated to be conservative for the transient to which it is applied; presemly available FLECHT heat transfer correlations ("PWR Full Length Emergency Cooling Heat Transfer (FLECHT) Group I Test Report," Westinghouse Report WCAP-7544, September 1970; "PWR FLECHT Final Report Supplement," Westinghouse Report WCAP 7931. 0ctober 1972) are not acceptable.

New correlations or modifications to the FLECHT heat transfer correlations are acceptable only after they are demonstrated to be conservative, by comparison with FLECHT data, for a range of parameters consistent with the transient to which they are applied.

During refill and during renood when reflood rates are hss than one inch per second, heat transfer calculations shall be based on the assumption that cooling is only by steam, and shall take into account any flow blockage calculated to occur as a result of cladding swelling or rupture as such blockage might affect both local steam flow and heat transfer.

6. Convective Heat Transfer Coefficients for' Bciling Water Reactor FuelRods Under Spray Cooling.

Following the blowdown period, convective heat transfer shall be calculated using coefficients based on appropriate experimental data. For reactors with jet pumps and having fuel rods in a 7 x 7 fuel assembly array, the following convective coefficients are acceptable:

(a)During the period following lower plenum flashing but prior to the core spray reaching rated flow,a convective heat transfer coefficient of zero shall be applied to all fuel rods.

(b)During the period after core spray reaches rated now but prior to refloodmg, convective heat transfer coefficients of 3.0,3.5,1.5, and 1.5 Btu-hr-8 ft-2 *F' shall be applied to the fuel rods in the outer corners, outer row, next to outer row, and to those remaining in the interior, respectively, of the assembly.

(c) After the two-phase reflooding fluid reaches the level under consideration, a convective heat transfer coefficient of 25 Btu-hr-ft-8 *F shall 8

be applied to all fuel rods.

7. The Boiling Water Reactor ChannelBox Under Spray Coo!ing. Following the blowdown period, heat l transfer from, and wetting of, the channel box shall be based on appropriate experimental data. For reactors with jet pumps and fuel rods in a 7 x 7 fuel assembly array, the following heat transfer coefficients and wetting time correlation are acceptable.

(a) During the period after lower plenum flashing, but prior to core spray reaching rated flow, a convective coefficient of zero shall be applied to the fuel assembly channel box.

1121

l (b) During the period after core spray reaches rated flow, but pnor to wetting of the channel, a convective heat transfer coefficient of 5 Btu hr 8 ft-8 *F"' shall be applied to both sides of the channel box.

(c) Wetting of the channel box shall be assumed to occur 60 seconds after the time determmed using the correlation based on the Yamanouchi analysis (" Loss of Coolant Accident and Emergency Core Cooling Models for General Electric Boiling Water Reactors," General Electric Company Report NEDO.10329, April 1971).

DISCUSSION Containment Pressure. One of the anomalies of the LOCA is that, although one would normally think that it would be desirable te have a low containment pressure to reduce leakage, a high containment pressure would be advantageous for the operation of the ECCS. A high containment pressure after the accident would terminate the blowdown sooner and would improve the capability of the ECCS because convective heat transfer coefficients are higher at high pressures. For the PWR's, a high ambient pressure would reduce the steam binding (through the higher steam density) and would lead to high reflood rates, which would further improve the heat transfer coefficients. Thus a guide is needed for the containment pressure that can be used in the analysis of the hypothetical LOCA.

De Interim Policy Statement specified that the maximum containment pressure allowed for the calculation of the effectiveness of the ECCS should be the original pressure (presumably atmospheric) plus 80% of the pressure increase estimated to be brought on by blowdown. For containment sy:tems used so far for PWR's (dry containment) this prescription was shown to be conservative by the Regulatory Staff (Exhibit 1113, Sec.15). However it is not too difficult to calculate the actual pressure, allowing for the various cooling devices placed in the containment structure to limit its pressure, and the new requirement is to trake such calculations. Improving the generality of the requirement has made it applicable to other types of containment, such as ice condenser containment.

The importance of containment pressure has been less for BWR's than PWR's because steam binding is less of a problem and because the heat transfer coefficients available for use in BWR analysis were derived at atmospheric pressure and so are conservative for elevated pressures.

In their response to the Staff's Concluding Statement, Combustion Engineering states a preference for the old formula. As the new rule is stated the old formula may still be used, provided it is shown that the pressure so calculated continues to be less than that obtained by a detailed calculation.

Both Combustion Engineering and Westinghouse pointed out in their responses to tne Staff's C ncluding Statement that the LOCA analysis is made assuming some loss of power, under which condition some of the pressure reducing systems might not be operable. They therefore suggest that for consistency one should not assume that all of the pressure reducing systems would be operating. Since it is possible for the LOCA to occur with no loss of power, the Commission rejects this suggestion.

Calcukution of Reflood Rate for Pressurized lyater Reactors. The reflooding rate for pressurized water reactors would be controlled to a large extent by steam binding, the phenomenon by which the resistance ta flow through the reactor system (steam generators, pumps, etc.) of the effluent from the r? actor core limits the rate of reflood and, indirectly, the rate of heat removal from the fuel rods. The pumps in their locked rotor condition would typically provide more than half of this resistance to flow so that the stipulation of their being locked is a serious limitation. If the pump rotors were not locked, their resistance ta flow would be reduced by 60%(Exhibit i113, p 1410). In their Concluding Statement, Combustion Engineering states that if the pumps were free running during reflood the calculated maximum temperature of the zircaloy cladding would be reduced by 75'F (CE Concluding Statement, p 3-61).

{

he stipulation oflocked pumps during reflood is unchanged from the Interim Policy Statement, and '

no new experimental information was provided during the hearing justifying a change in this part of the

]

rule. '

he Regulatory Staff in their Concluding Statement proposed the development of more sophisticated l refill-reflood computer programs, including those capable of predicting the expected oscillatory flow of I water into the reactor core. The Staff also proposed that the calculation should consider the carry-over of )

fluid from the top of the reactor core to be based on experimental data, principally the FLECHT tests, which were carried out with fixed flow rates. Combustion Engineering, who have a sophisticated code, PERC, that predicts oscillations, pointed out a . difficulty in their Comments on the Staff's Concluding Statement, in that use of the experimental carry over data would make major portior s of their i 1122 l '

l

______ __ _____ ________ _ ____J

? Y sophisticated code superfluous. Since the heat transfer coefficients must come from the FLECHT tests,at least for now,it seems logical and consistent to use those data to determine the amount of fluid leaving the core and passing out through the system. The Commission believes with the Staff that improved and more realistic models are desirable, but realizes that the full benefit of sophisticated models that predict the oscillatory flow cannot be obtained until there are more suitable experiments with which they can be compared. The rule, as written, allows for the use of new data and of more sophisticated codes when they are available.

As several participants have observed, the method chosen for calculating the mass of fluid leaving the top of the reactor overestimates it, at least for the FLECHT tests from witich it is derived, by the amount of two phase fluid retamed in the core above the quench front. To the extent that this carry-over fraction is overesumated, the amount of steam binding is overestimated and the reflood rate is underestimated, providing an additional modest conservatism. The method that we have chosen for calculating the carry over fraction is, however, as reahstic as the FLECHT data allow.

Steam Interaction with Emergency Core Cooling Water in PWR 's. The steam flow coming f:om the core through the cold legs of Combustion Engmeering and Westinghouse PWR's would be subject to increased resistance as a result of the high accumulator-injection flow rates (thousands of gallons per minute)into those legs. The restriction of no steam flow in the non-ruptured legs during ECC injection is conservative;it was part of the Interim Policy Statement. Combustion Engineering has already proposed a new prediction model based on experimental evidence (Concluding Statement, p 3-64; Exhibit i144, pp 810 to 818),and Westinghouse also proposed submitting a new model(Concluding Statement, p 80).

Reflood Heat Transfer for Pressuri:ed Water Reactors. The convective heat transfer coefficients used for calculating the cooling of the fuel rods during the reflood phase are derived from the PWR FLECHT test program (Exhibit 150). In these tests electrically heated rods, simulating fulllength fuel rods, were cooled by bottom flooding at various rates. Characteristically, after the very bottom of the heated section of the rods had been wetted, the rapid generation of steam led to an entrainment of water droplets and the generation of a two phase fluid that swept upward through the rod bundle. The mechanisms of heat transfer to this two phase fluid have been postulated to be (1) convective cooling to the steam,(2) droplet impingement on the rods and (3) radiation to the water droplets (and to the steam). Along at least part of the length of the rods the two phases would not be in thermal equilibrium,i.e., the steam would be at a higher temperature than the water droplets, and therefore the water droplets would S heated by the steam and evaporate as they pass up the column (Exhibit 150, pp 3-69-3 77). This complicated heat transfer mechanism was interpreted,by calculating heat transfer coefficients, which, when multiplied by the difference between the measured surface temperature of the rods and the saturation temperature of water at the test pressure, yielded the measured heat flux from the rods to the coolant.

The FLECHT tests and their calculated heat transfer coefficients have been criticized on several bases (Transcript, pp 6868,19,489; Exhibit 1041, Sec. 6). The principalitems questioned were: the effect of the flow housing, the use of stainless steel for the cladding of the heater rods instead of zircaloy, the use of steady flow instead of allowing the flow to oscillate as it would in a reactor,and the probability of errors.

Each of these will be discussed in turn.

The FLECHT ' tests were carried out with bundles of either 49 or 100 rods, arranged in a square array and surrounded by a steel housing about % inch thick. The peak ternperature of the housing was typically 750'F, even when the rods next to it were at temperatures in the vicinity of 2000*F. The concern was expressed that this housing did not suitably simulate the surrounding rows of fuel rods that would be present in the reactor. The Commission believes that this question has been adequately explored by experiment and examined in the record (Exhibit i113, pp 17 2; Exhibit 1078, pp 46 59), and concludes that the heat transfer measured for the inside rods of the bundles was not affected in any major way by the housing. The effect of radiation to the housing on the calculated heat transfer coefficients was estimated to l be less than 5% by Aerojet Nuclear Corporation (Exhibit 1113, p 17 3) and about three percent by I

Westinghouse (Exhibit 107, pp 46 52).

Stainless steel was used instead of zircaloy as the cladding material for nearly all of the FLECHT tests because it is more durable under the test conditions. Although it is not usual to e;tpect significant differences in convective heat transfer coefficients from different solid material surfaces, the possibility of such differences was considered, perhaps resulting from such factors as differences in thermal conductivity and differences in wetting properties. The reasonable conclusion was reached that the effect of the difference between zircaloy and stainless steel,if any. would be small. There is a difference, of course,in r

I 1123 l

T the rate of heat generation from steam oxidation, but this heat is deposited withm the metal under the surface of the oxide film. The presence of this heat source should not affect the heat transfer coefficients, which depend on conditions in the coolant outside the rod.

The few FLECHT runs made with zircaloy clad rods provide uncertain and conflicting evidence.

Westinghouse pointed out that all of the zircaloy runs except one (run 9573) yield higher heat transfer coefficients than were obtained with steel (Westinghouse Concluding Statement, pp C 74 to C 76; Exhibit 150, pp 3 98 & ff). Consolidated National interveners pointed out that most of these runs were made at unreasonably high flooding rates, and that a different result was obtained from run 9573 where the flooding rate was about one inch per second. In the first 18 seconds of tius run, before multiple heater rod failures occurred, the zircaloy clad rods heated up faster than predicted from the stainless steel based correlations (Exhibit 1041, pp 6.7 & ff). This anomalous result has been attributed to experimental error, or pessibly to an unusually skewed initial temperature distribution along the length of the rod (Exhibit 1113, pp 17 6-17 7).

On balance, the Commission sees no basis for concluding that the heat transfer mechanism is different for zircaloy and stainless steel, and believes that the heat transfer correlations derived from stainless steel clad heater rods are suitable for use with zircaloy clad fuel rods. It is apparent, however, that more experiments with zircaloy cladding are needed to overcome the impression left from run 9573.

At a number of places in the record mention is made of the oscillatory flow that would be expected in reflooding a reactor (Exhibit 1113, Sec.17, pp 1, 2,11,12,13; Exhibit 1144, Sec. 9)and of an enhanced heat transfer to be expected from it (Trans.p 6838). In the PWR FLECHT tests presently available, the water entered the bottom of the rod bundle at a predetermined rate, without provision for the development of an oscillatory flow. Westinghouse is carrying out another group of tests called FLECHT SET with a hydraulic system that more closely simulates a reactor, which is expected to allow oscillatory flow to take place. He results of one run, run 5, are discussed in some of the above references. This run did exhibit the oscillatory flow. The initial reflood rate, averaged over the oscillations, was quite high, but settled down to a value of about two inches per second within 20 seconds. After allowance is made for the effects of the high, initial flow, the heat transfer coefficient observed is very similar to that which would have been predicted from the prior FLECHT tests. (Exhibit 1113, p 1712). Thus, until more experimental evidence is available, heat transfer coefficients should continue to be based on the FLECHT tests,and caution should be exercised in asserting the existence of a conservatism because they are based on steady flow.

The accuracy of the FLECHT-determined heat transfer coefficients has been examined several times.

(Cf. the review in the Babcock and Wilcox Concluding Statement, pp 202 204.) Westinghouse estimated a possible uncertainty of 12% in the coefficients. (Trans. page 6878). The Aerojet Nuclear Company concluded "that the FLECHT data currently represent a best estimate of the heat transfer that will occur in a large undistorted core." They also concluded that an allowance of up to 20% may be needed "to bound the data due to experimental and inferential errors."(Exhibe !113, p 1714) The Commission approves of the use of the FLECHT data for calculating PWR re%od heat tunsfer, but notes that these will be more nearly "best estimate" calculations than bounding cal:ulations.

The PWR FLECHT Final Report Supplement, WlAP 7931, gives revised formulae for the calculation of heat transfer coef(icients, and,in a series of curves. compares both the old and new calculations with the experiments. These curves indicate that the calcu:ations, both old and new, predict greater heat transfer than would actually occur in the early part of tne reflood transients for low reflood rates.(Exhibit 1113, p 1714). As a result,we require that new coefficients be used in this region, together with a demonstration that they represent the data in a conservative manner.

The FLECHT tests simulated flow blockage in a number of runs by the insertion of perforated horizontal plateeMith reflood rates of one inch per second or higher, improvement was found in the rate of heat transfer as far as two feet upstream and four feet downstream of the blockage.The improved heat transfer was shown to be caused by break-up of the entrained droplets and increased turbulence.(Exhibit 1006s). The blockage in these tests rarged up to complete blockage over several channels with 75%

l blockage in other channels. For the flow blockage tests at a reflood rate of 0.6 inches per second, heat transfer was degraded by blockage. Presumably the poor results at the low reflood rate were the result of a lack of entrained water droplets, leaving only single phase steam cooling. (Exhibit i113, p 17 5).

The FLECIR flow blockage tests were criticized on the basis that the flat plates were not typical of bulging of the cladding. However, Davis tried blockage with sleeves versus plates and found little difference.

(Trans. p 4130).

1124

V T

l l

l

' As a result of these tests it appears that heat transfer coefficients based on undistorted rod geometry would provide a reasonable approach to estimating core temperature behavior during redood, for renood rates above one in./sec. For lower renood rates blockage would have a deleterious effect and one must resort to calculation with single phase steam cooling, taking into consideration the effects of blockage on core now distribution.

Convective Hear Transfer CoelTicients for BWR FuelRods UnderSpray Cooling. The time-history of a hypothetical design basis Boiling Water Reactor accident can be divided into six periods (Exhibit 1001, pp 4 7 & ff). A Dow coast-down pened would end when the water level fell below the jet pump inlet. Dere would then be a short period of flow stagnation in the core until more water escaped from the pressure vessel. When steam started to escape through the break there would be a rapid drop in the system pressure, causing the water in the lower plenum to Gash to steam.creatmg a two phase flow through the core. After the end of Cashing there is assumed to be a short " core heatup period" during which no core cooling would take place before the ECCS came into action. The ECCS would start first with a spray of water into the top of the core and later flood the core from the bottom.

Flooding the core is said to occur within three minutes of the piping break in the design basis accident, or within two and a half minutes after the start of the core spray. (Exhibit 1113, p 16-36). The reflood rate would be quite high, typically 3.7 inches per second (Exhibit 137, p 23) and would terminate the excursion quite rapidly once the renood water started entering the bottom of the core. (Exhibit 1069, p 15). Thus the function of the core spray is to keep the cladding temper;.ture from rising too rapidly during the approximately two and a half minutes between the core heatup period and the quenching action provided by the core reflood.

The BWR fuel. rods are in 7 x 7 arrays, with each array contained in its own channel box. With this arrangement each group of 49 fuel rods is largely independent of the rest of the core as far as c: , ling is concerned. The channel boxes would not become as hot as the fuel rods, since their only source of heat would be absorption of thermal radiation and gamma rays from neighboring fuel rods,and they would be more easily quenched by the core spray. They would serve as a convenient heat sink for radiation from the hotter fuel rods, especially those near the outside of the 7 x 7 array. There would be a diffusive now outward of heat from the inner rods, through radiation. The fuel rods would also lose some heat to the water droplets and associated steam by convection and radiation. Through these mechanisms the core spray would limit the rate at which the core heats up. (Exhibit 1069, p 15).

From the BWR FLECHT tests there is information on the heat transfer coefficients for both the convective heat now to the water droplets and steam and for the reflood phase. Le FLECHT tests were made with an electrically heated mock up of a 7 x 7 rod array complete with its channel box. The convective heat transfer coefficients were determined from the residue of a thermal balance after all of the l known inputs and outputs were calculated. The factors considered were the electrical heat input, the rate of change of the heat content of the rods as calculated from their temperature history, and the calculated radiation from the rods to each other and to the channel walls.The residue from these inputs and outputs was ascribed to convective heat transfer. The convective heat transfer coefficients so determined could not Se very accurate because their calculation involved taking the difference between two large numbers.The coefficients so obtained are small and are about what one would expect from the mechanisms of natural convection and radiation to steam. (Exhibit i 113, p 16-14). j The values of the calculated convective heat transfer coefficients depend to some extent upon the value  !

used for the thermal emissivity of the stainless steel. since the convective heat transfer is obtained after l subtracting the radiative heat transfer from the total. Theoretically a high value of the emissivity leads to a j low calculated convective heat transfer coefficient. Values of the emissivity measured after the tests ranged from 0.6 to 0.9 (Exhibit 461, p 81 and Exhibit 1113, p 16-14), and to add conservatism to the calculation, the Interim Policy Statement required the use of the highest measured emissivity,0.9, for the calculation of the convective heat transfer coefficients. However it turned out that this resulted in a higher coefficient ]

(less conr.ervative) for the criticalinner rods, with a higher estimated standard error. (Exhibit 461, Table 2.) l After reviewing the derivation of the coefficients as given in Exhibit 461, we believe that those originally l listed as best estimates by General Electric are the most credible and should be used. The effect of this change on the peak cladding temperature will be small, about five degrees according to Exhibit 461.  ;

There has been a great deal of criticism of the BWR FLECHT tests, particularly by the Consolidated l National Interveners (Exhibit 1041, Chapter 5), and both General Electric and the Regulatory Staff have l l defended them (Closing Statements). However, for the purpose of calculating the maximum cladding j j

1125 l I

I temperature, only the denved heat transfez coefficients are of any great importance. The ralues obtained haec always been known to have a high statistical error; furthermore the values are low and reasonable, and there seems httle to be gained by renewing the controversy over the manner of conducting and interpreting all features of the tests.

Et high but inevitable statistical error of the coefficient for the inner rods (1.5 ; 1.0 BTU /hr fta ,op) is bothersome and leads to an estimated error band of as much as 200 F in the calculated peak temperature m some circumstances (Exhibit 1113, p 16 36). The test bundle SS2N was used to derive the heat transfer coefficients; another test bundle, SS4N, resultad in claddmg temperatures 200*F higher than those of ahe bundle used as a standard; one half of this discrepancy could be explained by test differences, with the other half left to be attributed to statistical variations. (Exhibit i113, p 16-38). The problem of these large statistical errors in the convective heat transfer coefficients is compensated to some extent by the fact that the coefficients were determined at atmospheric pressure, whereas the reactor would be at some elevated pressure at which the heat transfer would be improved (Exhibit 1113, p.16-26).

The evidence for the value 25 BTU /hr ft'. F of the two phase refloodmg heat transfer coefficient is

. sketchy (Exhibit 1032, p.II-6.3-51), but it is applied for only a short time because the high reflood rate would quickly quench the core, and the exact value is of little significance.

The BWR Channel Box Under Spray Cooling. Radiation to the channel box would be an important mechanism for cooling the outer fuel rods of a 7 x 7 array, and the magnitude of the radiative cooling would depend to some extent upon the temperature of the channel box. Dunng the LOCA the channel box would be heated W radiation from the fuel rods and by absorption of gamma rays, and later it would be cooled by the core spray. Its temperature would drop sharply after it is wetted by the core sprav and subsequently it would become a better heat sink.

De time of wettmg has been calculated by an extension of a theory developed by Yamanouchi (Yamanouchi A., Journal of Nuclear Science and Technolog, 5, pp.547 558, Nov.1968). This theory calculates the progress of the wetting front for spray cooling by considering the longitudmal thermal conductance of the channel wall. Although the direct application of this theory over predicted the wetting times observed in the FLECHT tests,it was possible to correlate the data from the stainless steel runs with a group of channel parameters that were involved in the theory. The data for the runs with zircaloy cladding had more scatter, with both positive and negative deviations from the stainless steel correlation line. By adding one minute to the wetting times predicted by the correlation, all but one of the quench times observed in the FLECHT tests were encompassed. (Exhibit 461, p.6) Modified in this way, the calculated  ;

quench times seem to be adequately conservative.

With regard to the convective heat transfer coefficient to be used during core spray but prior to wetting, General Electric's present practice is to use a convective heat transfer coefficient of 20 BTU /hr fta ,op applied to only one side of the e hannel box. This number was derived from the FLECHT experiments as a best fit to the data. (Exhibit 461, p.8) The Regulatory Staff points out that the geometry in the FLECiff experiments differed from that in a reactor,in that there was an insulated outer channel surrounding the l channel box in the experiment. They calculated that radiation from the channel box to the outer channel l rnay have contributed up to half of the heat transferred from the channel box in the FLECHT experiments.

(Exhibit 1113, p.16-8) They therefore recommend reducing the convective heat transfer coefficient by a factor or two. Although General Electric objected (Responsive Ciosing Statement,Vol. 2, ppt 80-81) on the basis that the calculation is already over. conservative by virtue of the minute added to the channel wetting time, the Commission supports the position of the Regulatory Staff on the basis that no single cooling mechanism should be counted upon to exceed its expected performance. l IV. REQUIRED DOCUMENTATION 1.(a) A description of each evaluation model shall be furnished. The description shall be sufficiently complete to permit technical review of the analytical approach including the equations used, their approximations in difference form, the assumptions made, and the values of all parameters or the procedure for their selection, as for example,in accordance with a specified physicallaw or empirical correlation.

(b)The description shall be sufficiently detailed and specific to require significant changes in the evaluation model to be specified in amendments of the description. For this purpose, a significant change is a change that would result in a calculated fue! cladding temperature different by more than 20*F from the 1126

temperature calculated (as a function of time) for a postulated LOCA using the last previously accepted model.

(c) A complete listing of each computer program,in the same form as used in the evaluatica model, shall be furnished to the Atomic Energy Commission.

2.For each computer program, solution convergence shall be demonstrated by studies of system modeling or noding and calculational time steps.

3. Appropriate sensitivity studies shall be performed for each evaluation model, to evaluate the effect on the calculated results of variations in noding, phenomena assumed in the calculation to predominate, including pump operation or locking, and values of parameters over their applicable ranges. For items to which results are shown to be sensitive, the choices made shall be justified.

4.To the extent practicable, predictions of the evaluation model, or portions thereof, shall be compared with applicable experimental information.

5. General Standards for Acceptability-Elements of evaluation models reviewed willinclude technical adequacy of the calculational methods, including compliance with required features of Section I of Appendix K to 10 CFR Part 50 and provision of a level of safety and margin of conservatism comparable to other acceptable evaluation models, taking into account significant differences in the reactors to which they apply.

DISCUSSION:

Previous Experience with the Interim Policy Statement has shown that additional documentation would be useful,(Exhibit 1031, page 2; CN! Concludmg Statement, page 4.16; and Transcript pages 5643;6675; 6691; 10,879; to 10,883).

Conskierable hearing time was devoted to consideration of the adequacy of codes and analysis methods (see Exhibit 1043 and Transcript pages 8294; 8386; 11,06511,112; 11,156). Time would be saved in the hearing process,in generic reviews, and in case reviews, if for each evaluation model a detailed description were provided which defined the analytical approach and equations, the assumptions, the references, the selection and justification for the input parameters, and the mathematical symbolism used to establish the corresponding computer programs. A complete description and listing of the computer programs, in identical form to those approved and being used (at a specific time) for LOCA analyses,is needed by the Regulatory Staff so that they can be certain that the codes used for safety analyses always correspond to the approved, published evaluation model.

It is recognized that revisions in the evaluation models will be made from time to time, within the restrictions imposed by the section on re' quired and acceptable features. When these revisions constitute a "significant change" as defined in paragraph 1.(b) above, the changes must be described in detail and an updated revision of the computer codes provided.

The need for noding and sensitiuty studies for the computer programs is clearly reflected by the hearir}g record (e.g., Exhibits 1006, 1043, 1044, 1001, II13, 1148). This rule formalizes the scope and intent of such studies.

The need for comparisons of ti s calculations of analytical models with experimental data is discussed and the value is recognized in the written testimony of nearly all of the participants, including the Regulatory Staff (Exhibits 1001 & 1113). Westinghouse has stated the existence of some problems in interpreting the requirement for such comparisons. It is reasonable to restrict these comparisons to those that the proponents of evaluation rnodels have made of their own volition to check out certain features and to comparisons requested by the Regulatory Staff.

In their comments Babcock and Wilcox suggested omission of the technical review of the evaluation models. It is the Commission's opinion that, with the changes being made by this rule,it is necessary that a technical review of the evaluation models be made by the Commission; this review is the responsibility of the Regulatory Staff.

Both Westinghouse and General Electric objected that the subject of computer codes should not be part i

of this rule. As indicated in the soove references, the codes were discussed at length in the record in terms I of their adequacy and content; that there should be this much question is deemed sufficient reason to have the computer codes revealed to the Commission.

1127 l

T  !

CONCURRlNG OPINION OF COMMISSIONER ANDERS Though I join m the Commission's rulemaking decision issued today, a statement of th 9 basis for my concurrence may be helpful to the participants. As to the substantive rule itself.1 agree generally that the record leads to our conclusions. In my view, however, a weakness of the present record is that it does not provide an adequate basis for a thorough analysis of the benefits and penalties of actual compliance with the rule's operational requirements. I am nonetheless able to concur with the presenbed implementation procedure because it offers the potential for development of information now lacking.

Comphance with the new rule's operational limits within any short time period will entail certain penalties. At the same timt such compliance will achieve some degree of safety benefit. These factors were summarized by the Regulatory Staff (with respect to its recommendation) thusly:

(a) decrease in an already low radiological risk to the natural environment and public health and safety from postulated loss-of coolant accidents;(b) potential derating averaging about 5 percent of capacity of nuclear power plants for about two years, with attendant increase in air pollution and economic costs from increased operation of fossil-fueled generators;(c) increased nuclear generation costs due to cost of fuel design changes permitting resumption of full power operation. (Environmental Statement p.1).

While these benefits and some penalties were identified, they were not quantified to the degree of approximation that I believe possible and clearly warranted by this important issue. Moreover, the criteria by which the balance was struck were undefined. In short, the staff's cost-benefit analysis (Final Environmental Statement, pp 99149) was neither adequate nor persuasive. The Commission has properly rejected the precipitous implementation schedule recommended by the staff since the burden ofjustifying it cas not met.

Today's decision does not sxclude the possibility of an exemption from the requirement that a reactor be brought into compliance with the new rule six months after its effective date. Hopefully, few,if any, exemptions will be found warranted after careful review. In this regard, as stated in Section 50.46(a)(1)(vi, vii) of the new rule, a licensee must show to the satisfaction of the Commission that any such exemption wouH be in the public interest. This procedure leaves room for submission and Commission consideration cf appropriately detailed showings and quantification and analysis of factors affecting the implementation schedule such as those I now discuss briefly.

4 First,it is undisputed that the hypotheticalloss of. coolant accident is itself a highly unlikely event. The staff repeatedly acknowledged on the record that the probability of such an accident, coupled with a simultaneous failure of an ECCS conforming to the Interim Acceptance Criteria,is " negligible"(Ex. 2023 Tr. 22179; Transcript-Oral Argument Before AEC, p 20). The probability of these simultaneous occurrences has been estimated at approximately 10" per reactor year or one in 10 million reactor years (Tr. 22177, 22181, 22185). Even if such a highly unlikely coincidence would occur, the reactor containment structure would still provide substantial additional margins of safety for public protection. Of ,

course, there must be reasonable assurance of public health and safety, and the new rule unquestionably I affords a somewhat increased degree of assurance than the old. But, without attempting to prejudge the cutcome of any particulpt exemption request, the advantage of a further reduction in an already

" negligible" risk must be weighed critically against potential adverse impacts of rapid implementation. This b consistent with the Commission's statutory responsibility to regulate "in the national interest" 1 (Section 2e of the Atomic Energy Act of 1954, as amended). '

Second, compliance with the rule's operational requirements prior to fabrication and loading of redesigned fuel will appassetly require a derating of some plants. For example, the Regulatory Staff estimated that its proposed rule could produce a temporary average derating of nuclear generating capacity cf about 5% (Final Environmental Statement, pp 108,130). Such deratings could conceivably cause power reserves to fall below acceptable limits with increased probability of power outages. Occurring at a time when the nation faces an unparalleled energy crisis which already threatens to produce such disruptions, deratings would exacerbate a critical situation. The potential adverse impacts of these deratings certainly i would not be " negligible."

l Third, derating also would force utilities to resort to alternate fuels to make up the lost energy.

International conditions involving uncertain oil supplies strongly suggest the reduced availability of this fuel. To the extent that utilities must tirn to coal, derating could cause some .dverse environmental 1128

______._______..__s

T consequences, in the Chicago region, tot example, tne staff has stated that "there would be a significant impact on air pollution"if a 5% derating were to occur (FES, p 126). Air pollution increases could result in adverse effects on the health of persons living in affected areas, thereby tending to nullify the public health and safety considerations which were submitted to support the recommendation for rapid implementation.

De presence of these factors shows the need for more information and better analysis so as to provide the Commission a more rational basis for an assessment of the true impact of rapid compliance with the operational requirements of the new rule. It may be that the incremental safety benefit to be gained will justify the costs. On the other hand, it may be that the benefit of rapid compliance is outweighed by the various derating penalties-penalties which might be avoided by requiring compliance with the rule in accordance with a reactor's normal refueling cycle. In any event, showings can be made in support of-or in opposition to-requests for exemption submitted under the procedures described in our decision. Itis precisely because we have devised these procedures to obtain the information now lacking that I am able to concur in today's decision.

In addition, I wish to add a few words concerning the question of proprietary data. Protection of proprietary matters legitimately preserves for industry the benefits of its own research. This fundamental prindple of our free enterprise system fosters private investment in safet r-related R&D which in turn has resulted in enhanced safety margins for our nation's nuclear power plas . On the other hand, the public interest clearly requires some disclosure of data which provides the underpinning of a particular safety j requirement applicable to an entire industry. The public has a legitimate right to know and examine such data for the purpose of determining its validity. But this right must not be allowed to be abused,especially as U.S. industrial participants are able to provide substantial support to safety-related R&D in an environn*nt of rising domestic and foreign competition. Balancing these connicting considerations is a difficult, task. In the special circumstances of the case which is before us, I would agree that the Combustion Engineering item should be made public; and that this action should not be considered as precedent. Hopefully, procedures which can satisfy all legitimate interests will be developed in the near future.

Finally, as we recognize in today's decision, the inquiry does not end here. There are areas in which further research is necessary. In particular, the record shows conflicting estimates as to the acmptable maximum cladding temperature. Though I have accepted the recommendation for a limit of 2200*F (reflecting a conservative interpretation of the available experimental data), I am inclined to believe that there is a high probability that this interpretation is overly conservative. But, the limitations of the present record do not justify any course other than that which we have taken today. For the future, however,I  ;

emphasize our instruction to the Director of Reactor Safety Research (supra. page 1088) to give priority '

attention to this important area. In my view, the experimental procedures to be used should be developed with sufficient rigor to be acceptable to the nuclear manufacturing industry and intervenor experts as well  ;

as to the Commission. More information will be developed as this research continues. Where appropriate, these data will enable us to refine those assumptions now built into the rule which may prove to be overly conservative. The end result of this and the other factors noted above should be a fuller achievement of the  ;

benefits of nuclear power, while maintaining a policy of meticulous attention to matters of public health  !

and safety.

APPENDIX On November 30, 1971, the Atomic Energy Commission published in the Federal Register l (36F.R. 22774) a notice scheduling a legislative-type public rule making hearing on January 27,1972, l before a hearing board consisting of Nathaniel H. Goodrich, Esq., Chairman, Dr. lawrence R. Quarles, and l Dr. John H. Buck, conceming its interim statement of policy establishing acceptance criteria for emergency l core cooling systems for light water cooled nuclear power reactors, published June 29,1971 i

(36F.R.12247). Amendments to the interim criteria were published in the Federal Register on December 18,1971 (36 F.R. 24082) in a notice that stated that the amendments would also be considered at the rule making hearing.

Participation in the rule making hearing was extensive. The primary participants included the Commission Regulatory Staff, four reactor manufacturers, a consolidated group of electric utility companies, and the Consolidated National Interveners (CNI), a group of about 60 organizations and 1129

individuals, in addition, three states, the Lloyd Harbor Study Group, and several individuals participated to a lesser degree.The hearings lasted a total of 125 days and generated a record of more than 22.000 pages of transcript and thousands of pages of written direct testimony and exhibits. Oral argument from the seven principal participants was heard by the Commission on October 9,1973.

In implernentation of the National Environmental Policy Act of 1969, (P.L 91190), a Draft Environmental Statement concerning the proposed rule making was forwarded to the Council on Environtrental Quality on December 6,1972, and circulated for comment to participants in 'he hearing and interested Federal Agencies on December 7,1972. Notice of public availability of the Statement and an invitation for comment was also published in the FederalRegister at that time. Comments on the Draft Statement were received and a Final Environmental Statement was published on May 9,1973.

The Commission noted in the interim Policy Statement:

Protection against a highly unlikely loss-of-coolant accident has long been an essential part of the cafense-in-depth concept used by the nuclear power industry and the AEC to assure the safety of nuclear power plants. In this concept, the primary assurance of safety is accident prevention by correctly designing, constructing, and operating the reactor. Er. tensive and systematic quality assurance practices are required and applied at every step to achieve this primary assurance of safety.

Nevertheless, deviations from expected behavior are postulated to occur, and protective systems are installed to take corrective action as required in such events. Notwithstanding all this, the occurrence of serious accidents is postulated, in spite of the fact that they are highly unlikely, and engineered safety features are installed to mitigate the consequences of these unlikely events. The loss of coolant accident is such a postulated improbable accident; the emergency core cooling system is one of the engineered safety features installed to mitigate its consequences.

The Commission has adopted new regulations, set forth below, dealing with the effectiveness of ECCS.

In a 140 page opinion issued on December 28,1973, the Commission discussed the changes from the i interim acceptance criteria and the technical reason for them. Copies of this opinion are available for l inspection and copying at th6 Commission's Public Document Room,1717 H. Street, N.W., Washington, D. C.

The principal changes from the Interim Policy Staternent are as follows. The old criterion number one, specifying that the temperature of the Zircaloy cladding should not exceed 2300*F,is replaced by two criteria, lowering the allowed peak Zircaloy temperature to 2200*F and providing a limit on the maximum allowed local oxidation. The other three criteria of the IAC are retained, with some modification of the wording. Ihese three criteria limit the hydrogen generation from metal-water reactions, require maintenance of a coolable core geometry, and provide for long-term cooling of the quenched core, f The most important effect of the changes in the required features of the evaluation models is that swelling and bursting of the cladding must now be taken into consideration when they are calculated to occur, and that the maximum temperature and oxidation criteria must be applied to the region of clad swelling or bursting when the maximum temperature and oxidation are calculated to occur there. Another important change is the requirement that,in the steady state operation just before the postulated accident, the thermal conductance of the gap between the fuel pellets and the cladding should be calculated taking into consideration any increase in gap dimensions resulting fro'n such phenomena as fuel densification, and sh:uld also consider the effects of the presence of fission gases. When these effects are taken into consideration a higher stored energy may be calculated. Other changes in the evaluation models are mostly in the direction of replacing previous broad conservative assumptions with more detailed calculations where new experimentalinformation is available or where better calculational methods have been developed.

The wording of the definition of a loss of coolant accident has been modified to conform to its 1:ng4ccepted usay, limiting it to breaks in pipes. The new regulations also require a more complete

' documentation of the evaluation models that are used.

The Commission believes that the implementation of the new regulations will ensure an adequate margin of performance of the ECCS should a design basis LOCA ever occur. This margin is provided by conservative features of the evaluation models and by the criteria themselves. Some of the major points that contribute to the conservative nature of the evaluations and the criteria are as follows:

(1) Stored Hest. The assumption of 102% of maximum power, highest allowed peaking factor, and highest estimated thermal resistance between the UOa and the cladding provides a calculated stored heat that is possible but unlikely to occur at the time of a hypothetical accident. While not necessarily a margin 1130

over the extrerne condition, it represents at least an assumption that an accident happens at a ume which is not typical.

(2) Blowdown. The calculation of the heat transfer during blowdown is made in a very conservative manner. There is evidence that more of the stored heat would be removed than calculated, although there is not yet an accepted way of calculating the heat transfet more accurately,it is probable that this represents a conservatism of several hundred degrees F in stored energy after blowdown, most of which can reasonably be expected to carry over to a reduction in the calculated peak temperature of the Zircaloy cladding.

(3) Rate of Heat Generation. It is assumed that the heat generation rate from the decay of fission products is w% greater than the proposed ANS standard. This represents an upper limit to the degree of uncertainty. The assumption that the fission product levelis that resulting from operation at 102% of rated power for an infm' ite time represents an improbable situation, with a conservatism that is probablyin the range of 5 to 15%. The use of the Bakerdust equation for calculating the heat generation from the steam oxidation of zircaloy should also provide some conservatism, but the factor is uncertain.

(4) The Peak Temperature Criterion. The limitation of the peak calculated temperature of the cladding to 2200*F and the stipulation that this criterion be applied to the hottest regio i of the hottest fuel rod provide a substantial degree of conservatism. They ensure that the core would suffer very little damage in the accident.

Pursuant to the Atomic Energy Act of 1954, as amended, and Sections 552 and 553 of Title 5 of the United States Code, the following amendments to Title 10, Chapter I, Code of Federal Regulations, Part 50, are published as a document subject to codification to be effective on [30 days after publication in the FederalRegister} .

1. A new sentence is added to Section 50.34(a)(4) of 10 CFR Part 50 to read as follows:

50.34 Contents of applications: technicalinformation (a) * *

(4)"* Analysis and evaluation of ECCS cooling performance following postulated loss-of. coolant accidents shall be performed in accordance with the requirements of Q50.46 for facilities for which construction permits may be issued after December 28,1974.

2. A new sentence is added to Section 50.34(b)(4) 10 CFR Part 50 to read as follows:

Q50.34 Contents of applications;technicalinformation.

(a) " *

(b) " *

(4)'" Analysis and evaluation of ECCS cooling performance following postulated loss.of coolant j accidents shall be performed in accordance with the requirements of QSO.46 for facilities for which a j license to operate may be issued after December 28,1974. I

3. A new Q50.46 is added to 10 CFR Part 50 to read as follows:

Q50.46 Acceptance Criteria for Emergency Core Cooling Systems for Light Water Nuclear Pour Reactors.

(a)(1) Except as provided in subparagraphs (2) and (3) of this paragraph, each boiling and pressurized light water nuclear power reactor fueled with uranium oxide pellets within cylindrical zircaloy cladding shall be provided with an emergency core cooling system (ECCS) which shall be designed such that its calculated cooling performance following postulated loss-of. coolant accidents conforms to the criteria set forth in parapaph (b). ECCS cooling performance shall be calculated in accordance with an acceptable l evaluation snodst, and shall be calculated for a number of postulated loss.of coolant accidents of different sizes, locations, and other properties sufficient to provide assurance that the entire spectrum of postulated

, loss.of-coolant accidents is covered. Appendix K, ECCS Evaluation Models, sets forth certain required and

! acceptable features of evaluation models. Conformance with the criteria set forth in paragraph (b), with i ECCS cooling performance calculated in accordance with an acceptable evaluation model, may require that restrictions be imposed ora reactor operation.

(2)With respect to reactors for which operating licenses have previously been issued and for which l operating licenses may issue on or before December 28,1974:

(i)The time within which actions required or permitted under this subparagraph (2) must occur shall begin to run on [30 days after publication of the rule in the FederalRegister] .

(ii)Within six months following the date specified in subparagraph (i) of this subparagraph (2), an evaluation in accordance with subparagraph (1) of this paragraph (a) shall be submitted to the Director of l

1131

Regulation. The evaluation shall be accompanied by such proposed changes in technical specifications or license amendments as may be necessary to bring reactor operation in conformity with subparagraph (1) of this paragraph.

f (iii) Any licensee may request an extension of the six. month period referred to in subparagraph (ii)of I

this subparagraph (2) for good cause. Any such request shall be submitted not less than 45 days prior to expiration of the sivmonth period, and shall be accompanied by affidavits showing precisely why the evaluation is not complete and the minimum time believed necessary to complete it. The Director of l

Regulation shall cause notice of such a request to be published promptly in the FederalRegister; such notice shall provide for the submission of comments by interested persons within a time period to be established by the Director of Regulation. If, upon reviewing the foregoing submissions, the Director of Regulation concludes that good cause has been shown for an extension, he may extend the six-month period for the shortest additional time which in this judgment will be necessary to enable the licensee to furnish the subnussions required by subparagraph (ii) of this subparagraph (2). Requests for extensions of the six-month period, submitted under this subparagraph, shall be ruled upon by the Director of Regulation prior to expiration of that period.

(iv) Upon submission of the evaluation required by subparagraph (ii)of this subparagraph (2)(or under subparagraph (iii), if the six. month period is extended) the facility shall continue or commence operation only within the limits of both the proposed technical specifications or license amendments submitted in accordance with this subparagraph (2) and all technical specifications or hcense conditions previously imposed by the Commission, including the requirements of the Interim Policy Statement (June 29,1971, 36 F.R. I 2248), as amended (December 18,1971,36 F.R. 24082).

(v)Further restrictions on reactor operation will be imposed by the Director of Regulation if he finds that the evaluations subnutted under subparagraphs (ii)and (iii)of this subparagraph (2)are not consistent with subparagraph (1) of this paragraph (a) and as a result such restrictions are required to protect the public health and safety.

(vi) Exemptions from the operating requirements of subparagraph (iv) of this subparagraph (2) may be granted by the Commission for good cause. Requests for such exemption shall be submitted not less than 45 days prior to the date upon which the plant would otherwise be required to operate in accordance with the procedures of said subparagraph (iv). Any such request shall be filed with the Secretary of the Commission, who shall cause notice of its receipt to be published promptly in the FederalRegister; such notice shall provide for the submitsion of comments by interested persons within 14 days following Federal Register publication. The Director of Regulation shall submit his views as to any requested exemption within five days following expiration of the comment period.

(vii) Any request for an exemption submitted under subparagraph (vi) of this subparagraph (2) must show, with appropriate affidavits and technical submissions, that it would be in the public interest to allow the licensee a specified additional period of time within which to alter the operation of the facility in the manner required by subparagraph (iv) of this subparagraph (2). The request shall also include a discussion of the alternatives available for establishing compliance with the rule.

(3) Construction permits may be issued after December 28.1973 but before December 28,1974 subject to any applicable conditions or restrictions imposed pursuant to other regulations in this chapter and the Interim Acceptance Criteria for Emergency Core Cooling Systems published on June 29,1971 (36F.R.12248) as amended (December 18,1971,36 F.R. 24082): Provided, however, that no operating license shall be issued for facilities constructed in accordance with construction permits issued pursuant to this subparagraph, unless the Commission determines, among other things, that the proposed facility meets the requirements of subparagraph (1)of this paragraph.

(b)(1) Peak C1sdding Temperature. The calculated maximum fuel element cladding temperature shall not exceed 2200*F.

(2) Maximum Cladding Oxidation. The calculated total oxidation of the cladding shall nowhere exceed 0.17 times the total cladding thickness before oxidation. As used in this subparagraph total oxidation means the tot.! thickness of cladding metal that would be locally converted to oxide if all the oxygen absorbed by and reacted with the cladding locally were converted to stoichiometric zirconium dioxide. If cladding rupture is calculated to occur, the inside surfaces of the cladding shall be included in the oxidation, beginning at the calculated time of rupture. Cladding thickness before oxidation means the radial distance from inside to outside the cladding, after any calculated rupture or swelling has occurred but before significant oxidation. Where the cale ilated conditions of transient pressure and temperature lead to 1132

T a prediction of cladding swelling, with or without cladding rupture, the unoxidized cladding thickness shall be defined as the cladding cross sectional area, taken at a horizontal plane at the elevation of the rupture,if it occurs, or at the elevation of the highest cladding temperature if no rupture is calculated to occur, divided by the average circumference at that elevation. For ruptured cladding the circumference does not include the rupture opening.

(3) Maximum Hydrogen Generation. The calculated total amount of hydrogen generated from the chemical reaction of the cladding with water or steam shall not exceed 0.01 times the hypothetical amount that would be generated if all of the metal in the claddiig cylinders surrounding the fuel, excluding the cladding surround ng the plenum volume, were to react.

(4)Coolable Geometry. Calculated changes in cw geometry shall be such that the core remains amenable to cooling.

(5)Long-Term Cooling. After any calculated successful initial operation of the ECCS, the calculated core temperature shall be maintained at an acceptably low value and decay heat shall be removed for the extended period of time required by the long-lived radioactivity remaining in the core.

(c) As used in this section:

(1) Loss-of-coolant accidents (LOCA's) are hypothetical accidents that would result from the loss of reactor coolant, at a rate in excess of the capability of the reactor coolant makeup system, from breaks in pipes in the reactor coolant pressure boundary up to and including a break equivalent in size to the double-ended rupture of the largest pipe in the reactor coolant system.

(2) An evaluation model is the calculational framework for evaluating the behavior of the reactor system during a postulated loss-of coolant accident (LOCA). It includes one or more computer programs and all other information necessary for application of the calculational framework to a specifk LOCA, such as raathematical models used, assumptions included in the programs, procedure for treating the program input and output information, specification of those portions of analysis not included in computer programs, values of parameters, and all other information necessary to specify the calculational procedure.

(d)The requirements of this section are in addition to any other requirements applicable to ECCS set forth in this Part. The criteria set forth in paragraph (b), with cooling performance calculated in accordance with an acceptable evaluation model, are in implementation of the general requirements with respect to ECCS cooling performance design set'forth in this Part, including in particular Criterion 35 of Appendix A.

4. A new Appendix K is added to 10 CFR Part 50 to read as follows: Appendix K-ECCS Evaluation Models.

1. Required and Acceptable Features of Evaluation Models.

11. Required Documentation.

I. REQUIRED AND ACCEPTABLE FEATURES OF THE EVALUATION MODELS A. SOURCES,OF HEAT DURING THE LOCA For the heat sources listed in Paragraphs I to 4 below it shall be assumed that the reactor has been operating continuously at a power level at least 1.02 times the licensed power level (to allow for such uncertainties as instrumentation error), with the maximum peaking factor allowed by the technical specificatbna. A range of power distribution shapes and peaking factors representing power distributions that may occur over the core lifetime shall be studied and the one selected should be that which results in the most severe calculated consequences, for the spectrum of postulated breaks and single failures analyzed.

1. The Initial Stored Energy in the fuel. The steady-state temperature distribution and stored energy in the fuel before the hypothetical accident shall be calculated for the burn-up that yields the highest calculated cladding temperature (or, optionally, the highest calculated stored energy). To accomplish this, the thermal conductivity of the UO 2shall be evaluated as a function of burn-up and temperature, taking l into consideration differences in initial density, and the thermal conductance of the gap between the UOs

! and the cladding shall be evaluated as a function of the burn up, taking into consideration fuel densification i and expansion, the composition and pressure of the gases within the fuel rod, the initial cold gap dimension '

with its tolerances, and cladding creep.

2. Fission Heat. Fission heat shall be calculated using reactivity and reactor kinetics. Shutdown reactivities resulting from temperatures and voids shall be given their minimum plausible values, including l

l 1133 '

i

T

! allowance for uncertainties, for the range of power distribution shapes and peaking factors indicated to t studied above. Rod trip and insertion may be assumed if they are calculated to occur.

3. Decay of Actmides. The heat from the radioactive decay of actinides, including neptunium an plutonium generated dunng operation, as well as isotopes of uranium, shall be calculated in accordane with fuel cycle calculations and known radioactive properties.The actinide decay heat chosen shall be tha appropriate for the time in the fuel cycle that yields the highest calculated fuel temperature during th LOCA.

4. fission Product Decay. The heat generation rates from radioactive decay of fission products shall b assumed to be equal to 1.2 times the values for infinite operating time in the ANS Standard (Propose.

American Nuclear Society Standard " Decay Energy Release Rates Following Shutdown of Uraniurr Fueled Thermal Reactors". Approved by Subcommittee ANS-5, ANS Standards Committee, Octobe 1971). The fraction of the locally generated gamma energy that is deposited in the fuel (including the cladding) may be different from 1.0; the value used shall be justified by a suitable calculation.

S. Metal-Water Reaction Rate. The rate of energy release, hydrogen generation, and cladding oxidatiot from the metal / water reaction shall be calculated using the Baker.Just equation (Baker, L., Just, L.C.

" Studies of Metal Water Reactions at High Temperatures,111. Experimental and Theoretical Studies of the Zirconium. Water Reaction," ANL 6548, page 7, May 1962). The reaction shall be assumed not to be stearr limited. For rods whose cladding is calculated to rupture during the LOCA, the inside of the cladding shal also be assurned to react after the rupture. The calculation cf the reaction rate on the inside of the cladding shall also follow the Baker.Just equation, starting at the time when the cladding is calculated to rupture and extending around the cladding inner circumference and axially no less than 1.5 inches each way from the location of the rupture, with the reaction assumed not to be steam limited.

6. Reactor Internals Heat Transfer. Heat transfer from piping, vessel walls, and non. fuel internal hardware thall be taken into account. {

7. Pressurized Water Reactor Primary-to-Secondary Heat Transfer. Heat transferred between primary and secondary systems through heat exchangers (steam generators) shall be taken into account. (Not applicable to Boiling Water Reactors.) j 1

B. SWELLING AND RUPTURE OF THE CLADDING AND FUEL ROD THERMAL PARAMETERS  !

Each evaluation model shall include a provision for predicting cladding swelling and rupture from consideration of the axial temperature distribution of the cladding and from the difference in pressure between the inside and outside of the cladding.both as functions of time. To be acceptable the swelling and rupture calculations shall be based on applicable data in such a way that the degree of swelling and incidence of rupture are not underestimated.The degree of swelling and rupture shall be taken into account in calculations of gap conductance, cladding oxidation and embrittlement, and hydrogen generation.

The calculations of fuel and cladding temperatures as a function of time shall use values for gap conductance and other thermal parameters as functions of temperature and other applicable time. i dependent variables. The gap conductance shall be varied in accordance with changes in gap dimensions and I any other applicable variables. I C. BLOWDOWN PHENOMENA

l. Break Charactosiselas and Flow a.In analyses of hypothetical loss-of coolant accidents, a spectrum of possible pipe breaks shall be considered. This spectrum shall include instantaneous double ended breaks ranging in cross sectional area up to and including that of the largest pipe in the primary coolant system. The analysis shall also include the effects of longi.udinal splits in the largest pipes, with the split area equal to the cross-sectional area of the pipe.

b. Discharge Model For all times after the discharging fluid has been calculated to be two phase in composition, the discharge rate shall be calculated by use of the Moody model (F. J. Moody, " Maximum Flow Rate of a Single Component, Two-Phase Mixture," Journal of Heat Transfer, Transactions of the American Society of Mechanical Engineen, 87, No.1, February 1965). The calculation shall be conducted l

l 1134

. I with at least three values of a discharge coefficient applied to the postulated break area, these values sp-sing the range from 0.6 to 1.0. If the results indicate that the maximum clad temperature for the hypothetical accident is to be found at an even lower value of the discharge coefficient, the range of discharge coefficients shall be extended until the maximum clad temperature calculated by this variation has been achieved.

c.End of Blowdown. (Applies Only to Pressurized Water Reactors.) For postulated cold leg breaks,all emergency cooling water injected into the inlet lines or the reactor vessel during the bypass period shallin the calculations be subtracted from the reactor vessel calculated inventory. This may be executed in the calculation drring the bypass period, or as an alternative the amount of emergency core cooling water calculated to be injected during the bypass period may be subtracted later in the calculation from the water remaining in the inlet lines, downcomer, and reactor vessel lower plenum after the bypass period. This bypassing shall end in the calculation at a time designated as the "end of bypass," after which the expulsion or entrainment mechanisms responsible for the bypassing are calculated not to be effective. The end of. bypass definition used in the calculation shall be justified by a suitable combination of analysis and experimental data. Acceptable methods for defining "end of bypass" include, but are not limited to, the following: (1) Prediction of the blowdown calculation of downward flow in the downcomer for the remainder of the blowoown period;(2) Prediction of a threshold for droplet entrainment in the upward velocity, using local fluid conditions and a conservative critical Weber number, d.Noding Near the Break a.1d the ECCS Injection Points. The noding in the vicinity of and including the broken or split sections of pipe and the points of ECCS injection shall be chosen to permit a reliable analysis of the thermodynamic history in these regions during blowdown.

2. Frictionhl Pressure Drops. The frictional losses in pipes and other components including the reactor core shall be, calculated using models that include realistic variation of friction factor with Reynolds number, and realistic two-phase friction multipliers that have been adequately verified by comparison with experimental data, or models that prove at least equally conservative with respect to maximum clad temperature calculated during the hypothetical accident. The modified Baroczy correlation (Baroczy,C. J.,

"A Systematic Correlation for Two-Phase Pressure Drop," Chem. Enging. Prog. Symp. Series, No. 64, Vol. 62,1965) or a combination of the Thom correlation (Thom, J. R.S., " Prediction of Pressure Drop Dunng Forced Circulation Boiling of Water," Int. /. of Heat & Mass Transfer, 7, 709-724,1964) for pressures equal to or greater than 250 psia and the Martinelli. Nelson correlation (Martinelli, R.C., Nelson, D. B., " Prediction of Pressure Drop During Forced Circulation Boiling of Water," Transactions of ASME, 695 702,1948) for pressures lower than 250 psia is acceptable as a basis for calculating realistic two-phase friction multipliers.

3. Momentum Equation. The following effects shall be taken into account in the conservation of momentum equation: (1) temporal ch nge of momentum, (2) momentum convection, (3) area change momentum flux, (4) mornentum change due to compressibility, (S) pressure loss resulting from wall friction (6) pressure loss resulting from area change, and (7) gravitational acceleration. Any omission of one or more of these terms under stated circumstances shall be justified by comparative analyses or by  ;

experimental data.

i l

4. Critical Heat Flux j

Correlations developed from appropriate steady. state and transient. state experimental data are acceptabis for use in predicting the critical heat flux (CHF) during LOCA transients. The computer programs in which these correlations are used shall contain suitable checks to assure that the physical parameters are within the range of parameters specified for use of the correlations by their respective authors.

b. Steady state CHF correlations acceptable for use in LOCA transients inclede,but are not limited to, the following:

(1) W3. L S. Tong," Prediction of Departure from Nucleate Boiling for an Axially Non-uniform Heat Flux Distribution," Journal ofNuclear Energy, Vol. 21, 241 248,1967.

(2)BdW-2. J. S. Gellerstedt, R. A. I.ee, W. J. Oberjohn, R. H. Wilson, L J. Stanek, " Correlation of l Critical Heat Flux in a Bundle Cooled by Pressurized Water," Tw& Phase Flow and Heat Transfer in Rod Bundles, ASME,New York,1969.

1

! 4 1135

(3)Hench-Levy. J. M. Healzer, J. E. Hench, E. Janssen, S. levy, " Design Basis for Critical Heat Flux Condition in Boiling Water Reactors," APED SIM, GE Company Private report, July 1966.

(4)Macbeth. R. V. Macbeth, "An Appraisal of Forced Convection Burnout Data," Proceedings of the Institute ofMechanicalEngineers, l965 l966.

(5)Barnett. P. G. Barnett,"A Correlation of Burnout Data for Umformly Heated Annuli and its Uses for Predicting Burnout in Uniformly Heated Rod Bundles," AEEW R 463,1966.

(6)#ughes. E. D. Hughes,"A Correlation of Rod Bundle Critical Heat Flux for Water in the Pressure Range 150 to 725 psia,: IN 1412, Idaho Nuclear Corporation, July 1970.

c. Correlations of appropnate transient CHF data may be accepted for use m LOCA transient analyses if comparisons between the data and the correlations are provided to demonstrate that the correlations predict values of CHF which allow for uncertainty in the experimental data throughout the range of parameters for which the correlations are to be used. Where appropriate, the comparisons shall use statistical un,:ertainty analysis of the data to demonstrate the conservatism of the transient correlation.

~

d. Transient CHF correlations acceptable for use in LOCA transients include, but are not limited to, the following:

(1)CE Transient CHF. B. C. Slifer, J. E. Hench, " Loss-of Coolant Accident and Emergency Core Cooling Models for General Electric Boiling Water Reactors," NEDO 10329, General Electric Company, Equation C-32, April 1971.

e. After CHF is first predicted at an axial fuel rod location during blov-down, the calculation shall not use nucleate boiling heat transfer correlations at that location subsequently during the blowdown even if the calculated local Guid and surface conditions would apparently justify the reestablishment of nucleate boiling. Heat transfer assumptions characteristic of return to nucleate boiling (rewetting) shall be permitted when justified by the calculated local Guid and surface conditions during the reflood portion of a LOCA.

5. Post CHF Heat Transfer, Correlations

a. Correlations of heat transfer from the fuel cladding to the surrounding Guid in the post CHF regimes of transition and film boibng shall be compared to applicable steady-state and transient state data using statistical correlation and uncertainty analyses. Such comparison shall demonstrate that the correlations predict values of heat transfer coefficient equal to or less than the mean value of the applicable experimental heat transfer data throughout the range of parameters for which the correlations are to be  !

used. The comparisons shall quantify the relation of the correlations to the statistical uncertainty of the  !

applicable data.

b.The Groeneveld now film boihng correlation (Equation 5.7 of D.C. Groeneveld, "An Investigation I of Heat Transfer in the Liquid Deficient Regime," AECL 3281, revised December 1969), 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 the Fluid at Low Qualities," MIT Report Number 9079 26, Cambridge, Massachusetts, September 1963), and the Westinghouse correlation of steady-state transition i

boiling (" Proprietary Redirect / Rebuttal Testimony of Westinghouse Electric Corporation," U.S.A.E.C. 1 Docket RM 501, page ,251, October 26,1972) are acceptable for use in the post-CHF boiling regimes. In j addition the transition boiling correlation of McDonough, Milich, and King (J. B. McDonough, W. Milich, E.C King, " Partial Film Boiling with Water at 2000 psig in a Round Vertical Tube," MSA Research Corp., )

Technical Report 62 (NP-6976), (1958) is suitable for use between nucleate and film boiling. Use of all  ;

these correlations shall be restricted as follows: l (1) The Groeneveld correlation shall not be used in the region near its low pressure singularity, l (2)the first term (nucleate) of the Westinghouse correlation and the entire McDonough, Milich, and King correlation shall not be used during the blowdown after the temperature difference between the clad and the saturated fluid first exceeds 300*F, (3) transition boiling heat transfer shall not be reapplied for the remainder of the LOCA blowdown, even if the clad superheat returns below 300*F, except for the reflood portion of the LOCA when justified by the calculated local fluid and surface conditions.

6. Pump Modeling. The characteristics of rotating primary system pumps (axial flow, turbine, or l centrifugal) shall be derived from a 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 used for analysis should be justified. The pump model for the two phase region shall be verified by applica9 1136

i two-phase pump performance data. For isWR's after saturation is calculated at the pump suction, the pump head may be assumed to vary linearly with quality, going to zero for one percent quality at the pump suction, so long as the analysis shows that core flow stops before the quality at pump suction reaches one percent.

7. Core Flow Distribution During Blowdown. (Applies only to pressurized water reactors.)

a.The flow rate through the hot region of the core during blowdown shall be calculated as a function of tirne. For the purpose of these calculations the hot region chosen shall not be greater than the size of one fuel assembly. Calculations of average flow and flow in the hot region shall take into account cross flow between regions and any flow blockage calculated to occur during blowdown as a result of cladding swelling or rupture. The calculated flow shall be smoothed to eliminate any calculated rapid oscillations (period less than 0.1 seconds).

• b. A method shall be specified for determining the enthalpy to be used as input data to the hot channel heatup analysis from quantitles calculated in the blowdown analysis, consistent with the flow distribution calculations.

D. POST-BLOWDOWN PHENOMENA; HEAT REMOVAL BY THE ECCS 1.Singfe Failure Criterion. An analysis of possible failure modes of ECCS equipment and of their effects on ECCS performance must be made. In carrying out the accident evaluation the combination of l

ECCS subsysterns assumed to be operative shall be those available after the most damaging single failure of 1 ECCS equipment has taken place. I

2. Containment &cssure. The containment pressure used for evaluating cooling effectiveness during reflood and . spray cooling shall not exceed a pressure calculated conservatively for this pv-rose. The f f j calculation shall include the effects of operation of all installed pressure-reducin6 systems and pracesses. i 3.Calcukution of Reflood Rate for Pressurized Water Reactors. The refilling of the teanor vessel and the time and rate of reflooding of the core shall be calculated by an acceptable model that takes into consideration the thermal and hydraulic characteristics of the core and of the reactor system.The primary system coolant pumps shall be assurned to have locked impellers if this assumption leads to the maximum i calculated cladding temperature; otherwise the pump rotor shall be assurned to be running free.1he ratio of the total fluid flow at the core exit plane to the totalliquid flow at the core inlet plane (carryover fraction) )

shall be used to determine the core exit flow and shall be determined in accordance with applicable j experimental data (for example, "PWR FLECHT (Full Length Emergency Cooling Heat Transfer) Final Report," Westinghouse Report WCAP-7665, April 1971; "PWR Full length Emergency Cooling Heat Transfer (FLECHT) Group I Test Report," Westinghouse Report WCAP 7435, January 1970; "PWR FLECHT (Full length Emergency Cooling Heat Transfer) Group 11 Test Report," Westinghouse Report WCAP 7544, September 1970; "PWR FECHT Final Report Supplement," Westinghouse Report WCAP 7931, October 1972).

The effects on reflooding rate of the compressed gas in the accumulator which is discharged following accumulator water discharge shall also be taken into account.

4. Steam Interaction with Emergency Core - Cooling Water in Pressurized Water Reactors. The thermal 4(ydraulic interaction between steam and all emeryncy core cooling water shall be taken into account in calculating the core reflooding rate. During refill and reflood, the calculated steam flow in unbroken reactor coolant pipes shall be taken to be zero during the time that accumulators are dscharging water into those pipes unless experirnental evidence is available regarding the realistic thermalhydradic j interaction between the steam and the liquid, in this case, the experirnental data may be used to support w alternate assumption.  !

5.R4R med Reflood Hent Tranger for Pressurized Water Reactors. For teflood rates of one inch per l- second or higher, reflood heat transfer coefficients shall be based on applicable experimental data (c.

unblocked cores including FLECHT results ("PWR FLECHT (Full Length Emergency Cooling Heat

[~ <

Transfer) Final Report," Westinghouse Report WCAP 7665, April 1971). The use of a correlation derived from FECHT data shall be demonstrated to be conservative for the transient to which it is applied; I presently available FLECHT heat transfer correlations ("PWR Full Length Emergency Cooling Heat l Transfer (FECHT) Group i Test Report," Westinghouse Report WCAP 7544, September 1970; "PWR FLECHT Final Report Supplement," Westinghouse Report WCAP 7931, October 1972) are not acceptable.

New correlations or modifications to the FLECHT heat transfer correlations are acceptable only after they are demonstrated to be conservative, by comparison with FECHT data, for a range of parameters .

consistent with the transient to which they are applied.

During refill and during reflood when reflood rates are less than one inch per second, heat transfer calculations shall be based on the assumption that cooling is only by steam, and shall take into account any 1137

________-___________--________-_L

flow blockage calculated to occur as a result of claddmc swelling or rupture as such blockage nught affect both local steam now and heat transfer.

6. Convective Heat Transfer Coefficients for Boiling Water Reactor Fuel Rods Under Spray Cooling.

Following the blowdown period, convective heat transfer shall be calculated usmg coefficients based on appropriate experimental data. For reactors with jet pumps and havmb fuel rods in a 7 x 7 fuel assembly array, the following convective coefficients are acceptable:

a. During the period folio ving lower plenum flashing but prior to the core spray reaching rated flow,a convective heat transfer coefficient of zero shall be applied to all fuel rods.

b. Dunng the period after core spray reaches rated flow but prior to renoodmg. convective heat transfer coefficients of 3.0,3.5,1.5, and 1.5 Btu-hr ft 2 *F2 shall be applied to the fuel rods in the outer corners, outer row, next to outer row, and to those remaining in the interiot, respec;ively, of the assembly,

c. After the two phase refloodmg fluid reaches the level under consideration.a convective heat transfer coef0cient of 25 Btu hf'-ft-2 ppi halls be applied to all fuel rods.

7. The Boiling Water Reactor ChannelBox Under Spray Cooling. Followmg the blowdown period, heat transfer from, and wetting of, the channel box shall be based on appropriate experimental data. For reactors withjet pumps and fuel rods ir 7 x 7 fuel assembly array, the following heat transfer coefficients and wetting time correlation are acceptable.

a.During the period after lower plenum flashing, but prior to core spray reachmg rated flow, a convective coefficient of zero shall be applied to the fuel assembly channel box.

b.During the period after core spray reaches rated Gow, but prior to wetting of the channel, a convective heat transfer coefficient of 5 Btu-hr -ft 2 ?F8 shall be apphed to both sides of the channel box.

c. Wetting of the channel box shall be assumed to occur 60 seconds after the time determined using the correlation based on the Yamanouchi analysis (" Loss of Coolant Accident and Emergency Core Cooling Models for General Electric Boiling Water Reactors," General Electric Company Report NEDO 10329, April 1971).

II. REQUIRED DOCUMENT ATION 1.a. A description of each evaluation model shall be furnished. The description shall be sufficiently complete to permit technical review 'of the analytical approach including the equations used, their approximations in difference form, the assumptions made, and the values of all parameters or the procedure for their selection, as for example,in accordance with a specified physicallaw or empirical correlation.

b.The description shall be suf0ciently detailed and specific to require sigmficant changes in the evaluation model to be specified in amendments of the description. For this purpose, a significant change is a change that would result in a calculated fuel cladding temperature different by more than 20*F from the temperature calculated (as a function of time) for a postulated LOCA using the last previously accepted model.

c. A complete listing of each computer program,in the same form as used in the evaluation model, shall be fumished to the Atomic Energy ComrWssion.

2.For each computer program, sowtion convergence shall be demonstrated by studies of system l

modeling or noding an,d calculational time steps.

3. Appropriate sensitivity studies shall be performed for each evaluation model, to evaluate the effect l i

on the calculated results of vatiations in noding, phenomena assumid in the calculation to predominate, l

including pump operation or locking, and values of parameters over their applicable rariges. For items to l which results are shown to be sensitive, the choices made shall be justified. I 4.To the extent practicable, predictions of the evaluation model, or portions thereof, shall be compared with applicable experirnentalinformation.

5. General Standards for Acceptability-Elements of evaluation models reviewed will include technical adequacy of the calculational meth)ds, including compliance with required features of Section I of this Appendix K and provision of a level of safety and margin of conservatism comparable to other acceptable evaluation models, taking into account significant differences in the reactors to which tiiey apply.

l l l l

)

1

, 1138 l

l m

I i

i l

i i

l ATTACHMENT 6 10 CFR 50 STATEMENTS OF CONSIDERATION I

i

3 PART 50 o STATEMENTS OF CONSir ERATION US/ A-48. S/widowa Decoy Hoot be reevaluated before any actual Aamoral/teguiremeede 8. The potential impact of differences in modifications are made so that any facility type, design. or age on the

%e overallobjective of USA A-45 le to contemplated design changes resulting relevancy and practicality of the backfit.

evaluate the adeqeecy of ckrrent from the resolution of USA A-45 can be g, licensing design requirements to ensure considered at h same tune.

g ,p eat se nuclear power plants do not prosed mated boiling CenericIssue A-Ja. Adequacy of water reactors. However, in determining rio es g n unaccepta

_ ,. Sofety-RelotedDCPowerSupply an acceptable station blackout coping

[n The study includes an assessment of ne analysis performed for USI A-44 '*'

alternative means of shutdown decay re ts re 'u ac assumed that a high level of de powe' power reliability te g . number of heat removal and of divvse "<1edicated" system reliabdity would be maintamd emergency diesel rierstors. b systems for this purpose. Reedts will so that (1) de power system fadures include proposed rara ==aarlations reliability of the o site and onsite would not be a significant contributor to emergency ac power systems) could regarding the desirability of, and losses of all ac power and (2) should a possible design regelmments for, result la d Herent acceptable coping station blackout occur. b probability of capabilities. For example, plants with an improvements in existing systems a %s immediate de power systern faihare alternative dedicated decay heat already low risk from station blackout would be low. Whereas Generic Issue because of multiple, highly reliable ac

"'M*I ****d- A-00 focuses on cubancing bettery power sources are required to withstand 8 reliability, the resolution of USI A-44 is a station blackout for a relatively short a usta core c 1 un er ta n b out conditions can be considered a aimed at ensuring adotuate station period of time: and few if any, subset of the overall A-45 issue. battery capacity in the event of a staties hardware backfits would be required as

' blackout of a specified duration. a result of the rule. Plants with currently higher risk from station blackout are din n in oc n ha two nerefore. these two issues a,, required to withstand somewhat longer issues. USA A-44 deals with the consistent and compatible, Probability ofloss of ac power, the duration blackouts: and. depending on Fire Protection Program their existing capability, may need some capability to remow nacey heat using systems that do not reqdre ac power. Section 5&4e of to CFR Part 30 states " 0dificetione to ochiev the ionger that each operating nuclear power plant aladon Mout cape ty.

and the ability to restore acf power in a timely manner. USI A-45 deals with the must have a fire protection plan that 9. Whether the backfit is interim or final overallreliability of b decay n.at satisfies GDC 3. De fire protection and. ifinterim. the justification for removal function in terms of respoue to features required to satisfy GDC 3 are imposing the backfit on an interim basis.

transients, small-break loes-oldst specified in Appendix R to 10 CFR Part ne station blackout rule is the final accidenta, and special emergencies sadt 50. Hey include certain provisions resolution of USI A-44: it is not an as fires flowds, seismic events, and maarding alternative and dedicated interim measure.

sabotage. shutdown capability. To meet these Although the rerosueeadations that provisions some licensees have added.

• might result from the resolution of USI E %

• s/ 7/es or plan to add, improved capability to effestive 7/27/es A-45 are not yet final, some could affect restore power from offsite sources or the station blackout capability, while onsite diesels for the shutdown system. General Requirements for others would not. Racammaartationa A few plants have installed a safe **"'"'"i "i"8 *'I" f** *"

that involve a new or improved decay shutdown facility for fire protection that see ry, so tesumense of coneiaeretion heat removal system that is ac power includes a charging pump powered by dependent but that does not include its os sa PR seets

- debaia m .ow- m i, w-id its own of independent

• b event a sta bi ocoet.

power s.ome .-la e/te -

g,",,"*, system can provide makeup capability posew to/17/es ti as w as additional ac-independent decay heat M es primary cwlant syske as wd as 10 CFR Part 30

" "t' Idh reactor coolant pump seal cooling. His mod a$

Recommendations that involve sa could be a significant benefit in tenne of uhancing the ability of a planHe cope Emergency Core CooEng Systems; Revielens to Acoephnee Cr# erie additional decay heet removal systees with a station blackout. plants that have added eqelpment to schieve alternate menscv: Nuclear Regulatory with its own ac power supply weild g how a signiacant elfest se USI A-44. safe shutdown in order to meet Such a new additionalsystem would ApPendia R requirements could take .

receive the a asedle within the credit for that equipment,if available, messanev %rNuclearRegulatory USI A 44 uties by either changleg for coping with a station blackout event. Commission (NRC)is amending its the emergency ac power configuration regulations to allow the use of group or providing the ability to cope 7.hhand raarce heden on es alternative methods to demonstrate that with a station blackout for an extended NRC maciaWd wis b bckSt and 6e avanapy of a,g reemeces. thee core cooling system period of time.Well before plant (ECCS protect the nuc. lear modifications.if any.willbe ne estimated total cost for NRC reactor core during a postulated design implemer.ted to comply with the staties review ofindustry submittels required basis loss.of. coolant accident (LOCA).

blackout rule, it is anticipated that the by the station blackout rule is 91.5 De Commission is taking this action proposed e.,a-,aat resolution of USI A.- million based on submittels for 100 because research, performed since the es will be published for public comment. reactors and an estimated average of current rule was written has shown that Mose plants nuding hardwom 173 person-hours per reactor. calculations performed using current f modifications for station blackout could methods and in accordance with the current requirements result in estimates so-sc-s3 sepeernba ne,was

1 PART 50 o STATEMENTS OF CONSIDERATION of cooling system performance that are solicit the public's comments on whether that they would not derive an economic significantly more conservative than the existing rule should be estimates based on the improved benefit by performing realistic analysis

" grandfathered" indefinitely. That is: of ECCS performance. The position of knowledge gained from this research. 1. Should the conservative ECCS While the existing methods are an additional commenter is unclear evaluation method of Appendix K be concerning grandfathering. The conservative, they do not result in permitted indefinitely or should this remaining commenter was not opposed accurate calculation of what would aspect of the ECCS rule be phased out to grandfathering but thought the tctually occur in a nuclear power plant after some period of time? question is premature. This commenter during a LOCA and may result in less Commissioner Asselstine requested believes that indefinite use of existing than optimal ECCS design and operating the public's comments on the following:

procedures. In addition. the operation of ECCS evaluation methods should be

2. Should this rule change include an considered when significant experience some nuclear reactors is being explicit degree of conservatism that has been gained with the unnecessarily restricted by the rule, must be applied to the evaluation implementation of the new features of resulting in increased costs of electricity models7

3. This rule change would allow a 5 to the rule but makes no recommendation generation. This rule, while continuing as to what policy the Commission to allow the use of current methods and 10 percent increase m the fission should pursue in the meantime.

requirements, also allows the use of product inventory that could be released from any core meltdown scenario. The Commission agrees with the more recent information and knowledge Should this rule change explicitly majority of the commenters that existing to demonstrate that the ECCS would Appendix K evaluation models should protect the reactor during a LOCA. Tnis Prohibit any increase in approved power be permitted indefinitely. The emendment, which applies to alj levels until all severe a,ccident issues applicants for and holders of and unresolved safety issues are Commission also believes that the resolved? decision to permit continued use of such construction permits or operating licenses for light water reactors, also 4. Should the technical basis for this models can and should be made at this relaxes requirements for certain proposed rule change be reviewed by an time because it believes that both independent group such as the methods provide adequate protection of reporting and reanalyses which de not American Physical Society? the public health and safety. As contribute to safety. ,

described in the regulatory analysis, the arrecTive DAfs: October 17,1908. Summary of Public Comments probability of a large break is so low.

FOR PURTHER INFORMATION CONTACT: The comment period for the proposed that the choice of best estimate versus L.M. Shotkin. Office of Nuclear rule revision and the draft regulatory Appendix K has little effect on public Regulatory Research, U.S. Nuclear guide (52 FR 11385) expired on July 1, risk. 'Ihe TMI action plan calls for Regulatory Commission. Washington, 1987. Twenty-seven letters addressing industry to improve their small break DC 20555, telephone (301) 492-3530. the proposed rule were received by the LOCA evaluation models to be more expiration date, as well as nine realistic when evaluating the more svPPt.sMENTARY INFORs4ATIO*C probable small break accident scenario.

responses to the request for comments BackI "und on questions in the regulatory guide. A This has been done within the context of number oflate .:omments were also i 50.46 and Appendix K compliance and On March 3.1987, the Nuclear was entirely appropriate since small received. These wee also considered to Regulatory Commission published in the breaks are not limiting in design basis the extent that new and substantial Federal Register proposed amendments comments were provided. performance and a better understanding 152 FR 6334) to 10 CFR Part 50 and The public comment on the proposed of small break behavior is a desirable Appendix K. These proposed rule revisions have been divided into safety goal from a risk perspective.

tmendments were motivated by the fact thirteen categories and are summarized Therefore, the grandfathering provision that since the promulgation of I 50.46 of in the following paragraphs. Categories has been retained in the final rule.

10 CFR Part 50, " Acceptance Criteria for one through four represent the responses 2. Specification of Explicit Degree of Emergency Core Cooling Systems to the specific questions posed by tne Conservatism (Question 21. The majority (ECCS) in Light Water Power Reactors," ACRS and Commiss;oner Asselstine. In of the responses to this question eno 'he acceptable and required general, consideration of the pub'ic indicated that the proposed rule already feature. .cd models specified in comments resulted in no substantive contains conservatism in the required Appendix K e 10 CFR Part 50, revision to the proposed rule. uncertainty evaluation.

considerab!b reseerch has been 1. Grandfathering of Conservative The use of additional conservatism performed that has greatly increased the ECCSMethods of Appendix K(Question would be inconsistent with the objective understanding of ECCS performance ff. of the rule which is to provide a realiuic during a LOCA. It is now confirmed that Twenty-one of he t commenters evaluation of plant response during a the methods specified in Appendix K, specifically addressed the ARCS LOCA. The NRC has not included an combined with other analysis methode question concerning the grandfathering additional explicit degree of currently in use, are highly conservative of the current Appendix K approach. conservatism in this rule.

and that the actual cladding Seventeen of these commenters 3. Resolution of al/SofetyIssues Prior t;mperatures which would occur during recommended indefinite grandfathering to Allowing Power Leve/ Increases a LOCA would be much lower than of the existing Appendix K evaluation (Question 3). Some commenters pointed i those calculated using Appendix K models host cited the known out that fission product inventory is not l methods. In soliciting the public's consrmtism as the basis of their a direct function of total power, but comments on the proposed rule, the reccnunendation. In addition, several rather it is the rate of fission product NRC specifically requested its views on commenters stated that in light of the formation that is a direct function of questions posed by Commissioner known conservatism not allowing power. Fission product inventory Asselstine and the Advisory Committee continued use of existing Appendix K available for release during a core on Reactor Safeguards (ACRS). The evaluation models would be unfairly meltdown would be a function of ACRS requested that the Commission burdensome to licensees who determine burnup, not total power.

I I September 29,1995 50-SC 54

)

T PART 50 o STATEMENTS OF CONSIDERATION Actually, the inventory of fission audit tool, and to previde the necessary where errors are discovered in products is a complex function of both experience to audit licensee submittals. evaluation models, requests are made to time and power and not as simple as The staff does not believe that an NRC revise plant technical specifications, or desenbed by the commenters. Short demonstration of the methodology is a some other questions regardin the lived isotopes, such as xenon and prerequisite to this rulemaking. analyses are raised.The NRC elieves iodine, quickly reach an equilibrium Licensees wishing to adopt the best that shared responsibility for evaluation inventory and total steady state estimate approach permitted as a result models would not be in the best interest inventory of these fission products is a of this rule are neither required to use of the public health and safety and direct function of power. Inventories of this methodology nor to model their own therefore has not implemented the long-lived isotopes, such as strontium methodologies after it. This methodology suggestion of this commenter.

cnd cesium, are functions of total fuel will play an important part in the best- The NRC received two requests for an burnup, as described by the estimate model review process. The extension of the comment period to commenters. Intermediate-lived isotopic NRC has determined through twenty allow time for review of NUREG-1230.

inventories are complex functions of years of experience that independent which describes the research supporting time, power, and integrated power. In an analysis with independent the proposed rule revision.

independent study, documented in methodologies is the most effective way The NRC believes the comment period chapter XII of NUREG 1230. the staff to intelligently review new vendor or was sufficient since most of the research determined that she change in risk due licensee methodologies. it is therefore is not new and has been extensively to a 5% power increase is negligible. The appropriate that this new methodology reviewed in the past. Both commenters trBuments above do not alter the be subjected to stringent technical were contacted and told that comments Commission's position that the increase scrutiny, as directed by the Commission. received after the comment period in fission products available for release The NRC staff is committed to would be r.esidered if time permitted.

during a core meltdown caused by a 5% completing this demonstration by the Comrr.er.ts from both parties were power increase is negligible compared time that it will be needed to review receind late and were indeed to the uncertainty fn fission product licensee submittals and is confident that consiaered by the NRC.

release.The Commission has decided such a demonstration will be successful. 6. Reporting Requirements. Some not to delay the proposed rule revision Based on the paucity of negative commenters viewed the proposed pending resolution of all unresolycd response concerning the technical basis reporting procedures as new safety issues or severe accideftt issues for the proposed rule revision and requirements needing consideration in cnd therefore will proceed with this generally favorable review of the NRC the backfit analysis while others stated final rulemaking, as planned, uncertainty methodology, the that they are a major relaxation and

4. Independent Review of Technico/ Commission plans no further review of clarification of existing reporting the technical basis, requirements.

Basis (Question 4). Several commenters 5. General Comments on Pmposed The NRC position is that the reporting Indicated that the technical basis for the Rule. Twenty-one commenters made requirements are newin the sense that proposed rule has had adequate review ce the research was being performed. A comments of this nature. The majority of they will now appear in the Code of the comments came from the nuclear Federal Regulations. However,in number of commenters stated that it practice, these reporting requirements was the role of the ACRS to perform any industry of which 19 expressed support of the proposed rule. The industry also are indeed a clarification and relaxation review of the proposed rule revision because it is uniquely qualified due to strongly supports the specific ECCS rule over the current interpretation for the approach proposed by the NRC. One existing requirements and therefore the its familiarity with the research. net effect of these requirements will be commenter neither supported nor l The Commission agrees that the to reduce the frequency for reporting opposed the proposed approach. One technical basis has had adequate negative comment was received from en and reanalysis.

review, except for the uncertainty A number of commenters requested methodology which is new and untried an nym us individual within the nuclear industry who implied. without that only significant errors or changes in except for the General Electric the non-conservative direction or only specifics, that the ECCS rule is not Company,a use of an uncertainty sound and that public comment is not a those that result in exceeding the 2200*F evaluation of their SAFER code. As a fair hearing because expert insiders limit be required to be reported.In proof of principle and demonstration of would be afraid to comment. addition a number of commenters feasibility, the ACRS and a second Based on the absence of any suggested that the NRC require only t u certa y rne o oo d we oped suppom hohon b b negabe annual WW OWant em M response and the unprecedented amount changes.

by the NRC for use in quantifying the of research supporting the rule revision, The NRC considers a major error or uncertainty of NRC developed thermal the NRC does not consider this comment change in any direction a cause for hydraulic transient codes. Both the to be valid and has proceeded with this concern because it raises potential l ACRS and the peer group made rulemaking with no malor revisions. questions about the adequacy of the j generally favorable cosannente One commenter suggested that fuel evaluation model as a whole. Therefore.

concerning the methodology: however. reload suppliers should not be required the NRC requires the reporting of )

both groups recognized that a complete to complete fullI.OCA/ECCS analyses significant errors or changes. In either l demonstration (i.e., application to small because the hydraulics are not changed direction, on a timely basis so that the [

break I.OCA and the reflood portion of by a fuel change. Commission may make a determination I large break I,0CA) has not yet been Although this point is valid, the of the safety significance. Thus, t'.< **1 (

accomplished and certain reviewers Commission believes that it is an rule contains no change in this

questioned whether such a unworkable situation to allow fuel requirement.

l demonstration could be performed suppliers to make use of previous One commenter recommended that successfully.The only objectives of the analyses performed by others. It is the word "immediate" be deleted from i NRC methodology demonstration are to believed that serious questions of the requirement to propose steps to be I demonstrate feasibility, to develop an accountability would arise in cases taken to demonstrate compliance in the ,

l 50-SC-55 September 29,1995

i PART 50 o STATEMENTS GF CCNSIDERAT12N '

svent that the criteria in i 50.4e(b) are The requirement, which states that the Commission did not give consideration exceeded. 50*F criteria applies to the sum of the to altering them in the final rule.

The Commission considers this a very absolute magnitudes of temperature 10. Clodding Afoterials. nree urious condition in which the plant is changes from numerous error commenters requested that the corrections or model changes was l not in compliance with the regulations Commission consider broadening the cnd may be operating in an unsa;, formulated specifically because the language of the rule to allow the use of a minner. The word "immediate" reflects Commission requires knowledge of range of zirconium based alloys for this seriousness and is further defined serious deficiencies in evaluation cladding material.

by reference in other sections of Part 50. models in use by licensees. Allowing The Commission believes that this Several commenters questioned the errors or corrections which oft et one modification is beyond the scope of the need to report minor or inconsequential an ther to relieve a licensee of the current rule revision and should be arrors or changes, even on an annual thirty-day reporting requirement, would considered in a separate rulemaking basis, as required in the proposed rule. be counter to this objecuve. lf this action in which it would receive While errors or changes which result rec mmendation were accepted, two appropriate public review and comment in changes in caIculated peak clad err rs r changes, having a large impact pnor to implementation. In addition.

temperatures of less than 50*F are not on the calculated peak cladding zircaloy cladding material is specified in c:nsidered to be ofimmediate concern, temperature but in the opposite other portions of the Code of Federal thi NRC requires cognizance of such direction, would not be reportable if the Regulations, such as i 50.44. Making a

'i '

net magnitude of tteir difference was change of this type is more suitable in a less than 50*F. For tnis reason, and the broader regulatory context. Therefore, ut a eviation f om t previously has been reviewed and fact that no further action (beyond the Commission is not broadening the reposed annual reporting within thirty days) is required. de,finition of cladding materials within cceepted.The reporting is be beved to oe a fair nun as n Ma tha rulmaking.

requirement in the final rule. 11. OtherSuggestedExpansions to e rtYng a d 1 e a need to . C ntinued Use ofDougall- Rule Scope. One commenter believes be aware of changes and error Rohsenow. Five comments that that hydraulic loads occurring dunng a corrections being made to evaluation addressed tlus aspect of the proposed LOCA could cause steam generator models. Therefore, the annual reporting '"I' ** 'I".ed. One commenter tubes to rupture and that the NRC of minor errors remains in the final rule. believed that this correlation should not should resolve steam generator tube One commenter interpreted the use of be permitted without further verification integrity safety issues prior to publishing and should be phased out. Other this rule.

thi words "or in the application of such commenters supported continued use of Steam generator tubes are designed to a model as requiring reporting when f;cility changes (already reportable the correlation subject to the provisions withstand LOCA loads at allowed of the proposed rule. thinning, and there is no evidence to under i 50.59), resulting in model input changes, occur. The NRC position is that no safety contradict this. If anything, the problem concern is created by continued use of would be with inspection techniques to The regulatory language referred to is intended to ensure that applications of the correlation, as long as the evaluation detect the actual tube thinning and model is overall conservative. whether there is an unacceptably high models to areas not contemplated during Therefore, the Commission can not probability that a tube rupture during a lnitial review of the model do not result justify the burden of requiring licensees in errors by extending a model beyond LOCA due to tube thinning is in excess the range that it was intended. The to modify their evaluation models and to of the design basis. However, the risk perform reanalysis. As discussed in from LOCA with concurrent tube Ccmmission does not believe that further clarification of this requirement SECY 8M2, current evaluation models rupture will not be greatly affected by contain more conservatism than just the proposed rule change. As a result of is necessary tnd has not done so in the those required by Appendix K. the commenter's concerns, this issue has final rule. However, error corrections or changes been assigned as a generic issue (CI-Several commenters requested a could alter the conservatism of the 141) to be prioritized by the NRC staff.

further relaxation of the reporting model. Therefore, the Commission The results of the prioritization process requirement by changing the definition believes that it is necessary to ensure will determine if further action is cf significant code errors from 50*F to continued overall conservatism in the required.

s I

100*F. evaluation models as a basis for A second commenter believes that the While ; justification for the 50*F criteria continued use of the correlation. ECCS rule does not adequately address is largely judgmental the NRC believes 'Iberefore, the final rule does not modify a plant's long term decay heat removal l that it is sufficiently large to screen the this requirement except for the capability, and recommends a "short/

code error corrections and changes correction of a typographical error long term integrative analysis which have little safety significance identified by one commenter. approach." Both the existing l while providing a mechanism for timely 8. Uncertainty Evoluotion. The requirements and the proposed rule l reporting of more serious errors and comments received on the uncertainty contain the requirement to provide for changes. Since 50*Fis e threshold for evaluation support the proposed rule, long term cooling subsequent to a reporting and no further action is particularly the flexibility provided by a LOCA. Small increases in power that required pending NRC determination of non-prescriptive requirement. Therefore, may result from the proposed rule safety significance, the Commission has the Commission is publishing the final should not greatly change decay heat retained this criteria in the final rule. rule without modification of this removal requirements following a LOCA One commenter requested requirement. or any other accident or transient. Thus, consideration for allowing that the 9. Acceptance Criteria. The three the issue of decay heat removalis not cumulative effect of several errors and comments received on this topic were materially impacted by this rulemaking.

corrections be applied towards the 50*F all supportive of the existing criteria, as Moreover, any proposed increase in threshold. contained in i 50.46(b), and thus the power resulting from this rule September 29,1995 50-SC-56

4 l

PART 50 o STATEMENTS OF CONSIDERATION promulgation would be approved only realistic evaluation of actual plant of technical review of the issues, citer the licensee demonstrates that response. The large conservatism of enhances public participation in the decay heat removal capacities remain Appendix K served the public wellin process, and provides a complete public cdequate. The Commission is planning 1974 when there was great uncertainty record. Therefore, the Commission has no further action with regard to this in ECCS performance. However, these decided to proceed with the rulemaking issue. conservatism are now known to be as planned.

12. AcceptabilityofModels Approved very large, and there is ne need to "over Finally, this commenter questions the Under SECY-dA472. One commenter regulate" by maintaining this experimental basis for this rule because requests that the rule language be unnecessary margin.This type of full-scale ECCS bypass data is not yet modified to state explicitly that ECCS activity can often result in the available.

Evaluation models that have been expenditure of resources that would be The 2D/3D tests which will provide previously approved under SECY-.33 better spent improving safety in other this important data represent a small 472 continue to be acceptable under this areas. The benefits to safety, while portion of the total research upon which rule. difficult to quantify, are believed to be this rule relies. Significant research on SECY 83472 provides an alternative, substantial. While cost savings may ECCS bypass has already been ccceptable method for developing ECCS have been one factor resulting in the completed in small scale vessels and the cvaluation models. Licensees were still rule change, the Commission believes full-scale work is required only to required, however, to demonstrate that that the conservatism contained in ;he confirm the smaller scale results and svaluation models developed using the acceptance criteria themselves, as well quantify any uncertainty due to scale SECY-63-472 approach complied with as those required in the uncertainty effects. One full-scale ECCS bypass test the requirements of Appendix K to part evaluation required in this rule, are has already been completed under the

50. This final rule explicitly finds that adequate to protect the health and 2D/3D program which showed that more ECCS evaluation models, which have safety of the public. margin exists than expected from the This commenter also cites portions of small scale tests. Completion of the full.

been previously approved as satisfying the 1975 Ceneral Electric Company's scale tests only affects the uncertainties the requirements of Appendix K. remain Nuclear Reactor Study (Reed Report). in the calculations, and reduces them.

acceptable. Therefore, the Commission which claims that there is a lack of (Uncertainties must be addressed by sees no need for further clarification of understanding of phenomena and small this issue, a

licensees in any analysis under the safety margms. revised rule whether 2D/3D results are

13. Comments Received After Comment Period. Six letters dommenting May of the conclusions of the,,, Reed available or not. The Commission Report were valid in 1975 when it was concludes that there is no need to delay on the proposed rule were received written and due to this fact it was the final rule, while awaiting these data.

subsequent to the end of the comment

?hese omme t th ext nt thaYthe discussed in NUREG-1230 has been Section 50.48 Acceptance Criteriafor commentsprovided substantive conducted since the ' Reed Report' was Emergency Con Cooling Systems for information not previously considered. written and has resulted in significant Light WaterReactors.

One commenter believes that the 8 proposed i 50.46(a)(2) expands the I[o*,',*na p We o ow a Section 50.46(a)(1)is amended and discretion of the Director of the Office of redesignated i 50.46(a)(1)(1) to delete the significant margin to the ECCS Nuclear Reactor Regulation (NRR) by acceptance criteria exists, particularly requirement that the features of Section allowing imposition of immediate I f Appendix K to Part 50 be used to for the BWR/6 which was of concern in develop the evaluation model. This effective restrictions on reactor the " Reed Report." The contents of this operation without a prior determination section now requires that an acceptable report have been reviewed by the that such action is required to protect Commission on several occasions most evaluation model have sufficient the public health, safety, or interest. recently in NUREG-1285, and the finding supporting justification to show that the NRC's intent is not to alter the has been made that no new significant analytical technique realistically responsibilities of the Director of NRR safety issues are identified. For these desenbes the behavior of the reactor but to simply retain the description of reasons, the NRC is proceeding with this system during a LOCA. The NRC the scope of the authority that is rulemaking, as proposed. expects that the analytics! technique currently found in i 50.46(a)(1)(v). The same commenter also will, to the extent practicable, utilize Furthermore, the provisions of reatstic methods and be based upon recommends that credit for ECCS 150.46(a)(2) do not specify the margins be taken in the Individual Plant applicable experimental data. The procedure to be followed by the Director Examinations (IPE) and not through amended rule also requires that the of NRR. These procedures are set out in generic rulemaking. uncertainty of the calculation be Part 2 and remain unchanged by this The Commission agrees that plant estimated and accounted for when rulemaking. specific differences may justify the compedng the results of the calculation One commenter believes that the rule application of different margins and that to the temperature limits and other is illegal because it is based solely on these may be addressed through criteria of 5 50.4e(b) so that there is a cost savings considerations and that Individual Plant Examinations. high probability that the criteria would there is nothing wrong with large However, the requirement for licensees not be exceeded.The Commiscion conservatism. to evaluate ECCS performance and meet expects the realistic evaluation model to The Commission disagrees with this the acceptance criteria specified in to retain a degree of conservatism assessment. Safety factors are required CFR 50.46(b) is generic. The Commission consistent with the uncertainty of the to protect the health and safety of the believes that margins that may be calculation. The final rule does not public when uncertainties in plant reduced due to a better understanding of specifically prescribe the analytical a reactor's response to a LOCA should methods or uncertainty evaluation response exist. As these uncertainties are reduced. it is appropriate to modify be applied through a generic rulemaking techniques to be used. However, these safety factors to provide more action because it allows a broad range guidance has been provided in the form 50-SC-57 Septembe@ 1995

y i

l i

l l

PART 50

• STATEMENTS OF CONSIDERATION l

of a Regulatory Guide.8 in SECY-83-472. replaced as described in the following the NRC has found acceptable an report is to be filed within one year of paragraphs. discovery of the error and must be approach for estimating the 95th Section 50.46(a)(2) has been revised to reported each year thereafter until a p;rcentile of the probability distribution. Indicte that restrictions on reactor This percential is considered adequate revised evaluation model or a revised operation may be imposed by the evaluation correcting minor errors is is meet the high level of probability Director of Nuclear Reactor Regulation, approved by the NRC staff.

r; quired by the rule. It is also recognized if the ECC cooling performance Significant errors require more timely that the probability cannot be evaluations are not consistent with the attention since they may be important to determined using totally rigorous requirements of i 50.46(a)(1)(i) and (ii). the safe operation of the plant and raise m thematical methods due to the This section has been added to retain questions as to the adequacy of the complexity of the calculations. similar requirements that have been overall evaluation model. This final rule However, the NRC requires that any deleted from 5 50.46(a)(1)(i) by this rule defines a significant error or change as simplifying assumptions be stated so revision. This section does not specify one which results in a calculated peak that the Commission may evaluate them the procedures to be followed by the fuel cladding temperature different by to ensure that they are reasonable. The Director. These procedures are found in more than 50 *F, or an accumulation of NRC has independently developed and Part 2 and are unchanged by this errors and changes such that the sum of exercised a methodology to estimate the rulemahng. b awe magnM M b uncertainty associated with its own The c arrent rule contains no explicit m changes i gre er than 50 i

thermal-hydraulic safety codes. This r ents conc r r and p ] erst eport sys) is methodology is described in the ]uire g g; ,

Compendium of ECCS Research. 8 models are discovered or changes are changes.This definition of a significant This document also provides referenes made to evaluation models. However' change is based on NRC's judgment current practice has required reportins concerning the importance of errors and [

to the large body of relevant thermal- l of errors and changes and reanalyses changes typically reported to the NRC in '

hydraulic research, documents f4C studies on the effects of reactor power with the revised evaluation models. This the past. This final rule revision also increases on risk, and provides final rule explicitly sets forth allows the NRC to determine the background information on the 2CCS requirements to be followed in the event schedule for reanalysis based on the of errors or changes. The definition of a importance to safety relative to other rule. Whilw this method has not been apphcant or licensee requirements. l r; viewed for acceptability from the significant change la currently taken from Appendix K.Section II.1.b which Enors or changes that neult in the sinndpoint of safety licensing,it may calculated plant performance exceeding provide additional guidance on how the defines a significant change as one which changes calculated cladding any f the criteria of I 50.48(b) mean uncertainty may be quantified. In that the plant is not operating within the l temperature by more than 20 'F, addition to providing guidance to "9"N"#' I O' N'd "'

• Industry, this work was undertaken to The revised 6 50.46(a)(3) states specif;c requirements for reporting and require immediate reporting as required provide a proof of principle and a tool to independently audit submittals. reanalyses when errors in evaluation ' * '

• models are discovered or changes are imm a Hep 8t g plant Wo Appendix K.Section II. " Required mpliance with ! $0.46.

made to evaluation models. It requires Documentation," remains generally that all changes or errors in approved Appendix K ECCSEvaluation Models applicable, with only minor revisions evaluation models be reported at least mide to be consistent with the amended Amendmente have been made to rule.

annually and does not require any further action by the licensee until the Appendix K.Section I.C.5.b, to modify A new paragraph (ii) has been added the post CHF heat transfer correlations error is reported. Thereafter, although listed as acceptable. The "McDonough" to i 50.46(a)(1) to allow the features of reanalysis is not required solely because reference has been replaced with a more Section I of Appendix K to be used in of such minor error, any subsequent recent paper by the same authors evaluation models as an alternative to calculated evaluation of ECCS performing the uncertainty evaluation entitled "An Experimental Study of )

performance requires use of a model partial Filr2 Boiling Region With Water specified in the amended i 50.46(a)(1)(i). with such error, and any prior errors. l This method would remain acceptable at Elevated Pressures in a Round i corrected. The NRC needs to be Vertical Tube" which is more generally because Appendix K is conservative apprised of even minor errors or I

with respect to the realistic method available and which includes additional l changes in order to ensure that they data.

pr: posed in the amended I 50.46(a)(1)(1). agree with the applicant's or licensee's The heat transfer correlation of This would allow both current and assessment of the significance of the Dougall and Rohsenow, listed as ari l future applicants and licensees to use error or change and to maintain acceptable heat transfer correlation in cognizance of modifications made j

cxisting evaluation models if they did Appendix K paragraph I.C.S.b, has been  ;

n:t need or desire relief from current subsequent to NRC review of the removed, because research performed eperating restrictions. l evaluation model. Past experience has since Appendix K was written has j in i 50.46 paragraphs (a)(2) and (3) shown that many errors or changes to shown that this correlation overpredicts i

have been revised to eliminate portions evaluation models are very minor and heat transfer coefficients under certain '

the burden ofimmediate reporting of those paragraphs concerned with conditions and therefore can produce historicalimplementation of the current cannot be justified for these minor nonconservative results. A number of >

file. These provisions have been errors because they do not affect the applicants and licensees currently use immediate safety or operation of the the Dougall-Rohsenow correlation in I

plant. The NRC therefore requires approved evaluation models. The NRC 1

Fe enc Co 1 yo e : "l e [s etpcperiodic reporting to satisfy NRC's need has concluded thet the continued use,of

) 1.1s7. to be apprised of changes or errors this correlation can be allowed. This is  ;

} '"Compeneum of ECCS Research for Reshstic without imposing an unnecessary appropriate (even though parts of the

I,CCA Analysm." NUREG-1230. T13P.

burden on the applicant or licensee. This approved evaluation model. Dougall-i l

September 29,1995 50-SC-58 i

1 PART 50 o STATEMENTS OF CONSIDERATION l Rohsenow, are known to be the conservatism required to account for either to tailor the power shape within nonconservative) because the existing overall uncertainties in the calculations. the reactor or to increase the total -

evaluation models are known to contain Appendix K. Section ll.1.b. has been power. Changing the power shape a large degree of overall conservatism removed since this requirement has without changing the total power has a even while using the Dougall-Rohsenow been clarified in the amended negligible effect on the environmental correlation. This large overall l 50.46(a)(3). Likewise. Appendix K. impact. The total power could also be conservatism has been demonstrated Section 11.5. has been amended to i. eased. but is expected to be through comparisons between account for the fact that not all increased by no more than about 5% due evaluation models will be required to to hardware limitations in existing I svaluation model calculations and calculations using NRC's best-estimate use the features of Appendix K. Section plants. This 5% power increase is not computer codes. Thus, requiring that the 1. These minor changes to Appendix K expected to cause difficulty in meeting l do not affect any existing approved the existing environmental limits. The l rpplicants and licensees remove the Dougall-Rohsenow correlation from evaluation models since the changes are only change in non-radiological waste either " housekeeping" in nature or are will be an mcrease in waste heat their current evaluation models cannet rejection commensurate with any be justified as necessary to maintain changes to " acceptable features." not I

" required features." increase in power. For stations s fety.The stipulation that the Dougsll.

perating with an open (once through; Rohsenow correlation will cease to be Availability of Documents cooling system. this additional heat will teceptable for previously approved 1. Cop.ies of NURECs 1230 and 1285 be directed to a surface water body.

evaluation models applies only when i may be purchased from the Discharge of this heat is regulated under I changes to the model are made which Superintendent of Documents. U.S. the Clean Water Act administered by reduce the calculated peak clad Government Printing Office. P.O. Box the U.S. Environmental Protection temperature by 50 *F or more. However. 37082. Washington DC 20013-7082. Agency (EPA) or designated state the requirement to report any changes or Copies are also available from the agencies. it is not intended that NRC culmination of changes such that the ]

National Technical Information Service, approval of increased power level sum of the absolute magnitudes of the affects in any way the responsibility of

$285 Port Royal Road. Spnngfield. VA respective temperature changes is 22161. A copy is also available for the licensee to comply with the greater than 50 *F. still applies.

• public inspection and/or copying at the requirements of the Clean Water Act.

A new Section I.C.5.c has been added NRC Public Document Room. 2120 L The environmental assessment and to Appendix K to state the

• Street NW., Washington. DC 20555. finding of no significant impact on Commission,s requirements regarding 2. Copies of SECY-83-472. an which this determination is based are continued use of the Dougall-Rohsenow information report entitled " Emergency available for inspection at the NRC correlation in existing evaluation Core Cooling Systems Analysis Public Document Room. 2120 L Street models. Evaluation models which make Methods." dated November 17,1983. is NW., Washington DC. Single copies of use of the Dougall Rohsenow correlation available for inspection and copying at the environmental assessment and the and have been approved prior to the the NRC Public Documents Room. 2120 L finding of no significant impact are effective date of this rule may continue Street NW Washington, DC 20555. available from L M. Shotkin. Office of to use this correlation as long as no Single copies of this report may be Nuclear Regulatory Research. U.S.

changes are made to the evaluation obtained by writing L M. Shotkin. Office Nuclear Regulatory Commission, model which significantly reduce the of Nuclear Regulatory Research. U.S. Washington DC 20555. telephone (301) current overall conservatism of the Nuclear Regulatory Commission. 492-3530.

evaluation model. If the applicant or Washington. DC 20555.

licensee submits proposed changes to an 3. Regulatory Guide. "Best Estimate Paperwork Reduction Act Statement cpproved evaluation model. or submits Calculations of Emergency Core Cooling This final rule amends information corrections to errors in the evaluation Systems Performance." Task RS 701-4. collection requirements that are subject model which significantly reduce the may be obtained by writing to the to the Paperwork Reduction Act of1980 existing overall conservatism of the Division of Information Support (44 U.S.C. 3501 et seq.). These reporting model. continued use of the Dougall- Services. U.S. Nuclear Regulatory requirements were approved by the Rohsenow correlation under conditions Commission. Washington. DC 20555. Office of Management and Budget where nonconservative heat transfer 4. The Paraphrased Summary of (Approval Number 3150-0011).

coefficients result would no longer be Public Comments on the ECCS Rule is Regulatory Analysis acceptable. For this purpose, significant available for public inspection at the reduction in overall conservatism has NRC Public Documents Room. 2120 L The Commission has prepared a been defined as a " net" reduction in Street NW Wohington, DC 20555. regulatory analysis for this final calculated peak clad te- ture of at regulation. The analysis examines the ould have nading of No Sign 1Acant Environmental costs and benefits of the alternatives least 50*F from that w been calculated using axledag Impact: Availability considered by the Commission. The evaluation models. A reduction in The Commission has determined regulatory analysis is available fo-calculated peak clad temperature could under the National Environmental Policy inspection and copying for a fee at the potentially rssult in an increase in the Act of 1989. as amended, and the NRC Public Document Room. 2120 L cctual allowed peak power in the plant. Commission's regulations in Subpart A Street NW., Washington. DC. Single An increase in allowed plant peak of 10 CFR Part 51, that this rule is not a copies of the analysis may be obtained power with a known nonconservatism in major Federal action significantly from L M. Shotkin. Office of Nuclear the analysis would be unacceptable. affecting the quality of the human Regulatory Research, Washington DC This definition of a significant reduction environment and therefore an 20555, telephone (301) 492-3530.

in overall conservatism is based on a environmental impact statement is not Regulatory Flexibility Certification judgment regarding the size of the required.The primary effect of the rule existing overall conservatism in is to allow an increase in the peak local As required by the Regulatory evaluation model calculations relative to power in the reactor. This could be used Flexibility Act of 1980. 5 U.S.C. 005(b).

50-SC-59 September 29,1995 I i

I

l PART 50

• STATEMENTS OF CONSIDERATION the Commission certifies that this rule Whether or not a licensee or apphcant will not have a significant economic reduced. This effect is not expected to chooses to use realistic analysis, be very significant.

impact upon a substantial number of complete with an uncertainty analysis. 5. installation and continuing costs cmall entities. This mle affects only the each licensee must comply with the licensing and operation of nuclear associated with the backfit, including requirement to report changes to their the cost offacility down times or the power plants. The companies that own evaluation models (i.e., less than 50*F these plants do not fall within the scope cost of construction delay.

change in calculated peak cladding LOCA considerations resulting from of the definition of "small entities" set temperature) annually to the NRC. in forth in the Regulatory Flexibility Act or the present rule are restricting the addition, significant changes (those optimum production of nuclear electric the Smtll Business Size Standards set which have a greater than 50*F change cut in regulations issued by the Small Power in some plants. 'Ihese restrictions in calculated peak cladding can be placed into the following three Business Administration in 13 CFR Part temperature) have to be reported within categories:

121. Since these companies are 30 days.

dominant in their service areas, this rule (1) Maximum plant operating power,

3. Potentic/ change in risk to the (2) Operational flexibility and does not fall within the purview of the Act. publicfrom the occidentaloffsite operational efficiency of the plant, and release of radioactive materials. (3) Availability of manpower to work loc 1p er within t e act rc and eefc f E mle wd van from A backfit analysis is not required by possible increases in total power. Power plant to plant. Some plants may realize to CFR 50.109 because the rule does not savings of several million dollars per r2 quire applicants or licensees to make increases on the order of 5 will have an a change but only offers additional insignificant effect on risk. One effect of year m fuel and operating costs, increased power could be to increase Sign ant ster economic benent options and provides a clarification and w uld be rea zed by plants able to relaxation of existing reportin8 the fission product inventory. A five ercent power increase would result in a increase total power as a result of this c:.quirements. Nonetheless, the factors [ess than five percent increase in fission 0"'I'"I' "8" "I 7 ***'Y * #

in 10 CFR 50.100(c) have been a~ n alyzed *'

for the entire rule, products. Thus, less than five percent more fission products might be reler sed

",,{' , ,

' p '"

y pfa'c e cos

1. Statement of the specific dbjectives during core melt scenarios and savings for a five percent power upgrade that the backfit is designed to achieve. would vary between 18 and 127 million potentially released to the environment The objective of the rule is to modify during severe accidents. dollars depending on the plant.

10 CFR 50.46 and Appendix K to permit The rule still requires the fuel rod Additional information concendng these the use of realistic ECCS evaluation peak cladding temperature (PCT) remain potential cost savings are included in models. More realistic estimates of below 1200*F. Reactors choosing to the regulatory analysis.

ECCS performance, based on the The costs associated with the new improved knowledge gained from recent increase power by about five percent will be operating with less margin reporting requirements are deemed to be research on ECCS performance, may minimal. Although the existing between the PCT and the 2200*F hmit remove unnecessary operating than previously.The increased risk Appendix K has no official reporting restrictions. Also experience with the requirements, paragraph Ilt.b was previous version of I 50.46 has represented by this decrease in margin and increase in fission product interpreted by the staff to require a demonstrated that a clearer definition of reanalysis and report to NRC when inventory is negligible and falls within reporting requirements for changes and significant changes are made which errors is very desirable. the uncertainties of PRA risk estimates. change the peak cladding temperature In addition, other safety limits, such as by more than 20 'F. Therefore, this rule

2. General description of the activity departure from nucleate boiling (DNB).

that would be required by the licensee and operationallimits, such as turbine change, by changing the definition of or applicantin order to complete the significant changes to 50 *F. is actually a design, willlimit the amount of margin relaxation of current practices. The backfit. , reduction permitted under the rule. The The amendment allows alternative annual reporting of changes that are not rule could also potentially reduce the significant is not viewed by the NRC as m:thods to be used to demonstrate that risk from pressurized thermal shock by the ECCS would protect the nuclear allowing the reactor to be operated in a a ma}o burden since no other action is r; actor core during a postulated design required.

manner which reduces the neutron 6. Thepotentialsofetyimpact of b sis loss-of-coolant accident (IDCA). fluence to the vessel. changes in plant or operational While contiuuing to allow the use of

4. Potentialimpact on todiological complexity including the effect on other current Appendix K methods and exposun to facility employees. proposed and existing regulatory requirements, the rule also allows the Since the primary effect of the rule requirements.

use of more recent information and involves the calculational methods to be knowledge currently available to There are safety benefits derivable used in determining the ECCS cooling from alternative fuel management demonstrate that the ECCS would performance, it is expected that there perform its safety function during a schemes that could be utilized. The will be an insignificant impact on the higher power peaking factors that would 14CA. If an applicant or licensee elects radiological exposure to facility be allowed with the final rule provide ta use a new realistic model they will be employees. Because of the reduced required to provide sufficient supporting greater flexibility for fuel designers IDCA restrictions resulting from the when attempting to reduce neutron flux justification to validate the model and l new calculations it is possible for the at the vessel wall. This can result in a include comp 6risons to experimental plant to achieve more efficient operation corresponding reduction in risk from data and estimates of uncertainty,in and improved fuel utilization with pressurized thermal shock.

cccounting for the uncertainty, the improved maneuvering capabilities. As analysis would have to show, with a The reduced cladding temperatures a result,it is conceivable that there that would be calculated under the high level of probability, that the ECCS could be a reduction in radiological revised rule offers the possibility of performance criteria are not exceeded. exposure if the fuel reloads can be other design and operational changes September 29,1995 50-SC-60 l

t - - __

l PART 50 o STATEMENTS OF CONSIDERATION that could result from the lower 8. Thepotential mpact of differences suesasAmy:The Nuclear Regulatory c:lculated temperatures. ECCS in [acility type, design or age on the g gulations equipment numbers, sizes or televancy andpracticality of the m]idn a amen surveillance requirements might be backfit. emergency planning and preparedness r:duced and still meet the ECCS design The degree to which the rule would requirements are needed for fuel loading criteria (il not required to meet other affect a particular plant depends on how and low power testing of nuclear power limited the plant is by the LOCA plants.The rule itself will now require licInsing requirements). Another option restrictions. General Electric Company may be to increase the diesel / generator NRC findings on the licensee's (GE) plants do tend to be limited in stirt time duration.

In summary, the effe:t of this rule on ,P,'uld benefit f om re! e r m LhCA a c en d affect pe onson site. The Commission's prior practice of stfety wonel have both potential restrictions. However, this relief is positive and negative aspecs.The considering certain offsite elements of alread available for most GE plants licensee's plans has been modified and potential for reduction of ECCS system throu the recently approved SAFER codified in this regard to provide that c:pability in existing or new plants is evaluation model. Any additionalrelief NRC findings will be required before present. However, several positive due to a rule change would be of little further benefit. Westinghouse (W) fuelloading or low power testing in tspects may also be realized under the cocedination with offsite personnel and final rule. The net effect on safety wou'd lants would appear to directly benefit be plant specific. However, the from relaxation of LOCA limits, W agencies so that necessary resources can be applied on site for mitigating and probability of a large break LOCA is so plants represent the largest number of c ontaming accidents, and so that offsite low that the choice of best estimate plants, with 47 plants operatin;; and to additional plants being constructed. w agencies may be kept informed of plant vs,rsus Appendix K would have little events. The rule wf. c ao change the m tct on public risk. Indicates that most of these plants are limited by LOCA considerations. The prior practice, neve .acluded in the

7. The estimated resource burcen on prior rule itself, of reviewing plans for the NRC associated with theproposed potential benefit for plants of B&W and CE design is uncertain at this time, piompt public notification it the event backfit; and the availability of such of an accident.This practice of resources.

9. Whether theproposed backfit is interim orfinolandifinterim, the reviewing an offsite element oflicensee The major staff rescurces required emergency plans that has no onsite justification for imposing the proposed under the final rule are to review the bocAfit on on interim basis, application is being discontinued as not r:alistic models and uncertain'ty rxessary for public safety.The rule The rule, when made effective, will be ,

enalysis required by the reviss CCCS does not change the emergency planning i Rule. Based on previous expertace with in final form and not interim form. It will continue to permit the performance of requirements that must be satisfied the General Electric Company 4 SAFER ECCS cooling calculations using either before full power operation can be model and the learning that has resulted authorized. No new requirements are from these efforts,it is estimated that realistic models or models in accord with Appendix K. being imposed by the rule beyond those approximately one staff year would be that have been previously required by required to revLw each generic model Ust of Subjects in 10 CFR Part 50 rule and by prior NRC practice. The rule submitted.There are four major reactor Antitrust, Class.fied i information, Fire makes clear that no offsite elements of vendors (CE already has a revised prevention, incorporation by reference, the applicant s emergency plan. other evaluation model approved under the Interg vernmentalrelations, Nuclear than those set forth in this revised rule, 3 existing Appendix K for both le' oump power plants and reactors, Penalty, need be consideteo in connection with d I d' hadiation protection, Reactor siting low power licensing.

upda e t tr r t todo o y under this criteria. Reporting and Recordkeeping terscTive DATE: October 24,1988.

new rule) and several fuel suppliers and " **""'

utilities which perform their own Posi rustTHEn eseronesafices coe Tact:

cnalyses and potentially might submit For thc reasons set out in the Carole F. Kagan. Office of the General preamble and under the authority of the Counsel, U.S. Nuclear Regulatory generic models for review. However,it Atomic Energy Act of 1954, as amended.

is expected that only 3 or'4 generic Commission. Washington, DC 20555; the Energy Reorganization Act of1974, Telephone (301) 492-1632, or Michael T.

models would be subrnitted since not all as amended, and 5 U.S.C. 552 and 553.

plants would benefit from this rule. Jamgochten. Office of Nuclear  !

the NRC is adopting the following Regulatory Research, U.S. Nuclear  ;

Thus, about 3-4 staff years would be amendments to 10 CFR Part 50. Regulatory Commission. Washington.

r quired to review the expected generic DC 20555; Telephone (301) 492-3918.

33 Nt 3sess pprov d t e plan specWe ew is suPPLEssEsf7AnY lecrostesAftoet: ,

Putillahed 9/23/ss  !

very short. In addition, several vendors /2a/84 L Background tre currently planning to subaft realistic models in conjunction with the use of 10 CFR Port 50 On Me 9,1988, the Commission SECY-83-472. Therefore, staff resources publishe[In the Federal Register (53 FR e

mo e ein any vent Si t ese models Emergency % enq t

w3 o ,, ,gedoree r$y 1 PreparedneseReC mW i

would not change as a result of the what emergency planning and Nuclear Pourer Plant Fuel Loading and preparehese requirements are needed revised ECCS rule, there should be no Low Power Testing net increase in resources required over for fuelloading and low powc? testing of that already planned to be expended. In nucles power plants. As detailed in the Acesscv: Nuclear Regulatory notice of proposed rulemaking.10 CFR summary, while it is difficult to estimate Commission. 50.47(d) as promulgated on July 13,1982 acct. rate!y,it is expected that the rule change will have a small overall impact (47 FR 30232) provided that only a aerio,e pin,i ruj,, finding as to the adequacy of an on NRC resources, September 29,1995 50-SC-61

w ATTACHMENT 7 COMMISSION BRIEFING CN SAFETY EVALUATIONS, FSAR UPDATES, AND INCORPORATION OF RISK INSIGHTS (JUNE 4,1998) f l

L___________ __. _ . _ _ _ _ _ _ . _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ . . _ - - . _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - _ _ - _ _ _ _

~. .-- _ _ .

4 4

1 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3 ...

4 BRIEFING ON 5 SAFETY EVALUATIONS, FSAR UPDATES AND 6 INCORPORATION OF RISK INSIGHTS 7 ***

8 PUBLIC MEETING 9 ***

10 11 Nuclear Regulatory Commission 12 One White Flint North 13 Rockville, Maryland 14 15 Thursday, June 4, 1998 16 17 The Commission met in open session, pursuant to 18 notice, at 2:05 p.m., the Honorable Shirley A. Jackson, 19 C.'.a i rma n , presiding.

20 21 COMMISSIONERS PRESENT:

22 SHIRLEY A. JACKSON, Chairman of the Commission 23 GRETA J. DICUS, Member of the Commission 24 NILS J. DIAZ, Member of the Commission 25 EDWARD McGAFFIGAN, JR., Member of the Commission 2

1 STAFT AND PRESENTERS SEATED AT THE COMMISSION TABLE:

2 ANNETTE VIETTI-COOK, Assistaint Secretary of the 3 Commission 4 STEPHEN G. BURNS, Acting General Counsel 5 RALPH BEEDLE, NEI 6 HAROLD RAY, NEI 7 TONY PIETRANGELO, NEI 8 HUGH THOMPSON, NRC 9 DAVID MATTHEWS, NRC 10 SAMUEL COLLINS, NRC 11 MARK SATORIUS, NRC 12 13 14 15 16 17 18 19 '

20 21 22 23 -

24

• 25 3

1 PROCEEDINGS 2 (2:08 p.m.)

3 CHAIRMAN JACKSON: Good afternoon, ladies and 4 gentlemen. The purpose of this meeting is for the 5 Commission to be briefed by the Nuclear Energy Institute and 6 the NRC staff on proposed regulatory guidance related to the f 7 implementation of 10 CFR 50.71(e), which addresses updates 8 to final safety analysis reports, and proposed changes to 10 9 CFR 50.59 entitled Changes, Tests and Experiments.

10 The Commission recently approved making publicly 11 available a draft generic letter providing interim guidance 12 on the implementation of 10 CFR 50.71(e). The Commission is l IGf40 6/10'98 7:39 AM l

Wi&llbij{-12Cfp ---- - - - -

13 considering approving the staff's request to seek public 14 comment on this paper.

15 Concurrently, the staff is working to address 16 Commission direction on a revision to 10 CFR 50.59 detailed 17 in a staff requirements memorandum resulting from 18 SECY-97-205.

l 19 As a result of Co mission activity in this area, 20 NEI has requested an opportunity to brief the Commission on 21 its own activities in these areas and to offer ideas and 12 comment for Commission consideration.

23 Consistent with our stated commitment to involve 24 stakeholdtes in the regulatory process, the Commission is 25 interested in obtaining and considering the views of NEI on 4

1 these matters in an effort to develop the most fully 2 informed decisions possible.

3 We also look forward to hearing from the staff on 4 the status of their efforts, their opinions on the NEI 5 proposals made here, and the basis for their recent reply to 6 the Commission on 10 CFR 50.59 changes documented in a staff 7 memorandum dated May 27, 1998, which is publicly available.

8 It is cur hope there will be frank, cpen exchange 9 of the issues before us. Toward that end, I would encourage 10 both NEI and the NRC staff to provide real world examples of 11 the policy issues they discuss, but not just trivial or 12 anecdotal. Too often briefings on these and similar issues 13 become so philosophical and programmatic in nature that 14 connections between policy and field implementation is los-15 +

Unless any of my colleagues have any opening 16 comments they wish to make, Mr. Beedle or Mr. Ray, whoever 17 is leading the discussion.

18 MR. BEEDLE: Chairman Jackson, thank you very 19 much. Commissioner Diaz, Commissioner Dieus, Commissioner 20 McGaffigan. We appreciate the opportunity to talk with you 21 today. The matter of 10 CFR 50.59 and the FSAR update rule 22 are both very significant to the industry as well as the 23 Commission staff. A considerable amount of time and energy 24 is devoted to these two topics. As you well know, the 50.59 25 is probably the most exercised rule in the arsenal of 5

1 regulations that we deal with.

2 This afternoon we have Mr. Harold Ray, the 3 executive vice president, Southern California Edison. He's 4 also chairman of the NEI Regulatory Process Working Group 5 that was formed about a year ago to help us address and 6 focus on the issues of 50.59 and FSAR design basis, and so 7 forth.

8 We also have Tony Pietrangelo, director of 9 licensing with NEI, here with us today.

10 We are prepared to discuss the 50.59. We have 11 over the course of the last year had quite a bit of 12 interaction with both the Commissioners and the staff on 13 this topic.

14 We appreciate the publication of the draft 15 document on FSAR. That has certainly helped us put our 16 commente in perspective today. We think tt.at sets a good 17 tone on how to deal with those issues rather than waiting 18 for a public comment period, at lease to get them out and 19 make them available to digest and understand.

20 With that, I would like to turn to Harold for some l 21 comments from his perspective as an executive in the utility 22 to discuss the 50.59 FSAR issues.

23 Harold.

24 MR. RAY: Thank you, Ralph.

25 Chairman Jackson, in the interest of saving time kf40 6/10987:39 AM

__-~---~-u 6

1 and in the spirit of the dialogue that you invited, I'm 2 going to skip over three things on my talking points here, 3 namely, the introduction, the history of this sordid affair, 4 and the rationale for the industry initiative.

5 CHAIRMAN JACKSON: Is that s-o-r-t-e-d or 6 s-o-r-d-i-d?

7 [ Laughter.]

8 MR. RAY: The latter.

9 CHAIRMAN JACKSON: Just wanted to be sure.

10 MR. RAY: And also the rationale for the industry 11 initiative that exists. If any of the Commissioners have 12 questions about that, I'll be glad to comment on them, but I 13 think that we need to cut to the essence of what this is all 14 about as quickly as we can, and I would like to do that.

15 I don't want to skip over, however, recognizing 16 and, I hope you will accept, commending the Commission for 17 taking up this issue as you have.

18 I had the opportunity to make some comments at the 19 information conference not so long ago. Commissioner 20 McGaffigan was there, I believe. I tried to underscore the 21 fact that I thought that the Commission vote sheets on 22 3ECY-97-205 were very thoughtfully reflecting on the issues 23 that we all faced. Like us, I don't think any of you had 24 the magic solution, and so we are engaged now in a process 25 of trying to find what the best outcome is that we can l 1

7 i 1 cobble together here. j 2 It was vary instructive for us to eee your 1 3 deliberations on this. I commented at the time that the 4 Commission secretary doesn't often come in for comment, but ]

5 I thought a terrific job was done in trying to gather all l 6 that together into some summary of where this all stood 1 7 among the Commissioners.

8 Having said that, let me now dive into the essence 9 of this. Tony will be making the presentation that we have 10 prepared. Ne percicipated in its development and are 11 prepared to answer questions as we go or at the end.

12 I think the thing that would be most useful for me 13 to comment to you on before Tony speaks is the issue of 14 scope. In the May 27th memo that you referred to that is 15 something which is proposed to be deferred by the staff for 16 attention later. In a letter that Ralph had sent the 17 Commission we suggested that it was timely to take that 18 issue up now. Tony will indicate we are certainly prepared 19 to address it as a second step in the process that we face 20 here in the interest of addressing the other issues that are il on the table and ripe for decision.

22 But I do want to comment on it briefly in terms of 23 the importance of it, and particularly because it connects 24 to the generic letter that you mentioned, the 50.71(e) 25 issue.

8 1 Perhaps the easiest way for me to illustrate the 2 point that I want to leave with you is that in talking about 3 50.71(e) -- and I think I've made this comment to each of 4 you separately -- the role of the SAR in defining what the 5 scope of 50.59 is naturally arises. As you know, we are l 6 interested in not having the SAR constitute the scope of

} 7 50.59 and believe that it has traditionally not been the l

8 case that it did. But we find ourselves now at a point at 9 which that is in fact the case.

10 I want to just extract one sentence from the 11 forwarding letter to you o: the generic letter that you 12 referred to to illustrate the point. The staff is l3cf40 6/10'98 7:39 AM i

l _

.--__~- _ y 13- commenting in the context of the SAR that drawings might be 3d simplified in the SAR of the future. That is one of the changes that might be made. They comment that the effect of this guidance is to reduce the scope of 50.59 changes to

.. Some minor components would no longer be required to be 18 evaluated pursuant to 50.59 as they would no longer be "as 19 described in the safety analysis report."

20 So speaking to a detail on a drawing, you see, the 21 notions persist that by changing what 50.71(e) requires to 22 be in the SAR we are defining or eliminating, adding to, '

23 taking away from what is required to be addressed in 50.59, 24 and we just see that as a very significant problem and one 25 that I know you all recogrsize as well. We do need to be 9 l 1 committed, I think, to address that.

2 I can tell you as somebody who in past roles has 3 done a lot of writing of what is in a SAR that the intention 4 never was, I think even up to the present day, but surely i 5 over most of the time when SARs were being developed, that )

6 they serve as the definition of what was subject to the  !

7 control of 50.59.

8 They were in fact a convenient place to put 9 information. An excellent place as a matter of fact, rather 10 than have it distributed in many, many locations that were 11 hard to access, particularly when you are facing hearings 12 and other proceedings associated with issuance of a license.

13 We would just put it in the SAR. So a lot of things wound 14 up there.

15 I don't think that the regulatory guide that 16 defines what f.eeds to be in a SAR was written with that 17 intent either. Yet we find ourselves in that place.

18 You asked for a real world, not a trivial or 19 anecdotal example. This may be anecdotal. I hope it's not 20 trivial, but it's certainly real world.

21 I have found myself in the position of spending an 22 enormous amount of time, by my standard anyway, dealing with 23 noncompliance having to do with one of these details that I 24 think the staff is s gesting don't need to be on the P&ID; 25 in the SAR. I was dealing with it because it was a 10 1 violation of 50.59.

2 I won't go into the details unless you want me to, 3 but we had made a change that in our judgment did not 4 require Commission approval, did not involve an unreviewed 5 safety question, and so on, and that became a matter of 6 debate. It had to do with an orifice in a vent line and 7 whether it was removable entirely or whether it was 8 removable by use of a gate valve. And that's all it had to 9 do with.

10 Anyway, that is a tiny example, yet one that I 11 have personally been involved in, of why we are concerned 12 that the enormous amount of information that is in the SAR 13 can become a source of distraction to the Commission and to 14 us as licensees as we try and go about getting our job done 15 of assuring the safe operation of the plant.

16 I think that's all I need to say on that point. I 17 just wanted to underscore to you that it is a piece of this 18 overall picture that is equally important to the one of l 19 increase in consequences, and so on, that we are also l 20 dealing with here today and that I'm not going to speak to; t

21 I'll let Tony do that. But I wanted to underscore to you 22 the importance in our j'>dg.nent that if we defer giving 23 attention to that to a later time that it not be a too much 24 later time, because this issue is an area of uncertainty and

! 25 also one where I think improvement in the process can be i

I I

I 6/10 98 7:39 AM 4 cf 40

_____ _ _ - ~i l

l '

11 1 made.

2 CHAIRMAN .ACKSON: Thank you.

3 Commissioner McGaffigan.

4 COMMISSIONER McGAFFIGAN: I would like to ask a 5 question about the change. At the reg info conference you 6 were very strongly arguing for doing it in the first phase.

7 Is it that you now have seen these documents, particularly 8 the FSAR update guidance, the generic letter?

9 I know you are going to testify that you think 10 that it doesn't have to go out for formal public comment 11 because you are willing to change 98-03 and to accommodate, 12 and you think that the better use of resources, you're going 13 to testify, is that we use our resources to endorse your l 14 guidance.

15 Has the May 27th meeting with the staff and having 16 this document on the table and having some discussions 17 allayed some concerns and so now you are more comfortable 18 with the two-step process that was in the original 19 Commission SRM? i 20 MR. RAY: Commissioner McGaffigan, I would put it f 21 this way. At the margin we are persuaded that we are all 22~ committed to take this second step and therefore, since it l, 23 is the desire, I believe, of the Commission and certainly of '

24 the staff to take the first step, that we don't want to 25 appear as objecting to doing that unless it were to be 12 1 perceived that somehow that was going to be the only chance 2 we nad to address 50.59. <

3 On the other hand, I would say my concerns aren't j 4 allayed by this generic letter for the ery reason that I-L 5 said. It seems to underscore the notion that we find 6 troubling, that is, that the SAR, and as far as we know 7 everything in the SAR, is subject to 50.59 requirements. 1 8 At the limit, as I think Tony will say, we don't i 9 know how to keep that from making the SAR de facto just an 10 enlarged version of the tech specs.

11 MR. PIETRANGELO: That wasn't the reason. Last 12 week's meeting utsn't the reason why we are more amenable to 13 the two-step process. I think there are really two reasons 14 behind it.

15 One, I think we got a sense through individual 16 visits with you all that you are committed to do this.

17 Secondly, we think we have some mogentum 18 established by the Commission's actions to take some quick 19 action on some things that we think we are fairly close to.

20 In our view, trying to do scope at the same time would 21 probably prolong that, and we don't know for how long. So 22 we don't want to lose that momentum. We want to keep that 23 going. Again, with a commitment to look at that second step 24 in a very serious way, we are satisfied that that is the 25 right way to go.

13 1 MR. RAY: Yes. I think if anything the need is 2 more clear as a result of the generic letter statement that 3- I referred to. It was just an indication of where we have a 4 real concern.

5 MR. PIETRANGELO: We have to do the 50.71(e) 6 anyway. So tuat was really not the reason at all.

7 (Slide.).

8 MR. PIETRANGELO: First, I want to acknowledge we 1 9 received COM SECY-98-013 yesterday afternoon and got a j 10 chance to look at that a little bit and digest it. We'll l l 11 try to talk a little bit-abrn some of those views and 12 incorporate that during the : :esentation.

'5cf40 6/10/98 7:39 AM

--~.-__y 13 We do want to do a quick overview on the 14 Commission's SRM on SECY-97-205 and talk about the use of 15 acceptance limits.

16 One issue we added to this presentation that we 17 hadn't planned to talk about was the design basis 18 interpretation, but it does relate, we think, to some of the 19 issues we are going to discuss this afternoon.

20 Talk about the enforcement discretion provision in 21 the SRM.

22 Talk about the draft FSAR update guidance, the 23 draft generic letter, as well as Draft 99-03, and then 24 finish with the scope of 50.59.

25 [ Slide.]

14 1 MR. PIETRANGELO: The main bullets in the SRM were 2 to expedite this rulemaking, to clarify the threshold 3 criteria as we mentioned before; the enforcement discretion 4 prior to the rule change, and then to reconcile the two FSAR 5 update guidance documents.

6 We do want to mention one , art of COM SECY-98-13 7 dealing with accidents of a different type as well as 8 malfunctions of a di'ferent type. We stand by the proposal 9 we made in our November 14 letter on the language that is 10 appropriate for that criteria.

11 There is one p6rt of the COM SECY that we started 12 discussing this morning really and we're not sure we agree 13 with the staff on. I think we have to think about it more, 14 but f.t's really a higher lesel concern, and that's whether 15 50.5) is a procedural standard versus a safety standard. In 16 the sense that it's a standard for determining whether prior 17 Commission review or approval is needed, it is procedural in 18 that sense.

19 On the other hand, we don't think that something 20 that has no safety significance ought to be sent to the 21 Commission for prior review and approval. So there has to 22 be some safety content to that test. I don't think we are 23 prepared today to agree that it's only a procedural 24 standard, that there should be some safety content to the 25 decision. The Commission's direction in the March SRM on 15 1 minimal safety impact, we think we understood where you were 2 coming from, and I don't think it's a dichotomy either in 3 terms of those two questions. We just wanted to make that 4 point today.

5 . MR. RAY: This is the context specifically of the 6 subject of malfunction of equipment of a different type. ,

7 The statement is broader, seemingly. It says, in view of l 8 the use of 50.59 as a procedural standard rather than as a I 9 safety standard, the staff would not propose language of 10 minimal safety impact. I want to underscore that was in a j 11 specific context. Nevertheless, it raises the notion of j 12 50.59 that we are not sure we understand at this 71nt.

13 [ Slide.]

14 MR. PIETRANGELOi One of the issues where we think i 15 we do disagree with tht staff is on the use of acceptance )

16 limits in determining the increases in consequences and 17 reductions in margin of safety threshold lines. We think 18 both of those kinds of limits can ce found in NRC safety 1 19 evaluation reports as well as other NRC guidance, like the J 20 standard review plan in Part 100. {

21 We did discuss this issue at length with the staff I i 22 on April 23. I think the result of that was we pretty much j 23 agreed to disagree.

I 24 CHAIRMAN JACKSON: Whether you like the criteria '

25 or not, do you believe that the staff has established clear 6cf40 6/10 % ' 19 A M L--_____

16 1 and objective criteria? You may not like the criteria, but 2 do you believe the staff has established such criteria?

3 MR. PIETRANGELO: We have not seen what minimal in 4 terms of consequences means yet. So it's hard to answer 5 that. Conceptually, if there is a line out tnere that is 6 based on the regulatory framework, I think we would prefer 7 that to some construction of what minimal means vis-a-vis 8 where you stand versus the line. I think that would be our 9 preference.

10 CHAIRMAN JACKSON: Commissioner McGaffigan.

11 COMMISSIONER McGAFFIGAN: Going back to the 12 Chairman's admonition that we talk about this in practical 13 terms, as I understand the NEI Guide 96-07 and Mr. Collins' 14 letter of January of this year on this issue of increases in 15 consequences, what you say in that guide is if you find that 16 the agency said something in accepting it relevant to some 17 other document, a Part 100 limit or another document, saying 18 we are accepting it because it's less than that, then that 19 is the limit. If on the other hand we say we accept it --

20 you'll have to correct me here -- with reference to those 21 things, then that is the limit.

22 MR. PIETRANGELO: That's correct. That's what our 23 guidance says.

24 COMMISSIONER McGAFFIGAN: What is the practical 25 effect of that difference at a nuclear plant?

17 1 MR. PIETRANGELO: One would have to come in from 2 the* change and the other one wouldn't. What we have been 3 advising licensees to do when they are caught in that 4 dilemma -- there are some that have in the SER -- the SAR 5 value is the only value that you would find. We say, well, 6 you'll have to go in then based on our guidance. The advice 7 we have given is when you get your next SEE from the staff, 8 try to get the acceptance limit in the SER so that this is a 9 one-time exercise and you won't have to continue to do that 10 in the future, and then eventually you will have everybody 11 consistent across the industry.

12 COMMISSIONER McGAFFIGAN: But does the staff 13 understand that that is what you arc advising?

14 Obviously the staff has disagreed with you. If 15 they disagree with you, then one way not to ever provide 16 that flexibility you are looking for is to make sure you 17 don't do what you have just announced you've told your 18 licenseec to try to do.

19 Aside from when you have to come in and declare it 20 a USG and come in for a license amendment or other approval, 21 what are we talking about specifically in terms of the types 22 of things that we end up having to deal with that you think 23 we shouldn't have to deal with?

24 MR. PIETRANGELO: I can give you an example, and I 25 think it's a current one. South Texas plant has received a

- 18 1 level 4 violation on a very small increase in consequences, 2 from 22-3/4 rem to 23-1/4. The acceptance limit is 30 per 3 GDC-19, control room habitability. Because it was more than 4 zero, there is the level 4 violation. Yet in a previous one 5 the limit was clearly established as 30. That would be one

' 6 where we would think that you shouldn't have to go in for 7 something like that.

8 COMMISSIONER McGAFFICAN: In that particular case ,

9 the minimal could cover it.  !

10 MR. PIETRANGELO: Right. I 11 COMMISSIONER McGAFFIGAN: I guess the staff's 12 concern, as I've understood it over the years, is they are 7cf40 6/10/98 7:39 AM E _ _ __ _ .._ __ _ __ _ ___

13 concerned about going from 22-3/4 to 29-3/4 and being right 14 up against.the limit. That's of concern if we haven't i 15 routinely approved 29-3/4 in other places; if we have L 16 routinely approved 29-3/4, it may not be.

17 I'm just trying to understand.

18 MR. PIETRANGELO: My understanding of the staff's

19 is the closer you get to the limit, the more interested they l 20 get. It's very similar to the Reg Guide 1.174 discussion.

( 21 CHAIRMAN JACKSON: Do you have a comment on this?

l 22 MR. RAY: Yes, I do, or point. I don't want to i 23 make it sound like we are talking past each other relative  !

24 to staff, Commissioner. So let me aad to what Tony has .

L 25 said. I 19 1 I think we need to pay attention to the statement 2 staff makes, which I will read one sentence of here.

3 However, the degree of margin remaining to the

-4 limits might be less as viewed by the staff than the i 5 licensee. Therefore if a licensee subsequently made changes t 6 that would have the effect of increasing the calculated 7 doses up to the limitr,- it is possible that the staff {

l -8 -conclusion would be that tne limits were actually exceeded.

9 So in this case, the example that Tony just gave, 10 . we understand that the staff may have a different view about 11 what the increase was, that it wasn't from 23, or whatever  ;

12 it was, that it was something else. That concern that they j 13 have*does need to be addressed.

14 Another example. You may be well aware of the 15 Niaghra Mohawk case where there was a blowout panel. The 16- thing was set for 80 pounds. The bolts were supposed to be 17 at 45; t,ey were at 53 or something. The argument is made,

[

18 well, as long it's far away from the limit, then we don't L l 19 look at it as carefully as if it's closer to the limit, and 20 therefore when you move from being far away to being closer 21 you need to tell us so that then we can weigh in and see if 22 there is something we want to do in terms of our own 23 perception. We understand that.

! 24 That's why I made the point I did in passing that, 1

25 okay, given that, though, if you take that philosophy to the l

20 b 1 limit, you basically convert everything that has been said 2 Into a tech spec in the sense that we are concerned about 3 it.

4 So we need to be aole to sort through and separate 5 .the things that are not in the tech specs but which have t

6 this character to them that there is some margin that needs )

7 to be maintained against the acceptance limit or the staff I 8 wants to re-review the analysis. We need know where those

'9 are. We don't know which they are. That's the dilemma. I 10 wanted to add thee.

11 CHAIRMAN JACKSON: Let me ask you a different 12 question. If a SER found a facility to be well below the 13 Part 100 guidelines, would you conclude that Part 100 is the 14 acceptance limit?

15 MR. RAY: It's the acceptance limit. I would not 16 conclude that you could approach the acceptance limit 17 without prior NRC approval.

18 CHAIRMAN JACKSON: 'But you know thers is a 19 footnote to Part 100 that specifically states that Part 100 20 guidelines are not acceptance limits. So are you of a view 21 that rulemaking would be required to have Part 100 22 guidelines as acceptance limits?

23. MR PIETRANGELO: We are very familiar with the 24 footnote, and we are trying to get a context for that. I 25 think how we read that is you don't have it acceptable to 8 cf 40 6/10/98 7:39 AM

_____~____y 21

, 1 release radiation to the environment. That's not what that 1 2 means, and I think that's what the footnote was directed at.

! 3 If you go back through the stancard review plan and all the 4 sections where you do have accident analyses and look at the 5 criteria, that is what is referred to in all the cases we 6 looked at.

7 CHAIRMAN JACKSON: This relates actually to the 8 earlier comments by Mr. Ray. Do you agree, though, that the 9 SERs are not part of the scope of 50.59?

10 MR. PIETRANGELO: Right now they are not mentioned 11 in 50.5G. We have some additional proposal language that 12 would get those in play, and I'm going to speak to that in a 13 second.

14 CHAIRMAN JACKSON: Right. Because how are we 15 having a discussion about acceptance limits that are in SERs 16 for purposes of determining USQs under 50.59 if the SERs are 17 not part of the secpe of it? That has always been my 18 problem with this.

19 MR. RAY: I think your comments acknowledge the 20 industry has accepted that SERs are in fact things that must 21 be included within the scope of 50.59 notwithstanding the 22 fact that they are not mentioned.

23 I think the next slide that Tony is going to go to 24 lends itself to talking about Commissioner McGaffigan's 25 question, which is, what is the practical application of 22 1 this? What does it matter?

2 *

[ Slide.]

3 COMMISSIONER McGAFFIGAN: May I ask one more on 4 this?

5 CHAIRMAN JACKSON: Sure.

6 COMMISSIONER McGAFFIGAN: By all means go to the 7 slide.

8 Did I just detect in 96-07, what Mr. Ray just said 9 about as you get close to the acceptance limit, you 10 understand that the staff wants to take a look? Maybe not ,

11 if it's going from 22-3/4 to 23-1/4, but as it gets close to I 12 30. That isn't in 96-07, that concept, at the current time. )

13 So tnat is something that, trying to work out something, I 14 you'd be willing to talk about. I 15 The other way to get at it is this question of 16 minimal. What is minimal? Maybe minimal is something that 17 isn't one percent or something, or 10 percent, or whatever, 18 but maybe it's something that depends on, relative to the 19 acceptance limit, am I making more than an X percent .

! 20 approach to the acceptance limit? Is that what you are l 21 going to be suggesting?

22 MR. RAY: Let me clarify one thing and then answer 23 the question.

24 I didn't say it quite the way I think you repeated '

l l 25 it back to me, Commissioner. I said if the Commission said j l

23 I that this is acceptable because it is far away from the i 2 acceptance limit, which I think was what the Chairman l 3 suggested, and you now make it close to the acceptance 4 limit, does that make a difference? And I said yes, I 5 believe it did, because the Commission said this was okay 6 because it was far away from the acceptance limit. Which is 7 a little different than saying it you are getting close to 8 the acceptance limit, that's a problem in and of itself.

9 MR. PIETRANGELO: And I'm not sure how many SERs 10 say that.

11 CHAIRMAN JACKSON: The real point is, does minimal 12 have a definition, or should it, in and of itself, or can it 9ef40 6/10 98 7 39 AM

. ._. _--_______a

---_~~m 13 only be defined relative to the distance from some boundary?

14 That's really what it boils down to. So I am interested in 15 understanding what you think the boundaries are and what in 16 fact your recommendation is.

17 Yes, Mr. Ray.

18 MR. RAY: Can I ask a clarifying question? Is 19 minimal in that context -- I guess I thought about it 20 differently, which is minimal means minimal change.

21 CHAIRMAN JACKSON: I didn't make a statement. I 22 asked a question.

23 MR. RAY: Okay. Therefore, I don't want to leave 24 it as if that's agreed. I'm understanding minimal to mean 25 minimal change, not minimal away from some limit.

24 1 MR. PIETRANGELO: Your reading of 96-07 was 2 correct.

3 (Slide.]

4 MR. RAY: To say it another way, I believe minimal 5 applies to the boundaries of the white square.

6 CHAIRMAN JACKSON: Let's go back to Part 100 7 guidelines.

8 MR. PIETRANGELO: Part 100 is the darker gray 9 square where the tech spec limits are.

10 MR. RAY: You bet. And minimal we don't apply to 11 that boundary at all.

12 MR. PIETRANGELO: Right. This is similar to what 13 ve said at the reg info conference. If you accept how we 14 view it here, we don't believe there should be any reduction 15 in the margin of safety; we don't think you apply minimal to 16 that line. If the consequence acceptance limite are used,

^

17 i.e., Part 100 or whatever else was in the SER, then you 18 don't use minimal for that either.

19 I think when we viewed the Cer ssion's SRM you 20 were talking minimal up from the whito >

t the inside.

21 MR. RAY: Yes, that we can c white box a 22 little bigger, and that would be okay. ,re not suggesting 23 we can go a little bit over the line hen it comes tc the 24 limits that have been set by the Commission in the tech 25 specs, in the regulations, or any other place.

25 1 MR. PIEYRANGELO: I guess our central point here 2 is and what makes this really germane to not only the design 3 basis discussion but the consequence, margin of safety, and 4 even probability discussion to some extent, is that the FSAR 5 was submitted when a plant went to get its operating 6 license. It had all this information in it, including the 7 technical specification information lifted out of the FSAR 8 and made part of the operating license.

9 There is a hierarchy established with that 10 information that was selected to be the technical 11 specifications. Then you apply 50-90; you need to get prior 12 review and approval before you change any of those values, 13 but the rest of that you shouldn't have to get prior review 14 and approval unless it's a similar char.ge to one cf those 15 limits.

16 i think the effect of how we have been treating 17 50.59 is to make all the information in the SAR a tech spec 18 limit, and I think that Niagara Mohawk case is another ,

, 19 example of where there was basically a degraded condition or l l 20 nonconforming condition that cnanged it from the white box l 21 that was requiring a one-hour report like it was a tech spec 22 violation. We think it's very problematic if you are going 23 to treat all the FSAR information like technical 24 specifications. That's in -- one's interest.

25 MR. RAY: The sen -res that I read earlier apply l 10 of 40 6'10:98 7:39 AM ,

i f

L--_-___________________-_

_ _ _ - ~ ~ - _ _ ,

1 2o 1 to this light gray area surrounding the white area. They 2 are basically saying that when the license is granted, it's 3 based upon that being a big area, and if you do comething to '

4 make it smaller, you need our prior approval. That's what 5 the staff is saying here.

6 Although Tony said that' the way it is, I guess I 7 would change that to say that's the way it has becomo. It 8 certainly wasn't that way for a long time. I've been in 9 this business a long time, as most everybody else in the 10 room has been, and I can say that many things were put in 11 the SAR without sny idea that that in fact was going to 12 become the case. j 13 CHAIRMAN JACKSON: How would the boundaries of 14 your ciagram fit with the ASME code?

15 MR. RAY: The ASME code is in the area lying 16 outside, in the most dark band. In other words, the ASME 17 code, having sat on the committee there too, I can say 18 addresses where the limits should be on stress on other ,

19 things given that construction is imperfect, that there are 1 20 defects in the material, that the loadings are going to be i 21 uncertain, and many other things. You put in the ASME code  !

22 a margin against the true breaking strength of the material, 23 'let's say, which then allows you to deru.e the next box in, 24 wnich are the limits of the ASME ss they are adopted by the 25 Commission; the Commission establ.shed for design. Nobody 27 1 can go beyond those.

2 MR. PIETRANGELO: In fact the source documents for 3 much of the design basis information are from the codes.

4 MR. RAY: Then we back down further to say, okay, 5 here's how we are going to operate the plant, and for those 6 really important things, we are going to put them in tech 7 specs and make sure that you guys focus on them and don't go

-8 outside that box without getting our approval. What we are 9 now doing is talking about other things that aren't in the 10 tech specs but are in the SAR, and thus to the issue of 11 scope that I addressed to you in the beginning.

12 CHAIRMAN JACKSON: Suppose we had a hypothetical 13 plant that had a containment with an internal desian 14 pressure of 50 psia and ultimate failure pressure of 100 15 psia, but the accident analysis says that you can have 46 16 psia peak containment pressure. Where do those fit on this 17 box?

18 '

MR. PIETRANGELO: The 50 would be the acceptance 19 limit.

20 MR. RAY: It would be from the inside out, 46, 50 21 and 80, I think you said.

22 . . .

MR. PIETRANGELO: Right, 46, 50 and 100 from the 23 inside out.

24 ,, CHAIRMAN JACKSON: So where is 46?

25 HR. PIETRANGELO: Forty-six is on the perimeter of 28 1 the white box.

2 MR. BEEDLE: If I can take that a step further, 3 you picked one that is very likely to be one that you can't 4 go to 46-1/2 or 47 without getting the Commission's approval i

5 even though it's not in the tech specs. Our dilemma As, how 6 do we pick those out from the zillions of other pieces of 7 information in the SAR if we are going to use the SAR?

8 That's the problem with the SAR. If we change to 9 some other measure or some other definition of what we are 10 concerned about, then it's easy to capture the 46 in that, 11 and say, I don't want you to change the 46; I don't want to

12 make it 46-1/2 or anything else without my prior approval.

1ief40. 6/10.98 7 39 AM

13 As long as we continue to use the SAR, we are in 14 the dilemma that there is so much in there that doesn't have 15 that importance that I think we have a hopeless task.

16 CHAIRMAN JACKSON: Yes.

17 COMMISSIONER DIAZ: I was Just going to say that I 18 hate to use the word right now, but it appears to me that j 19 you are trying to risk rank the design and operating l 20 envelope.

I 21 MR. RAY: I wouldn't hate to use that word, 22 Commissioner Diaz, but I thought that this was a binary risk 23 ranking here we are talking about.

24 CHAIRMAN JACKSON: It's a tier.

25 MR. RAY: One of my committee's responsibilities 29 1 is to support from the industry side risk information, risk

. 2 ranking, application of risk, and I certainly want to 3 endorse the idea.

4 CHAIRMAN JACKSON: Then you should have endorsed 5 cption 5 of the staff's paper.

6 MR. RAY: You have me at a disadvantage.

7 [ Laughter.)

8 MR. PIETRANGELO: Can we go to the next slide, 9 please.

10 (Slide.]

11 CHAIRMAN JACKSON: We're not through yet.

12 COMMISSIONER McGAFFIGAN: Following up on that 13 example, if those numbers had been 20, 50 and 100, what you 14 are saying is you understand why we would be concerned if we 15 are'already at 46, the operating envelope, but if it's at 16 20, then going to 21 or 22 or even 25 -- let me just ask 17 you. If the inside envelope is 20, the next one is 50, the 18 next one is 100, where should we get concerned?

19 MR. RAY: You illustrate the problem we have, 20 which is that there hasn't been, to use Commissioner Diaz' 21 notion, a categorization of these things in terms of risk 22 ranking.

23 I can imagine a situation in which 20 would be 24 even something that the Commission wouldn't want you to 25 exceed without their prior approval for the reason, as the 30 1 staif argues here, they just didn't review the analysis very 2 thoroughly because it was so far away. Well, then you have 3 to make that clear, because we can't guess where that is 4 true without, as I said, converting everything into a tech 5 spec type limit.

6 We have got to somehow solve this problem. I 7 understand that, and I don't have any silver bullets.

8 CHAIRMAN JACKSON: I want you to keep in mind your 9 risk categorization.

10 MR. PIETRANGELO: We did mention that we would try 11 to say how we would incorporate risk insights during this 12 particular briefing. I think in this case, besides the 13 example Harold gave you, we think there is an opportunity, 14 and right now it's on very much an evolutionary path, to 15 change the perimeter of t!.e gray box through risk informed 16 tech specs ..ud applying PRA even to the design basis 17 accident analyses. NEI has a pilot project to do that, and 18 I think there is one pilot that is looking at coincident 19 LOCA and loop and all that kind of thing from a risk 20 perspective.

21 CHAIRMAN JACKSON: Commissioner Diaz.

22 COMMISSIONER DIAZ: Even if we back away from 23 risk, and I'm going to go back to Commissioner McGaffigan's 24 20, there is an engineering judgment that is applicable to 25 those cases. Engineering judgment tells us that there is no 12 cf 40 6/10'98 7:39 AM

_ __. ____ ~ _

31 1 difference in calculation accuracy or in measurement of 2 response of equipment between 20 and 21, and that's minimal.

3 MR. RAY: Commissioner Diaz, I certainly agree 4 with you. I don't want to prolong this part of the 5 discussion beyond what you all wish to do, but I do want to 6 say that we are in an era that is different than the past, 7 for whatever reason. There is no point in debating why we

8. are here, but we are here. One of the things I was going to 9 say in the history discussion was, I think we have learned, 10 all of us, that we have got to deal with the literal 11 application of these words, like it or not, and we're here 12 to try and figure out how to do that.

13 I would just suggest to you that the exictence of 14 the tech specs was in fact a binary risk ranking. Stuff 15 went in there or it didn't go in there. Now we are into a 16 different.world. I won't opine on that further.

17 .

CHAIRMAN JACKSON: It was a binary ranking. We 18 could argue all afternoon about to what extent it is risk 19 ranked. One could argue that within the FSAR there is a 20 risk ranking. Then if you were looking at relative risk, 21 you might have things in the ESAR and things in the tech 22 specs that cross each other one way or the other.

23 My basic point of view is that it's really a new 24 paradigm, but we will continue within the context of trying 25 to tinker at the edges, which is really where we are, in my 32 1 epinion.

2 MR. RAY: Regrettably.

3 MR. PIETRANGELO: Could we go to the next slide, i 4 please? We're running way behind hero.

5 (Slide.] l 6 CHAIRMAN JACKSON: No, you're not behind, because .

7 we asked the questions.

8. MR. PIETRANGELO: Design basis interpretation.

9 There is some history to this one.

10 CHAIRMAN JACKSON: There is history to all of 11 this.

12 (Laughter.]

13 MR. PIETRANGELO: Eirst of all, it is an important 14 issue because it's critical to both operability and 15 deportability determinations. We put together guidance in 16 .the late 1980s or early 1990s. We revised it last year. It F1 has been with the staff since November.

16 '

Interpretation that was provided to a licensee on i 19 a particular issue gave a new interpretation of what the 20 50.2 definition of design basis entails, and that is any 21 information you used to determine the acceptability of the 22 design. It's very much what we were just talking about.

23 CHAIRMAN JACKSON: Have you seen guidance 24 documents that say that?

25 MR. PIETRANGELO: It's in a letter.

33 1 CHAIRMAN JACKSON: A letter from?

2 MR. PIETRANGELO: The agency to a licensee.

3 MR. RAY: Yes. As a matter of fact, it's 4 September 1997.

5 MR. PIETRANGELO: September 12, 1997.  ;

6 MR. RAY: The say it's expressed, it says the j 7 guidance in NUREG-1022 and this letter in September -- l 8 MR. P!ETRANGELO: There is a little bit more 9 history. When we saw that interpretation, we immediately 10 wrote.to the agency saying, wait a minute, there is a ,

11 Commission policy statement from 1992; there are other  ;

12- regulatory guidance documents that we don't think are ]

13 cf 40 6/10/98 7.39 AM  !

l.

c___----________-_-_-_:____-_______________-_-____-____

.-----__~n 13 -consistent with this interpretation of what design basis 14 information entails.

15 There was more interaction with the licensee in a 16 subsequent letter that came out this past March which 17 mentions the consistency of this September 12th letter with 18 .the deportability guidance in NUREG-1022. I think that's 19 part of the problem. We don't think those two things are 20 consistent at all.

21 MR. RAY: It is precisely what I was alluding to 22 in the scope context. Again, it is one sentence. Let me 23 read it:

24 It would be inappropriate for the NRC staff at l 25 this juncture *, provide any new or different guidance 34 -

1 regarding the definition of design bases provided in 10 CFR 2 50.2 beyond that already provided in NUREG-1022, revision 2, 3 .and the NRC's letter of September 12, 1997.

4 The problem is those two things are not >

5 consistent.

6 MR. PIETRANGCLO: And if there was a new -

7 interpretation, it.was made in September, b CHAIRMAN JACKSON: This is one example and it's 9 something obviously, if in fact what you say is true is 10 true, that we need to follow up on. Has this been a 11 broad-based change that you have seen in guidance documents 12 or numerous communications between the NRC and licensees?

13 .

MR. PIETRANGELO: There is no NRC guidance on 14 this. The deportability guidance which was just issued, I 15 believe in February of this year, was the NUREG-1022, 16 revision 1. We know the staff asked some activities to look 17 at 50.72 and 50.73 reporting. Our point is that that is 18 kind of trying to get at the symptom versus the root cause 19 of what is the appropriate interpretation of the --

20 MR. RAY: You asked a question I don't think we 21 have a good answer for: how prevalent is this? I can't 22 answer that.

23 CHAIRMAN JACKSON: That is what I mean. There are 24 two questions. One is, is it your understanding that the 25 definition of design basis has remained static since the 35 1 days of the Atomic Energy Commission, or do you feel there 2 has been some evolution as the industry and we have 3 responded to events such as TMI, Browns Ferry, et cetera?

4 That is one question.

5 e The second is whether there is either specific 6 change in guidance or there is some widespread de facto 7 change in guidance that exists through correspondence.

8 MR. PIETRANGELO: I think it's the latter, 9 Chairman Jackson.

10 CRAIRMAN JACKSON: Then you need to bring us that

'll data.

12 MR. RAY: We understand.

13 .

MR. PIETRANGELO: Wait. I got calls this week j" 14 from a Region IV utility group, and they read everything 15 that comes out of this agency. They're afraid about being 16 in willful noncompliance for not repc" ting in a similar 17 . instance,-and this has a destabilizing effect.

18 CHAIRMAN JACKSON: I'm n;t here to argue with you, 19 Mr. Pietrangelo. I'm saying to you we are trying to reach

20. some point of where we can and should go on this. If you 21 want to be helpful to NRC, then what you can do is provide 22 us with the information in a constructive way. That's all 23 I'm trying to tell you.

24 MR. BEEDLE: We will provide the Commission with a 25 letter.

' *4 cf'40 6/10/98 7:39 AM

_____ ________________ - _ _ _ _ _a

_________y 36 1 COMMISSIONER McGAFFIGAN: This is an issue that I 2 hadn't been up on, but we have some draft report language 3 that may have been changed since. I assume it reflects this 4 issue.

5 In order to resolve this design basis uncertainty,

6. NRC needs to. reaffirm its interpretation of design basis 7 information consistent with NUMARC 90-12 or the proposed NEI 8 97-004 revision of NUMARC 90-12, 9 You said earlier that 97-004 had been submitted to 10 us last year sometime?

11 MR. PIETRANGELO: Last November.

12 COMMISSIONER McGAFFIGAN: With a request that we 13 endorse as a guidance document? How is that transmitted to 14 the Commission? ,

15 MR. PIETRANGELO: That particular letter, I

{

16 believe, was sent to Mr. Collins or Mr. Callan. I can't j 17 remember which. Previously the NUMARC 90-12 document, we l 18 did-get a letter of acknowledgement from the director of NRR )

19 at that time. Subsequent to that there was a Commission 20 policy statement. I can read you the language that we {

21 quoted in the letter from the staff, but we'll provide that i 22 later. It basically said, our rationale for the design 23 basis was consistent with the 50.2 definition. Then you see 24 the NUREG-1022 guidance as well.

25 Our point is we thought we had a fairly good trail 37 1 and guidance path, and then this letter came out that 2 seemingly was a new interpretation of that. We tried to 3 bring it to the agency's attention.

4 MR. RAY: As I think we said at the beginning, 5 this is a bit of a sidetrack from the 50.59 and 50.71(e) 6 subjects that the Chairman indicated we are here to talk 7- about, but it bears on it to some extent. So we bring it to 1 8 your notice and we'll follow up with a letter.

9 MR. PIETRANGELO: Let's go to the next slide, 10 please.

11 (Slide.]

12 MR. PIETRANGELO: Part of the SRM from March 24 13 dealt with enforcement discretion for 50.59. We talked 14 about this a little bit. There seemed to be a disagreement 15 following the reg info conference session about what that 16 direction from the Commission meant with regard to 17 enforcement discretion until the rule was changed to l 18 incorporate the minimal standard versus the zero increase 19 standard.

20 Our perspective was that we don't know how long 21 the rulemaking is going to take. We hope it's a going to be 22 a fairly quick one, but that in the interim, to avoid 23 examples like the South Texas one I went over before, there 24 shouldn't be enforcement action taken when the clear intent 25 of the Commission on minimal with regard to probability 38 1 increases or consequences or t.alfunction with a different 2 cause but the same-result occurred out there.

3 The staff was appsrently interpreter.g that as, l 4 .well, there still has to be enforcement action but instead i 5 of a level 3 it's a level 4, or Anstead of a level 4 it's a 6 non-cited violation.

7 We thought this is very similar to two-year 8 discretion on the FSAR that ends this October. We wanted to 9 raise this because we think there are some interpretation 10 differences between how we flew what was in the SRM versus 11 the staff.

12 CHAIRMAN JACKSON: How do you propose that we I

'l.?cf40- -6/10/98 7:39 AM

i 13 proceed in a way to ensure consistency if there isn't a 14 {

definition of minimal? i 15 MR. PIETRANGELO: Our guidance, and we have the 16 initiative that the deadline is the end of this month, is i 17 really less than minimal; it's negligible; or where there is 18 a discernible trend.

19 COMMISSIONER DIAZ: We are not going to get into 20 that.

21 MR. PIETRANGELO: That's the industry guidance.

22 My point is that our standard is already higher in the sense 23 of less than minimal than what the Commission said in the 24 SRM. So we think there will be consistency in that regard.

25 COMMISSIONER McGAFFIGAN: The increase in 39 1 consequence within acceptance limits. You get into that 2 same issue we just spent 20 minutes on.

3 MR. PIETRANGELO: That's right, subject to 4 whatever the Commission decides, obviously.

5 CHAIRMAN JACKSON: Are you basically suggesting a 6 blanket exemption to the industry?

7 MR. PIETRANGELO: If it has the minimal consistent 8 with the intent of the SRM until the rule is changed.

9 CHAIRMAN JACKSON: I'm just saying there is an 10 issue having to do with what guidance one is operating off 11 of. Otherwise, what you are saying is that you want a 12 blanket exemption until the rule is done. Is that what you 13 are saying?

14 MR. RAY: I would be perfectly comfortable with 15 the notion that the staff simply needed to assert that 16 something wasn't minimal and thereby say that they had 17 satisfied the Commission's direction if they felt it was 18 necessary to do so.

19 Very often we see things identically in terms of 20 their significance. It's the compelling need to go ahead 21 nevertheless that is the problem. So I don't think we need 22 to debate as much as we think we do what is minimal and what 23 is not, because I'm comfortable with any of the NRC managers 24 that I know making a judgment about what is minimal. I just 25 would like them to be able to say, well, yeah, I agree it's 40 1 minimal, and therefore enforcement action is not required.

2 MR. PIETRANGELO: Next slide, please.

3 (Slide.]

4 MR. PIETRANGELO: Switching gears now to the draft 5 FSAR update guidance. This is an overview slide for the 6 next few. I want to talk about focus of what the update 7 ought to entail, some of the reconciliation issues that we 8 read in the SECY that transmitted the draft generic letter, 9 talk about enforcement discretion on 50.71(e) aise, and give 10 our perspective on that.

11 Before we move to the next slide we want to thank 12 the Commission for issuing the draft generic letter. I 13 think that wps very helpful for us to get with the staff. I 14 think we are on a positive track here, and I think you will 15 sce as we go through the issues that this one is on a good 16 path to resolution.

17 Next slide, please.

18 [ Slide.)

19 CHAIRMAN JACKSON: I detected a degree, shall we i 20 call it, of excitedness in the April 16th letter from NEI.

I 21 You recognize that the guidance is interim.

22 MR. RAY: Remembering only that the subject of 23 implementation of 50.71(e), if we can separate it from tne 2? tide of 50.59, is one that I think has the character that 25 Tony just described to you. It's on a track that is 16 cf 40 610 98 7 F) AM m

, _ ______ ~. _

41 1 reasonable.

2 MR. PIETRANGELO: I think the staff did a good job 3 in the draft generic letter o: spelling out what was 4 originally required under 50.34 in terms presentation of the 5 design basis for 50.2, the safety analyses, the operating 6 limits, and then what we call a contextual description of 7 those things.

8 Our guidance document basically had the same 9 focus. I think our only point we have been discussing 10 lately is the limits on operation we would equate with the 11 tech spec values that were in the original SAR that were 12 lifted out and became the tech specs. So they may or may 13 not be in the SAR, but in any event whatever is in the SAR 14 ought to be consisttent with the tech spec values.

., 15 Next slide, please.

16 [ Slide.}

17 MR. PIETRANGELO! In the SECY te the Commission 18 the staff said there were three reasons why without change 19 they would be unable to endorse NEI 98-03. The first had to 20 do with removal of historical information; second, removal 21 of obsolete and less meaningful information; and third, 22 treatment of detailed drawings.

23 In our presentation material that we discussed 24 with the staff on May 27 we proposed some changes to 98-03 25 that were very consistent, we think, with the draft generic 42 ,

1 letter that would resolve those issues. t 2 In addition, we understand there are going to be 3 some comments made part of the meeting summary from May 27 4 that will provide additional comments that the staff has on 5 98-03. We think we will be able to turn our document around 6 by the end of this month to continue the discussion. So it 7 is very positive.

8 In terms of the three issues that would preclude 9 endorsement, we were very comfortable that we could address 10 those concerns, 11 CHAIRMAM JACKSON: Could I get you to go to slide  !

12 10. l 13 MR. PIETRANGELO: Yes.

14 CHAINdul JACKSON: Based on your interpretation of 15 design basis, would this approach that you are talking about 16 exclude updates for nonsafety-related issues involving 17 station blackout, ATWS, or safe shutdown under Appendix R?

18

• MR. PIETRANGELO: Absolutely not. Those are 19 required under 50.71(e). But I think the types of 20 information about those things you cited would fall into 21 these categories.

22 CHAIRMAN JACKSON: What about FSAR supplements l 23 submitted under the license r6aewal?

24 MR. PIETRANGELO: It's required by Part 54.

25 CHAIRMAN JACKSON: That's right, but is it 43 1 captured? ,

i 2 MR. PIETRANGELO: Those are really talking about 3 programmatic descriptions, and I might have to refer to the 4 PM for license renewcl, Doug Walters here. But my 5 understanding is that is what the rule calls for, to I 6 supplement the SAR with programmatic descriptions made as a 7 result of the license renewal review.

8 CHAIRMAW JACKSON: Would you like to comment?

9 MR. WALTERS: The position we have taken is that 10 under license renewal the FSAR supplement would be to the 11 same level and same detail that you have today, and it would 12 be the incorporation of, let's say, programs that you are 17 ef 40 6/10/98 7:39 AM

_~

13 crediting as aging management programs if they are not 14 already described. So there are really two issues: What do 15 you put in the SAR and then what is the level of detail?

16 CHAIRMAN JACKSON: Can you give me a contextual 17 description of a hypothetical accident analysis? '

18 MR. PIETRANGELO: Not off the top of my head I l

19 can't.

20 MR. RAY: I think it's a redundancy, isn't it, a 21 hypothetical accident analysis? All of them hopefully are l 22 hypothetical.

23 MR. PIETRANGELO: By contextual, we mean how does 24 it fit into the safety analysis. This is the safety 25 analyses report. By contextual, we mean how does that 44 1 information --

2 CHAIRMAN JACKSON: Can you give me an example?

3 Frame it out for me.

4 MR. PIETRANGELO: I think in terms of presenting 5 the design basis and the description of the system and how 6

7 it functions and all that, I would say what were the input assumptions and parameters that were used and the input back 8 to the safety analyses. I would try to keep that 9 description contextual to the safety analyses.

10 We know, though, that over time it got broader 11 than that. I used to work for a vendor and we had separate 12 documents that were system descriptions that had the 13 nameplate data en the motors and the pumps, and all that was 14 eventually included in the SARs, and the initial hazard 15 summary report, I don't think, had a lot of that kind of 16 information.

17 CHAIRMAN JACKSON: On slide 11, how would NEI 18 98-03 change as a result of the draft generic letter's 19 content?

20 MR. PIETRANGELO: I will do them one by one. I 21 think I can do the first two, and I may need help on the 22 third .

23 On the treatment of historical information and the 24 revision that the staff has, we suggested removal of 25 historical information that wasn't going to change. I think 45 1 the draft generic letter made some points about what was 2 required by 50.34. What we suggested to the staff on the 3 27th is rather than remove that historical information, it 4 could be reformatted into an appendix or some other part of 5 the document where it would not be subject to change or 6 subject to update. That's our definition of historical.

7 On the second bullet, with regard to removal of 8 obsolete and less meaningful information, the draft generic 9 letter suggested that the licensee needed to have a process 10 to establish by some criteria what information was obsolete 11 and less meaningful.

12 I think it would go back to the previous slide on 13 what the focus ought to be. That process along with 14 providing a rationale for the update of why that information 15' came out, p kind of documentation trail that with that 16 process there would be flexibility to remove and less 17 meaningful information.

18 I think it's basically the same on the detailed 19 drawings. As long as there was a process of paper trail to 20 say why the drawing went from very detailed to a schematic, ,

'21 for example, I think we would tie it back to whether those 22 components listed on the drawings or in that detail were 23 credited in the accident analysis.

24 MR. RAY: Let me interject here and say I believe 25 a lot of this activity is driven off from the need to 18 cf40 6/10/98 7:39 AM

l 46 1 conform the SAR to an acceptable scope for 50.59. If you  !

2 once break that link, I think the question is, well, why not  !

3 have detailed drawings in the SAR? It's not that big a 4 deal. You just take reduced size P& ids; and put them in j 5 'there. That's what people had done, and they thought that I 6 was okay. (

7 The reason that we are driven back to take out all l 8 of these details and slim it down to something that doesn't 9 have a lot of details in it is really in an effort to make 10 them not subject to 50.59, which is exactly what the staff 11 said it was for.

12 I just think we need to first break that link and 13 then decide what to do with the SAR, because we might come 14 out with different answers.

15 CHAIRMAN JACKSON: My position is well known. 1 16 MR. PIETRANGELO: Could we go to slide 12, please.

17 (Slide.]

)

l 18 MR. PIETRANGELO: With regard to the enforcement l 19- discretion on 50.71(e), we are still in the middle of a 20 period of enforcement discretion that ends October 18 of ]

i 21 this year. That enforcement discretion required the )

22 licensee put a program on the docket to describe how they 23 would go back to validate and verify the information that is 24 in the SAR is accurate.

25 Given that this is the first stab at regulatory 47 1 guidance for 50.71(e), there is also an issue with regard to 2 completeness. We've had a lot of discussion on whether 3 completeness and accuracy are mutaally exclusive or not. My 4 own opinion is they are not. Sometimes you are not entirely 5 accurate if you don't have all the information there.

J 6 The sub-bullets here. '

7 There is no safety urgency for this information.

8 This is information that is already on the docket. It's a 9 location problem per the regulation, and we recognize that 10 under 50.71(e) there will be a need for many licensees to go 11 back. They may not have captured some of the information 12 that was required. j 13 CRAIRMAN JACKSON: Are you talking completeness or j 14 accuracy?

15 MR. PIETRANGELO: I think we are talking both, 16 Chairman Jackson. Again, in my own mind, it's hard to 17 separate the two. But we understand that what is in the 18 document needs to be consistent with what is in the plan and 19 the procedures. There are different shades of this also. I 20 think for a lot of licensee, based on the feedback we have 21 received, they will have identified a lot of the 22 discrepancies in the FSAR, but they may not have closed them  ;

23- out yet; they may be in the corrective action program.

24 _

CHAIRMAN JACKSON: Do you support the staff's 25 approach to risk informed enforcement discretion periods?

48 1 MR. PIETRANGELO: I think that is a way to do it.

2 We did have a discussion about this in last week's meeting.

3 You don't want to get in a situation where you are off for a 4 couple of years and you don't know at the end of that period .

5 whether everybody is going to be finally done with this or 6 not, and it's appropriate to get some feedback at some point 7 or some intermediate milestone that would give the 8 Commission a sense that the licensees are progressing with 9 this, and it makes sense to use risk ranking to focus on 10 those.

L 11 Right now, given that the FSAR, as I think Harold 12 has underscored, is net basically a risk-significant j

19 cf 40 6/10/98 7:39 AM

____ _ __ _ _ .. yi 11 13 document, there are certain systems you could pull out by 14 the maintenance rule guidance to say, yes, we can focus on 15 those first.

16 CHAIRMAN JACKSON: Wasn't that the original 17 go-forward direction? It's certainly the accuracy issue on 18 the FSARs, that it should have been done on that risk ranked 19 basis. So if in fact it hasn't been done on that risk 20 ranked basis, why should there be two more periods? If the 21 most risk-significant things haven't been done, why should 22 there be two more years?

23 MR. PIETRANGELO: I think with regard to accuracy 24 it hasn't mattered at this point. I said only a way, 25 because I may want to go back to that focus slide and make 49 1 sure all my design basis information is accurate and make 2 sure all my safety analysis inputs are accurate, because 3 there has been a lot of activity on those too, and that is 4 not with regard to risk significance. That's another way to 5 approach the which ones I do first argument.

6 COMMISSIONER DIAZ: Your recommendation is two 7 years?

8 MR. PIETRANGELO: That's the normal cycle for an 9 FSA update period.

10 CHAIRMAN JACKSON: Right. It already will have 11 been two years in October, right?

12 MR. PIETRANGELO: Yes, and I think there has been 13 extensive effort. I think most people are there with regard 14 to accuracy, but the completeness part and given the new 15 guidance, we think it's appropriate to extend that.

16 COMMISSIONER McGAFFIGAN: Could I clarify?

17 CHAIRMAN JACKSON: Sure. Go ahead.

18 COMMISSIONER McGAFFIGAN: I guess I'm having 19 trouble with the accuracy and completeness as well. Your 20 recommendation, as I understand it, is to not try to make 21 the distinction between accuracy and completeness.

22 If I were to take the staff's proposal and try to 23 merge it with yours, give you six months for accuracy and 24 completeness with regard to the systems and the maintenance 25 rule or the most risk significance and give you two years 50 1 for the rest of it, but don't try to make this distinction 2 between accuracy and completeness on October 18, 1998, where 3 you would have to be accurate -- I don't want to get into 4 semantics over whether it was inaccurate because it was 5 incomplete, and maybe we just need a time period for both.

6 That's what strikes me as I listen to this for the first 7 time.

8 CHAIRMAN JACKSON: Right, but the real issue is 9 that in the end, whatever the time period is, we're probably 10 guaranteed that you are going to come back and say we should 11 have two more years, right?

12 COMMISSIONER McGAFFIGAN: Get the Bibles out.

13 (Laughter.]

14 MR. RAY: Is that a question requiring an answer?

15 CHAIRMAN JACKSON: No.

16 MR. PIETRANGELO: I think it's a question of 17 degrees. As Harold underscored, there is a lot of other 18 information in the SAR. We know there is more important 19 information than others. We may want to use that as the 20 stick to measure progress versus some other.

21 The final slide on SARs is the outcome slide.

22 [ Slide.]

23 MR. PIETRANGELO: Our conclusion based on the 24 draft generic letter and t.e eeting we had with the staff 25 on the 27th. We don't see 4 need to issue the draft generic 20 cf 40 6/10:98 709 AM

--. _ y.

51 1 letter, and purely from, we think, an efficiency and speed 2 standpoint, we can save a step in this process.

, 3 We are comfortable that we are on converging paths )

l 4 with the staff based on the meeting. It is more efficient

^

i 5 to seek public comment on a final draft reg guide endorsing 6 our guidance versus having to get formal public comment on 7 two separate documents.

j 8 We have talked about a tentative schedule for 9 closure with the staff.

10 CHAIRMAN JACKSON: What is that?

11 MR. PIETRANGELO: I'm about to go through that.

i ~12 CHAIRMAN JACKSON: Tell me the ultimate drop-dead l 13 date.

! 14 MR. PIETRANGELO: By the end of the year.

! 15 Our other conclusion is we don't think there is a

! 16 need for rulemaking on 50.71(e). That language is pretty l 17 straightforward, and I think we are comfortable with it.

18 [ Slide.]

l. 19 MR. PIETRANGELO: Finally, the last set of slides 20 is on the scope of 50.59. We have already discussed the 21 industry proposal at some length, about decoupling the scope 22 from the SAR, trying to define in A-1 of the rule what scope l 23 is, and our April 16 letter suggested a focus on the safety l 24 analyses in the SAR.

25 (Slide.)

52 I 1 MR. PIETRANGELO: We think there are a number of 2 benefits to doing this. In the interest of time, we won't 3 do this, but we could give you several examples on full 4 safety evaluations that really have little or no safety or 5 regulatory value.

6 We did a survey last year as part of our

7 commenting on NUREG-1606. The average full safety

l. 8 evaluation time takes about 27 hours3.125e-4 days <br />0.0075 hours <br />4.464286e-5 weeks <br />1.02735e-5 months <br />, and that does not 9 include review time. We think there could be a substantial 10 benefit in terms of reducing the number of these full 11 evaluations. We think it would improve the consistency 12 between the rule and implementation.

13 . Finally, it gets at trying to define what is 14 important in the SARs from a 50.59 standpoint and would have

15 the effect of leveling this playing field on big SARs versus l 16 small SARs, and we know there has been a concern about that. j 17 Last slide, please.

l- 18 +

(Slide.]

19 MR. PIETRANGELO: We continue to believe there is i 20 a need per the Commission's SRM to expedite the rulemaking 21 on the threshold criteria, and that's the best way to get l

22 long-term regulatory stability in 50.59. I should have said 23 before that part of the rationale for the enforcement 24 discretion until the rulemaking is complete is to get that 25 kind of stability in the short term, but that's no l 1 substitute for the rule language being correct.

2 We have not changed our mind about the need for a ]

3 rule change on the scope of 50.59. We are prepared to work j 4 in the two-step process. We went over the rationale before,  ;

5 Commissioner McGaffigan. I think the primary reason was  !

6 after the individual visits we were convinced there was a 7 commitment on the part of the Commission to follow through 8 on this.

9 .

We are ready to go on this. We had some

-10 preliminary contractor work done on this concept of safety 11 analyses. We are encouraged by the results we are getting 12 thus far.  !

1 21 cf 40 6/10/98 7 39 AM  !

13 CHAIRMAN JACKSON: When you talk about safety 14 anslysis, do you includa shutdcwn safety, ATWS, station 15 blackout, Appendix R safe shutdown?

16 MR. PIETRANGELO: Yes. It's all the required 17 analyses as well as some of the requested ones for 50.71(e) 18 that had an effect on the analyses or were new.

19 CHAIRMAN JACKSON: Does that include human 20 performance and operational safety issues that are currently 21 covered in the programmatic sections of the ESARs?

22 MR. PIETRANGELO: Our initial work that we have 23 asked the contractor to help us with went back through all 24 the generic letters and bulletins and tried to find where 25 there was a request for a safety analysis to be submitted by 54 1 the licensee. That could be with regard to some human 2 performance, but I'm not sure. We did find out of a 3 population of about 300 generic letters about 21 that did in 4 fact request the licensee to submit a safety analysis; in a 5 population of about 100 bulletins there are about 29 that 6 requested a safety analysis.

7 CHAIRMAN JACKSON: Commissioner McGaffigan.

8 COMMISSIONER McGAFFIGAN: I want to underst_.id how 9 the 50.59 process works in a real plant. I had an 10 interesting conversation a month or two ago with a young man 11 who I won't name but who had worked in plants. We got into 12 a discussion as to whether we approve changes in plant 13 managers and whether a safety evaluation has to be done to 14 determine whether there is an unreviewed safety question 15 whem Joe replaces Tim.

16 I said to him, Oh my God, they can't be doing 17 that.

18 But do you? When Joe replaces Tim as head of the 19 operations department, is there a multi-thousand dollar 20 evaluation made as to whether that is an unreviewed safety 21 question?

22 MR. RAY: No. It's a fair question, but the 23 answer is there is nothing in any set of reference documents 24 that I could thi.nk of that would mean that was a change to 25 the facility.

55 1 COMMISSIONER McGAFFIGAN: That was the example 2 this person used.

3 MR. PIETRANGELO: That was a title change.

4 COMMISSIONER McGAFFIGAN: The title change one I'm 5 well aware of. We have a license amendment in at the moment 6 because it's a tech spec change. This particular utility 7 did not, as the staff recommended a long time ago, get all 8 these titles out of the Administrative section of the tech 9 specs. So we have a license cmendment in at the moment to 10 change plant manager to vice president, and we a:e going to 11 have to go through a license amendment process, o 12 I assume, Mr. Ray, that Southern California Edison 13 probably took the staff's advice in the late 1980s and got 14 all of that stuff out of its tech specs. Or maybe you never 15 change titles.

~

16 MR. RAY: We were a lead plant for standard tech

, 17 specs, and I don't believe that the titles ars in the lL 18 standerd tech specs.

19 MR. BEECLE: The situation you are referring to, 20 though, the plant had in their tech specs and in their FSAR 21 titles. When they changed the title of their senior manager 22 on site, then they ended up doing 50.59s for that change as 23 well as tech spec changes in order to accommodate that.

24 Some plants do 50.59s when they change people; when they 25 change a plant manager, they do a 50.59 on it.

l' l

! 22 of 40 6/10/98 7:30 AM i

h .

56 1 A lot of that is driven by the request or comments I 2 by resident inspectors, and in some cases regional-based 3 inspectors, and so the plant reacts to that and says it's a 4 change in the facility, whether it's people or process or 5 equipment, and they effectively come to ALSAP where tney do 6 50.59s.

7 COMMISSIONER McGAFFIGAN: We aren't supposed to 8 vote in public, but I suspect there is no Commissioner who 9 wotid ever ask you to do that in the history of the agency. ,

10 I get a little bit frustrated. I used to work on Pentagon l 11 type issues. We would sneeze in the Congress and they would j 12 catch pneumonia at the Pentagon. So I understand. Some of j 13 this stuff is self-imposed. That's the only point I'm 14 trying to make.

15 MR. BEEDLE: I would agree with that, and I think 16 that just points out the significance of the work that is i 17 ongoing right now on 50.59 and 50.71(e). The complexity of 18 the process is such that you have several thousand people 19 out there trying to utilize this rule and thqy all look at 20 it a little bit differently, and where there is ambiguity or 21 a vague definition, then they all interpret it a little bit 22 differently. I'm not saying that they are wrong, but I'm 23 telling you that it yields strange results, this being one 24 of them.

25 MR. RAY: We need to get up and give the staff a 57 1 chance to address this issue.

2 CHAIRMAN JACKSON: Yes.

3 MR. RAY: I just want to say one thing in 4 conclusion. I'll just say it very briefly. I perceive that 5 we will engage in a continuing quest for the unattainable, 6 and that is an objective definition of minimal, and so on.

7 I think we ought to get beyond that. I believe the managers 8 in the NRC are responsible public officials who, if they are 9 .given the latitude to decide that something is okay because 10 it's minimal, will make the right decisions, and we don't 11 need to try and find some ruler to give them all that they 12 can apply to everything out there. '

13 CHAIRMAN JACKSON: Then you will head off the next 14 Tower's parent study.

15 MR. RAY: Chairman Jackson, if I had the 16 opportunity to head it off, 1 would make that commitment. I 17 don't believe that's in my purview.

18 '

CHAIRMAN JACKSON: Thank you very much. We 19 appreciate it.

20 Let's hear from the NRC staff.

21 Good afternoon.

22 Mr. Thompson.

23 MR. THOMPSON: Thank you, Chairman Jackson,

.. Commissioners. This is a very important issue both for the 25 NRC and for the industry. I think you've heard today that 58 1 we have made a lot of progress working well together, and we 2 certainly intend to continue that.

3 Before I turn the remarks over to Dave Matthews, 4 who will be making our presentation today, also at the table 5 we have Mark Satorius, who is the deputy for the Office of l 6 Enforcement, and then Sam Collins, who is has a few opening 7 remarks.  !

8 MR. COLLINS: Madam Chairman, Commissioners, the ,

9 majority of my opening remarks have been covered in some i 10 context. I would just acknowledge that today's meeting is 11 part of the continuing dialogue and interaction with the 12 Commission on these important elements of the regulatory I'

i 23 of 40 6/10/98 7:39 AM

__- - - - _ ~

13 process.

14 The staff is and has demonstrated in tne past it 15 is willing to continue discussions with NEI on guidance 16 documents for implementation. As Hugh so noted, we are here 4 17 today to provide the Commission progress reports since the '

18 last meeting that was held on this topic in December and to 19 respond to the Commission's questions. With that, I will 20 turn the briefing over to David Matthews.

21 MR. MATTHEWS: Good afternoon 22 [ Slide.)

, 23 MR. MATTHEWS: In the interest of time, I do have 24 a slide on background which I will have you look at briefly, l l 25 and then I am going to discuss the central issues that we l 59 1 have today on the updated FSAR guidance and 10 CFR 50.59.

2 Recommendations were provided to the Commission in 3 these areas, among others, in 97-205 in September 1997. We  !

4 had an immediate action shortly after that to address a 5 problem relative to regulatory stability in terms of the 6 treatment of USQs during periods when a plant might be shut 7 down and needing to restart and their relationship to safety 8 and operability.

9 The Commission approved and we issued a revision 10 to Generic Letter 91-18 to address that issue. I've heard 11 feedback from many arenas that that was long overdue, well 12 needed, and has resulted in an increased amount of stability 13 in terms of the treatment of USQs end their relationship to 14 operability.

15

• We provided a briefing tu the Commission in 16 December, which this is in effect an update to. We have 17 provided the Commission a proposed generic letter which 18 would address interim guidance on updating of ESARs in 19 accordance with 50.71(e).

20 Prior to that time, because of concerns associated j 21 with the enforcement policy and its relationship to '

22 treatment of violations under 50.59 we established an 23 enforcement panel. We did that in November of 1997 We did 24 it by the instrument of an enforcement guidance memorandum jl l 25 which is publicly available that we issued at the end of l 60 i 1 October. It guides the enforcement discretion the staff 2 exercises under the current policy when dealing with issues l 3 of 50.59 and attempts to resolve concerns associated with l 4 relative safety significar e. i 5 . We also have a draft rulemaking proposal on 50.59 6 that has matured to the point that it is out for office 7 concurrence throughout the NRC and is with the Office of 8 General Counsel. This is with a goal of providing the l 9 Commission a draft rulemaking package by the requested date 10 of July 10th. At the present time we are on track to i 11 provide you that rulemaking package as requested.

I 12 (Slide.)

l 13 MR. MATTHEWS: Turning now to the FSAR, I wanted

14 to provide a little context for future discussion and l

15 reminding everybody that the FSAR serves several purposes in 16 our regulatory structure.

17 The requirements are outlined in 50.34(b) relative 18 to its contents, and 50.71(e) relative to its updating.

19 However, the last four bullets on this slide 20 indicate that it is relied upon in many contexts, one of 21 which is the scope of $0.59 in that 50.59 limits its scope 22 to the facility as described in the safety analysis report.

23 But it is also relied upon as a reference for the vast 24 majority of licensing actions that the NRC undertakes in 25 response to licensee requests for amendment.

24 cf 40 6/10/98 7:39 AM

.-_---_~-

61 1 It is also used as a document for NRC inspectors 2 in that it describes the facility and is utilized in many 3 different ways to implement our inspections procedures 4 associated with an examination of the conformance of the 5 facility with the agreed upon licensing basis of the plant 6 as reflected in the FSAR.

7 I mentioned 50.59. It also, as was mentioned 8 earlier today, is a document that forms a basis document as

, 9 we move into a renew -d license arena, and the license 10 renewal rule in Part 54 envisions that it would be

! 11 supplemented as described in that rule and then continue on 12 as one of the foundations for describing the licensing basis 13 for a renewed license.

14 CHAIRMAN JACKSON: Let me ask you a question.

15 Does 50.71(e) apply directly to the license renewal 16 supplement?

17 MR. MATTHEWS: Yes. Not by its word, but the 18 license renewal rule indicates that 50.71(e) applies as well

19 to the supplement. So indirectly I would arlue 50.71(e) l 20 app]ies.

( 21 CHAIRMAN JACKSON: The scope of 50.59 is described 22 as the " facility -- in the safety an.31ysis report." Does 23 that mean the FSAR, the FSAR and other documents, the 24 updated FSAR7 25 MR. MATTHEWS: It means the facility's FSAR as 62

! 1 described in 50.34(b) as updated in accordance with 2 50.71(e).

3 CHAIRMAN JACKSON: Right.

l 4 MR. THOMPSON: That's clear.

l 5 MR. MATTHEWS: Let me now turn to slide 5.

6 (Slide.]

l 7 MR. MATTHEWS: In developing guidance on the l 8 updating of FSARs to provide additional guidance beyond that 9 which was available with regard to the implementation of 10 50. 71 (e ) we examined alternative approaches to providing 11 this guidance. The staff concluded that guidance could be 12 provided that would provide enhanced recognition of the 13 requirements of 50.34 and 50.71(e) and at the same time 14 would provide some needed stabilization to the issue 15 surrounding uncertainties as to what should and shouldn't be 16 in a FSAR.

17 We also concluded that there was benefit to going l 18 f6rward with this guidance at this point in time and there 19 wasn't a need for rulemaking to address the problems that 20 had been identified.

( 21 That's a long way of saying that the issues that  !

l 22 had been raised associated with information that was l l 23 contained in the FSAR whose safety significance might not be 24 all that evident or reflected obsolete, outdated or 25 historical information could be dealt with under the current i 63 l 1 regulations by treating it in a different category, putting I l 2 it in a different appendix or a different portion or 3 formatting of the FSAR, and therefore increase the utility 4 of the FSAR for the purposes which I described in a prior i 5 slide.

6 Yes, rulemaking could have been undertaken to 7 eliminate the need for some of that information from even 8 ueing in the FSAR. The staff didn't feel that it was really 9 a worthwhile use of the Commissien's resources to undertake 10 that rulemaking given that we think the purposes could be 11 served by this interim guideae in that regard.

12 CHAIRMAN JACKSON: ;ould our consideration of the 25 ef 40 6/10/98 7:39 AM

13 information required be limited to the information 14 originally required by 50.34(b)?

15 MR. BURNS: No.

16 MR. MATTHEWS: I think our consideration of 17 information_always ought to be confined by the description 18 of information in 50.34(b). That information may have 19 changed over time and therefore the updating requirements 20 would possibly include more information, but not information 21 of a different type.

22 CHAIRMAN JACKSON: So that cevers ATWS and all 23 these other things that we talked about?

24 MR. MATTHEWS: Yes.

25 COMMISSIONER McGAFFIGAN: Are you going back to 64 1 that slide you were on?

r 2 MR. MATTHEWS: Yes. I did just want to indicate l 3 that the guidance would be applicable to plants undergoing 4 decommissioning in terms of updating.

5 We did propose in connection with that revised 6 guidance to the Commission that enforcement discretion be 7 applied in the following way, and we have discussed this 8 already to some degree.

[ 9 In light of the fact that we have already had a 10 longstanding involvement in the issue of improving the 11 accuracies of FSARs stemming way back to a policy statement 12 that the Commission issued and an attendant or related 13 enforcement discretion that was granted relative to that 14 accuracy, our view is that with regard to the information 15 that is in an FSAR or should have been there relative to the 16 plant as it stands today that the FSAR should be accurate by 17- the deadline that the Commission imposed by virtue of 18 granting the original discretion.

19 Although accuracy and completeness we could argue 20 semantically, the staff adopted those terms really for 21 convenience. The concept, I think, is generally accepted 22 that the FSAR should describe the plant as it is built and 23 being operated. That's accuracy. If there is information 24 that should have been included in the FSAR even though it 25 might exist somewhere else, we think it ought to be included 65 1 within the FSAR at one location, and that's completeness.

2 With regard to the issue of completeness, f 3 particularly in light of the enhanced guidance we are 4 proposing to be provided, we proposed a two-step process, l 5 that the material of the highest safety significance, and we 6 would use as a descriptor of that the description that was 7 utilized in the regulatory guidance we published associated 8 with the maintenance rule, ought to be included within the 9 FSAR within six months of issuance of the final generic 10 letter, and we think that an additional period of time could 11 b's provided for the information of less significance.

12 i COMMISSIONER DIAZ: Could you clarify for me this 13 enforcement discretion? Is there a compelling health and 14 *afety issue by which the accuracy is demanded by 10/18 and 15 d.9 completion of high safety significance six months later?

16 Is there any reason why we should maintain that?

17 MR. MATTHEWS: I have a personal view on the 18 significance of accuracy. Given the use of the FSAR and 19 that it is relied upon as a description of the plant, 20 sometimes to the exclusion of actually going out and 21 looking, I think accuracy does have a significance, and I .

22 think it probably goes beyond completeness in terms of that 23 significance.

24 MR. COLLINS: Commissioner, broadly looked at, the 25 staff's main focus would be that if it's being used, it 26 cf 40 6/10.98 739 AM c

l 66 1 should be accurate. To what extent we are amenable to 2 enforcement discretion is in fact, I believe, a resource 3 question; at what point do we believe it is necessary to '

4 focus the industry's resources on this type of a g 31 within 5 a defined period. I think that in and of itself is a matter 6 of some discretion by the Commission to determine how 7 exactly do we want to dictate the industry use those limited 8 and vital resources, because *Fis is an important topic.

9 But day to day, gib that this document is 10 utilized, it should be viewea as being accurate when it's 11 used. So I think there is a window in there. What that is 12 is probably a matter of some discretion and Commission s

13 guidance. I 14 CHAIRMAN JACKSON: This is what you are proposal I

. 15 is in terms of what you call the risk informed.

16 MR. THOMPSON: That reflects the staff's current 17 proposal. There are some judgments in that.

18 MR. MATTHEWS: This does reflect a phased I 19 approach, which I think is responsive also to the staff's i 20 concern that we not come upon another two-year deadline, 21 then look, find we are not there, and our choice has become 22 very limited, and then you have to ask the safety 23 significance question, and if you can't answer that it's i 24 overwhelmingly safety significant, what action do you take l 25 at that point?

67 1 I think the idea of periodic checking and feedback 2 is ibportant, which this does.

3 COMMISSIONER McGAFFIGAN: I want to ask a 4 technical question as to how all this relates. Even if we 5 take the accelerated NEI approach, which I think in answer 6 to the chairman they said by the end of year they would hope 7 you would be in a position to endorse 98-03, you get into a 8 situation where the final guidance may not be out, whether 9 it's by the generic letter, which NEI would say is a slower 10 approach, or this convergence that appears to be occurring 11 where you would endorse in a reg guide their language.

12 Should we pragmatically exercise enforcement 13 discretion at least to the point where a document gets out 14 that everybody agrees on? Does that bear on the accuracy 15 issue?

16 MR. SATORIUS: One thing that I think is important 17 to point out is that the enforcement policy as written today 18 provides for discretion as long as licensees are finding and 19 fixing these problems in the FSAR. So we have discretion 20 available to us beyond what we would propose here. That 21 would continue to be available for the staff to utilize.

22 COMMISSIONER McGAFFIGAN: So what you have here is 23 a blanket enforcement discretion which you propose to 24 terminate at some point, and then you have remaining within 25 the policy some discretion to use even after the blanket 68 1 discretion terminates.

2 CHAIRMAN JACKSON: Right, but that would be true 3 even after 10/18/98.

l 4 MR. THOMPSON: That's true. If it's 5 self-identified, if they have an aggressive program that 6 they are looking hard and identifying the errors, they get 7 credit for that.

8 CHAIRMAN JACKSON: But does not the current 9 enforcement policy also have a risk gradation built into it 10 also?

11 MR. SATORIUS: It utilizes risk-informed 12 information in order to make our determinations.

27cf40 6/10/98 7:39 AM

ll 13 COMMISSIONER McGAFFIGAN: But then you are 14 expending resources of your own and the licensee, saying 15 it's a 3 that deserves to be a 4 or a non-cited, et cetera.

16 MR. THOMPSON: That's a process that we would not 17 necessarily have to go through.

18 CHAIRMAN JACKSON: You could ask this question.

19 Is there enough of a distinction between accuracy and 20 completeness, in your minds, to be able to draw this line at 21 10/18/987 22 MR. SATORIUS: I am a member of the 50.59 review 23 panel, and myself and other members of the staff hear every 24 proposed 50.59 violation, and we do determine that there are 25 some that you can say there is an accuracy issue here or 69 1 there is a completeness issue here. I think the answer to 2 your question is, yes, we can determine the difference 3 between the two, and we have been able to do that.

4 CHAIRMAN JACKSON: But in your opit. ion, as you 5 have gone through it, are licensees able to consistently 6 draw a distinction between the two so that we aren't just 7 creating problems for them and problems for us?

8 MR. COLLINS: I think it's probably not 9 appropriate to ask licensees to have a program that is 10 formulated that way such that they would have to focus 11 resources on accuracy versus completeness. I'm not smart 12 enough, for example, to be able to dictate that to happen.

13 I think the goal would be for the documents to be both 14 accurate and complete at a given point in time, which is at 15 the discretion of the Commission, given the licensee's best 16 use of resources, with a caveat that if the document is to 17 be used to make regulatory decisions. then that portion of 18 the document needs to be accurate.

19 CHAIRMAN JACKSON: That's why you really want to 20 put this 10/18/98 here.

21 MR. THOMPSON: Yes.

22 MR. COLLINS: That was the original thought.

23 MR. THOMPSON: It might be helpful if we take a 24 look at and give you some examples of the types that fall 25 into the two categories so at least you could have available 70 1 to you how we bend those.

2 CHAIRMAN JACKSON: Okay.

3 COMMISSIONER DIAZ: If I may go a little bit 4 further in time and look at 2/28 and the fact that that is 5 al'so a deadline date that the Commission has set to get 6 clarification on the scope and how all the things are coming 7 together, and the fact that, as we all understand frem the 8 50.59, the real issue is definition, how do you define

! 9 things so that people can actually work with them?

10 Would it be appropriate to be as strict as we want 11 on a date in which we have really defined what the 12 requirements are, be it 10/18 or 2/28 or six months later, j 13 whatever it is that is appropriate, but without ambiguity l 14 and "Oh, I didn't understand it was accuracy or this was

( 15 completeness" or we just frame it at one point and say this i 16 is it?

17 MR. THOMPSON: That's certainly one approach that 18 we could do. We gave you our recommendation. For the 19 reasons we said, it's our best recommendation right now, but 20 that doesn't mean that there is not merit in other 21 approaches. We have worked with this issue a fairly long 22 time. There is enough information available, enough 23 guidance available that the dates that are spelled out in 24 our proposal are, I think, doable in most cases.

• 5 i COMMISSIONER DIAZ: Okay. l I

i i 28cf 40 6/10/9R 7:39 AM

71 1 MR. THOMPSON: There may be some people that 2 started late and didn't have it, or they may have a bigger 3 p oblem than we originally anticipated, but as I said 4 earlier, if they are really working at it and they are 5 self-identifying it, we think that the current enforcement 6 policy gives us flexibility and gives them flexibility not 7 to face escalated enforcement.

8 CHAIRMAN JACKSON: Okay.

9 [ Slide.] ,

10 MR. MATTHEWS: On page 6 I wanted to summarize the 1 11 staff's review to date of NEI 98-03. We shared the 12 substance of these reservations with NEI the last time we 13 met with them, and that is why they indicated the- were well 14 aware of what the staff's views were.

15 In addition, of course, they had the benefit of 16 seeing the draft generic letter, which would have also 17 articulated to them what differences there were between that 18 and 98-03.

19 We did perform a preliminary review in response to 20 their request for our review and endorsement of 98-03, and 21 we initiated that review in November when we received that 22 document.

23 The document that we received in draft form as 24 originally proposed the staff would not be able to endorse l 25 short of also proposing changes to our rules with regard to 72 1 tne content of FSARs. We are receptive to the possibility 2 of ehdorsing a revised 98-03 if it is revised to address 3 those particular shortcomings.

4 We think there is a path to resolution in terms of 5 coming to an agreed upon NEI document, but to some degree 6 that is very heavily dependent upon their ability to respond 7 to us with a document that reflects those changes, and we 8 don't have an estimate right now of how soon they will be 9 able to do that, although they have committed that they will 10 try to get us back a document very shortly.

11 We committed to provide them our preliminary 12 comments such as they have been developed to date. We even 13 proposed that since we shared them with them orally in a 14 public meeting, we may attach the description of those 15 concerns to the back of the meeting summary so they would 16 have that, and that we would endeavor to deliver a letter to 17 them very shortly thereafter articulating the concerns with 18 more specificity to give them a basis for making changes 19 that will have the effect of hopefully being as close as 20 possible. 1 21 In terms of overall schedule, though, I think you 22 heard from them an estimate of December as a possible 23 schedule for bringing that to closure. Recognize that the 24 staff, if it is to endorse a document like that, would need 25 to endorse it by means of a reg guide, and we have to have 73 1 public participation in that process. So we would be faced 2 with coming to resolution with NEI, then issuing a draft 3 regulatory guide proposing to endorse'NEI's document, 4 receiving the public comment on same, and then going through 5 the final steps associated with a final reg guide.

.6 That process is one that we can proceed on. We 7 outlined that process, by the way, in the Commission paper 8 and did so in some detail.

9 The staff believes, however, that issuing the 10 draft generic letter for public comment as it has been 11 proposed to you and then issuing that in final form could 12 take place as soon as four months after your agreement with l

I I 29ef40 6/10/98 7:39 AM

13 the contents of that generic letter. This would also have 14 provided public comment during the summer on that document 15 and would have met that need.

16 At a later date, then, that generic letter could 17 in effect expire as far as its utility is concerned once we 18 endorse a reg guide that would have been developed in 19 parallel with that process.

20 That's a long way of saying that we think we can 21 move in a parallel process, and would succeed with interim 22 guidance being out there sooner.

23 COMMISSIONER McGAFFIGAN: I don't totally 24 understand that. I'm looking at the paper. From the date 25 of Commission approval of the generic letter, which hasn't 74 1 occurred yet, it's 150 days to issuing the final generic 2 letter. So we are talking close to the end of the year in 3 any case; we are talking November. And we have run a 4 parallel process that could be resource intensive. You are 5 talking about workshops and all that.

6 We will have a public process if you get to the 7 reg guide, as you said, a reg guide endorsing 98-03, but we 8 don't have two competing documents out there, 98-03 as they 9 try to adjust it to meet the staff's desires, and this 10 generic letter simultaneously out there.

11 MR. MATTHEWS: I have on the one hand a generic 12 letter that has already been prepared that you've heard from 13 NEI they have no problem with in terms of its content. It's 14 ready for issuance. So it can get out very quickly. I 15 have, on the other hand, the possibility that we are going 16 to be able to reach closure with NEI on a document I haven't 17 seen the nature of yet. So I have a little difficulty in 18 being as certain with one date as I am with the other.

19 I feel more comfortable with the staff's ability 20 to issue the generic letter and go through that process of 21 public participation and workshops than I do on setting a 22 date for when I'm going to be able to issue a draft reg 23 guide.

24 COMMISSIONER McGAFFIGAN: What process did we 25 follow in the maintenance rule? There we endorsed an NEI 75 1 reg guide, right, 94-01, or something like that?

2 MR. MATTHEWS: Tom Bergman was intimately involved 3 in that process.

4 COMMISSIONER McGAFFIGAN: Did we start with a 5 generic letter and have competing documents?

6 MR. BERGMAN: There was a parallel reg guide very 7 early in the process. NUMARC 93-01 eventually took it over 8 and the staff never issued that original regulatory guide.

9 Going from one revision to another even of NUMARC 10 93-01 is a great deal of work. If you look at Reg Guide 11 16160, Rev 2, after several years of work we still had about l 12 a dozen exceptions or augments to 93-01 that we took in that i 13 reg guide.

l 14 That process to go from Rev 0 to Rev 2 -- Rev i 15 was withdrawn shortly before Rev 2 came out of NUMARC 93-01 16 -- was a good year of work between the staff and NEI to come 17 up with a workable Rev 2 to 93-01, and we still had to put a 18 lot in the reg guide. The scope of this FSAR thing is at 19 least as comprehensive as 93-01.

20 MR. MATTHEWS: If there are no more questions on 21 the FSAR updating, I'd like to turn to a discussion of 22 50.59.

23 (Slide.)

24 MR. MATTHEWS: Sl:1e 7 summarizes in bullet form a

[

25 paraphrasing of the Commissi:n'.s SRM on 97-205 as it l

l I 30 cf 40 6/10/98 7:39 AM

76 1 pertains to 50.59. There were many other issues in that 2 SRM.

3 (Slide.}

4 MR. MATTHEWS: I wanted to summarize what the 5 staff has done in response to that SRM in connection with 6 50.59.

7 We have prepared a proposed rule package which 8 addresses the following issues:

9 It adopts an approach for allowing for minimal 10 increases in probability and consequences.

11 It establishes a definition for acceptance limits 12 for defining margins.

13 It introduces the possibility that we would allow 14 equipment malfunctions with a different result, which is 15 consistent with an NEI view that that is more important than 16 malfunctions of a different type, and we agree with them.

17 It also addresses collateral changes to Part 72 18 because the rules are parallel with regard co ISFSIs and 19 spent fuel storage.

20 We have before you in a COM SECY of a number I 21 can't recall right now three remaining questions that 22 stemmed from that SRM.

23 That is, your suggestion that we consider 24 including a provision that would permit accidents of a 25 different type with minimal safety impact; if they were 77 1 identified, that the change, if it were to result in that, 2 would still be an acceptable change without NRC approval.

3 That we reconsider acceptanc" limite for 4 consequences. Given that he staff had viewed that 5 acceptance limits for consequences ought to be that 6 reflected as the acceptance limit in the FSAR, NEI has 7 proposed that acceptance limits either established by the 8 SRP, the SER or regulatory limits represent the degree of 9 freedom that they would be permitted to have without NRC 10 involvement. The staff has been opposed to that.

11 I'm going to comment on that because I want to 12 make a summary statement about this issue in a moment after 13 I mention minimal redections in margin of safety.

14 The Commissian also asked us to consider 15 regulatory language that would permit minimal reductions in 16 margin of safety provided you put some limit on the word 17 " minimal." This was a proposal that the Commission made to 18 the staff in that SRM.

19 The staff had not proposed minimal decreases in 20 margin of safety to be permitted. We viewed that permitting 21 that in the light of 50.59 was translating 50.59 from 22 essentially a procedural regulation or a process-related 23 regulation into a safety-related regulation in that we were 24 now going to permit changes to margins of safety that had 25 been established through the licensing process.

78 1 We went into more detail in our memorandum as to 2 the reasoning behind that, but I wanted to then reflect in 3 terms of that issue and the one on acceptance limits for 4 consequences that also raises that same issue in that we 5 felt that that was proposing a change in philosophy with 6 regard to the role that 50.59 plays in our regulatory 7 framework; that we were moving it into an arena that it was 8 starting to become a safety regulation as opposed to what 9 the staff had traditionally viewed 50.59 as buing, and that 10 being one that controlled process and regulatory process in 11 terms of setting thresholds for when the agency needed to 12 become involved and whether to agree with a change or not.

31cf40 6/1048 7 39 AM w____ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ .

13 Many changes that might exceed the threshold for 14 needing staff review still ultimately are acceptable, but 15 when we deal with issues that have started to threaten the 16 margin of safety established through the licensing process, 17 we feel the NRC needs to be involved in those decisions 18 before we agree with that change.

19 CHAIRMAN JACKSON: Thinking outside the box, is 20 there a way to reconcile those two, or should it become a 21 safety regulation in the sense that you are describing it?

22 I know I am putting you on the spot.

23 MR. MATTHEWS: You've heard this phrase. It's a 24 matter of degree and how much flexibility and freedom that 25 the Commission chooses to want to give the industry.

79 1 COMMISSIONER DIAZ: Minimal.

., 2 [ Laughter.]

3 MR. MATTHEWS: If it's minimum, then I would argue 4 the best way to do that would be to confine it to being a 5 procedural regulation, not a safety regulation.

6 COMMISSIONER DIAZ: That might not be responsive 7 to the Commission's original intention. It might be that we 8 might want you to think outside of the box and see not what 9 is traditional, but what is effective, what is protective, 10 and what can be really done.

11 MR. MATTHEWS: I understand that. I think the 12 staff took a hard look at it from that perspective. I don't 13 believe our answer that we provided you in May was 14 conditioned on it just being traditional. I think we 15 actually viewed that there was a potential that margins of 16 safety might be reduced in ways that were unintended, and 17 that the staff, upon having the opportunity to review them, 18 might not have agreed with.

19 Let me go back to the issue of consequences.

20 We've seen charts and boxes, but the essential issue as I 21 understood it, as NEI has described it to us in meetings 22 that predated this one, was that they would like the 23 utilities to have the ability to move from the acceptance 24 limit that they established in their FSAR and that the staff 25 agreed with to another limit if they so chose or as the 80 1 result of a facility change that would be an SRP established 2 limit or a regulatory limit without NRC involvement.

3 Those changes could be significant. They might be 4 so significant as to engender, if they were submitted as a 5 license amendment, our inability to make a no significance 6 hazard claim on that license amendment.

7 If you were to accept the NEI proposal as I've 8 understood it in the past, there was a great deal of change 9 that could be made on their own volition without NRC 10 involvement, and the staff is concerned that it's a greater 11 change than we would want to approve, and may include 12 changes that we wouldn't approve were they submitted.

13 COMMISSIONER McGAFFIGAN: I want to stay on this 14 consequence thing. I'm going to bo back to the first one 15 too at some point, accident of a different type.

16 You heard NEI earlier today say it may be a matter l

17 of degree, that they understand if the limit is 50 and they 18 are at 46.5 that you might want to know about it. What they 19 don't understand is if the limit is 50 and you are 21.9 20 going to 22.3, whether that's a big deal and whether we are 21 wasting our and their resources on that. I heard that there 22 is some middle ground here, that a line in the sand sort of 23 approach may not be the right thing.

24 What has confused me all along is why what they 25 say in NEI 96-07 with regard to the basis for the tech spec, 32cf40 6/10/98 7:39 AM

81 1 where they say the same thing, that they find acceptance l 2 limit either in regulations or SERs or standard review 3 plans, and that's okay with you guys; that's what Sam i 4 Collins' letter said in January; but when it comes to the 5 consequences where they use almost verbatim verbiage it's 6 not okay.

7 I think Mr. Ray said occasionally 20 going to 22 8 might be significant. How do we define that without having 9 everything submitted, without having every change of what 10 may be a trivial nature submitted to us?

11 I remember Commissioner Rogers when we dealt with 12 tritium and the .2 millirem increase when the acceptance 13 limit was rems, a .2 increase in consequences under some 14 design basis accident scenario at the site boundary. When 15 we get through this process, that can't be, and I know it 16 won't be, because that can't be a 50.59 unreviewed safety 17 question. That is one metric by which we can judge whether 18 we have succeeded.

19 Is there a middle ground there?

20 MR. MATTHEWS: I think the staff in the Commission 21 paper, and I'm ready to discuss it here, was going to 22 propose a middle ground.

23 I have to be honest. The statements of Mr. Ray 24 with regard to that issue on consequences was the first time 25 that we heard that kind of view out of 1EI, because previous 82 1 to this point in time it has been on consequences, an issue 2 of whether or not they could move from the acceptance limits 3 identified in the FSAR up to those values you just 4 described, SRP, SER, or regulatory limits, without staff 5 review.

6 Our view is that a minimal change, as the 7 Commission had suggested, is the way to put a limit on that.

8 We have a view that is still undergoing staff review of what 9 would put some limits on minimal increases. What I am 10 giving you is the opposition to a position we've heard to 11 date from NEI that they be allowed to move to the regulatory 12 limit or to the acceptance limit in the SRP irrespective of 13 the licensing review that was conducted on the acceptance 14 limit they offered in the ESAR originally.

15 We think the FSAR value provides a very sound 16 basis for determining the baseline from which we ought to 17 measure minimal, and we would suggest that is a good 18 baseline for a regulatory process in terms of when the NRC 19 ought to get involved.

20 COMMISSIONER DIAZ: Would you say that again? I'm 21 sorry.

22 MR. MATTHEWS: We think the value that the

! 23 licensee offered in the FSAR with regard to the consequence 24 attendant to a given design basis accident is a good 25 baseline from which to measure a minimal increase which will l 83

! I attempt to put boundaries around through the rulemaking L 2 process, and that will provide the flexibility.

3 L What we are opposed to is a position that would 4 allow increases from'that FSAR value up to regulatory 5 limits, SRP limits, or if the staff had done a SER 6 evaluation that came up with a consequence that exceeded the 7 FSAR value. We view that since this is a regulatory 8 threshold that we should only hold them accountable for the 9 analysis that they did themselves, namely, the FSAR value, 10 and the changes about that are what need to be examined.

11 COMMISSIONER McGAFFIGAN: When I once asked the 12 question, why is 96-07 okay when it discusses acceptance 33 cf 40 6/10/98 7:39 AM i

_ _ _____. . - - ~ -

13 limits for margin of safety as defined in the basis for any 14 tech spec and why it isn't okay for consequences, one of the 15 answers I got was a legal answer: because (2) (i) mentions 16 the words "previously evaluated in the safety analysis 17 report" and (2) (111) , where the margin of safety is defined 18 as the basis for tech spec doesn't talk about "as previously 19 evaluated."

20 Is it the legality that leads you to the 21 rejection, or is there a substantive reason as to why NEI is

, 22 okay in looking to acceptance limits as they have defined l 23 them in (iii) but they are wrong on (i)?

24 MR. MATTHEWS: It's a good question. What I have 25 been speaking to is the issue of minimal increases in 84 1 consequences attributable to that portion of 50.59 that 2 describes that.

3 On the other issue, margins of safety, our view is 4 that the margin of safety, which is the difference between 5 the value proposed and accepted by the staff and some 6 ultimate design value or some regulatory limit, that is an 7 established differential that should be maintained.

8 Otherwise we are in effect changing the philosophy of 50.59 9 and allowing margins of safety to be decreased voluntarily.

10 So we think you need to hold the margin of safety.

11 However, there are instances where calculations 12 are done associated with consequences related to design 13 basis accidents for which a plant change that would result 14 in a minor change in those accident consequences would be 15 acceptable.

16 COMMISSIONER McGAFFIGAN: Maybe I misinterpreted 17 Mr. Collins' letter from January. Didn't you endorse NEI 18 96-07 as it dealt with acceptance limits for margin of 19 safety?

20 MR. MATTHEWS: Yes. That's not a problem. They 21 didn't propose any flexibility on that point and NEI has 22 never proposed any flexibility.

23 COMMISSIONER McGAFFIGAN: I thought they did.

24 MR. MATTHEWS: Not that I'm aware of.

25 COMMISSIONER McGAFFIGAN: Maybe I'm misreading 85 1 96-07.

2 MR. MATTHEWS: Can you clarify, Tony?

3 MR. PIETRANGELO: I think what David just 4 described we would disagree with. The example I would cite 5 is what has been in NSAC 125 since 1989 and is still in NEI 6 96-07 It's a containment heat pressure example that the 7 Chairman went over. Clearly in that we would not call the 8 margin of safety the difference between the calculated value 9 in the acceptance limit.

10 It's back to our box chart again from the 11 acceptance limit to the failure point. He referred to that 12 as the margin of safety, and we would disagree with that.

13 That has never been the industry position.

14 COMMISSIONER McGAFFIGAN: They are endorsing you 15 but they are using a different definition.

16 MR. PIETRANGELO: That's correct.

17 COMMISSIONER McGAFFIGAN: So you didn't really 18 endorse them on that.

19 CRAIRMAN JACKSON: You endorsed the words but you 20 have to give definitions.

21 MR. MATTHEWS: Yes, and I've indicated that here, 22 that we had proposed that we would provide a definition of 23 acceptance limits to be applied to calculations of margin of 24 safety. We may disagree with NEI with regard to what margin 25 we are talking about. They would like it to be a margin 34 cf 40 6/10/98 7:39 AM

-- - -- _ _ n l

l l 86 1 outside of a box that we would choose, but I don't think 2 there is any disagreement with regard to the existence of a

! 3 terminology of minimal decreases in margin of safety.

4 COMMISSIONER McGAFFIGAN: One last question. On 5 the accident of different type with minimal safety impact, I 6 read the staff paper as at least being willing to go to "is 7 created" as opposed to "may be created," which is what we 8 say in Part 60; is that correct?

9 MR. MATTHEWS: Yes. Understand there is an 10 attendant consequence to that, 11 COMMISSIONER McGAFFIGAN: I understand. But it 12 has the beauty of at least being consistent with what we did 13 and defendable, and we're not asking for speculation on a 14 licensee's part.

15 MR. MATTHEWS: You are correct in that regard.

16 The possible unattractive consequence is that if they were 17 to fail that test and bring that to us as an amendment, it 18 is not an amendment that we could issue as a no significance 19 hazard amendment. If a hearing were requested and accepted, 20 we would have to hold that hearing before we could issue the 21 amendment because of the nature of 50.92.

22 CHAIRMAN JACKSON: Why don't we go on.

23 (Slide.]

24 MR. MATTHEWS: On the next slide we provided some 25 preliminary views. I put it that way because this is just 87 1 that, a proposal on how to deal with issues associated with 2 minihal increases of probability and consequences.

3 These are conditioned somewhat by our interaction 4 with the industry and the Commission and the public on the 5 development of the reg guides associated with risk-informed 6 licensing decisions.

7 In there, as you know, there is a terminology of 8 "very small" wherein the staff would entertain license 9 amendment requests that would allow very small increases in 10 probability relative to core damage frequency and large 11 early release fraction. So to some degree we are talking 12 about potential changes being in what I would refer to, as 13 you've heard it before, the negligible category, below those 14 levels when you are dealing with similar parameters.

15 We have proposed that we permit increases of 16 probability of accidents in the order of one percent without 17 the need for any NRC review.

18 +

With regard to reliability or probability of f 19 equipment malfunctions, we think that there could be a 20 graduated establishment of threshold based on safety 21 significance.

22 With regard to consequences, we have attempted to 23 address this issue that consequences, when they are very far 24 below regulatory limits, we are probably more flexible on 25 than consequences that start to approach regulatory limits.

88 1 CHAIRMAN JACKSON: How are licensees going to 2 reach these numerical conclusions? Are we saying they have 3 to have at their disposal the way to numerically determine 4 the changes in risk to support compliance?

5 MR. MATTHEWS: It's going to be a challenge. I 6 think we would expect that they would invoke the same level 7 of precision on this problem as they did with regard to the t 8 initial calculation that formed the basis for the value in 9 the FSAR. You only have the tools that you have.

10 CHAIRMAN JACKSON: How do you go about verifying 11 the adequacy of a licensee's assessment? Is the inspection 12 staff going to do that, and uw are they going to do that?

35 ef 40 6/10/98 7:39 AM l

13 MP. KATTHEWS: That is going to be a daunting task j

14 as well, but the inspection staff is going to have to at l

15 least recognize that this is going to form a basis for licensee decisions and be able to appreciate the 16 17 reasonableness with which they have done that. We are going ~ l 18 to have to have inspection guidance. That will have to be a 19 companion piece. l 20 CHAIRMAN JACKSON: You're going to have to kind of 21 take selected systems based on risk and have the SRAs take a 22 look at it. l 23 MR. KATTHEWS: Yes.

24 MR, THOMPSON: It would be that type of approach 25 and probably a sampling basis.

89 1 CHAIRMAN JACKSON: Okay.

2 [ Slide.]

3 MR. MATTHEWS: You've heard from NEI with regard 4 to their view with regard to the concern they have in that 5 they think that the scope of 50.59 ought to be moved away 6 from the FSAR.

7 We think that the criteria changes that we have 8 proposed ought to go forward now. We haven't examined 9 alternatives for scope beyond those we discussed with you 10 last year with regard to going to some sort of risk informed 11 perspective associated with essential information or less 12 essential information.

13 -

NEI has offered us the outline of a proposal in a 14 meeting a month or so ago which we are willing to go further 15 with*and examine. Our idea was that during the course of 16 the year, as they flesh that proposal out, we'll be very 17 receptive to hearing that and seeing whether or not it 18 provides a feasible alternative.

19 (Slide.)

20 MR. MATTHEWS: With regard to 50.59 enforcement 21 discretion, you asked us to exercise discretion during the 22 period of any rule change associated with 50.59. We would 23 propose to continue the current policy, which does have 24 provisions for discretion.

25 We have added to that, as you know, the 50.59 90 1 enforcement panel, which meets as needed on every 50.59 2 enforcement action. We would propose as a result of our 3 experience gained in that panel to bring back to you in our 4 July rulemaking package the criteria that have evolved and 5 that we would propose we would embrace in more concrete form 6 for this purpose.

7 COMMISSIONER McGAFFIGAN: I understand what you 8 are doing is consistent with the reading of the SRM based on 9 what you had originally proposed.

10 3rly What we did on the FSAR update where we had this 11 _ket time period during which we were trying to solve a 12 . lem until October 18th of this year, or whatever other 13 time we decide, and then have discretion after that time ,

14 clock is over with, why is that approach not what the staff 15 originally proposed in 97-205, and why shouldn't we at least 16 consider the approach that has been suggested at least for 17 minimal increases in probability? .

18 I think for consequences Tony was trying to slip I 19 one in there on that one.

20 There is a lot of procedure that gets into. This 21 enforcement panel is presumably resource intensive. We are 22 spending a lot of time thinking about whether something is 4 1 23 or non-cited. Maybe the right thing to do is to sort of say 24 they are all minor until proven otherwise and therefore we 1 25 are not even going to bother to write them up if they are in 36 cf 40 6/10/98 7:39 AM

91 1 this absolutely minimal category.

2 MR. MATTHEWS: Let me turn to Mark first.

3 MR. SATORIUS: The short answer is that, quite frankly, they are not all minor. It does take some staff 4

5 review to determine there are a lot of them that are minor.

6 COMMISSIONER McGAFFIGAN: If they fit the category 7 where they are going to be after the rule goes through, they 8 are going to be not only not minor, they are not going to be 9 rules violations at all. That's what the NEI proposal is.

10 For those things that meet a minimal test, or in their view, 11 a negligible test, we shouldn't be spending a lot of 12 resources on them, at least as regard to probability. With 13 consequences there may be some.

14 CHAIRMAN JACKSON: I think, though, in order for 15 you to do that -- I'm not the lawyer -- you hav- to exercise 16 discretion with respect to something. It strikes me that 17 you then basically have to consider whether you want to tell 18 the staff to, on an interim basis, adopt NEI guidance 19 vis-a-vis what a minimal increase would be, barring modulo 20 working through what the ultimate situation is going to be 21 with the rule. They have to operate off of something, if I 22 take Mr. Sartorius' point of view that not everything is 23 necessarily trivial. So you have to have some guidance that 24 you operate off of. The Commission in principle could say, 25 okay, on an interim basis use NEI's guidance and then go 92 1 forth and do the rule.

2

• COMMISSIONER McGAFFIGAN: We could also amend the 3 enforcement policy in some way.

4 How many of the level 4 violations that we are nce 5 criticized for increasing from 500 to 1,400 over the last 6 two years are in this area?

7 MR. SATORIUS: Prcbably about 40 or 50. Since we 8 started the panel process, I think in October or November, 9 we have probably considered 60 or 70 issues total. We 10 average about two or three a week.

11 CHAIRMAN JACKSON: So that's not where the problem 12 is.

13 MR. COLLINS: Commissioner, to respond generally, 14 the staff is not opposed, with proper guidance, EGMs, 15 enforcement guidance memorandums, and training to the 16 inspectors, to providing for a period of implementation and 17 stabilization. During that period I think it's a learning 18 psocess both for the industry and for us as regulators, 19 including the inspectors, to understand how the process is 20 to be implemented, and there is clearly a transition cost 21 with that, and there is a savings in resources over a graded 22 period of time.

23  ; I would propose tt.-t during that period, though, 24 us still go through at least a phased manner of 25 understanding the industry's implementation and testing our 93 1 inspection and our enforcement guidance. That would be 2 probably with a panel very similar to the one that Mark is a 3 member of now. The disposition of those issues, however, 4 would be a matter of discretion.

5 CHAIRMAN JACKSON: Right. J 6 (Slide.] j 7 MR. MATTHEWS: I just wanted to conclude with what )

8 we saw as the next steps. I think they are probably  !

I 9 obvious.

10 We were looking for Commission direction on the 11 proposed generic letter on FSAR updating. Attendant to

12 that, of course, is the acceptance or at least the response

! 37 cf 40 6/10/98 7:39 AM l

l 13 to the staff's proposal with regard to enforcement 14 discretion in the area of ESARs.

l 15 We were looking for a response back to the issues f 16 we raised in our recent May memorandum with regard to l 17 clarification or possible modification of the SRM in several l 18 areas.

j 19 We have, as I mentioned, a rulemaking package l

20 embracing those elements that I referred to. That is very 21 far along. We met with the ACRS this morning. We are 22 meeting with the ACRS again, I believe on the 17tn of June.

23 We are scheduling CRGR review of that rulemaking package, 24 and we expect that to come to closure quickly.

25 And we are going to continue interactions with NEI l 94

1 with regard to all of these topics. They have several l 2 guidance documents on our plate that relate to this issue.

3 They've got 97-04 with regard to design basis issues;

(

4 they've got 96-07 with regard to 50.59 issues; and they've 5 got 98-03 with regard to FSAR issues. So we are actively 6 reviewing and working with them on these documents in the l 7 hopes that we can come to agreement on industry guidance 8 that could conform to our current regulations or those that 9 have a likelihood of being imposed.

10 COMMISSIONER McGAFFIGAN: Can I ask a process 1] question? It came up at the reg info conferences, putting 12 these documents out as they come to us and the generic 13 letter which you got permission to do, and you had the May 14 27th meeting, the 50.59 memo that you've given to us for 15 resolution. I've noticed over the time I've been here this 16 disconnect between what we allow you all to do and what we 17 allow NMSS to do. NMSS is off doing Part 35 rulemakings and 18 putting straw men out on the Web and coming to us 19 occasionally for a little guidance, and a lot of guidance 20 this month. There are various and sundry other quite open 21 processes they run.

22 One of the criticisms that we get is we oftentimes 23 aren't as open on the reactor side, and I understand the 24 Commission over the years has kept you on short leashes on 25 the reactor side, like the design reviews on the modern 95 1 reactors, et cetera.

2 MR. COLLINS: I'm not sure I like that analogy, 3 but I understand your point.

4 MR. MATTHEWS: My image of a short leash is two 5 links.

6 (Laughter.]

7 COMMISSIONER McGAFFIGAN: A metaphor came from one 8 of the staff I talked to.

9 CHAIRMAN JACKSON: Why don't you look into  :

10 creating a 50.59 chat room?

11 .. COMMISSIONER McGAFFIGAN: Would the staff

12 appreciate the greater flexibility in their interactions?

i 13 Public interactions. Not in closed doors, but public l 14 interactions with the regulator on what you all call 15 pre-decisional documents I don't know what Carl calls 16 them, because he gets awa, with a lot more flexibility.

17 CHAIRMAN JACKSON: Let's let him answer it. l 18 MR. THOMPSON: Obviously it is very helpful to l 19 have an open and frank dialogue. What you have to be 20 comfortable with is what the stage and level of maturity of 21 these documents is. What I guess I would like to propose is 22 that we would come back and maybe propose some guidelines 23 for you. We have done that in the NMSS area. We have told 24 you when we are going to put things up on the Web. We have 25 told you when we are going hold public workshops.

38 cf 40 6/10/98 7:39 AM  ;

_________y 96 1 Maybe just put some guicelines out. I think we 2 can do that and give us some more flexibility as well as 3 give you an understanding of how we would decide that.

4 CHAIRMAN JACKSON: The only question and probably

5. why it has been on such a short leash, aside from the issues 6 involved, is to ensure that it is not just one channel, that 7 if it is public and you are dealing with the stakeholder, 8 that you deal with all the stakeholder. There are 9 different constituencies, and NEI is a critical one, but 10 it's not the only ene.

11 MR. THOMPSON: We have special arrangements with 12 Agreement States. They are kind of co-regulators, and we 13 have a certain degree of flexibility there.

14 COMMISSIONER McGAFFIGAN: Another document that 15 some of us, because it's about six inches thick, have been 16 slow to vote on -- the Chairman, give her credit, has --

17 CHAIRMAN JACKSON: That's because I'm a fast 18 reader.

19 COMMISSIONER McGAFFIGAN: It's the decommissioning 20 reg guide. Even as we voted on it it was on the Web page.

21 We must have a pretty big Web page, by the way.

22 MR. COLLINS: Your point is well taken. It's a 23 worthy pursuit.

24 CHAIRMAN JACKSON: Chat Room 50.59.

25 (Laughter.)

97 1

CHAIRMAN JACKSON: On behalf of the Commission, 2 let me thank NEI and the staff for presenting to the 3 Commission the results of their respective evaluations and 4 recommendations for improvements in the areas of FSAR 5 updates and 10 CFR 50.59.

6 The staff's Commission papers on these areas and 7 today's presentations are helpful in describing the options 8' available to addressing these two important issues. While 9 obvious differences remain between the staff and NEI on 10 issues related to 10 CFR 50.59, it's encouraging to note 11 that clarity and agreement are being reached in the area of 12 FSAR update requirements.

13 The conclusions reeched in this area appear to be 14 appropriately focused on meet ng and properly enforcing the 15 existing regulations, ensuring that information is 16 maintained current and that new information is appropriately 17 and accurately included.

18 At the same time, these conclusions allow 19 licensees the latitude to reformat to some degree, to slim 20 down and to simplify their FSARs.

21 With respect to 10 CFR 50.59, things are in a 22 state of flux, but it is clear that the staff is working 23 hard to responsibly implement Commission direction, and the i 24 estent to which the staff's conclusions are adopted will be 25 considered obviously by us in the near future. NEI's l 98 l 1 comments in this area have been helpful in presenting

! 2 alternative approaches to the changes we seek to make, l

3 particularly with respect to 50.59, including the scope 4- issue.

5 Unless there are further comments, we are 6 adjourned. Thank you.

7 [Whereupon at 4:30 p.m. the briefing was 8 concluded.)

9 10 11 l 12 t 39ef40 6/10/98 7:39 AM I