ML20135F130

Transcript of 961206 Public Meeting in Rockville,Md Re Meeting W/Acrs.Pp 1-73.Supporting Documentation Encl

ML20135F130
Person / Time
Issue date:	12/06/1996
From:	NRC COMMISSION (OCM)
To:
References
REF-10CFR9.7 ACRS-GENERAL, NUDOCS 9612120377
Download: ML20135F130 (177)

Text

. _. - _ _ _ __ _. .. . . - _ _ . _ . _ _ . _ . .. _ __

P i

TRANSMITTAL T0: t//ocumentControlDesk,05C-12 D i

- ADVANCED COPY T0: Public Document Room, L Street  :

I

. DATE: /o2 )// /7/

FROM: SECY, Operations Branch Attached is a copy of a public Commission meeting transcript and related meeting document (s). They are being forwarded for entry on the Daily Accession List and placement in the Public Document Room. No other distribution is requested or required.  ;

Meeting Titie: l0] pet h c, p3jyh AcQtAr6yy [cygni 7f& e. On 5 AaacTor s&guarDr 44 bas ) - AL/k ,

/Dee7/n a a

Meeting Date: /02 /6 /9/ ,

Item.

Description:

• Copies  !

Advanced DCS to PDR CoDY 1 1

1. -TRANSCRIPT

'l d5N 1 h dnb/! /

/

2. h 2h l9(, Ah?thn 10 3. [b4 l2 I l

&m T. Jarkch e 3.

4.

5.

6.

120074

• PDR is advanced one copy of each document, two of each SECY paper.

Original transcript with attachments is filed in Correspondence & Records Branch.

9612120377 961206 T PDR 10CFR PT9.7 PDR

\

i

, I

,,g;

.L 1

, UNITED STATES OF AMERICA l NUCLEAR REGULATORY COMMISSION 1

i l

Title:

MEETING WITH ADVISORY COMMITTEE ON

} REACTOR SAFEGUARDS (ACRS) - PUBLIC MEETING Ie l Location: Rockville, Maryland i

i 1

i Date: Friday, December 6,1996 i

i i

Pages: 1 - 73 i

ANN RILEY & ASSOCIATES, LTD. p 1250 I St., N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034

)

I o  !

DISCLAIMER This is an unofficial transcript of a meeting of j the United States Nuclear Regulatory Commission held on December 6, 1996 in the Commission's office at One White Flint North, Rockville, Maryland. The meeting was open to public attendance and observation. This transcript has not been reviewed, corrected or edited, and it may i contain inaccuracies. j The transcript is intended solely for general informational purposes. As provided by 10 CFR 9.103, it is not part of the formal or informal record of decision of the matters discussed. Expressions of opinion in this transcript do not necessarily reflect final determination or l beliefs. No pleading or other paper may be filed with the l Commission in any proceeding as the result of, or addressed to, any statement or argument contained herein, except as the commission may authorize.

l e

4

O 1

1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3 - - -

4 MEETIEG WITH ADVISORY COMMITTEE 5 ON REACTOR SAFEGUARDS (ACRS) 6 ,

7 PUBLIC MEETING  !

l 8 - --

l 9

10 Nuclear Regulatory Commission 11 One White Flint North l 12 Rockville, Maryland i

13 Friday, December 6, 1996 14 l

15 The Commission met in open session, pursuant to 16 notice, at 9:35 a.m., Shirley A. Jackson, Chairman, 17 presiding.

18 19 COMMISSIONERS PRESENT:

20 SHIRLEY A. JACKSON, Chairman of the Commission 21 KENNETH C. ROGERS, Commissioner 22 GRETA J. DICUS, Commissioner 23 NILS J. DIAZ, Commissioner  ;

l 24 EDWARD McGAFFIGAN, JR., Commissioner '

25 l

l l

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

2 1 STAFF AND PRESENTERS SEATED AT THE COMMISSION TABLE: ,

2 JOHN C. HOYLE, Secretary of the Commission 3 THOMAS KRESS, Chairman 4 ROBERT SEALE, Vice Chairman 5 GEORGE APOSTOLAKIS, Member

~

6 ,

IVAN CATTON, Member ,

7 DON MILLER, Member 8 DANA POWERS, Member -

9 WILLIAM SHACK, Member 10 11 12 13 14  !

15

• 16 17 18 -

19 {

l 20 21 22 23 24 25 I

l i

l ANN RILEY & ASSOCIATES, LTD.

Court Reporters )

j 1250 I Street, N.W., Suite 300 '

Washington, D.C. 20005 (202) 842-0034

l

\

. 1 3 ,

1 PROCEEDINGS 2 CHAIRMAN JACKSON: Good morning, Indies and 3 gentlemen. It is a pleasure to meet once again with 4 Dr. Kress and the members of the NRC's Advisory Committee on 5 Reactor Safeguards, who plan to discuss a number of topics 6 of, interest with the Commission at today's session.

7 However, since this is the last Commission meeting 8 for ACRS member Dr. Ivan Catton, I want to pause for just a 9 moment to express the Commission's appreciation for your )

10 eight years of dedicated service to the ACRS and to the i

11 Commission. In fact, I would like to present to you this l 12 letter expressing our sincere appreciation, signed by all 13 members of the Commission, thanking you on behalf of each 14 and all the Commissione s. Thank you very much. l i

15 MR. CATTON: Thank you very much.

16 [ Applause.]

17 CHAIRMAN JACKSON: I also have a plaque.

18 MR. CATTON: Thank you very much. {

19 [ Applause.]

20 CHAIRMAN JACKSON: Over the years the ACRS has 21' provided valuable and timely advice to the Commission on the 22 safety aspects of proposed as well as existing nuclear 23 facilities. I know you are considering a number of topics 24 that are of critical importance to the Commission today. So 25 we are fortunate to be able to draw on your expertise and ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

4 1 the views of you who are a selected group of technical ,

2 experts as we try to solve and resolve various issues in 3 licensing and regulation.

4 During today's briefing we will cover a number of 5 topics and I will just detail them for the record.

6 .. Digital instrumentation and control systems.

7 The plan developed by the Office of Research for 8 upgrading thermal-hydraulic codes.

9 Risk-informed, performance-based regulation and 10 related matters.

11 The potential use of IPE and IPEEE results to 12 compare the risk of the current population of plants with 13 the safety goals.

14 Use of the Commission's safety goals on a 15 plant-specific basis.

16 The use of RuleNet in the regulatory process.

17 Dr. Kress and members of the Committee, my fellow 18 Commissioners and I welcome you to this meeting and 19 anticipate another candid and informative session with the 20 Committee.

21 I understand that copies of the briefing material 22 are available at the entrances to the room.

23 If my colleagues don't have any opening comments, 24 Dr. Kress, why don't you proceed.

25 MR. KRESS: Thank you once again. It is a ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

5 1 pleasure to meet with the full Commission in a way of 2 communicating and exchanging the views and maybe clarifying 3 some of our views for you.

4 I don't have much else to add other than that. So 5 I think we ought to jump right into the agenda. The first 6 it,em is the digital I&c. That, of course, is under the 7 cognizance of Dr. Miller.

8 MR. MILLER: Thank you, Tom.

9 I'm pleased to report that a regulatory framework 10 for digital I&C systems has been issued, and that has been 11 issued for public comment in the form of a Standard Review 12 Plan Chapter 7 update, various regulatory guides, a large 13 group of branch technical positions, and an SER on EMI/RFI.

14 The ACRS met in three different meetings, in 15 March, May and August, to discuss these documents with the 16 members of the staff. I might say that dialogue has been 17 very constructive on both sides. We issued reports to the 18 Commission in June and October following those meetings.

19 We do expect final review of the overall 20 regulatory framework sometime in either April or May, 21 depending on various schedules, public comments and the 22 National Academy study. The Committee also expects 1

23 recommendations made by the Academy study will be factored 24 into the regulatory framework at that time in addition to 25 public comments and an SER on commercial, off-the-shelf P

, ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

1 software. '

2 CHAIRMAN JACKSON: When do you expect the National 3 Academy Phase 2 report to be available?

4 MR. MILLER: The current date that I understand is ,

5 the middle of December. So just in a few weeks. We will, 6 of, course,.have that very quickly, and I'll be reviewing it, 7 as well as other members, during the month of January.

8 COMMISSIONER DICUS: That report was clearly 9 overdue. There is a little bit of information, I think in  !

10 part, at least, because of some of the complexities 11 involved. Is there something else why this is running 12 behind schedule? I had heard perhaps it was going to be  !

13 into next year before it was available.

14 MR. MILLER: Members of the ACRS have really not 15 been too involved in their debates. I think only one member 16 attended one meeting, as I recall. l 1

17 CHAIRMAN JACKSON: I don't think you can really i 18 control that.

19 MR. MILLER: It's their study.

20 COMMISSIONER DICUS: I just wondered if you had 21 anything else on it.

22 MR. MILLER: There are difficult issues. I know 23 some members of that study panel and I know their views, and 24 there are very diverse views on several issues. They are 25 trying to reach a consensus, and that's a challenge at ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

7 ,

L 1 times.

2 MR. APOSTOLAKIS: I attended the meeting. The 3 differences of opinion were significant. So I can see how 4 the report can be very late. Very significant.

j 5 MR. SEALE: I think there is one other point. The 6 Committee was very much involved in the process, working 7 with the staff to initially decide to do the study, that is, 8 to refer the issue to the National Academy, and so on.

9 There was a rather conscious decision that it would be 10 inappropriate for the ACRS then to try to influence the 11 direction of that study. So we've tried to keep our hands 12 off of it and strictly act as observers to the extent that 13 we did have any involvement at all.

14 CHAIRMAN JACKSON: Thank you.

15 MR. MILLER: Thanks for the question. We are 16 anxious to see the report. It will certainly be very 17 important to the regulatory framework.

I 18 There have been a number of issues and concerns j 19 we've expressed in our letters that I would like to review 20 for you. Members of the Committee will probably make 21 comment.

22 One is the level of detail provided in the 23 regulatory guides.

24 The second is the balance and guidance between the 25 review of design process versus assessment or product.

ANN RILEY & ASSOCIATES, LTD.

Court Repor'.ers 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

8 1 And then the linkage of the Standard Review Plan ,

2 Chapter 7 update with other parts of the standard review 3 plan, specifically probably Chapter 18, which is PRA.

~

4 A graded approach is based on importance of 5 safety. I might make a comment. As the various parts of

~

6 the Standard Review Plan PRA, which I know you are all 7 interested in also hearing about, are evolving it is 8 becoming evident that various parts of that might be 9 important to digital I&C systems. So there may be an 10 obvious linkage and obvious supporting situation between 11 those parts.

12 Finally, we had a review of the I&C research 13 program yesterday, and I'm pleased to say a number of issues 14 there addressing both current research programs and planned 15 research programs are important to I&C systems, particularly 16 digital systems. Specifically, software and design, which 17 are a noted weak link in the process of software 18 development, and environmental stressors, which we have 19 various interest in. Specifically, lightning, smoke and 20 EMI/RFI, all being addressed by a research program.

21 I always like to close with a comment. As we look 22 at digital I&C systems, particularly as they relate to 23 safety systems, we always need to look back in history. I 24 always like to look back at IEEE 279, which is part of 25 regulation, which very clearly says safety systems in ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

9 1 nuclear power plants should be and must be maintained very 2 simple. That means digital I&C systems with respect to 3 safety systems have to be simple. In that respect, they 4 should be able to be made safer and very effective in their 5 use in nuclear power plants.

6 ,' CHAIRMAN JACKSON: Let me ask you this question.

7 I'm aware, and you've just made the point, that the 8 Committee has raised concerns about the level of detail 9 provided in the reg guides. Do the guidance documents 10 contain acceptance criteria for things like software reviews 11 and defense in depth and diversity, and if so, how would you 12 characterize those criteria as they are currently laid out?

13 MR. MILLER: I think there is even diversity of 14 opinion among the Committee members. First of all, the 15 regulatory guides fundamentally endorse a set of industrial 16 consensus standards, which really represent current software 17 engineering practice.

18 In that sense, there is some debate about the 19 level of acceptance criteria. I know my colleague 20 Dr. Apostolakis will jump in on that one, because he has a 21 lot of concern about that where the acceptance criteria are 22 correctly specified.

23 I think in the area of diversity and defense in 24 depth, those issues are more appropriately spoken to in 25 branch technical positions rather than regulatory guides, i ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

10 1 George, I'm certain you want to make comments on ,

2 that one.

3 MR. APOSTOLAKIS: I really don't like what I see.

4 I tried to follow the detai]s of the guide. One of the 5 things that strikes you is that you are continually referred 6 to,some IEEE standard. By the time you are done you have a 7 whole pile of standards, each one sending you to another and 8 another and another. I am exaggerating a little bit, but 9 you follow all these references to another guide, another 10 guide, another guide, and at the very end the ultimate 11 advice is make sure you do a good job. To me that's not an 12 acceptance criteria. You could have told me up front that I 13 have to make sure I do a good job.

14 Example. You have to have a plan to review or to 15 develop the software. That's great. Let's have a plan. So 16 you go a little more deeply to see what does that mean. All 17 it says essentially is the acceptance criteria is "here's my 18 plan." Then you say, okay, you have a plan. To me that 19 means nothing. That means absolutely nothing, just to say 20 here is a plan, thank you very much; you do have a plan; 21 let's go on to the next item.

22 I don't see any requirements or any criteria as to 23 what constitutes a good plan. The mere existence of the 24 plan seems to be the acceptance criteria. And that's a 25 problem I'm having.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

1

~

11 1 CHAIRMAN JACKSON: Any other members of the 2 Committee have any comments?

3 MR. MILLER: I think the counterpoint is that the 4 standards do specify the quality of the plan in various 5 aspects. They do specify requirements for testing 6 th_roughout the plan. Each step you go through the software l 7 development process requires testing. l 1

8 In the end the staff has to audit whether the plan 9 is carried out or not. Maybe Professor Apostolakis is 10 worried about how do you audit it and verify the plan has 11 been carried out in a high quality manner. It does require 12 you te carry out a plan.

13 MR. APOSTOLAKIS: Obviously there is a 14 disagreement here.

15 MR. MILLER: One of the strengths of this 16 Committee is there is not always total agreement.

17 MR. APOSTOLAKIS: Trying to understand how these 18 documents work, what is a regulatory guide, what is this, an 19 SLP and so on, I went back and I found some other regulatory 20 guides for other issues that dealt with more, let's say, j 21 technical issues, like thermal-hydraulics and so on. And I 22 found acceptance criteria that are acceptance criteria: 'You j 23 should make sure that the humidity is below this value. To 24 me that's an acceptance criteria. To say "make sure you 25 have a plan" is not an acceptance criteria in my book.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

12 1 On the other hand, Professor Miller has argued ,

2 that there is a certain culture among software developers.

3 To them this kind of advice means a lot, which perhaps means 4 that we are at the mercy of the reviewer. If the reviewer 5 is very good, perhaps you will get a good review. If the 6 reviewer is mediocre, you will get a mediocre review, 7 because certainly the documents will not force him to do a 8 good job.

9 It's probably an extreme position, but it was also 10 very frustrating, frankly.

11 MR. KRESS: I think part of the problem here is 12 ycu're trying to put together a process to assure a very 13 reliable set of software, software that does what you want 14 it to do. An ideal acceptance criteria would be that it has 15 a certain reliability and it has a certain fidelity. It's 16 just impossible, I think, at this stage of the state of the 17 art of evaluating software to come up with such definitive 18 criteria. I think one has to back off to controlling and 19 defining the process in developing the software, which is 20 what the standards and the reg guides do.

21 I just don't think there was any choice other than 22 to write the review plan and the reg guides in a way that 23 you focused on process and not the final fidelity and 24 reliability of the product. This has been a debate among 25 the Committee itself, process versus product, for quite a ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

13 1 while.

2 MR. MILLER: Each stage of development does test )

3 the product in the sense of that stage. What is missing and 4 what is frustrating to some members is a final test of the 5 final product to give you a number which says this product

~

6 wi,11 perform with this amount of reliability. That's 7 difficult. That's not able to be done with current software l 8 engineering practice.

9 MR. APOSTOLAKIS: That's a different issue, 10 though. I was arguing about the quality of the process  !

l 11 itself. The issue of process versus product is another l 12 issue. There I tend to agree with you that in this l

13 particular case it's very hard to test the product itself.

l 14 So you have to control the process. What does controlling ,

1 l 15 the process mean? Give me a plan? To me that is not i l 16 controlling the process.

1 17 By the way, hardware also we have a problem when 18 it comes to design errors, because that is essentially what 19 we are talking about.

20 CHAIRMAN JACKSON: Is your difference of opinion 21 due to what you think the IEEE standards that are endorsed 22 in these reg guides actually accomplish, or you just don't 23 like the kind of structural layout of the reg guide?

24 MR. APOSTOLAKIS: I would say both, but especially 1

25 the former. I don't know what the standards actually i

. ANN RILEY & ASSOCIATES, LTD.

, Court Reporters 7

1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 (202) 842-0034 I

~

r -

14 1 accomplish. In fact, Dr. Powers raised a very interesting ,

2 question as to who actually puts together these standards, 3 but maybe he can tell us about that.

4 Dana, I put you on the spot here.

5 MR. POWERS: It strikes me it's not the best 6 example to raise on this issue of how you use expert 7 opinion. In an area where you don't have a vast amount of 8 experimental data to calibrate that expert opinion, you are ,

9 really asking experts to prognosticate the future. We have 10 a fair history that experts do a very poor job in 11 prognosticating the future.

12 MR. KRESS: Present company excepted.

13 [ Laughter.]

14 MR. POWERS: I've got several little skeletons in 15 the closet that I've not told you about.

16 The NRC is frequently in the position where it has 17 to prognosticate the future in its probabilistic risk 18 assessments because it can't develop an experimental 19 database on everything that we want to know about reactor 20 accidents, and they have set a standard for how you use i

21 expert opinion in those circumstances. It's a fairly 22 detailed, prescriptive approach.

23 The question I posed is, when we set these 24 standards like this where we don't have a lot of 25 experimental data, ought we not use this fairly detailed ANN RILEY & ASSOCIATES, LTD.

Court Reporters .

1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l 1

15 1 approach that the NRC has set for using expert opinion?

2 Instead of saying to the community, in the case of 3 software engineers, I'm going to write a standard and you 4 are invited to comment on this, you say, no, I want to be 5 sure I get comments from the width and the breadth and depth 6 of that community. I want comments not from those people 7 that attend the meetings and those people that are active in 8 the community alone, but I want the conments from those 9 people that are the Lone Rangers, the vild thinkers, the 10 non-joiners as well, because they may actually have the 11 insight that I need to prognostic the future here.

12 In fact, a study on expert opinion as it has been 13 used in Great Britain when they frequently formulate royal 14 commissions to look into various things like -- I believe 15 one of the more famous ones is whether heavier than air 16 aircraft could exist or not, and more recently, on utility 17 of pressurized water reactors in nuclear power. When people 18_ went back and looked at the history of these committees, 19 they found-that if you were to bet on the most outlandish 20 view as opposed to the majority view, you would be more 21 right than wrong.

22 I think that is why you want a breadth of opinion 23 on these areas where you just don't have experimental data 24 to calibrate experts. The consensus standard approach that 25 is used I think confines you to getting the much more ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

~

16 l

l 1 conformist view, because you tend to get members of the ,  ;

2 societies; you get the members who are employed by companies 3 that are large enough that they can afford to have people >

4 devote time to this; and you don't get as extensive 5 participation of those people in the startup organizations 6 that are struggling to make a living but who may have the 7 real insights for the future. The wilder opinion.

8 The question I posed is, is this consensus process 9 correct for those circumstances where there is not a wealth 10 of experimental data, or would it be better to use the 11 process that in fact the NRC has developed and is now being 12 academically researched? It has always been a great  ;

13 amazement to me that there are actually experts and expert 14 opinion in the world. I'm very interested in the kinds of 15 things that they are trying to do, because we do have to use 16 expert opinion a lot.

17 This is one of those areas where I'm simply

UB raising the question. Is the consensus process adequate for 19 the NRC's needs here?

20 CHAIRMAN JACKSON: Let me make sure I understand 21 one point coming out of all that you have said. Are you '

22 essentially positing that it is impossible to have a balance 23 between the design process or the quality of the design 24 process and the assessment of the final digital product in 25 the end, the software?

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 i Washington, D.C. 20005 '

(202) 842-0034

l 17 l i i i 1 MR. KRESS: No, of course not. There has to be a 2 connection between those two. The quantification of the 3 final product is a difficult problem. ,

4 CHAIRMAN JACKSON: Is that a particular problem in 5 this specific context, or is it one that is throughout many 6 industries?

7 MR. KRESS: I think it is in the specific context  !

8 of software reliability.

9 MR. MILLER: But that still applies to every 10 industry.

11 MR. KRESS: Yes.

12 MR. SEALE: It's generic in the software in the 13 I&C sense, not in the regulatory sense.

14 CHAIRMAN JACKSON: In the nuclear sense.

15 MR. APOSTOLAKIS: It may be in fact simpler for us i 16 in some respects, because according to the experts we are I 17 not really using systems that are too complex. All the 18 horror stories about software that are in the literature i 19 come from very complex systems, like a shuttle, and so on.

20 We don't have those. We don't have those. So for simpler 21 systems probably you can do a better job. There is no 22 question that the answer to your question is, yes, we can 23 have a nice balance.

24 COMMISSIONER ROGERS: You say do a better job. Do 25 a better job on what, the process?

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

18 1 MR. APOSTOLAKIS: The product.

2 COMMISSIONER ROGERS: I seem to hear you saying 3 you could focus more on the product for simpler systems 4 rather than the process.

5 MR. APOSTOLAKIS: Yes.

~

6 ,

CHAIRMAN JACKSON: Or equally.

7 MR. APOSTOLAKIS: Or equally, yes.

8 COMMISSIONER ROGERS: Or some combination.

9 MR. APOSTOLAKIS: Yes. You can test it. If it's 10 simple, you can test it. You can apply simple methods for 11 analyzing it. If you have a huge system that is controlling 12 a shuttle, that's a different story.

13 COMMISSIONER DIAZ: Is it possible to really make 14 a very specific QA program for the process and the product, 15 so specific that it will actually give you guidance of 16 acceptability of both?

17 MR. APOSTOLAKIS: By specific, you mean for the 18 specific product?

19 COMMISSIONER DIAZ: Right. Sometimes we tend to 20 take OA programs as very broad things. Can we really be 21 more specific and actually apply it to the process and the 22 product with some common points that enables you to conduct 23 the procese.?

24 MR. APOSTOLAKIS: I believe that would help a lot.

25 That would help a lot. If you look now at the guides, they ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

19 1 really don't address, as far as I can tell, the specific 2 systems we are using or are about to use.

3 Again, when I visited the Academy committee, it 4 was the same thing. You say something and you are hit with 5 three weird incidents someplace in a very complex system.

6 Then you scratch yourself: Does that apply to me?

7 I th4.nk the fact that our systems are not very I

8 complex has to play a very central role here.  !

9 MR. MILLER: The standards themselves are 10 developed for the software industry, not specifically for 11 our particular type system.

12 MR. APOSTOLAKIS: That's correct.

13 MR. MILLER: As a consequence, they were developed 14 generically for complicated systems like you find in the 15 space shuttles and airplanes, and so forth. As long as we 16 keep focusing on the fact that nuclear power plant safety  !

17 systems should be simple, then agreeably we should be able 18 to maybe do better than we are doing with these standards.

{

19 But what we are endorsing is these standards which are 20 current software practice, and we just need to keep looking 21 for better ways to do it.

22 CHAIRMAN JACKSON: I think, though, that there are 23 two things that one could ask, one of which Commissioner 24 Diaz has already asked, and that is, can an appropriately 25 focused QA program or effort as part of the design process ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

20 1 itself track to certain comfort about the quality of the .

2 product?

3 The second issue is, if you are already saying 4 that the existing standards are for more complex systems 5 than one typically is focused on in a nuclear plant, that 6 th,ese are the accepted standards, but therefore a more 7 complex universe, but we are somehow saying that our 8 universe is more simple, then it strikes me that should not 9 embedded in those be the ability to try to develop something 10 like what Commissioner Diaz is talking about, and why can 11 there not be focus in that particular area?

12 MR. APOSTOLAKIS: Do you want an answer?

13 CHAIRMAN JACKSON: A short one.

14 MR. APOSTOLAKIS: I think what you said can be 15 done. I think the next round it would really be useful if 16 the staff summarized or took what they felt was important 17 from these IEEE standards and not keep referring to one 18 standard after the other, after the other. It can be done 19 in a simple document.

20 CHAIRMAN JACKSON: In fact, then, it would be 21 fair, as you call it, in the next round, however that is 22 defined, to ask the staff to do that and to think about how 23 that can be used in the context of Commissioner Diaz' focus 24 and then to have you gentlemen come back and' comment again.

25 Let me see if any of the other Commissioners have ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l 21 1 any questions. Commissioner Dicus.

2 COMMISSIONER DICUS: No.

3 CHAIRMAN JACKSON: Commissioner McGaffigan.

4 COMMISSIONER McGAFFIGAN: No.

5 CHAIRMAN JACKSON: Commissioner Rogers. do you

~

6 have any further questions?

7 COMMISSIONER ROGERS: No.

8 CHAIRMAN JACKSON: Commissioner Diaz. l 9 COMMISSIONER DIAZ: No.

I 10 CRAIRMAN JACKSON: Let's move on.

l l 11 MR. KRESS: Very good. Let's move to the next 1

l 12 item, which is the research plans for upgrading the l 13 thermal-hydraulic codes.

l l

14 Ivan, you're on.

15 MR. CATTON: I believe the path that was initiated 16 by RES is a good one and the ACRS supports it. I basically 17 agree with the views expressed in the letter but would like 18 to balance the initial euphoria with a few of my own views.

19 In the past, the ACRS has recommended that further 20 development of codes like TRAC and RELAP5 be curtailed and 21 that some resources be focused on a fast running code for a I

22 PWR and like the BNL code for BWRs. At this time the BNL 23 code has disappeared and a fast running code for PWR has 24 never appeired.

25 There are a number of reasons why I still lean f

I i

l ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l

22 1 towards this view. .

2 First, the two-fluid modeling that is the basis 3 for both TRAC and RELAPS is one of the more challenging 4 problems in heat transfer and fluid mechanics. Because of 5 its complex closure relations, a great deal of detailed 6 experimental data is needed, and for the most part it does  !

7 not exist. As a result, there are many ad hoc relations '

8 that are tuned to macroscopic data in these codes. ,

9 Some examples are simple things. Like a droplet i

10 in a decelerating flow versus an accelerating flow. What '

11 happens to it? A good example is, if you have fog in a room 12 and the velocity goes to zero, these codes will predict the 13 water falls on the floor.

14 There are others. Dynamic flow regimes. How do 15 these things change as the conditions of the problem change?

16 Unfortunately, in the past, agency support for 17 producing the needed data and understanding has not been 18 forthcoming. And I don't see it now. Without it, two-fluid 19 modeling has evolved about as far as it is going to.

20 The three facilities, an AP600 at Oregon State, an 21 SBWR at Purdue, and a B&W plant at the University of 22 Maryland, along with the relationship with the French will 23 yield a great deal of what is called integral system data.

24 This is the macroscopic data. And because they are very 25 well instrumented, in fact they are far better than anything ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

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

, 23 l

1 we've had in the past, some of the detail type information i

2 needed for support of codes like TRAC and RELAP may be s 3 forthcoming. Without special emphasis, however, I think it 4 will fall short.

5 This doesn't mean that codes like TRAC are not i -

6 va.,luable. They are, and they have been terribly neglected 7 and for the most part the user has been ignored, leaving

, 8 their use to the dedicated code jock. I can define that if 9 you want.

10 The plan to revitalize TRAC by updating the 11 FORTRAN and development of graphical interface and input 12 along with consolidation of all that is good is long overdue 4

13 and should be encouraged. You have several people in RES 1 14 who are highly qualified to do so. Further, the involvement 15 of users from NRR and AEOD will add another dimension to the

, 16 process. I believe this is essential.

17 In the past, the problem has been that once the

18 code process started, it was somewhere else and the user 1 19 came later. As a result, these codes were not very user 20 friendly, and I think if you want to make good use of the 21 computational capabilities within the agency, they damn well 22 better be user friendly. They have to be other things too, 23 but that's important.

24 What else should you do?

25 Codes like TRAC have a history of not performing ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

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

24 1 well when presented with new problems. This is not .

2 surprising when you realize that the code was developed for 3 a large-break LOCA, which is a fast transient. That makes 4 many facets of the two-phase flow and heat transfer 5 relatively unimportant. And that there was sufficient data 6 to tune the codes. We have a great deal of large-break LOCA 7 data. It essentially became the biggest empirical fit 8 you've ever seen.

9 There is nothing wrong with this, and much of the 10 complexity in the code was needed in order to get it done.

11 A simpler tool, however, would also have done the job.

12 There are lots of examples of this where a much simpler tool 13 can accomplish the same thing as the big code. It requires l

14 skilled people to make the judgments as to when you can do 15 what.

16 The problem was we were carried away with the view  !

17 that given the right set of equations we could solve any 18 problem, and it took a long time for that euphoria to wear 19 off. This started in the early 1970s. It was a very i 20 heavy-duty committee that was put together to help start 21 this thing off and running, and when all those equations 22 were up on the board, we all got excited.

23 There are many difficulties with the codes. For 24 example, one cannot do time and space conversion. The basic 25 part of the code is quite weak. This is one of the reasons ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

25 1 you have to tune them. You can't do the conversion studies.

2 A lot of what is done with them is just because that's the 3 way it was done before; we've interpreted data to fit the 4 way we've done it. This is a problem.

5 Yet the code did meet its target mission, and much 6 more. I think these codes are very good for problems where 7 we know how to use them, and they will continue to be 8 useful, but how far you can push it is another question.

9 During the past 20 years we have seen a number of 10 issues arise and found our computational tools to be 11 defective. The problem was they were too inflexible and 12 rigid to allow changes needed to focus more on a particular 13 phenomena. I can name a few of these if you wish.

14 The new plan includes a task that will supposedly 15 deal with this, and it is to make the code modular so that a 16 separate model can be incorporated. For example, if you are 17 treating the pressurized thermal shock, you can have a CFD 18 code that could be inserted somewhere to deal with it.

19 There are a lot of things you have to think about 20 before you do that. If you are going to start interfacing 21 different kinds of tools, you damn well better make sure 22 that whatever the structure is can accommodate this. If 23 it's successful, it will go a long way towards dealing with 24 my concerns. A demonstration that this can be done 1

25 supposedly is underway. I've not heard about it. I don't ANN RILEY & ASSOCIATES, LTD. 4 Court Repcrters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

~ '

26 1 know when the results will be put on the table. ,

2 Another aspect is computational time. I think you 1 3 need to have codes that you can get answers quick and you 4 can look at a lot of parameter variations. If you don't 5 have that, you don't properly evaluate the problem you have ,

6 at, hand.

7 Again, the large-break LOCA was a fast transient ,

8 and required a particular kind of time advancement 9 algorithm. It was also forgiving. So one could be 10 numerically careless and not worry too much about things 11 like numerical dissipation. If you want to treat reactor 12 instabilities of some kind or another, you don't want 13 something built into your code that damps disturbances. You 14 just don't want that. And you have to go to special lengths 15 to make sure it doesn't happen.

16 With the AP600 and the small-break LOCA on the 17- table, we are now faced with long duration transients.

18 AP600 long-term cooling is an example. And as Westinghouse 19 is finding out, they can't afford to do the calculations.

20 So they have to play all sorts of games to pick up pieces of 21 it as it moves along.

22 A problem like this is a quasi-steady problem.

23 Why the hell are you using a transient code? I don't 24 understand, but some aspects of this business are beyond me.

25 To summarize, I believe the cleanup, i ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 l (202) 842-0034

l I l

27 1 standardization and consolidation plans are essential.

2 Right now you have a tool that is difficult to use, 1

3 inflexible, and it costs you probably more money in trying )

4 to use it than it will cost you to fix it.

5 Before further development, however, I think RES 6 should take a serious look at future needs of the agency and l i

[ 7 assure you that the bases and resources are available.

8 There must be sufficient reason and resources coupled with a 9 commitment to pursue the basic understanding needed to 10 support the underlying two-fluid modeling concept before 11 marking on a new two-fluid code program.

12 The needs have been delineated repeatedly for the 13 past ten years. Almost every meeting you go to. You were 14 at the CSNI meeting, and I bet somebody could show you a 15 viewgraph from that meeting that doesn't look very much 16 different than one we would have put up ten years ago.

17 There are just difficult problems, and nobody has come forth 18 with the resources to eliminate them.

19 Flexibility and modularity are another aspect. If 20 they cannot be accomplished, I think the effort should be 21 downscaled and consideration should be given to different 22 codes for different problems.

23 Sort of as a final note, the computer code should 24 not be more detailed than your understanding and. data will 25 allow. Don't make the same mistake we did in 1974.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

28 1 CHAIRMAN JACKSON: Let me ask you a couple of ,

2 quick questions. When you talk about the need to deal with i

3 issues having to do with computational time with an ability 4 to have variation of parameters, are you saying that it's an 5 issue having to do with modeling, the kind of algorithms 6 us.ed, or the platforms, or all of the above?

7 MR. CATTON: All of the above. I think that 8 Wolfgang Wolf did that at BNL. He set out to put together 9 what he called a plant analyzer, but he set out to put l  :

10 together a program that was based on data as he had it to 11 analyze the BWR. His ground rules were fast, reasonably 12 accurate, and an ability to address a wide range of types of 13 problems. When the LaSalle instability incident occurred, 14 it was really good tha* 'he agency had that capability, 15 because GE said it's not a problem.

16 Wolfgang, because of the kind of program he had, i 17 was modeling the entire plant. The TRAC-GE really hadn't l i

18 donc that because it was too expensive. The result was i l

19 people who lean more towards the PRA view took a bunch of l l

20 sequences, said these are possible things that can happen,

! 21 and calculated the end point. He did 60-plus calculations l

22 in a very short period of time and clearly demonstrated 23 where the problem was. Whereas the TRAC-B, which was the l

24 agency's program, they never got it running. They couldn't

. 25 solve the problem. They couldn't make it work.

l ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l

, i 29  ;

1 Eventually that code wound up at Penn State and l

l i

2 some students managed to make it work, but that's a separate 1 3 issue.

4 So I think it is sort of all of the above. You i 1

l 5 really need to focus on what you want. If what you want is 6 a fast running, highly reliable code, you know when you 1

7 start you are going to give up something. You need to try l l

8 to figure out what it is you are going to give up. Maybe 9 you want to give up being able to solve the large-break i 10 LOCA, because it's a relatively low risk thing anyway.

11 You want to address other kinds of problems: What i 12 kind of transients do people like Caruso deal with for NRR?

13 Maybe compile a list and then ask yourself, what do I need?

14 Most of the time you don't need very much of the horsepower 15 in these big codes.

16 CHAIRMAN JACKSON: Let me ask you this.

17 MR. CATTON: I think I answered more than you 18 asked.

19 CHAIRMAN JACKSON: That's right. It's like the l

20 algorithm that has more than the data.

21 tLaughter ]

22 CHAIRMAN JACKSON: I take that back.

23 I noted that you had indicated that you felt that 24 financial constraints forced NRC to allow the codes that 25 would be predictive of fuel behavior to kind of wither on l

1 ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

30 l 1 the vine. Do you think that was because of an undue focus , ,

2 on severe accidents, and issues related to, say, high burnup 3 fuel hadn't been identified?

l 4 The real question I have is, the codes that we 5 have and that are not state of the art, the ones that have 6 to,do with prediction of fuel behavior, are you saying that 7 they can't adequately predict fuel and clad behavior at the i

l 8 burnups now being used by licensees?

9 MR. CATTON: One thing I very deliberately did was 10 avoid discussing the fuel codes. I'm starting at the clad 11 working ot" l 12 really don't know, but maybe Dana could help, 13 because Dana has been paying a little more attention to the 14 fuels problems.

15 To me, as far as the t.hermal-hydraulics is 16 concerned, if you change the burnup, you change some of the 17 conditions that the code hrss got to operate under.

18 MR. POWERS: The agency will admit that indeed its l

19 codes aren't capable of treating fuels at the very high '

l 20 burnups right now, that they have been allowed to atrophy.  ;

21 The research program is trying to amend that problem.

22 I don't think we have reviewed the length and the 1

23 breadth of their attempts to amend that, but certainly it is l 24 my impression that we do not now have a program that carries 1  :

25 our understanding of the way fuels behave, especially when l

4 ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

31 1 you go to burnup sufficient to develop a rim effect; that we 2 have a physical understanding of all that goes on in the 3 fuel.

4 Perhaps of more importance is we don't have a good 5 understanding of all the degradation that occurs in the clad 6 as,we go to very high burnups. A lot of the concerns about 7 high burnup fuel have been prompted by some experiments l 8 dealing with reactivity insertion accidents.

l 9 Those are interesting, but I don't think that is 10 where the big difficulties are going to arise with very high l

l 11 burnup fuel. I think it is really the degradation of the l 12 cladding and other kinds of operational accidents where that l

13 clad failure is going to pose difficulties to you.

14 I don't think we have a good understanding of all 15 the metallurgical processes that take place at this high l

16 burnup, and they are going to be very stochastic type of 17 processes. Much of the difficulties that arise with the 18 cladding occur because you get high hydride precipitation at 19 local fluctuations in the temperature in the clad. Those 20 are caused by discontinuities in clad thickness, 21 discontinuities in the fuel clad gap. Small perturbations 22 in the manufacturing process lead to localized deposition of l

23 large hydrides that make the clad brittle.

24 You can understand that brittle clad now affects 25 everything else that you are going to do with this fuel.

l ANN RILEY & ASSOCIATES, LTD.

, Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1

l

~

32 1 It's going to affect accident situations. It's even going ,

2 to affect handling and storage subsequently.  ;

3 I don't think we have predictive tools in this 4 area right now. I think the staff has now embarked on a ,

5 research program that they are fairly enthusiastic about.

6 Wg have not as a committee or as a subcommittee reviewed 7 that research program, though our intention is to do that 8 when they are ready to come forward.

9 The research program is interesting because it is 10 not the NRC going it alone; it is the NRC joining with the 11 world in this area, because all reactors are interested in '

12 using fuel to longer and longer burnups. It contributes 13 enormously to the economics of nuclear power. So the 14 research program is a consortium of efforts between NRC, 15 France and Japan in particular, and it may be a broader 16 community than that. I'm uncertain.

17 CHAIRMAN JACKSON: How close are we to the edge 18 relative to the current burnups? .

19 MR. POWERS: Our codes were developed for  !

20 predicting fuel behavior under things like reactivity 21 insertion and were prepared and validated against a database 22 that extended no higher than 33,000 megawatt days per ton.

23 We now approve fuels going up to, I believe, 55,000 megawatt 24 days per ton. So it's not a case of being close to the 25 edge. We are now beyond our validation limits.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

33 l l

1 That probably is not terribly important. You can 1

2 probably extrapolate the behavior that we saw up to 33,000 l 3 megawatt days per ton up to around 50,000 or 55,000 megawatt 4 days. When you go over that and you start developing rim 5 effects and hydride precipitation is when you get into the )

6 prpblem . I believe right now that licensees need to make a 7 very special and elaborate case to go beyond that. I don't i

8 know of anybody that has tried to go beyond that.

9 So our codes aren't validated into this regime, 10 but it's the next step to go beyond the 55,000, where I 11 think that would be an unacceptable situation.

12 CHAIRMAN JACKSON: Commissioner Rogers.

13 COMMISSIONER ROGERS: Coming back to the 14 thermal-hydraulic questions, I know in your letter, your 15 comments, Dr. Catton, you said that you thought a broa, der 16 approach is needed where different modeling schemes would be 17 tied together. I take it this relates to your comments 18 about modularity and flexibility. Is that correct?

19 MR. CATTON: That's correct, and that's not going 20 to be an easy task.

21 COMMISSIONER ROGERS: But you also go on to say  !

22 that a skilled code user who is also knowledgeable in the 23 field of thermal-hydraulics is needed to decide what is 24 important and how do we implement it in a code. Where do 25 you see us standing with that in-housa capability now or our ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D C. 20005 (202) 842-0034

34 1 ability to tap that? ,

2 MR. CATTON: In the past few years there has been  :

i 3 a significant change. I don't remember when one of the ,

4 Commissioners decided that it might be nice if your own  !

5 people could run your codes, but I think it has happened.

~

6 You have people like Caruso in NRR, who is really very good 7 at using the code. You have people like Joe Kelly in 8 Research, who is really very good. ,

9 There was a very nice paper by Mr. Caruso where he 10 talks about three kinds of users. One is the guy who 11 understands all of these things and knows how to run the i 12 code; the second is just a good engineer; and the third is a >

13 systems kind of guy.

14 I think you should develop some sort of L

15 administrative controls on how you do business so that if l 16 it's a run of the mill kind of problem the systems engineer ,

17 is welcome to do whatever he wants to do. If the problem is j 18 a new one, you need to get your category 1 person involved 19 and maybe develop some sort of a sign-off system.

20 Too many times people get too complacent about the 21 results of the code. If that printer, or whatever, the 22 screen, runs the numbers up and they are like you've seen 23 them before, you tend to believe it. If it's new and you're l

24 not experienced with actually touching some of the data, you i 25 don't recognize good from bad, and usually you can generate ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

. _. .. . . ~ . _ _ . _ _ _ _ _ - _ . _ . - . - . _ . _ __-._ ___..___._ _ _ _ - - _ _

35 1 all kinds of arguments as to why it's good: the code did it; 2 and this code compared with that code. That can lead you i

3 into troubles.

4 The category 1 type, there are not very many of 5 them in the agency, and I think you have got to maintain 6 th,em somehow, and you need to establish a prvo dure so that 7 anytime they move into new problems, like boron mixing or 8 whatever, the category 1 person is involved with the process 9 at least in a review capacity.

10 CHAIRMAN JACKSON: Commissioner Dicus.

l 11 COMMISSIONER DICUS: Let me follow up on that.

l 12 Not so much in terms of the capabilities of the individuals, f

13 but you seem to be implying -- and correct me if this is a 14 wrong implication -- that perhaps three people is adequate 15 for staffing for what we need to accomplish here. Or would 16 you suggest we need additional staffing?

17 MR. CATTON: I really haven't seen a plan showing 18 what all the tasks are. That aspect of our interaction was 19 somewhat superficial. I don't know all that they are going 20 to do. I think the three people in Research are very good.

21 What kind of workload they are going to pick up relative to 22 what they send out the gate, I don't know. So I really j 23 can't address that question.

24 I think within NRR you have a very good team, and i

25 as near as I can tell it's adequate, but you probably have l

ANN RILEY & ASSOCIATES, LTD.

l Court Reporters i 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i __ _ _ _ _

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

36 1 1 to ask them if they are overworked. .

2 CHAIRMAN JACKSON: Commissioner Diaz. f 6

3 COMMISSIONER DIAZ: Long time no see.

l 4 MR. CATTON: About ten years, I guess. {,

5 COMMISSIONER DIAZ: First, a generic comment.

6 I've been looking at the past history. I realize that ACRS i j

7 really in a very consistent way has emphasized the use of (

8 systematic, practical, auditable, flexible and traceable 9 methods to integrate the experiments with the codes and 10 develop the capabilities. I believe that is a very, very 11 worthwhile sense of direction that we need, and I hope we l

l 12 keep doing that. I have a small interest in the area from {

13 past experiences.

i 14 I think that some of the smaller issues that l i

15 always keep coming up, those we need to determine that we 16 have the staff to solve them. I do agree with Dr. Catton f 17 that there is a time in which flexibility is important and  !

18 there is a time in which we need to assess our capabilities i

19 to develop codes or change them and to use them properly, f 20 I believe that we have the capability to use them 21 properly, and if we are going to really take this five year  !

22 plan, which I guess everybody agrees is basically and l 23 fundamentally a good plan, we need to really focus on what ,

i 24 capabilities you have to have, code developers inside,  !

25 people that can interact with the community on a one-to-one l I

7J04 RILEY & ASSOCIATES, LTD. ,

Court Reporters l 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i

l 37 l, 1 basis and not be lost when Dr. Catton comes out with a new l 2 project, which he is quite capable of doing very quickly, as i 3 I well know.

4 I have a question in this area which may be 5 addressed to the Committee. I have done some experiments.

6 I _always like to see that we have some experimental 7 verification of thermal-hydraulic codes. We know that codes 8 are one-dimensional or are dynamic, and going into static or 9 vice versa might have some problems.

10 Are there financial constraints that NRC has in 11 the international arena, in this experimental base that we 12 are trying to get that are really affecting our ability to 13 guide the experimental programs to the things that we need?

14 MR. CATTON: We need to first separate the kinds 15 of experimental programs and their purpose. The large 16 facilities are integral facilities. You don't get the 17 detailed physics that you need to address the issues about 1

18 the internals of the code. So what do you do?  !

19 You're a person who has been tasked with creating 20 thi, computer program. You just make sure it gets the right 21 answer that you measured. But you have a lower level. You 22 have droplet sizes. Are they ligaments? What's their 23 shape? You have all sorts of constitutive relations.

24 Nobody really agrees on what they all are. But once you 25 pick a set of equations, then you tune thens. This is very ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

38 ,

1 true in two-phase flow. Even'in porous media where the bed- ,

~

2 is fixed there are lots of disagreements on what the 3 . equations should be, t

4 Under these kind of circumstances you need to do l t

5 the necessary work to understand the basic physics or else 6 back off on your expectations from the code. I think it's a i

7 mistake to build a model that requires data input at a level 8 that you don't have it unless you plan ~to go get it.  !

9 If you go back through the history of the ACRS'  !

10 views, when the program first started this was a complaint,  !

11 that you get a major group like the people at Los Alamos,  !

12 Frank Harlow and those guys, who are just super at  !

13 developing codes, but unless you feed them the bottom '

14 physics, they are not going to get-it done right. But i

15 they're going to make it work. i 16 So now you have this on the table. What do you do  !

17 with it? As long as you are working.within the macroscopic l 18 data you have at hand, you know. In Japan they-had the 1

19 SCTF. I forget what SCTF stands for, but it was a reflood

• i 20 facility. The first thing they found is that if you got the 21 friction factor right, the interfacial friction factor, the 22 heat transfer was wrong. If you got the heat transfer 1

23 right, something was wrong with the other. What it was is a 24 basic inconsistency, but the code had been tuned to deal  !

25 with it.

ANN RILEY & ASSOCIATES, LTD. }

Court Reporters 1250 I Street, N.W., Suite 300  !

Washington, D.C. 20005 (202) 842-0034  :

I I

o 39 1 Some of these things have been taken out, but you  ;

2 still have, for example, the heat transfer packages. Most

, 3 data is taken by measuring a heat flux from a boundary into 4 this mix. You don't know whether the heat goes to the  ;

1 5 vapor, then the droplet, or the droplet to the vapor, or l

6 wherever. Now the code person has to do something with 7 this. So they somewhat arbitrarily split them. Sometimes l 8 good, sometimes bad. A guy with a lot of insight might even 9 get it right.

10 But you're never really sure until you try to 11 measure these things. If you don't want to measure these 4 12 things, then rewrite the way you work your code so it deals l

13 with this thing that you measured. What this allows you to 1

! 14 do is to have good traceability. You can make a good I i 15 statement about uncertainty in the results.

16 I don't know if I answered your question.

l

. 17 COMMISSIONER DIAZ: I don't think so, but that's 18 okay.

i 19 (Laughter.]

1 l 20 CHAIRMAN JACKSON: Commissioner McGaffigan.

j 21 COMMISSIONER F 1AFFIGAN: I'd like to ask about 22 the balance of what the appropriate role is of the different 23 actors that we can call on, the staff, the three people you 24 talked about in Research, the labs. I know there has been

25 some history of problems in dealing with at least one of the a

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

1 e

40  !

i 1 ~1 abs. The universities and the university community, which , )

2 you have built into the plan, and the private sector. How 3 do you get the sort of continuity, the data so that the l 4 codes don't get more complex than the basic physics? What 5 is the role of the different entities as you see it, and are i 6 thpre dangers?

7 My original concern was we might be trying to do .

8 too much in house. What is the appropriate role of the 9 different institutions?

10 MR. CATTON: Let me try. I'think, first, you have 11 to have some kind of an in-house effort going on. If you 12 don't, you won't have that category 1 person -- you just 13 won't --

who is interested in these things enough, that 14 understand them well enough to help you. That's number one.

15 You need some people. They need to be somewhat unencumbered i 16 with all of the bureaucratic management stuff that is a 17 necessary part of federal government. I was going to use 18 some other words, which I did not do.

• 19 CHAIRMAN JACKSON: We are not bureaucrats. I 20 don't know what you are talking about.

21 [ Laughter.] t 22 MR. CATTON: And I think at present the management 23 within RES seems to have done that. These three people are

24. pretty unfettered.

25 What kind of support do they need? In the past ANN RILEY & ASSOCIATES, LTD. f Court Reporters  :

1250 I Street, N.W., Suite 300 '

Washington, D.C. 20005 (202) 842-0034

l 41 1 they have just gone to the national labs, and I think there  ;

2 there has been an uncoupling, because they don't manage the 3 labs. I don't know whether this is a characteristic of 4 government or what, but the programs just have not been '

5 managed well at any of the labs in the sense that there is a 6 cl, ear definition of what you want and a clear statement of 7 when you got it.

t 8 I think the labs can play a role. I think places 9 like Los Alamos have the people who could take some of these 10 things and put it into good programming. They certainly ,

l

11 could do that. But when it comes to trying to understand 12 the basic science of two-phase flow and heat transfer and 13 these kind of things, you are not going to get it from the 14 labs. The labs are professional people who have a job to do 15 from eight to five. Somehow you need that graduate student 16 who can't get out unless he understands it.

17 There is a problem with getting that set up. If 18 you just do a grant, the professor and a student can go off 19 on a different tangent, spend your money, and thank you very 20 much. I think what you need is an institute, because the 21 institute then is staffed by some full-time professional 22 people, and they can lean on the professor or cut him off if 23 whatever it is is headed in the wrong direction, and you 24 also have somebody that you can reach out and touch. If you j 25 set this up properly, it can be the focal point.

, JJRJ RILEY & ASSOCIATES, LTD.

l Court Reporters i 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 l (202) 842-0034 l

42 1 I was very impressed with the French and the way ,

2 they do business. In France, typical?.y the lab, or 3 whatever, is directly attached to the university. A French 4 professor usually has a managerial role in that, and he's 5 got his interest in what the students do. The French felt 6 that if we do this in the thermal-hydraulics area, which is 7 Grenoble, we're going to maintain the capability that we r 8 need; new people are going to come into it because, let's 9 face it, the professor is going to hammer students wherever i

10 he is.

11 I think the process works. Not necessarily as 12 direct grants, although there is a way you can deal with 13 that. Other agencies do it through workshops, and the 14 professors who go to them are not stupid. They know if they 15 want your money, they had better propose something within  :

16 that spectrum.

t 17 COMMISSIONER McGAFFIGAN: This institute or 18 center, would it involve potentially a consortium of  ;

19 universities? It wouldn't necessarily be tied to a single 20 one? Or would you see it as a consortium that had an t

21 institute at one but the ability to tap universities t

22 nationwide if it was needed?

23 MR. CATTON: If I were doing it, that's the way I 24 would do it. I think there is a nice example of what was 25 done in the thermal spray area at Stony Brook. They sort of

{

I ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 i

43 1 sit in the lead position. However, there it's an NSF 2 center. I wouldn't do that. It would have to be an 3 institute so that I could get some assurance that whatever i

) 4 the path is I'm trying to follow will be followed.

5 They have a whole range. I think Sandia and Idaho 6 both have very good thermal deposition, spray deposition

7 kind of laboratories. They're involved with them. So they t

8 sit at the top, and you have the national labs associated 4

9 with them.

10 There is another thing that I have been bothered i

11 by too.

, 12 CHAIRMAN JACKSON: I think we are going to have to 13 move along.

14 MR. CATTON: Sure.

i 15 CHAIRMAN JACKSON: Finish your sentence.

16 MR. CATTON: I was just going to say that other i 17 government agencies seem to be able to control the labs. I 18 don't know why NRC cannot.

19 CHAIRMAN JACKSON: Okay. l I

i 20 MR. KRESS: I resist the temptation to try to l

9 21 defend the labs. 1 I i

22 CHAIRMAN JACKSON: Why don't we move on to the 4 23 next topic, i

24 MR. KRESS: Let's move on to a subject that is of 25 much interest and one which we have paid a great deal of

, ANN RILEY & ASSOCIATES, LTD.

J Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 4

. _ . _ ._ . _ _ . . .-- _._.m_____... . . . _ . _ _ _ _ _ _ _ _ _ . . . _ . _ _ _ . - . _ _ _ . . _ . . _ . . _ . _ _ . _ _ _ . _ _ _

44 ,

1 attention to recently, and that's the risk-informed, , .

2 performance based regulation and related matters.

3 Dr. Apostolakis.

l 4 MR. APOSTOLAKIS: We have been meeting with the l

5 staff on a fairly regular basis. The PRA subcommittee had  !

6 twp meetings last summer. The full Committee heard from the  !

7 staff in August. ,

i 8 I thought we were proceeding well and according to' i 9 -schedule. There were cuae differences on specific guidance, l 10 but the discussions were technical, and so on, until we  !

11 found yesterday that things.are now up in the air, that .

12 there are some questions that have been raised at the high i 13 levels regarding allowed risk increases and whether they i

14 should be allowed at all.

15 The original schedule now is not valid anymore.  :

1 16 We have tentatively scheduled a supplemental meeting with 17 the staff for, I think, the third week of February to review 18 the documents that will be delivered to us by the first week 19 of February. This will be the final review, and we will 20 write a letter to you during the March meeting the first 21 week of March. That seems to wrap it up.

22 Do you have any questions?

23 CHAIRMAN JACKSON: .What progress is the staff

~

24 making in addressing the issue of uncertainty in the use of 25 PRA results?

ANN RILEY & ASSOCIATES, LTD. l Court Reporters .

'1250 I Street, N.W., Suite 300 I Washington, D.C. 20005 l (202) 842-0034 I 1

1

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

45 1 MR. APOSTOLAKIS: You mean in the regulatory 2 guides, how they are dealing with uncertainty?

i 3 CHAIRMAN JACKSON: Correct.

l 4 MR. APOSTOLAKIS: That is one of many issues.

1 5 There is a disagreement there. I think the staff wants to l 6 give prescriptive guidance as to what kind of PRA one should l

i 7 have, depending on the application and the change in risk.

l l 8 For example, a rough point estimate calculation 9 would be acceptable if the change was very, very, very t

10 small, like ten to the minus six or something, without l 11 external events, and so on. And there is some point to 12 that.

13 But I've always felt that we are trying to overdo 14 it and be too prescriptive. It seems to me the guidance  !

15 should be that we should be using the models and analyses  :

16 that are appropriate to the situation. People feel that '

17 that is too general, that we have to be more specific.

18 The thing that is missing is how do you handle ,

19 uncertaint.ies that have not been quantified. In fact, we 20 had an interesting discussion at the last meeting as to 21 whether the ten to the minus four core damage frequency '

22 subsidiary goal was set as an absolute goal or was set with 23 respect to what can be quantified. In other words, I can go 24 as close to the ten to the minus four goal as I can if I can 25 prove with my calculations and analytical tools that I have l t

l l

l l

ANN RILEY & ASSOCIATES, LTD.

I Court Reporters

! 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 l

(202) 842-0034

~

46 j i done a good job.

2 If, on the other hand, that subsidiary goal is an 3 absolute goal for the core damage frequency period, then I 4 should not be allowed to go very close to it because I know

! 5 that there are certain things that are not in the PRA that 6 contain the risk to the frequency of core damage. We were 7 told that the original intent was to exclude these.

8 I'm not sure the things that have not been 9 quantified have attracted the attention they deserve. On 10 the other hand, I don't think that that is because people 11 feel that these are not important. The staff has had its 12 hands full trying to develop all these regulatory guides in 13 the last several months. But these are certainly things 14 that both sides are aware of.

15 CHAIRMAN JACKSON: I'm not sure how happy I am 16 with your answer.

17 What are your views on the plant-specific ,

t 18 application of PRA results?  !

19 MR. APOSTOLAKIS: First of all, you are not happy 20 with my view or with the way things are?

21 CHAIRMAN JACKSON: I'll make it explicit. I asked 22 the question of what progress you feel the staff is making 23 in addressing the issue of uncertainty in the use of PRA '

24 results, and you said that is one of many issues. Then you  !

25 talked about uncertainty that had not been quantified. l l

ANN RILEY & ASSOCIATES, LTD. ,

Court Reporters  !

1250 I Street, N.W., Suite 300 Washington, D.C. 20005 t (202) 842-0034 i

47 1 I think what is missing here is, if there really 2 are some issues of points of vulnerability, et cetera, in 3 terms of what the staff is doing, what would be needed to be 4 able to give comfort to make use of PRA and how it's handled 5 in these guidance documents without being what you consider 6 to be too prescriptive, it would be useful for the 7 Commission to get a listing of that.

8 As long as we kind of talk out in space, it's very 9 hard to pin down just where the problem or problems seem to 10 be. Perhaps that is what you will be addressing in your l 11 letter to the Commission once you have reviewed these 12 documents in February.

13 I like PRA and I know on a rudimentary basis how 14 to do PRA calculations, but that's not my job. However, in 15 order for me to do my job and the Commission to do its job, 16 we need to have more understanding and specificity about 17 where you think the problems really are, because that forms 18 the basis of giving guidance back to the staff in terms of 19 what needs to happen. That's really what I was trying to l

20 talk about.

21 MR. APOSTOLAKIS: My answer to that would be, yes, 22 they are making good progress. There are some disagreements 23 but the disagreements are not fundamental. In particular, 24 they are trying to be more prescriptive than I would like.

25 CHAIRMAN JACKSON: Are there any advantages or ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

48 1 benefits to our assembling a group of specialists to review ,

2 the PRA guidance documents beyond the review that'ACRS is 3 itself providing?

4 MR. APOSTOLAKIS: The documents that will be 5 produced in February?

6 ,, CHAIRMAN JACKSON: Right.

7 MR. APOSTOLAKIS: I would say no. You will be 8 getting general comments that you already have.

9 CHAIRMAN JACKSON: Commissioner Rogers.

10 COMMISSIONER ROGERS: Nothing else.

11 COMIiISSIONER DICUS: Nothing else.

12 CHAIRMAN JACKSON: Commissioner Diaz.

13 COMMISSIONER DIAZ: You just brought up the issue 14 a while ago of allowing small increases in risk under 15 certain conditions. Would you elaborate why that is a 16 problem to you?

17 MR. APOSTOLAKIS: It is prcblem in the following 18 sense, in my mind. If we declare in advance that we would 19 have a risk-informed and performance-based system that will i

20 not allow increases in risk, then we might as well forget 21 about trying. Why are we doing it?

22 You have to be able to allow increases, because in 23 essence you are saying, well, keep it the way it is, or play 24 games and package changes in such a way that the net 25 increase appears to be zero.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters ,

1250 I Street, N.W., Suite 300 l Washington, D.C. 20005 )

(202) 842-0034 )

49 l

1 It seems to me that when the Commission states 2 quantitative health objectives and you are well below the 3 objectives in the subsidiary goals, you should be allowed to l 4 increase a little bit. How fast, whether we should allow  ;

I 5 all the units around the country to come just close to the 6 goal, these are questions that certainly deserve 7 consideration.

8 But to say that no increases are allowed when in 9 fact we are doing this every day without quantifying risk 10 -- the staff told us that we have about 1,000 requests for 11 changes in the licensing basis every year that are not done 12 using risk assessment. So when we look at the risk number, 13 we say, no, we don't want it to go up. But when we don't 14 have a risk number, it's okay?

15 COMMISSIONER DIAZ: I do agree with you that we 16 need to visit that area very carefully.

17 CHAIRMAN JACKSON: Isn't it a question of the 18 context within which you talk about increase in risk? l 19 MR. APOSTOLAKIS: Sure.

20 CHAIRMAN JACKSON: And the elestion is, what is l l

21 the goal and what margin has one relative to that goal? l 22 When you then talk about risk increase, you are talking  !

23 about risk increase relative to something.

24 MR. APOSTOLAKIS: Sure.

25 CHAIRMAN JACKSON: I think that is where some i

ANN RILEY & ASSOCIATES, LTD. l l

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 1

50 1 clarification is needed in terms of that you just can't talk ,

2 about it in a vacuum. It strikes me that that is where 3 there has to be some convergence of the discussion here.

4 MR. APOSTOLAKIS: And the assumption is, of 5 course, that the quantitative health objectives and the 6 sub'sidiary goals have already been met. We're talking about 7 increases without constraint. If you are safe above ten to 8 the minus four in core damage frequency, the question 9 doesn't even come up, because you have violated the law.

10 CHAIRMAN JACKSON: That's the point. One has to 11 put the discussion in the proper context.

12 MR. APOSTOLAKIS: Sure.

13 COMMISSIONER DIAZ: I have one minor question.

14 The current set of pilots that you are running to try to 15 check the applications, the ISI, the ISD, the graded QA, are 16 those providing you with sufficient feedback to address the 17 adequacy of the program?

18 MR. APOSTOLAKIS: We know what is going on. So 19 far, at least my personal opinion is that I haven't seen 20 anything that has helped me understand things better.

21 COMMISSIONER DIAZ: Okay.

22 MR. APOSTOLAKIS: But that may be coming. I don't 23 know.

24 COMMISSIONER DIAZ: There might be a little bias.

25 MR. SEALE: I would say, though, that my ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l .

I 51 I  ;

1 impression is that the pilots have provided me with a 1

2 considerable confidence that the people in the utilities are i 3 rapidly learning how to use PRAs in what they feel is a very I l i

! 4 constructive way. So it's not a tool that is going to be l

5 thrust into the hands of neophytes or anything like that.

l

6 , The other point is that the review that you will l 7 get from those people on the SRP when it goes out for public 8 comment will be a very competent review from their point of 9 view. You should expect to get comments from them which are 10 very practical, very focused on the issues that may remain 11 due to perhaps difficulties in articulating the review plan, 12 or whatever. But they're applying PRA methodology and doing 13 a pretty good job of it.

14 MR. APOSTOLAKIS: One last comment. I think the 15 interesting thing to see in this whole process is how 16 difficult it is for people who are used to doing things in a 17 highly prescriptive way, how difficult it is for them to 18 free themselves from that and move more towards a  !

19 performance-based system. 4 20 There are phrases in the guides that in the hands 21 of someone clever who wants to undermine the process are 22 killers. For example, one of the principles is " maintain 23 adequate defense in depth." Give that to me and I would not 24 approve any change, because you will never meet adequate 25 defense in depth, in my mind, if I want to act that way.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

I 52 1 What is the alternative? We all seem to like .

2 defense in depth even though that is a concept that is up in 3 the air. I don't think anybody has ever defined it. You do 4 have to have defense in depth. What is adequate defense in j 5 depth? That kind of fuzziness has to be there, but I must I

~

6' say it makes me very uncomfortable, because we have also l

7 other principles of that kind.

8 I appreciate the fact that you cannot just drop 9- everything and say make sure the core damage frequency is 10' below the goal. That's the'other extreme. But this is, I i

11 think, the primary difficulty in writing a good guide right 12- now.

I 13 CHAIRMAN JACKSON: In some sense that's-why it's a '

14 longer window than the end of this year and in fact it's the 15 end of next year that presumably there is going to be 16 iteration and re-normalization when there is' review on the i 17 outside not only by those who look at these things from an 18 intellectual perspective, but those'who actually are trying I 19 to make practical application of them. I i

20 Part of the reason I asked the question about i 21 assembling a group of specialists is that many times if we 22 study our own navels that's as much as we see. That's true 23 in terms of how soon we propagate things to the outside, to 24 let the world take a look, but also in letting the world

'25 take a look that there are other industries that use risk ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l

=

l

\

53 .

1 assessment methodologies and PRA, and presumably they could 2 share some of their wisdom. It may not be directly specific 3 in terms of application to nuclear plants, but there are 4 some fairly sophisticated uses of it at other places. I i 5 don't know if you have any reaction to that, but it's a bias l

{ -

6 th,at I have. l i

7 MR. APOSTOLAKIS: I don't think we will get much 8 help from other industries. I think we are at the 9 forefront. They may be using it or they may say they are {

10 using it, but I don't think anybody is using PRA to make l

11 decisions in other industries. Look how long it took in our {

12 industry. The reactor safety study was published in 1974, I

13 22 years ago. Now we are talking about risk-informed 14 regulation, not even risk-based, 22 years later.  !

I 15 CHAIRMAN JACKSON: Better late than never.

16 Commissioner McGaffigan. l l

17 COMMISSIONER McGAFFIGAN: I'm going to pass.

18 CHAIRMAN JACKSON: Dr. Kress.

I 19 MR. KRESS: The next item on our agenda is the 20 potential use of IPE/IPEEE results to compare the risk 21 status of the current population of plants with respect to l

22 the safety goals. It was my initial thought that this was 23 something that should be done, and it sounded like a very 24 good idea at the time I brought it up.

25 We then took a closer look at the IPEs and the ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

54 1 IPEEEs to see how this might actually be implementing. As ,

2 it turns out, in our view, these ac PRAs are just too uneven 3 and incoislete. In order to compare with the QHOs, for 4 example, you really do have to have some form of a 5 full-scope level 3 PRA or some surrogate that approximates 6 it,or bounds it.

7 Most of the IPE/IPEEEs cannot be characterized 8 that way, as full scope. None of them did shutdown risk.

9 They did a margins analysis for the seismic. Most of them 10 did a FIVE analysis. Many of them didn't go to level 2 11 even, especially with the fission product transport and the 12 source term part of it.

13 After looking at those and actually doing a great 14 deal of work of trying to figure out how to bound the 15 missing parts, for example, how to bound the consequences 16 that would come out of level 3 on a site-specific basis and

, 17 how to bound the missing parts in the level 1 and 2, we 18 thought that the results you would get by such bounding 19 analyses would just be too uncertain and too questionable 20 for the purpose, and that it really wasn't worth the effort 21 that it would take just for the purpose of seeing what the 22 status is with respect to safety goals.

23 The feeling is, looking at full-scope PRAs that do 24 exist, that we do meet the safety goals by considerable 25 margin. That's not a definitive answer because of the ANN RILEY & ASSOCIATEF LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

55 j 1 limitations in these studies, but that feeling is there.

2 Ne really think it ought to wait until sometime in 3 the future when better, more complete full-scope level 3s 4 are available, and then one can make a definitive statement.

5' We just didn't think it was worth the resources and the 6 expenditure to try to do that now.

7 CHAIRMAN JACKSON: Let me ask you these two 8 questions which kind of relate to that. Do you know what 9 percentage of the IPEs would meet the PRA review guidance 10 criteria being developed in the guidance document and the )

I 11 Standard Review Plan? '

12 MR. KRESS: The answer is no, I don't know. We 13 may have some other opinions.

14 MR. APOSTOLAKIS: Zero or perhaps one or two.

15 CHAIRMAN JACKSON: Dr. Powers.

16 MR. POWERS: Zero.

17 MR. KRESS: That was my opinion too. We have to 18 remember that the IPEs and IPEEEs weren't intended for that 19 purpose. So it's not a criticism.

20 CHAIRMAN JACKSON: I know they weren't, but the 21 second question bears on that. Do you have a feel for how 22 much use is currently being made, is trying to be made, or 23 should be made of the IPE results in the regulatory 1

24 decision-making process? That is, are they being made use l 25 of, should they be made use of, and how much are they being ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

56 1 used? .

2 MR. KRESS: Very interesting question. I can't 3 answer the part of it that says how much they are being 4 used. I can address the "should."

5 CHAIRMAN JACKSON: Are they being used as far you 6 knpw?

7 MR. KRESS: I think they are, yes.

8 CHAIRMAN JACKSON: In spite of what you've just 9 said?

10 MR. KRESS: Yes. Some of them are quite good on II the level 1 in addressing the core damage frequency. That 12 is a fairly appropriate and frequent use of them, I tnink, 13 in regulatory decision-making. That is being used.

14 CHAIRMAN JACKSON: What is our metric for 15 determining their acceptability in that context?

16 George, do you want to address that?

17 MR. APOSTOLAKIS: It pains me to say this, but you 18 really don't need to do a good uncertainty analysis and a 19 full-scope PRA to get excellent insights about your plan. I 20 used to think that unless you did that you didn't have a 21 good PRA. Of course you don't have a good PRA, but the 22 basic results that are of great use to the licensee and the 23 staff, namely, ranking the accident sequences, identifying 24 important systems, you can get those with very simplistic 25 analysis, point estimates and so on.

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

57 1 When in doubt, if you are a bit conservative, in 2 other words, should I include or exclude this particular 3 sequence, is it important or not, don't play games. Include 4 it. You know your analysis was crude. Keep it. But it 5 turns out that you really get a lot of good information that

~

6 way. In fact, some of the risk meters, the risk monitors 7 that some of the utilities are using now are using these 8 simple models.

9 CHAIRMAN JACKSON: I'm trying to make a separation 10 between the use that licensees make of it for their own 11 purposes within the current regulatory context and changes 12 in that regulatory context or regulatory decisions being 13 actually made based on them. There is a difference.

14 MR. APOSTOLAKTS: Yes.

15 MR. SEALE: There is a comment on the borderline, 16 though, and that is that a lot of the licensees are using l 1

17 their PRAs as a part of the process of the pilot studies I 18 which are input to regulatory decision-making.

19 I think almost across the board a common feature 20 is that most licensees are finding deficiencies in the IPEs l 21 or IPEEEs as they were originally cast and are going back i

22 and reexamining certain issues in coming up with the input 23 that they then have to their pilot evaluations.

24 That sort of demonstrates the other part of the 25 purpose of the IPE program, which was to get the utilities ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

58 1 to assess the showstoppers, and so forth. They are getting .,

2 used to using them, and when the questions arise, apparently 3 they are prepared to examine the issue in more detail.

4 As I say, that is an input to the regulatory 5 system, I think.

6 ,

MR. KRESS: Back to your question of the 7 acceptability for regulatory uses, I think our biggest 8 problem with them is their incompleteness, particularly that 9 they don't deal with the shutdown risk, and their incomplete 10 characterization of the effects of seismic. There are some 11 questionable parts in how they treat common cause failures, 12 and some of the reliability numbers that come out of 13 different databases don't seem to be consistent.

14 In terms of acceptability, the biggest problem we 15 have with them is their completeness.

16 CHAIRMAN JACKSON: How much progress is there in 17 terms of being able to incorporate into PRAs degradation of 18 equipment systems, components and equipment within certain 19 systems? i 20 MR. KRESS: I would defer to George on this one V

21 too, but my own personal opinion is it's hardly in the PRAs 22 at all.

23 CHAIRMAN JACKSON: In a certain sense, if one 24 wants to look at performance and the effect of maintenance 25 and effectiveness of maintenance and examine it within a ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

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

59 1 risk perspective, one has to be able to incorporate that, 2 and most of the PRAs and level 1 PRAs and the kind of l

accident sequences that are modeled do not incorporate that.

1 3 -

4 It's a success / failure, a binary approach.

5 MR. KRESS: Presumably some of the reliability 6 nu,mbers ought to implicitly reflect the status of 7 maintenance and QA and how that affects reliability, but 8 once it's in the PRA it's not changed. There is no time 9 variation; there is no differentiating between a good 10 maintenance program and a bad one.

11 George, you might want to talk about this. You 12 know a Jot more about the PRA.

13 MR. APOSTOLAKIS: It's true that aging effects are 14 not in the PRA right now, but the real question is whether 15 they would make a difference. The studies that I have seen 16 are very inconclusive. In other words, you cannot conclude 17 that the failure rate really goes up for particular 18 components.

19 CRAIRMAN JACKSON: I've seen plant-specific data 20 that indicates an ability to discern the effectiveness 21 within a given system, which has a certain reliability or 22 unavailability, the relative importance of certain 23 components within that system, which can change dramatically 24 the risk profile and in certain cases goes against the 25 conventional wisdom of what components or subsystems within ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

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

60 1 the larger one are affecting that overall system ,

2 reliability, but it requires a treatment that has to do with 3 a non-binary approach to condition or degradation.

4 MR. APOSTOLAKIS: Sure.

5 CHAIRMAN JACKSON: So it's a different dimension 6 in, terms of the statistical approach and the kind of i

7 statistical modeling that goes into the PRAs, but it is one '

8 that, at least based on a couple of examples that were 9 shared with me -- and I'll be happy to talk with you about  ;

10 those --

11 MR. APOSTOLAKIS: I would like to see those.

12 CHAIRMAN JACKSON: They have a very dramatic l

-13 effect. So if you are really talking about efficacy of 14 maintenance, and so forth, it's an issue. If you are .

15 talking about real plant-specific applications and 16 understanding how the risk profile changes in a given plant 17 as a consequence of a maintenance program, I think this is 18 relevant.

19 MR. KRESS: I think it would be a quantum 20 improvement in the PRAs to incorporate that sort of variable 21 probability of failure or variable reliability that is not 22 just binary, yes or no, fail, and relate it to plant 23 specific items. That is a good idea. I hope somebody is 24 approaching that.

25 CHAIRMAN JACKSON: I've seen some work that is l

e ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

61 1 along that line.

2 MR. APOSTOLAKIS: It seems to me that before we 3 jump to conclusions we should really look at how that work 4 was done. I have seen some of that work too. I don't know 5 if it's the same work.

~

l 6 , CHAIRMAN JACKSON: I don't think the issue is to 7 debate the specifics. I think the issue has to do with an l 8 ability to migrate into this framework some capability of

! 9 really understanding on a plant-specific basis what 10 equipment degradation means in terms of the risk profile of i 1

11 the plant and what that can or cannot say about the efficacy 12 of maintenance and maintenance programs, because that in 13 fact is relevant to us in terms of the implementation of 14 things like the maintenance rule, and particularly if one is 15 going to marry these PRA and risk-informed approaches to 16 real life situations.

17 MR. APOSTOLAKIS: I agree with you. but I think 18 right now, based on the evidence I have seen -- not the 19 models, the evidence -- I would say that it's inconclusive.

20 The evidence is not telling us that this is an urgent issue.

21 Let me put it that way. I'm getting now awfully close to 22 being conflicted, by the way. We should have that 23 capability, but I don't think it's an urgent issue.

l 24 MR. KRESS: It's another good reason to be in

, 25 favor of the reliability database program.

l l

l l ANN RILEY & ASSOCIATES, LTD.

t Court Reporters i 1250 I Street, N.W., Suite 300

( Washington, D.C. 20005

} (202) 842-0034 l

62 I i

1 CHAIRMAN JACKSON: Why don't we go on. We are ,

l t

2 running out of time.

{

3 MR. KRESS: The next issue we have touched on a 4 little is also mine. It's the use of safety goals on a i

5 plant-specific basis.

l

~

6 ,

It's our view that the safety goals were not 7 originally intended for that purpose, but if the desire is 8 to move toward a more risk-informed or risk-based regulatory 9 program, one will have to deal with specific plants, because 10 that's what you are doing with them. They all come in with ,

11 requests for changes to the licensing basis and a decision 12 will have te made, for example, as to whether it should be 13 granted, and you will want to be risk-informed en that. You 14 will have to have some sort of, I guess we could call them, ,

i 15 acceptance criteria as to whether you will grant 16 particularly, say, increases in risk, small increases in  !

17 risk, or grant a change at all, or whatever.

18 We felt that it's not necessary that the safety 19 goals be these acceptance criteria but that that would be a  :

20 good place to start since we do have defined things intendcd 21 to tell us how safe is safe enough. l 22 We think that it's a good place to start, that you 23 will need to be plant specific in application in a 24 risk-informed, performance-based world, and that in order to 1

25 do it with the safety goals there is the subsidiary goal of q l

)

1 ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 1

Washington, D.C. 20005 i (202) 842-0034

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

63 i

1 ten to the minus fourth, vhich is the most useful one 2 because it's directly obtainable from PRAs that we now have.

3 The tough ones to deal with are the QHOs, which 4 are the more important ones. Well, I think they are the 5 more important ones, but ten to the minus fourth is just as  !

6 important. In order to really deal with that as an 7 acceptance criteria you do have to have some surrogate for a 8 full-scope level 3 PRA with consequences that are site 9 specific.

10 In our letter we said that that would be the 11 ultimate that you are looking for, but it's going to be some 12 time before each plant, or at least a substantial traction 13 of them, have such a full-scope capability, and it would be 14 appropriate to back off from that and develop surrogates for l 15 the QHO that would be bounding.

16 We feel very strongly that such currogates can be 17 developed. They will take the form of a combination of core 18 damage frequency and a large early release, which has to be 19 defined, but we have some ideas on how to define that. Or 20 an even lower tier than that would be a core damage 21 frequency combined with a. conditional containment failure 22 probability, which might be different for different classes 23 of plants, BWRs versus PWRs.

24 We have endorsed the concept of developing these 25 lower tier surrogates that can be used as acceptance ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

1

. I 64 l i

1 criteria. We are in the process now of putting together , !

, 2 what I guess I would call a white paper that would more j i

4 3 fully define what we mean by things like an LERF and a 4 conditional core damage frequency for this use, how they 5 would be derived directly from the QHOs, and how one would t 6 de, rive them in such a way that you're sure they are bounding ,

t s

i 7 but yet are still useful tools in such an acceptance 6

! 8 criteria. i f'

9 We are not quite through with that white paper.

, 10 It's under discussion by the Committee, but I think it would l

l 11 be useful in defining and further clarifying our views on 12 this subject.

l ,

. 13 CHAIRMAN JACKSON: When do you think you would i

j 14 have that?

, 15 MR. KRESS: I expected to have it this ACRS  !

!' 16 meeting, but it didn't quite get there. It will surely be i 17 finished by next week and circulated.  !

} 18 CHAIRMAN JACKSON: So you're talking within weeks.  !

4 19 MR. KRESS: Oh, yes. It's not that far away.

20 CHAIRMAN JACKSON: Commissioner Rogers.

21 COMMISSIONER ROGERS: There are lots of questions 22 here. How the safety goals relate to an adequate protection i 23 standard. What's the relationship between these things? If '

24 you want to start using safety goals for regulatory I 25 purposes, what does that really mean if we already have an  !

i ANN RILEY & ASSOCIATES, LTD. [

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l 65 )

l 1 adequate protection standard that is being met? '

2 I think one has to try to sort this out. I know 3 it's a very, very difficult question. I certainly have i

4 struggled with it in my own mind for years. But if you want 5 to use safety goals, surrogates or not, for regulatory 6 pu,rposes, I think you are going to have to come to grips 7 with how will a regulatory decision based on safety goals, 8 which will be a probabilistic statement of affairs, be 9 related to the adequate protection standard, which is not a 10 probabilistic standard, I don't think.

11 MR. KRESS: It's not.

12 COMMISSIONER ROGERS: I'm not even sure what it 13 is. We know it's there, but what is it?

14 I think we are going to have to come to grips with 15 that if we want to start regulating using safety goals.

16 MR. KRESS: I think that is a wonderful question.

17 COMMISSIONER ROGERS: I don't know if it's 18 wonderful, but it's certainly --

19 MR. KRESS: I've thought about it considerably.

20 COMMISSIONER ROGERS: I think we need your best l 21 thoughts on this, because it seems to me this is where the l 22 treacherous territory is going to come.

23 MR. KRESS: It really is. The real question is,  !

24 where do the safety goals lie with respect to adequate l 25 protection in terms of level of risk? It is my personal ANN RILEY & ASSOCIATES, LTD.

Court Reporters 2250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

66 1 feeling that adequate protection is a level of risk, if you ,

2 could translate it into risk, that is below the safety  !

3 goals.  !

4 . I say that because what we have now is an adequate 5 protection type of concept, and'from all the PRAs I've seen f

6 and everything that.I can get from the IPEs to see what the f i

7 status of risk is, we.are well below the safety goals, which 8 tells me that adequate protection is a level of safety that  ;

9 is better than the safety goals.

l 10 COMMISSIONER ROGERS: I think we could argue that, j 11 because if in fact you believe that all the plants meet i

12 safety goals, they have gotten there through application of l 13 an adequate protection standard.  !

14 MR. KRESS: But adequate protection is a different 15 level of safety for each plant; it's plant specific; and it  :

16 has never really been quantified in terms of what.is a risk.

17 If we were to now say we want to move to another criteria, 18- which is safety goals, you are treading on very thin ice, '

19 because if it is a level above that, you really have to be 20 careful how you are going to define these acceptance l 21 criteria. This is one of the reasons I think, as you saw in 22 one of our letters, that the safety goals are not quite well 23 formulated for this purpose of plant-specific application.

24 That is one of the main reasons that we mentioned, because 25 you are probably putting forth a set that is not as good in  ;

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 4

- - p-,-. - , - - n , -

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

67 1 terms of risk level as adequate protection, and that is a 2 real concern that I think has to be dealt with.

3 COMMISSIONER ROGERS: I'm sure we are not going to I 4 settle it right here, but I do think that it is a very key j 5 question in thinking about the use of safety goals for l j

6 regulatory purposes.

7 MR. KRESS: It is, definitely.

1 8 COMMISSIONER ROGERS: Yes, Dr. Apostolakis.

9 MR. APOSTOLAKIS: I think there are a couple of 10 thoughts here. First of all, I don't think that what we 11 have now, the deterministic system we have now defines 12 adequate protection. I think the clear statement about i

13 adequate protection was the QHOs the Commission approved a 14 number of years ago.

15 CHAIRMAN JACKSON: Why don't you define for the 16 Commission the QHOs. He knows. We have new Commissioners.

17 MR. APOSTOLAKIS: I'm sorry. Those were approved 18 in the mid-1980s, I understand.

19 MR. CATTON: Quantitative health objectives.

20 MR. APOSTOLAKIS: They are both qualitative and 21 quantitative, and the quantitative part says, I think, in 22 terms of individual risk that the risk from nuclear power 23 plants should not exceed one tenth of one percent of the 24 risks from all other causes, and similar kinds of things for 25 delayed deaths, and so on. It's one tenth of one percent of ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

68 1 societal risks. .

2 It seems to me that really defined adequate 3 protection. The other one was sort of haphazard: let's do 4 this, let's do that, and then that's adequate protection.

5 It's interesting to remember, by the way, that 6 wh,en the reactor safety study was published a lot of the 7 old-timers were surprised that the core damage frequency was 8 so high. They didn't think it was going to be close to ten 9 to the minus four and five. They were very surprised. So 10 when you quantify things you see them from a different 11 perspective.

i 12 One other thing that maybe is not directly related 1

13 but I really think I ought to tell you is I'm extremely 1 i

14 uncomfortable with this notion that the safety goals, the 15 quantitative health objectives apply on the average. I 16 don't know what that means to the industry, on the average, 1 17 and the sooner we get out of it the better off we'll all be.

18 COMMISSIONER ROGERS: It's really just hand 19 waving. It isn't really established when one says that. So 20 I agree with you. I'm not disagreeing.

21 The other point that I would like to raise, a

22 Dr. Kress, is you seem to imply that somehow there would be 23 eventually the development of level 3 PRAs by all plants. I 24 don't understand what would drive anybody to that at this a

25 point. Why would you do it?

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

l 69 1 MR. KRESS: I believe that if we do move into a 2 risk-informed, performance-based or risk-based system, then 3 what that means to the industry is -- one of the things it 4 means is that they can come in now with requests for relief 5 from burdensome regulations that don't add to their safety

~

6 bu; really causes them a great deal of problem in terms of 7 resources and time and actually may detract from safety. l 1

8 If they see the probability or good possibility of 9 getting some relief like this and the acceptance criteria 10 that we have on granting such relief has to do with meeting 11 QHOs, which require a level 3, that would be their incentive 12 to develop. The owners of these level 3s will have to be 13 the plants, not NRC. It will have to be in their hands to 14 come forth with sufficient justification in terms of their 15 position with respect to the OHOs and how a change changes 16 that position. I think the incentive is a relief of burden.

17 COMMISSIONER ROGERS: Provided that there is some 18 clear regulatory statement from NRC that it would accept 19 something and allow some relief, if you want to call it 20 that. I don't think we have done that.

21 MR. KRESS: No. I think that is part of what you 22 will see in the Standard Review Plan for implementation of 23 PRA and the reg guides that we are reviewing now. That was j 24 one of the concepts that is built into that. It does 25 involve relief and allowing small increases in risk in the l

ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 I (202) 842-0034

J 70 I 1 interest of overall improvements in the whole process. I ,

2 think that's the incentive, and without that, I don't think 3 we will see level 3 PRAs.

4 COMMISSIONER ROGERS: I think you have to be very 5 explicit in your comments on this.

1 6 .

MR. KRESS: I think we intend to be on that 7 'particular one. ,

i 8 COMMISSIONER ROGERS: Thank you.

9 CHAIRMAN JACKSON: Commissioner Dicus. [

10 COMMISSIONER DICUS: No.

l t

i 11 CHAIRMAN JACKSON: Commissioner Diaz.

I

' i 12 COMMISSIONER DIAZ: No. I support strongly J  !

13 Commissioner Rogers' question.  !

1 r 4 .

14 MR. KRESS: There is one more item. We probably  !

j 15 can finish it very quickly. That's the use of the RuleNet 16 in the regulatory process.

17 Dr. Shack.

18 MR. SHACK: Several of our members have looked at 19 the RuleNet site, although none of us managed to register in l

20 time for formally participating in the process, f 21 I believe that most of us think that the 22 technology represented by RuleNet does offer a great promise 23 as a way to involve the public, licensees, intervenors, and 24 staff in a more effective interaction in the regulatory 25 process. This improved interaction can lead to greater 7007 RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

71 1 confidence in agency decisions and greater public confidence i

2 in the way the agency deals with the industry.

i 3 The whole process will obviously have growing 4 pains and evolve as we are doing it. It's hardly 5 transparent to use in some ways. There is some concern that 6 Internet access is kind of still an elitist thing, although  !

7 I think that is widely becoming much more accessible. From 8 my point of view, even in the current state it still offers 9 the easiest public access to the regulatory process that I i 10 can envision.

11 I have certain biases as a computer jock. I think 12 the give and take of an electronic forum is far superior to 13 sending in a comment or a letter that sort of goes into a 14 black hole and sometime later comes out as a resolution of 15 public comment. There is a real give and take in near real i

16 time in electronic forum. And it may even offer greater 17 opportunities for reflection than a public meeting. One 18 just reads the threads in the letters and there is a 19 discussion and you begin to think about it. I think it 20 offers a great deal of promise.

21 I was a little disappointed in how few people 22 participated in RuleNet, and that may be partly just the 23 uncertainties involved with it. They certainly didn't make 24 it very easy to find. I could never understand why there 25 wasn't a direct link from the NRC home page to RuleNet, and l

a ANN RILEY & ASSOCIATES, LTD.

I Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034 l l

__. _ _ __ _ .. _. _ _ _ _ . - . _ . _ _ _ _ .- __ ___ __ . _ _ . = _ .

72 (

L 1 there is still no direct link from the home page to the LSS ,

2 Net, which is the sort of successor attempt at this. If you 3 are going to make this publicly accessible, it should be a 4 little more transparent to the public in how to get there.

5 Again, we have no formal ACRS position on this, 6 but I think those of us who looked at it thought it was a 7 very interesting attempt to increase public involvement.

8 CHAIRMAN JACKSON: Thank you.

9 Commissioner Rogers. f 10 COMMISSIONER ROGERS: No thank you.

11 CHAIRMAN JACKSON: Commissioner Dicus. j 12 COMMISSIONER DICUS: No.

13 CHAIRMAN JACKSON: I'd like to thank the ACRS very 14 much for another informative briefing and again to thank 15 particularly Dr. Catton.  !

16 The topics of today's presentation obviously l 17 focused on a number of issues that are critical for 18 maintaining and improving the NRC's ability to regulate h 19 effectively. I want to encourage the ACRS to continue to 20 provide the Commission its perspective on issues important 21 to our mission and to be forward looking in bringing 22 developing concerns to the Commission's attention in order 23 to help ensure that we are prepared to meet the future 24 challenges.

25 Clearly there are a number of follow-up issues  !

l ANN RILEY & ASSOCIATES, LTD. l' Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034  ;

73 1 here that we have discussed, several of which have been 2 explicitly discussed here today and are of concern to the 3 Commission. Two examples. Some of our discussion in the 4 digital I&C area and then the later discussion of the 5 relation of the safety goals to the adequate protection 6 st,andard. I think we all know where we have to go from 7 here.

8 Unless my fellow Commissioners have any further 9 comments, I'm about to adjourn. In adjourning, I am going 10 to ask that we clear, because we have an affirmation to do, 11 and then we can follow up and have any follow-on discussions 12 that anyone would like to make. Thank you. We're 13 adjourned.

14 [Whereupon, at 11:29 a.m., the meeting was -

15 adjourned.] '

16 17 e

18 19 20 21  !

22 23 24 25 ANN RILEY & ASSOCIATES, LTD.

Court Reporters 1250 I Street, N.W., Suite 300 Washington, D.C. 20005 (202) 842-0034

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

CERTIFICATE l

i l

This is to certify that the attached description of a meeting i of the U.S. Nuclear Regulatory Commission entitled:

TITLE OF MEETING: MEETING WITH ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS) - PUBLIC MEETING 1

i PLACE OF MEETING: Rockville, Maryland l

DATE OF MEETING: Friday, December 6, 1996 was held as herein appears, is a true and accurate record of the meeting, anc that this is the original transcript thereof l taken stenographically by me, thereafter reduced to typewriting by me or under the direction of the court reporting company Transcriber: /7f)tpltera[ M//tffn i Reporter: Michani Paulus l

l I l

___. . ~ . ._- __ __ . _ _ _ _ _ _

y

/ 'o,,

n UNITED STATES NUCLEAR REGULATORY COMMISSION 4 $

o

,$ ADVISORY COMMITTEE ON REACTOR SAFEGUARDS j WASHINGTON, D. C. 20555 November 26, 1996 MEMORANDUM TO: John C. Hoyle Secretary of the Commission FROM:

~

John T. Larkins, xec boIY ive Director Advisory Committee on Reactor Safeguards

SUBJECT:

ACRS MEETING WITH THE NRC COMMISSIONERS, DECEMBER 6,1996-SCHEDULE / BACKGROUND INFORMATION -

The ACRS is scheduled to meet with the NRC Commissioners between 9:30-11:00 a.m. on Friday, December 6, 1996, to discuss the items listed below. Background materials related to these items are attached.

A. Introduction - NRC Chairman 9:30-9:35 a.m.

B. 1. Digital Instrumentation and Control 9:35-9:45 a.m.

Systems - Dr. Miller (pp. 1-13) )

2. Office of Nuclear Regulatory Research 9:45-10:10 a.m.

Plan for Upgrading Thermal-Hydraulic Codes - Dr. Catton (pp. 14-37)

3. Risk-Informed, Performance-Based 10:10-10:30 a.m.

Regulation and Related Matters -

Dr. Apostolakis (pp. 38-49)

4. Potential Use of IPE/IPEEE Results 10:30-10:40 a.m.

to Compare the Risk of the current Population of Plants with the Safety Goals - Dr. Kress (pp. 50-66)

5. Use of Safety Goals on a Plant- 10:40-10:50 a.m.

Specific Basis - Dr. Kress (pp. 67-77)

6. Use of RuleNet in the Regulatory 10:50-10:55 a.m.

Process - Dr. Shack (pp. 78-86)

C. Closing Remarks - NRC Chairman 10:55-11:00 a.m.

Attachment:

As stated i i

cc: ACRS Members ACRS Technical Staff  ;

1

W 4

W ITEM B.1:

DIGITAL INSTRUMENTATION AND CONTROL SYSTEMS (DR. MILLER) 4 l

l

ITEM B.1: DIGITAL INSTRUMENTATION AND CONTROL SYSTEMS During the May 24,1996 meeting with the Commissioners, the ACRS discussed the issues associated with the use of digital instrumentation and control (I&C) systems in nuclear power plants. The ACRS discussed the status of its review of the proposed update to Standard Review Plan (SRP), Chapter 7, " Instrumentation and Controls," which includes SRP sections, Branch Technical Positions (BTPs), and regulatory guides for digitalI&C systems. The ACRS also discussed the results ofits review of the National Academy of Sciences / National Research Council (NAS/NRC) Phase 1 study repon and its expectation to review the Phase 2 study when available. The NAS/NRC Phase 2 study report was initially scheduled to be completed on September 30,1996. However, it has been delayed and the exact date of availability of this report is uncenain at this time.

On October 10,1996, the ACRS completed its review of the proposed SRP Chapter 7 update.

ACRS Subcommittee on I&C Systems and Computers also met with representatives of the NRC staff and its contractor, the Lawrence Livermore National Laboratory, to discuss proposed SRP sections, BTPs, and regulatory guides on October 8,1996.

The Committee issued letters dated June 6, and October 23, 1996, regarding the draft SRP sections, BTPs, associated regulatory guides, and a safety evaluation repon (SER) on a topical repon by the Electric Power Research Institute (EPRI) on electromagnetic /radiofrequency interference (EMI/RFI). In those letters, the ACRS identified several issues regarding the level of detail provided in the regulatory guides, the balance in the guidance between the review of the design process and the assessment of the product, the linkage between Chapter 7 and other SRP Chapters, and graded approaches based on importance to safety. As stated by the EDO in the November 1,1996 letter to T.S. Kress, the ACRS plans to discuss these matters with the staff during future meetings. The Committee did, however, inform the staff that it has no objection to issuing the draft SRP Chapter 7 and associated guidance documents for public comment. The staff is scheduled to brief the ACRS on the proposed final SRP, BTPs, and regulatory guides in April-May 1997, after reconciliation of public comments.

The ACRS expects to hear a briefing on the NAS/NRC Phase 2 study report in Febmary 1997, depending on the availability of this report. The Phase 1 study report identified safety {

and reliability issues arising from the introduction of digital I&C technology. The Phase 1 >

study report identified the following eight key issues - six technical and two strategie:

Technical l l

e software quality assurance e common-mode software failure potential e system aspects of digital I&C technology l e human factors and human-machine interfaces j e safety and reliability assessment methods l

• dedication of commercial off-the-shelf hardware and software i

1 I

_ _ _ . . ~ _ . . _ _ _ _ . _ _ . _ . _ _ . _ _ _ _ . _ _ _

Strateeic e case-by-case hcensmg t e adequacy of technicalinfrastructure The Phase 2 study repon is expected to identify criteria for review and acceptance of digital systems in retrofits of existing plants and in new standard plant designs. This repon is also i expected to characterize and evaluate alternative approaches to cenification of digital systems and to rec'6mmend guidelines for reviewing licensee submittals. The staff plans to incorporate the insights from the Phase 2 study into the proposed final SRP, BTPs, and regulatory guides.  ;

i The ACRS also expects to review an SER on an EPRI topical repon on commercial off-the- i shelf (COTS) software. The staff is working closely with an EPRI working group on this l matter. The use of COTS software and hardware is an imponant issue. The staff plans to j incorporate appropriate guidance into the proposed fm' al SRP and associated guidance '

documents, as appropriate. The ACRS also plans to review the draft regulatory guide for equipment protection against lightning and other guidance for environmental stressors. The Committee will review these matters when the documents become available.

Attachments:

e Letter dated October 23,1996, from T. S. Kress, ACRS Chairman, to James M. Taylor, Executive Director for Operations,

Subject:

" Draft Update of Standard Review Plan, ,

Chapter 7, ' Instrumentation and Controls'" (p. 4) e Letter dated November 1,1996, from James M. Taylor, Executive Director'for Operations, to T. S. Kress, ACRS Chairman,

Subject:

" Draft Update of Standard Review Plan, Chapter 7, ' Instrumentation and Controls'" (pp. 5-6) e Letter dated June 6,1996, from T. S. Kress, ACRS Chairman, to James M. Taylor, Executive Director for Operations,

Subject:

" Regulatory Guidance Documents Related to Digital Instmmentation and Control Systems" (pp. 7-9) e Letter dated June 21,1996, from James M. Taylor, Executive Director for Operations, to T. Kress, ACRS Chairman,

Subject:

" Regulatory Guidance Documents Related to Digital Instrumentation and Control Systems" (p.10) e Repon dated October 13, 1995, from T. S. Kress, ACRS Chairman, to Shirley Ann Jackson, NRC Chairman,

Subject:

National Academy of Sciences / National Research Council Study on " Digital Instmmentation and Control Systems, Safety and Reliability Issues" - Phase 1 (pp.11-12) 2

?

e i

• Letter dated October 31, 1995, from James M. Taylor, Executive Director for Operations, to T. S. Kress, ACRS Chairman,

Subject:

"The National Academy of  !

Sciences' Report on Digital Instrumentation and Control, Safety and Reliability Issues" (p.13) l i

t i e

)

l l

l l

3 i

j

/pMthy'og UNITED STATES

! '.  ! n NUCLEAH REGULATORY COMMISSION j $ I ADVISORY COMMriTEE ON REACTOR SAFEGUARDS I WASHINGTON, D, C. 20$55 3 .....

1 October 23, 1996 l

d j Mr. James M. Taylor '

Executi've Director for Operations j U.S. Nuclear Regulatory Commission ,

j Washington, D.C. 20555-0001 l

. 1

Dear Mr. Taylor:

j

SUBJECT:

DRAFT UPDATE OF STANDARD REVIEW PIAN, CHAPTER 7, i " INSTRUMENTATION AND CONTROLS" I

During the 435th meeting of the Advisory Committee on Reactor l l Safeguards, October 9-12, 1996, we reviewed portions of the draft j update of Standard Review Plan (SRP), Chapter 7, " Instrumentation

! and Controls." We heard presentations by and held discussions with

! representatives of the NRC staff and its contractor, the Lawrence l Livermore National Laboratory (LLNL), regarding proposed SRP

! sections and Branch Technical Positions (BTPs) related to digital i instrumentation and control (IEC) systems. In addition, our l Subcommittee on Instrumentation and Control Systems and Computers i

met with the NRC staff and LLNL on October 8,1996, to discuss this j natter. We had previously met with the staff and ILNL in March and j May 1996 to discuss draft SRP nections, BTPs, and associated j regulatory guides, and provided comments in a letter dated June 6, j 1996. We also had the benefit of the documents referenced.

We have no objection to the staff's proposal for issuing the draft i update of SRP Chapter 7 and associated BTPs for public comment.

4 However, in the June 5,1996 letter, we identified issues regarding the level of detail provided in the regulatory guides, the balance

! in the guidance between the review of the design process and the

, assessment of the product, the linkage between Chapter 7 and other

. SRP chapters, and graded approaches based on importance to safety.

4 In a letter dated June 21, 1996, you responded to our letter of 1

June 6,1996, stating that the staff will continue its discussions with the ACRS on these issues. We plan to discuss these matters with the staff during our future meetings. ]

1 Sincerely, j

h.

T. S. Kress j Chairman 4

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

N p  %. uumso CTArms j NUCUEAR REGUL.ATORY COMMISSION l

WAGMitf8 TON. D.C. MEEB GOM

~

November 1, 1996 Dr. Thomas S. Kress, Chairman Advisory Committee on Reactor Safeguards U.S. Nuclear Regulatory Commission Washington, D.C. 20555

Dear Dr. Kress:

$UBJECT: DRAFT UPORTE OF STAleARD REVIEW PLAN, CHAPTER 7 "IltSTRUMENTATION Als CONTRDLS" The ACRS, in its letter dated October 23, 1996, commented on the staff's proposed draft update of Standard Review Plan (SRP), Chapter 7

'Instrumsntation and Controls." The ACRS stated that it has no objection to the staff's proposal for issuing the craft SRP Chapter 7 for public comment.

The staff plans to issue the draft SRP Chapter 7 update for public comment in November 1996.

The staff will continue its discussion with the ACRS inconjunctionwiththe resolution of public comments onthelevelofdetailintheregulatoryguides referenced in the SRP, the balance in the guidance between the review of the design process and the assessment of the product, the linkage between Chapter

, 7 and other SRP chapters, and graded approaches based on importance to safety.

l The staff plans to meet with the ACRS in the first quarter of 1997, when the final SRP Chapter 7 update is completed.

]

$1ncerely, J

3ame mes M. aylor ecutive Director for Operations '

l cc: Chatruan Jackson l Commissioner Rogers Commissioner Dicus l

Commissioner Diaz Commissioner McGaffigan SECY l

5

?

?

Mr. James M. Taylor )

j l

References:

1. Memorandum dated September 16, 1996, from Frank J. Miraglia, Jr., Office of Nuclear Reactor Regulation, to Edward L.

Jordan, Committee to Review Generic Requirements,

Subject:

Request for Review of Updated Standard Review Plan Chapter 7, Instrumentation and controls (attached) ,

2. Letter dated June 6,1996, from T. S. Kress, ACRS Chairman, to l j James M. Taylor, Executive Director for Operations,

Subject:

Regulatory Guidance Documents Related to Digital ,

i Instrumentation and control Systems

! 3. Letter dated June 21, 1996, from James M. Taylor, Executive Director for Operations, to T. S. Kress, ACRS Chairman,

Subject:

Regulatory Guidance Documents Related to Digital Instrumentation and Control Systems i

6

_ _ _ _ .._ _ _ _ _ . - _ _ _ _ _ . . _ ~ ___._..___.__.__-.._.___._m.- _ _ . . -.. _ . _ .

1

[ 'g, UNITED STATES

) *.-~ 8 o NUCLEAR REGULATORY COMMISSION

) $ ADVISORY COMMITTEE ON REACTOR SAFEGUARDS

$ WASHINGTON, D. C. 30586 h~

j - k*****

i

June 6, 1996 i

j Mr. James M. Taylor i Executive Director for Operations j U.S. Nuclear Regulatory Commission i Washington, D.C. 20555-0001 l

Dear Mr. Taylor:

SUSJECT: REGULATORY GUIDANCE DOCUMENTS RELATED TO DIGITAL INSTRUMENTATION AND CONTROL SYSTEMS During the 429th and 431st meetings of the Advisory Constittee on Reactor Safeguards, March 7-9 and May 23-25, 1996, we reviewed portions of the proposed Standard Review Plan (SRP), Branch i

Technical Positions (BTPs), and Regulatory Guides related to digital instrumentation and control (IEC) systems. We held discussions with representatives of the NRC staff and its a contractor, the Lawrence Livermore National Laboratory (LLNL) . In

{ addition, our Subcosaittee on IEC Systems and Computers met with ,

1 the NRC staff and LutL to discuss these documents on March 6 and i

{ May 22,1996. We also had the benefit of the documents referenced.

l l The staff requested ACRS to review the SRP Chapter 7 update in the early stages of development to aaaa==adate the schedule set forth j; in the Digital IEC Task Action Plan. The staff expects to complete

development of the SRP Chapter 7 update and associated guidance in i september 1996, integrate the recossendations from the National i i Academy of Sciences / National Research Council (NAS/NRC) Phase 2 1 i study report in Cictober 1996, publish the Draft SRP Chapter 7 and i associated guidance for public comment in December 1996, and issue the final SRP and related guidance in May 1997.

The staff is revising the SRP, adding two new sections, developing new BTPs, and preparing six regulatory guides that endorse eight industry s+=adaw15s. The staff presented a safety evaluation report l (SER) on an Electric Power Research Institute (EPRI) topical report

for electromagnetic /radiofrequency interference (EMI/RFI) design i requirements and testing. A planned BTP on cosaarcial off-the-I shelf (C0TS) software may be replaced by an SER on a topical report i being developed by an EPRI working group. We concur with the staff l conclusions in the SER associated with the EPRI topical report on i EMI/RFI and encourage the staff to complete an SER for the EPRI topical report on COTS.

1 i

i 7 t

~

i l l

} Considering the fact that the staff is using generally accepted

) U.S. software engineering practices, it appears that the staff j

approach is appropriate to update the SRP and associated guidance l to codify the current regulatory framework for digital IEC. We i raised several issues (e.g. , the linkage between SRP Chapter 7 and l other SRP chapters, and graded approaches based on importance to

] safety) that were subsequently clarified by the staff. The staff j agreed to document these clarifications.

i We have raised other issues that include the level of detail j provided in the regulatory guides and the balance in the guidance j between the review of the design process and the assessment of the

product. We plan to report on these and other digital I&C issues j at a later date. -

We plan to review the staff's remaining SRP sections, the BTPs, and

the SER on the EPRI topical report on COTS when they become 4

available.

4 sincerely, k ^

N J

)

J. 5. m= _

T. S. Kress j Chairman l

Rafarances:

1. U. S. Nuclear Regulatory Commission, Standard Review Plan, J Section 7.0, " Instrumentation and Controls-Overview of Review

! Process," Draft Version 3.0, February 12, 1996 4

2. U. S. Nuclear Regulatory Commission, Standard Review Plan, Section 7.1, " Instrumentation and Controls-Introduction,"

i Draft Version 7.0, February 14, 1996 l 3. U. S. Nuclear Regulatory Commission, Standard Review Plan, 1 Section 7.2, " Reactor Trip System," Draft Version 6.0, April

17, 1996 1 4. U. S. Nuclear Regulatory Commission, Standard Review Plan, Section 7.9, " Data Communications," Draft Version 4.1, April 18, 1996 j 5. U. S. Nuclear Regulatory Commission, (Proposed) Branch i Technical Position HIC 8-14

" Guidance on Software Reviews for I Digital Computer-Based Instrumentation and Control Safety i Systems," Version 9.0, February 14, 1996

6. U. S. Nuclear Regulatory Commission, (Proposed) Branch

Technical Position HICB-16

" Guidance on the Level of Detail

Required for Design certification Applications Under 10 CFR

! Part 52," Version 7.0, April 12, 1996 4

8

. -3

7. U. S. Nuclear Regulatory Commission, Draft Regulatory Guides, transmitted by memorandum dated February 9, 1996, from M.

Wayne Rodges, Director, Office of Nuclear Regulatory Research, NRC, to John T. Larkins, ACRS:

e U. s. Nuclear Regulatory Commission, Draft Regulatory Guide DG-XXXX, Version 2.7.2, " Verification, Validation, Reviews, and Audits for Digital Computer software Used in safety systems of Nuclear Power Plants" e' O. S. Nuclear Pegulatory Commission, Draft Regulatory Guide DG-XXXX, Version 2.0.7, " Configuration Management Plans for Digital Computer Software Used in safety systems of Nuclear Power Plants"

8. U. s. Nuclear Regulatory Commission, Draft Regulatory Guides, transmitted by memorandum dated April- 26, 1996, from M. Wayne Bodges, Director, Office of Nuclear Regulatory Research, NRC, to John T. Larkins, ACRS:

e Draft Regulatory Guide DG-XXXX, Version 2.0, " software Unit Testing for Digital Computer software Used in safety systems of Nuclear Power Plants"

• Draft Regulatory Guide DG-XXXX, Version 2.0, " Developing software Life Cycle Processes for Digital Computer software Used in safety systems of Nuclear Power Plants" e Draft Regulatory Guide DG-XXXX, Version 2.0, " software Requirements specifications for Digital Computer software Used in safety systems of Nuclear Power Plants" e Draft Regulatory Guide DG-XXXX, Version 2.0, " software Test Documentation for Digital Computer software Used in safety systems of Nuclear Power Plants"

9. Memorandum dated January 30, 1996, from F. Miraglia, Office of Nuclear Reactor Regulation, NRC, to E. Jordan, Committee to I

Review Generic Requirements, NRC, subject: Request for i Endorsement of the safety Evaluation Report on Electric Power i Research Institute Topical Report, TR-102323, " Guidelines for Electromagnetic Interference Testing in Power Plants" j

l i

4 i

i

' 9

( '4j y-

% UNITED STATES

! 2 NUCLEAR REGULATORY COMMISSION l.-

,  ! wasniucrow. o.c. somewooi

% .,... f June 21, 1996 Dr. T.S. Kress, Chairman Advisory Committee on Reactor Safeguards U.S. Nuclear Regulatory Comission '

Washington, D.C. 20555 l

SUBJECT:

' ~

REGULATORY GUIDANCE DOCUMENTS RELATED TO DIGITAL INSTRUMENTATION AND CONTROL SYSTEMS i

Dear Dr. Kress:

The ACRS, in its letter on this subject dated June 6, 1996, commented on the staff's proposed Standard Review Plan (SRP) update, Branch Technical Positions (BTPs), and new Regulatory Guides related to digital instrumentation and l

control systems. The ACRS stated that it appears that the staff approach is appropriate to update the SRP and associated guidance (BTPs and Regulatory i

Guides). Additionally, the ACRS concurred with the staff conclusions in the Safety Evaluation Report (SER) associated with the Electric Power Research Institute (EPRI) topical report on electromagnetic /radiofrequency interference.

The ACRS encouraged the staff to complete an SER for another EPRI topical report on comercial off-the-shelf software. The staff will continue to work toward an SER on the EPRI document.

The staff will continue its discussion with the ACRS on the level of detail in the proposed new guidance and the balance between review of the digital system design process and the assessment of resulting digital system products during meetings with the ACRS in August and October of this year.

Sincerely, s . /

\

CC M.Tg[or l Executive Director for Operations e

cc: Chairman Jackson Comissioner Rogers Comissioner Dicus SECY l

l l

?

10

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

1. 8to

',  % UNITED STATES

! t NUCLEAR REGULATORY COMMISSION -

{*

l ADVISORY COMMITTEE ON REACTOR SAFEGUARDS WASHINGTON, D. C. 20005 e

October 13, 1995 l

l

~

l l The Honorable Shirley A. Jackson Chairman U.S. Nuclear Regulatory Commission l Washington, D.C. 20555-0001 {

Dear Chairman Jackson:

s

SUBJECT:

NATIONAL ACADENY OF SCIENCES / NATIONAL RESEARCH COUNCIL I STUDY ON " DIGITAL INSTRUNENTATION AND CONTROL SYSTEMS IN NUCLEAR POWER PLANTS, SAFETY AND RELIABILITY ISSUES" -

PHASE 1 I

i During the 425th meeting of the Advisory Committee on Reactor Safeguards, October 5-7, 1995, we reviewed the National Academy of Sciences / National Research Council (NAS/NRC) Phase 1 report on Digital Instrumentation and Control Systems in Nuclear Power .

Plants, Safety and Reliability Issues. The NAS/NRC Committee '

Chairman described the results of the Phase 1 report. We also had the benefit of the documents referenced.

The objective of the Phase 1 study was to define the important safety and reliability issues concerning hardware, software, and human-machine interfaces that arise from the use of digital instrumentation and control technology in nuclear- power plant operations. The report identifies eight key issues six technical and two strategic. It notes that these issues are common to other industries where software is required for dependable operation of systems. The report succinctly presents the issues that the NAS/NRC Committee found to be important.

We agree that the issues identified in the Phase 1 report will be important considerations as digital technology is used more extensively in nuclear power plants. In the past, we have called attention to the effects of environmental stressors. The NAS/NRC Chairman stated that the NAS/NRC Committee considered, but decided not to raise this issue to the level of a " key technical issue." l l We continue to believe this is an important issue that the staff  ;

must address as it develops its regulatory guidance for digital  !

systems. However, this is part of the broader issue of environmental qualification of safety-related equipment and does not need to be a key issue of the Phase 2 study.

11 l

l

o ,* ,

t

. Honorable Shirley A. Jackson 2 i We have concerns regarding a potential conflict between the Phase 1

' 2 completion schedule and the staff's schedule for issuing the Standard Review Plan (SRP) and associated regulatory guides. We l believe it is important that the SRP and other regulatory guidance

! benefit from the insights in the Phase 2 report.

Sincerely, I

i t

i J S. W T. S. Kress

, Chairman

References:

1. Report dated 1995, from the Committee on Application of Digital Instrumentation and Control Systems to Nuclear Power PJant Operations and Safety, Board on Energy and Environmental Systems, Commission on Engineering and Technical Systems,  !

National Research Council,

Subject:

Digital Instrumentation '

and control Systems in Nuclear Power Plants, Safety and Reliability Issues - Phase 1

2. Memorandum dated December 2,1993, from Ivan Salin, Chairman, .

NRC, to NRC Commissioners,

Subject:

Computers in Nuclear Power Plant Operations

3. Letter dated July 14, 1994, from T. S. Kress, Chairman, ACRS, to Ivan Selin, Chairman, NRC,

Subject:

Proposed National Academy of Sciences / National Research Council Study and Workshop on Digital Instrumentation and Control Systems

4. Letter dated August 23, l'94, from Ivan Selin, Chairman, NRC, to T. S. Kress, Chairman, ACRS, regarding ACRS letter of July 14, 1994 on National Acade.my of Sciences / National Research Council Proposnl for a Study and Workshop on the " Application of Digital Instrumentation and control Technology to Nuclear Power Plant Operations and Safety" e

12

i p

i p f ** % ,\"

s UNmED STATES i

~

NUCLEARwasmworon, REGULATORY COMMISSION o.c. men.asm

%g

! October 31, 1995 Dr. Thomas S. Kress, Chairman Advisory Comeittee on Reactor Safeguards

{ U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001 SUBJECTr AIS CONTR0L, TIE NATIONAL SAFETY Als RELIABILITY ACADDIY ISSUES OF SCIENCES' RE

Dear Dr. Kress:

4 i ,

[ I as responding to your letter on this subject, dated October 13,1995, in i which the Advisory Casunittee on Reactor Safeguards i National Acadesty of Sciences / National Research Counc(ACRS) causented on the

il (NAS/ldtc Phase 1 Safety And Reliability Issues." Report, " Digital Instrumentation And Contr ,

of safety-related digital instrumentation and control system 1

i Environmental as a " key technical issue." stressors are defined as an issue in the NAS/Imc s staff is conducting confirmatory research to investigate and chara 4 failure use in nuclear modes andplants.

power degradation mechanisms of digital technologies proposed Furthermore, this research is assessing the  ;

i impact power plants. of smoke on advanced instrumentation and control hardware in nuclear j

i a regulatory guide on the environmental qualification of digita instrumentation and control systems.

our activities on this key issue in an OctoberWe informed the NAS/IEtc Committee about 17, 1995, meeting with them. .,

We share your concerns regarding a potential conflict between the Phase 2

! completion schedule and the staff's schedule for issuing the Standard Review 4

Plan (SRP) and associated regulatory guides.

! for the completion of the study by September 30, Our1996, contract withincludes which NAS/INtc thecalls delivery of the Phase 2 report. The staff has expressed its concern and will

continue to encourage a timely completion of the NAS/IEtC study.

insights expected from the Phase 2 report.it is important that the SR i

Sincerely, e . A,,.

3' N s M. T lor ecutive tractor for Operations cc: Chairman Jackson

' Commissioner Rogers SECY ,

13 ,

',I

/

l 1

l

- i i

l l

I ITEM B.2:

i OFFICE OF NUCLEAR REGULATORY RESEARCH PLAN l FOR UPGRADING THERMAL-HYDRAULIC CODES l l

(DR. CATTON) j l

I

l

, 1 l

l l

t ITEM B 2: OFFICE OF NUCLEAR REGULATORY RESEARCH PLAN FOR l I

UPGRADING THERMAL-HYDRAULIC CODES l l

! I In response to concerns expressed by Chairman Jackson and the ACRS, the Office of Nuclear ,

Regulatory Research (RES) has undertaken a Five-Year Plan, the central goal of which is to l upgrade the NRC thermal-hydraulic codes. Key points regarding this Plan include the following-e There are four goals of this Plan: (1) improve and maintain the reactor safety codes, (2) i improve the two-phase flow formulation and constitutive relations used in the codes, I as supponed by detailed experimental data, (3) maintain experimental facilities and conduct tests for code validation, and (4) develop and maintain in-house capabilities.  !

e The work to improve the thermal-hydraulic codes will progress in two Phases. In Phase I, the capabilities of the current suite of codes will be consolidated into one code.

RES proposes to use the TRAC-P code as the base architecture for this effort. In Phase II, pilot studies of a few promising schemes will be performed that may lead to an improved code. Phases I 6c II will be performed in parallel.

. A supponing experimental program will be conducted at three test facilities available to NRC: OSU (Oregon State University), PUMA (Purdue University), and THECA (University of Maryland). In addition, cooperative programs are expected to be established with JAERI (Japan) and the CEA (France) to obtain test data and analysis l

expenise.

Detailed information concerning this matter was provided to the Committee via a

-(predecisional) memorandum, dated September 6,1996, from James M. Taylor, Executive Director for Operations (EDO), to the Commission.

On September 18-19,1996, the ACRS Thermal Hydraulic Phenomena Subcommittee reviewed RES' scope and approach for the conduct of the Five-Year Plan. This matter was subsequently reviewed by the ACRS during its October 1996 meeting. Formal comments on the results of the Committee's review were provided in an October 21,1996 report to the Commission. In a letter dated November 12, 1996, the EDO responded to the ACRS comments and recommendations, stating that "we are pleased that the ACRS finds this plan holds much promise to revitalize the NRC Thermal-Hydraulics Research Program. We look forward to constructive interactions between the staff and the ACRS Thermal Hydraulic Phenomena Subcommittee in implementing the different elements of the prograw."

In a November 12,1996 memorandum to the ACRS Executive Director, Commissioner Diaz requested a summary of the specific recommendations made by the ACRS over the past two years that penain to NRC thermal-hydraulic codes. The requested information was provided via a November 14,1996 memorandum.

14

i l

[

RES Code Upgrade Plan:

Commission Meeting In addition to the above, a copy of a November 20, 1996 memorandum is attached that transmitted a summary of ACRS comments on the NRC research program pertaining to thermal-hydraulic codes, which were made between 1972 and 1991.

The ACRS plans to continue its review of the RES Five-Year Plan in tandem with the staff's progress on this matter.

Attachments:

• Report dated October 21, 1996, from T.S. Kress, ACRS Chairman, to Shirley Ann Jackson, NRC Chairman,

Subject:

Thermal-Hydraulics Research Plan (pp.16-18) e Letter dated November 12, 1996, from James M. Taylor, Executive Director for Operations, to T. S. Kress, ACRS Chairman,

Subject:

Thermal-Hydraulics Research Plan (pp.19-20) i e Memorandum dated November 12,1996, from Commissioner Nils J. Diaz, to John T.

Larkins, ACRS Executive Director,

Subject:

Thermal Hydraulic Codes, and, Memorandum dated November 14, 1996, from John T. Larkins, ACRS Executive Director, to Commissioner Nils J. Diaz,

Subject:

Previous ACRS Comments on NRC Thermal-Hydraulic Codes and Programs (pp. 21-26)

• Memorandum dated November 20,1996, to ACRS Members, from Paul A. Boehnert, l ACRS Staff,

Subject:

ACRS Review of NRC Thermal- Hydraulic Research Program - l Recommendations Regarding NRC's Thermal-Hydraulic Codes, 1972-1991 (pp. 27-37) l i

l 15 l

l

%*g UNITED sfATas

'/ NUCLEAR RESULATORY COWilSSION

{* g , ADVlaOnY 000mAITTEE ON CEAcTOn saps UAnOs maswassrow.o.c.ssses I October 21, 1996 i

i The Honorable Shirley Ann Jackson Chairman i

U. S. Nuclear Regulatory Commission i Washington, DC 20555-0001 L

Dear Chairman Jackson:

]

SUBJECT:

THElul4L-HYDRAULICS RESEARCH puW i

i During the 435th meeting of the Advisory Committee en Reactor Safeguards, October 9-12. 1996, we reviewed the scope and approach of the Thermal-1 Mydraulics Research plan of the Office of Nuclear Regulatory Research (RES). Our subcommittee on Thermal-Itydraulic phenomena met on September 18-19, 1996, to review this matter. During this review, we had the

benefit of discussions with representatives of the IIIC staff. We also had

the benefit of the documents referenced.

The overall plan developed by RES to consolidate existing camputational tools into a single computer code is timely and should be implemented. We agree with its objectives of standardized progrmuning, better physics,

nexibility (modularity), computational efficiency, a graphical user

interface, and thorough documentation. The RES plan to review the past 22 years of experience with codes like TRAC and RELAp, as well as the successful Code scaling Applicability and Uncertainty evaluation methodology, should help avoid some of the problems of the past. We

' recommend that this review also consider the French code. CATHARE, and its uncertainty evaluation methodology.

The RES staff espects to identify the key physical processes that the new l 4

code must endel. Also, RES plans to determine whether TRAC-p has an I architecture that will allow it to provide flexibility with respect to l tr.sertion of new models or modules and whether it has the capability to '

interface with other codes like CONTAIN and SCDAp. We concur in these plans and emphasize that highest' priority should be elven to the

development of sufficient flexibility to facilitate medlfications in l response to future modeltag challenges. l The INtc Office of Nuclear Reactor Regulation and the Office for Analysis and Evaluation of Operational Data are primary users of thermal-hydraulic  !

codes. They should be a part of this process from the beginning.

Consequently, we recommend that a code users group be instituted early in the development program, l

i 16  ;

i

. -z- '

! - We concur in the RES plan to incorporate the integral effects test

! programs at Gregon State University, the University of Maryland, and i

Purdue University into the overall verification and validation program.

l The cooperative agreement with the French authorities to obtain analytical and experiments 1 data developed at the Grenoble facility should also prove i to be valuable for validating the code. The present RE$ relationship with 1

the aboeas three universities and the French authorities should significantly enrich the proposed thermal-hydraulics research Plan.

These, along with other cooperative agreements, should be pursued,

) independent of the final direction of the RES Plan.

i We commend the staff for the development of this plan which holds much premise to revitalize the NRC Thermal 4tydrau11cs Research Program.

t j Additional comments by ACRS Member Ivan Catton are presented below. I sincerely, i

i j y. A '

T. 5. Kress '

! Chairman 1 i '

l Additional r-- r.tt by Ivan Cattan. Acet ' " r I i I agree with the views of my colleagues expressed above but would like to l emphasize the need for careful planning at the outset of the RES Thermal-Hydraulics Research Plan. The research program that led to the present i suite of thermal-hydraulic codes was initiated in 1974 to address the i large-break loss of coolant accident. The mission was well defined and l the agency met its objectives, i

l At the outset, it was thought that a properly designed thermal-hydraulic i code would be able to model all related problems. Over the years,

however, experience has shown that the codes did not meet this objective; i 1.e., whenever we needed solutions to a new problem that was a little different, the codes were inadequate, because they could not be readily i modified to accommodate the special circumstances demanded by the new j problem.

l The inability of the codes to address numerous new problems emphasizes the l need for a different approach. There is no single code that can model all

the different physical phenomena that occur in a nuclear power plant. A J broader approach is needed where different modeli schemes can be tied

together to successfully address the problem at . Further, a skilled l code user, who is also knowledgeable in the field of thermal-hydraulics,

! is needed to decide what is important and how to implement it in a code.

l A code, no matter how good, will never substitute for a capable thermal-1 hydraulic analyst.

I i

i i

i i

17 i

i 1 4

k.

.. i. Some of these problems will be heavily dependent on the use of what is

' commonly kasm as computational fluid dynamics (CFD), some on the use of the kind of modelta j empirical approach. Thereg found will in betoday's codes, and same problems some that may willrequire even requirethe an use of stand-alone CFD codes. Further, the development of a single code l for all users may not be a realistic goal. A skilled user needs a

different level of computational power than does a less-skilled user.

i Ensuring adequate flexibility in a single code to accommodate the needs of i

both computational power and user skills will require a great deal of 1

thoughtful planning; this plannigi should take place at the beginning of j the development of the RES Therma -Hydraulics Research Plan.

(

Referas eat:

! 1. Demorandum dated September 6,1996, to the Ceemission from James M. 1

! Taylor, NRC Executive Director for Operations,

Subject:

Thermal- I j Hydraulic Five-Year Research Plan, (Predecisional - For Internal 1

! 2. Use Nuc U.S. Only) lear Regulatory Commission, Office of Nuclear Reactor

i lation, Analytical Support Group Technical Analysis Reports, l

! M SASG- M-05; SASG-95 SASG-95-07; 5ASG-96 ! SASG-96-07 (Proprietary Information)

! 3. ACRS report dated June 15, 1989, from David A. Ward, ACRS Chairman, j to Lando W. Zech, Jr., INtc Chairman,

Subject:

NRC Thermal-Hydraulic Research Program I 4. ACRS report dated June 7, 1988, from David A. Ward, ACRS Chairman, i to Lando W. Zech, Jr., NRC Chairman,

Subject:

NRC Research Related j to Heat Transfer and Fluid Transport in Nuclear power Plants t

, w 1

> 1 1

1 i

l 1

i i

]

i l 18 I ~

i

' p k UNITED STATES l g S NUCLEAR REQULATORY CCMMISSION j  ;

• WASHINGTON, o.C. 2 056 4001

l. \,,*****

l November 12, 1996 i

i Dr. Thomas S. Kress, Chairman j Advisory Comittee on Reactor Safeguards t

U. S. Nuclear Regulatory Commission

! Washington, DC 20555 SUBJECTi THERMAL-HYDRAULICS RESEARCH PLAN

Dear Dr. Kress:

The staff met with the ACRS Subcommittee on Themal Hydraulic Phenomena on  !

4 September 18-Ig,1996, and the Full Committee on October 11, 1996, to discuss the I scope and approach of the Thermal-Hydraulics Research Plan. The result of the ACRS review is reported in a letter to me dated October 21,1996, " Thermal- ,

i hydraulics Research Plan." We are pleased that the ACRS finds this plan " holds i

much promise to revitalize the NRC Thermal-Hydraulics Research Program." We look i

forward to constructive interactions between the staff and the ACRS Thermal Hydraulic Phenomena Subcosmittee in implementing the different elements of the program.

The plan to consolidate the existing computational tools into a single code will

! use a modernized TRAC-P architecture with FORTRAN-90 programmiing, better physics, l flexibility (modularity), computational efficiency, graphical user interface, and thorough documentation. The current modularity of TRAC-P will form the base of

! the consolidated code and will pemit the use of different correlations, models j and data for different components. It will also provide an improved user i interface to the code input deck and output display. As part of the code i

requirements specification we will identify the appropriate interfaces to link j the consolidated code with other codes and to facilitate modifications in response to future modeling challenges. Interface with the CONTAIN code is ,

i straightforward and can be accomplished using existing software e.g., Nessage i Passing Interface (MPI). However, interfaces to the SCDAP code might not be

! straightforward and may require developing a cleaner SCDAP package with

appropriate interface to the consolidated code. We will continue our assessment I

of the need to link the SCDAP code to the consolidated code and will discuss the j rationale for and the staff finding regarding this issue with the ACRS before j final implementation. _

An NRC Thermal-Hydraulic Code User Group is being formed to facilitate l communication between the various NRC offices (NRR, AE00, and RES) with

responsibilities in the area of thermal-hydraulic analysis and the RES staff who have the responsibility of developing and assessing the NRC thermal-hydraulic codes. Enhanced communication between representatives of the themal-hydraulic 3 code user cosasunity and those charged with the development of the themal-

hydraulic code and the associated experimental program will enable RES to better j focus its efforts and carry out its responsibility to ensure that the future i analysis needs of the agency and the NRC thermal-hydraulic code user community j will be met.

1

19 i

Dr. Thomas Kress, Chairman 2 Finally, we will continue our efforts to increase collaboration with othp, ,

international organizations that are performing thennal-hydraulic reseattn, j similar to the existing collaborations with CEA of France and JAERI of Japan. i i

Sincerely, ,

1 I

s .'I or E cutive irector for Operations cc: Chairman Jackson Commissioner Rogers Commissioner Dieus Commissioner Diaz Commissioner McGaffigan SECY 20

c _ . . .., . a <

. . _ - . . a_. - .- . . - -.x_. . . .-.-..-- .

l l

l

• W COMPILATION OF ACRS COMMENTS PERTAINING TO NRC THERMAI, HYDRAULICS CODES (1994-96)

(INFORMATION REQUESTED BY COMMISSIONER DIAZ) i l

l l

i

' f NUCLEAR RE2ULATORY COMMISSION g WAaMIN87oN D.C.s0054 one,* November 12, 1996

--an MEMORANDUM TO : JohnImaiae. Executive Director Advisory Comuninee on Reactor Safeguards (ACRS)

~ i l FROM:

ra==ia=i==r Nils J. Diaz .

SUBJECT:

THERMAL HYDRAULIC CODES Since understanding the thennal hydraulic rd=aa==== in nuclear power plants is essential to ensuring reactor safety, I en interested in our current analytical capabilities in thennal hydraulics and in our plans to maintain and enhance NRC's thennal hydraulic codes. In addition to reviewing the NRC's "Ihennal-Hydraulic Five-Year Research Plan," I would t appreciate it ifyou could provide me a sununary of the speciSc ruca===ndarinna rnade by the  : ACRS in the last two years regarding NRC's thennal hydraulic codes. l cc: Chainnan Jackson Cornmianoner Rogers Ca==ienioner Dicus l Ca==ienioner McGafEgan SECY l I l l l  ; i I ! l ! i j 21

UNITO STATES. -- - -- - -- 4 - NUCLEAR RECULATORY COMMISSION c

      'f        'l                    ADVtSORY COMhMTTEE ON REACTOR SAFE!UARDS a m esserom.o.c.ssues                            ;

] November 14, 1996  ; i I MEMORANDUM TO: Commissioner Nils J. D l FROM: John T. Larkins,) [^n ecutive Director , Advisory Committee on Reactor Safeguards

SUBJECT:

PREVIOUS ACRS COMMENTS ON NRC THERMAL-HYDRAULIC j CCTMS AND PROGRAMS I 4 In response to your November 12, 1996 memorandum, attached is a

compilation of specific recommendations and relevant comments made by the Advisory Committee on Reactor Safeguards on the subject of 1

the NRC thermal-hydraulic codes and programs over the past two ! years. The material provided was directly excerpted from the , relevant ACRS letter reports. l

Attachment:

ACRS Comments Pertaining to NRC Thermal-Hydraulic  ; i Codes l l l i cc: Chairman Jackson l Commissioner Rogers ! Commissioner Dieus , commissioner McGaffigan 1 SECY f 4 ) 1 i i a

 +
~

22

COMPILATION OF ACRS COMMENTS PERTAINING TO NRC , THERMAI, HYDRAULIC CODES (1994-1996) 1 1 As noted below, the ACRS provided four letters over the past two years that pertain directly l to the subject matter. In some cases, the comments and recommendations provided do not l appear to apply specifically to code development. However these items [e.g., development of Phenomena Identification and Ranking Table (PIRT), scaling analyses] are considered by the Committee to be vital to ensuring that the codes will be able to adequately model the phenomena in question. October 21.199612tter: " Thermal-Hydranlin R-ech Plan"

                            . The overall plan developed by RES to consolidate existing computational tools into a single computer code is timely and should be implemented. We agree with its objectives of standardized programming, better physics, flexibility (modularity), computational efficiency, a graphical user interface, and thorough documentation.

The RES plan to review the past 22 years of experience with codes like TRAC and RELAP, as well as the successful Code Scaling Applicability and Uncensinty evaluation methodology, should help avoid some of the problems of the past. We recommend that this review also consider the French code, CATHARE, and its uncertainty evaluation methodology. The RES staff expects to identify the key physical processes that the new code must model. Also, RES plans to determine whether TRAC-P has an architecture that will allow it to provide flexibility with respect to insertion of new models or modules and whether it has the capability to interface with other codes like CONTAIN and SCDAP. We concur in these plans and emphasize that highest priority should be given to the development of sufficient flexibility to facilitate modifications in response to future modeling challenges. The NRC Office of Nuclear Reactor Regulation and the Office for Analysis and Evaluation of Operational Data are primary users of thermal-hydraulic codes. They should be a part of l this process from the beginning. Consequently, we recommend that a code users group be  ! instituted early in the development program. l l We concur in the RES plan to incorporate the integral effects test programs at Oregon State , University, the University of Maryland, and Purdue University into the overall verification l and validation program. The cooperative agreement with the French authorities to obtain j analytical and experimental data developed at the Grenoble facility should also prove to be valuable for validating the code. l 23

l

• 1 j March 19.1996 Letter: "NRC Staff Prorran on the -Adeauacy A==* ament of the

! RFT AP5/ MOD 3 Code for Simn1=rion of AP600 Passive Plant Rehavior" We have been asked to comment on the approach and methodology for demonstrating the ! adequacy of the RELAP5/ MOD 3 code to calculate AP600 passive plant behavior in suppon i of the design certification review. We believe that the overall approach and methodology j being employed by RES for this assessment is acceptable. Most of the necessary elements are j in place. A substantial amount of work remains, however, and we believe that the schedule for successful completion cannot be met. l' RES should perform a more robust and complete top <iown system scaling analysis for ROSA, SPES, and OSU. An entire transient should be evaluated to quantify the effects of various

distonions in the three facilities and to demonstrate that the experimental database is sufficient 4

to validate the code. Any additional distortions or anomalies identified should be added to l the list of distortions compiled by RES in late-1994, and that remain to be addressed. The j scaling effon should be integrated with the Phenomena Identification and Ranking Table. The thermal stratification in the Core Makeup Tank (CMT) observed in the tests needs to be

studied. Its effects on core inventory have to be understood because neither RELAP5/ MOD 3 l nor the Westinghouse computer codes can, at present, reliably predict thermal stratification.

1 RELAPS is still undergoing significant and rapid modifications. A calculation has not yet been j performed with a version of the code that contains all the planned changes. Numerous j calculations will need to be performed to mature the code and validate it using data obtained l from various separate effects and integral facilities tests. 1 i Overall, the approach and methodology for qualifying RELAP5/ MOD 3 for AP600 simulation

appear to be adequate. However, two possible "show stoppers" remain

1) simulation of the j CMT thermal stratification and 2) simulation of long-term cooling, which is still an issue.

j Serious consideration should be given to addressing these obstacles. i April 12.1995 letter: "NRC Test and Analysis Prorram in Suonorr of AP600 Advanead T ieht - Water Passive Plant Denien Review" l 1 l Completion of the Phenomena Identification and Ranking Table (PIRT) for the AP600 remains an important task. It was much easier to develop the PIRT for the current operating l plants because a great deal of relevant test data were available. This is not the case for the 1 AP600 and SBWR passive plants. Development of the PIRT should be concurrent with a scaling analysis and review of test results to provide quantitative suppon for the engineering juA-ments that must be made. The RES approach appears to be systematic and well organi'*A. ] We recommend, however, that RES fully document the development of the PIRT. ! We are concerned about the applicability of the present thermal-hydraulic codes (TRAC, j RELAPS) for analysis of plants like the AP600. These codes have to predict types of thermal-

hydraulic behavior for which they have been shown to be weak; i.e., prediction of condensation, thermal stratification, and water level. We recommend that RES consider  :

]1

24

        .._s -- .                 e,.= - ,                                                                             ,_. - . . . , .

i l

  =

developing a contingency plan in the event that the codes cannot adequately predict these key j , phenomena. l t i Although the focus of our meetings with RES was on the development of the PIRT, some j

reference was made to determination of computational uncertainty. The uncenainty i

parameter of choice is peak clad temperature for the laq;e-break LOCA while reactor vessel j

l primary system inventory is the choice for the small-break LOCA. With resources being I ! reduced, we recommend that RES focus its attention on the more safety-significant small-break  ! } a LOCA. - I November 10. 1994 Letter: "NRC Test and Analysis Proernme in Sunoort of AP600 and 3 SBwK Advanc~l Iieht Water R**cror Passive Plant Damien Cenification Reviews" l In the absence of a full-scale test facility, an understanding of the thermal hydraulic behavior j of a passive plant design will depend on the use of computer codes. The NRC staff has , decided to modify RELAP5/ MOD 3 for its confirmatory thermal hydraulic analysis of the ! AP600 and SBWR designs. The important phenomena the code must simulate should be j delineated in the Phenomena Identification and Ranking Table (PIRT), thus allowing one to formulate integral and separate effects experiments that will yield appropriate data for code ' validation. Code validation should be an integrated process involving code development, experimentation, and an understanding of the physics of two-phase flow and heat transfer. 1 i The major objective of the thermal hydraulic code development effort should be to produce ! a code capable of predicting the behavior of a full-scale nuclear power plant with acceptable l uncenainties. For existing nuclear plant designs, we have had the benefit of many integral and

separate effects experiments at a wide variety of scales to help arrive at an estimate of the i

uncenainties in the code predictions. We are now dealing with two passive plant designs j which evidence more complex thermal hydraulic system dynamics, and for which there is a paucity of relevant experimental data. There are several causes for this more complex dynamic

behavior

(1) steam condensation at low pressure, (2) use of gravity. driven coolant injection, l and (3) the existence of many components and complex hydraulic paths that give the system j many degrees of freedom. Understanding this dynamic behavior requires evaluation of scale j distonion effects and dynamic characteristics in the various test facilities. In this regard, two questions should be addressed and resolved: (1) is the evolution of a panicular transient

) influenced by configurational and/or scale distortions, and (2) do configurational and/or scale

distonions in the various test facilities preclude simulation of some important dynamic effects while introducing other dynamic effects that may not be important in a full-scale plant design?

To address these questions, a top down scaling analysis must be performed. i ! Direct counterpart tests in ROSA, OSU, and 5 PES are not possible. This make< it difficult j to extrapolate the observed thermal hydraulic behavior to full scale. A well-planned effort to l integrate experiments with code improvement and assessment is needed to quantify j uncenainties. At present, RELAP5/ MOD 3 predicts strong oscillations both when they are

}                       observed in tests and when they are not. Consequently, the calculated behavior can neither       l
!                      be attributed conclusively to numerical nor physical effects. The mechanisms by which the y                       various observed modes of oscillation are initiated and maintained need to be understood so 4

j 25  ! 1 _.,.....e._ . ., ,,

that their potentialinfluence on the thermal hydraulic behavior of the AP600 can be evaluated. The judicious selection of test conditions for the facilities, together with the conduct of a careful data analysis and scaling, should provide a satisfactory solution. Again, a PIRT has not been completed. The PIRT effon should be brought to a close so that a proper evaluation of PUMA 8 and the GENE test facilities (GIST, GIRAFFE, and PANDA) can be made. To preclude atypicalities in the interactions of the various systems and to help determine an appropriate set of initial and operating conditions for the PUMA system, the scaling of the global dynamic component interactions (among the reactor vessel, drywell, wetwell, PCCS, ICS, and GDCS) should be completed before the facility design is frozen. l 1 l l 2 The PUMA test facility was initially constructed in suppon of the SBWR design cenification review. Since GE Nuclear Energy has, in effect, terminated the SBWR design cenification review, RES has redirected the PUMA program to suppon its current Thermal-Hydraulic Five-Year Research Plan. 26

a *,J.A.e.de.a.4.--a& wa 4 J m..ue,..ms .,m. y m _.e.mJ.a.L--Jm am adh -- A_ A4w..A4As%4.,4+4_46_4.#4-.4aMe. 4 4; _.M-_ 5.ame.mA 5d _u asa..ha p 4 3, 4_ _mm -64#4__e . O I i l r l I 1 l l COMPILATION OF ACRS COMMENTS PERTAINING TO NRC THERMAL-HYDRAULICS CODES (1972-1991) l i

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

     .                                                                                                                      l November 20,1996 i

l MEMORANDUM TO: ACRS Members FROM: P. Boehnen, Senior Staff Engineer /s/ l

SUBJECT:

ACRS REVIEW OF NRC THERMAL HYDRAULIC i RESEARCH PROGRAM - RECOMMENDATIONS REGARDING NRC's THERMAI, HYDRAULIC CODES, 1972-1991 i In response to inquiries from Commincioner Diaz's Office, I undenook a review of the Committee's recommendations relative to NRC/AEC programs for development and use of thermal-hydraulic codes. Subsequently, the Commiccion requested that the ACRS discuss its review of the Office of Nuclear Regulatory Research (RES) Five-Year Thermal-Hydraulic Research Plan during the ACRS Meeting with the Commission, scheduled for Friday, December 6,1996. Commissioner Diaz also made a written request for a summuy of the Committee's recommendations penaining to RES thermal-hydraulic codes over the past two years. A copy of this memorandum and this Office's response are attached. As background to the Committee's December 6 meeting with the Commission, a compilation of the Committee's recommendations and comments on the subject mat:er, as excerpted from its relevant repons dating back to 1972, is attached as separate copy. I have divided these comments into two subsets: comments made in ACRS repons (prior to 1994 - the post-1994 reports are subsumed into the response to Commissioner Diaz's request noted above), and, comments made in the ACRS Annual Repons to Congress on the NRC Safety Research Program. i Attachments: As stated cc: J.Iskins , R.Savio S. Duraiswamy ACRS T/H Phen. Sub. Consultants ACRS Fellows 27 l

1 1 COMPILATION OF ACRS COMMENTS PERTAINING TO !* NRC THERMAleHYDRAULIC CODES l (1972-1991) i r l Below, is a compilation of the comments made by the ACRS relevant to the NRC thermal-l hydraulic code development and assessment program. The excerpted comments are subdivided , into two groups: those made in ACRS letter repons, and those made in the Committee's , annual report to Congress on the NRC-RES safety research program. 1 i ACRS Istter Recons ' ) e Daenmentation of Comouter Codes". dated May 17.1991 and "NRC Comonter Codes ' l and Their Doenmentation". A=r A October 11.1990 4

.                         The guidelines for code documentation supplied to us by RES should be fleshed out

] and cited by reference in all code development work statements. Programs to maintain l existing codes should include a task to bring code documentation into compliance with the proposed guidelines. } j A similar set of guidelines should be developed for use by NRR in its review of i industry codes used for safety evaluations. l A ponion of the regulatory process depends heavily on the results of calculations done - i for the NRC by the national laboratories or other contractors. The codes used for j these calculations range from thermal hydraulic codes like RELAP5 or TRAC to severe  ; accident codes like SCDAP or MELCOR. Many of these codes are poorly documented, thus leaving one unable to determine either their capabilities, or perhaps

                                                                                                                                                )

] j more imponantly, their limitations. i l Many millions of dollars have been spent on the development of the computer codes j used by the NRC, nearly $20 million for RELAP5 alone. The NRC should make

sufficient funding and resources available to ensure that the documentation associated with the development of the agency's codes is adequate.

e. "NRC Thermal-Hydraulie Ra==rch Prorram". Aatad Inne 15.1989 -

i Maintain the present large systers codes, TRAC-PFI/ MOD 1, RELAP5/ MOD 2, TRAC-l BWR, and RAMONA-3B, for an indefinite period. Limit improvements only to those required by: (a) the discovery of important errors, or (b) crucial new information from j the foreign experimental and assessment programs or the B&W testing program. Do ! not undenake major new restructuring or 'zero-based' improvements to the j constitutive equations or numerical algorithms in these codes. We are not convinced j by the arguments given for the need to develop TRAC-PF1/ MOD 2 and 1 RELAPS/ MOD 3. It is our view that the proposed modifications will not substantially ] improve the codes. l 28 1 1 I

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

b l a

1 2

l i

ACRS Comments on T/H Codes  !

! l ! Instead, consideration should be given to the development of a new type of systems ! code that will be more useful for analysis of extended plant transients involving l interactions of plant systems. The Committee also made this recommendation in its i June 7,1988 report. TRAC and RELAP were originally designed to analyze the LB-j LOCA. A rapid and severe reactor transient, in great detail. There is a need for a more empirical and efficient analytical tool. We envision a code that would be able,

for example, to make a rapid and sufficiently accurate analysis of the power oscillations l observed last year at the LaSalle County Station, Unit 2 plant. Such a code would be i more akin to advanced simulator codes than to TRAC and RELAP. The BWR code (HIPA) now in use at Brookhaven National Laboratory is an example of the type of code we are suggesting.

l, e f "NRC Research Related to Heat Tran=fer and Fluid Tran=non in Nuclaar Power ! Plants". Tune 7.1988 l The designs for so-called evolutionary LWRs and especially the " passive" LWR being developed by the Electric Power Research Institute and DOE, will require research by i-the NRC to confirm censin favorable characteristics being claimed. The DOE Advanced Reactor Severe Accident Program is not sufficient for this purpose. The l NRC should use existing codes to review these designs so there is sufficient lead time j to conduct more experimental or code development work, if necessary. i There is some uncenainty about applicability of the RELAP-5 code to BWRs and to j LBLOCAs. This should be resolved. i j Full documentation should be completed for the NRC codes that are maintained for

active use. This should include not only user manuals but developmental assessment repons and models and correlations documents. Ideally, these would be published as NRC documents in the NUREG series to ensure widespread availability.

                              *                "ACRS Comments on the CaAa bling Aoplic kility and Uncerra'nty Methodolory for Determination of Uncensinty Associ*aA With the Use of Ramlietic ECCS Evaluation Models". dated Scotember 16.1987 Before Code Scaling, Applicability and Uncensinty (CSAU) evaluation methodology can be applied to a given code, complete documentation (e.g., code manual, model and correlation quality assurance daenment, and assessment reports) is necessary. In the past, such thorough documentation has not always been available for licensing codes.

We recommend that steps be taken to ensure that future development of codes for licensing activities be performed in a manner that ensures completion and availability I of needed documentation before the code is released. 29

i l j . 4 1 ACRS Comments on '

T/H Codes  ;

i , { The codes used to analyze thermal-hydraulic behavior are very large and complex. l Validity of calculated results is dependent on the competence of the code user and the way in which the code is used. For CSAU to be effective, the code developers, assessors and users must use the code consistently. We recommend the NRC Staff take the necessary steps to ensure that proper controls are established. , I e "AbRS Commente on the Safety Fa=-arch Proeram and BnAget for Fieral Year 1987". , ! dated lune 11.1985 ) I We believe that a major capability for thermal-hydraulic research should be maintained. I It is especially imponant to ensure the continued availability of experienced spacialiers.  ; l We suppon the concept for a new or modified integral test facility to be co-located ) l with a technical staff capable of analysis and code development. The budget proposal  ; of $4.5 million for this effort is appropriate if firm phnning for such a facility is begun immediately and if it leads to a decision to proceed. l The advanced thermal-hydraulic codes, such a TRAC and RELAP, have contributed j greatly to an understanding of reactor operation and risk. To ensure their continued j availability and utility, they must be effectively maintained. We endorse the NRC ) program to maintain these codes and to continue their assessment and improvement

by a program of cooperation with foreign research institutions. We caution, however, j that a strong domestic capability for maintaining and asing these codes must continue.

The nuclear plant databank (NPDB) and nuclear plant analyzer (NPA) programs are intended to extend the utility of the major thermal-hydraulic codes. The NPDB program, while important to users in both NRR and RES, is not research. We ! recommend that it be funded from the NRR budget. The NPA program has the ! potential for providing an extremely flexible and useful tool for nuclear systems ! analysts. We believe this development program is important and should be continued. ! However, one part of the NPA program involves development of parallel- processing l using smaller, special-purpose computers as a substitute for mainframe computers used j for the major thermal-hydraulic codes. While this work is interesting and may be of ! general use, it is not directly related to the NRC's thermal-hydraulic research program. i We recommend that it be funded elsewhere. ! e "Comm,nts on the NRC Safety F-ech Proeram Bndent for Fi= cal Years 1985 and

1986". dated inne 16.1983

The main focus of effort is on the development of a Plant Analyzer and a Plant Data  ;

Bank. The Plant Analyzer effort is divided between the Brookhaven National i Laboratory (BNL), and a cooperative program at the Los Alamos National Laboratory

(LANL) and the Idaho National Engineering Laboratory (INEL). The former is to i develop a parallel processor approach for a BWR Plant Analyzer and the latter is to I j 30 i i

l-

= i ACRS Comments on T/H Codes  !

I develop a pressuriud water reactor (PWR) Plant Analyur using fast-running versions of the TRAC and RELAP-5 codes. We believe that the goal of the Plant Analyur , l research is not sufficiently well focused. NRR, the intended customer for a Plant ' Analyur, has recommended a low priority for this program. While we believe that

the Plant Analyur could become an important and useful tool for NRR, the ,

! development will be expensive. This program should be deferred until NRR becomes ' j more enthusiastic and involved in defining the research goals. Further, it is not clear that the two programs are coordinated well enough to lead to i equally competent systems for both BWRs and PWRs. It appears that the parallel processor concept, if effective, could apply to both BWRs and PWRs. Broader aspects , ! of the BWR analyur are apparently not being developed in the current BNL effort. l i A vital part of the Plant Analyur and Plant Data Bank Programs is the review of  ;

actual plant data on normal and transient operating parameters to validate codes. RES l

has noted that it has been extremely difficult to obtain these data from U.S. licensees. ' We recommend that NRC explore the possibility of a cooperative program, perhaps

with EPRI, to obtain needed information.

e "Recort on Water Reactor Safety Ra=aerch". Aared November 20.1974 The RSR advanced code development program hu been extended from one to three contractors. The Committee concurs in obtaining a broader-based participation for technical input into the AEC's advanced code, Reactor System Transient (RST). The RST code is to have the capability of utilizing inputs based upon realistic estimates, as well as conservative estimates, so that safety margins can be more quantitatively determined. Special effort is needed to develop analytical predictive methods for realistic evaluations. Major benefits from a successful RST code would be in resolving questions relating to the significance of modeling techniques and parameters, and in allowing the use of scaled experiments in place of full-scale tests. e "Recort on Water Reactor Safety Ra=== ch". AataA February 10.1972 Imnroved ECCS comouter codes. The mathematical modeling of the reactor s, stem should include both improvement in current capabilities and the development of new l and improved integrated computer codes which better represent the actual phenomena I involved. ACRS Reoorts to Confess on NRC R**eror Safety Raeaarch From 1977 through 1985, the Committee provided comprehensive stand-alone reports to the Congress, pursuant to an amendment of the Atomic Energy Act of 1954 by Section 5 of 31

ACRS Comments on T/H Codes Public Law 95-209.2 Below, are relevant excerpts from these repons pennining to the NRC programs on thermal-hydre dic codes. e " Review and Evaluation of the Nuclame Reen1= tory Commie ion Safety Rae nech Proeram for Fi= cal Years 1986 and 1987 - A Report to the Coneress of the Unit A Statac of America". dat.A February 1985 (NUREG-1105) The need for a future integral test program and companion code development capability should be evaluated as pan of the NRC long-range planning. Such an evaluation should consider the merits of low-pressure integral facilities (such as that being operated at the University of Maryland for B&W plants); or, NRC might become one user of a general-purpose facility sponsored by DOE and the nuclear power industry. Funding for the Separate-Effects Testing and Model Development program should be maintained approximately at the present level. Funding for the code assessment work should be substantially reduced. A critique should be performed of the base assumptions and other relevant aspects of the existing methodology to evaluate possibly significant deficiencies in the codes. A review should be made also of additional thermal-hydraulic phenomena, such as water hammer, that may warrant inclusion in the code models. The program related to international agreements should be continued to obtain foreign experimental data in exchange for overseas use of the TRAC and RELAP codes. Sufficient funding should be committed in maintaining these codes to ensure their continuing utility. Funding for the Plant Analyzer and Plant Data Bank can be cunailed, if necessary, to provide funds for other essential safety research. e " Review and Evaluation of the Nucl =r Reen1* tory Commi== ion Safety F-rch Proeram for Fiscal Year 1985 - A Report to the Coneress of the UnitaA Stata= of

                                                                               ~

America" dated February 1984 (NUR5G-1039) We support the NRC effon towards increasing international cooperation in the area of code assessment. Funding for the code assessment work should be maintained approximately at its present level. 2 Since 1985, the Committee has met the intent of Public Law 95-209 by the annual issuance of a repon consisting of all relevant ACKS repons written in a given year that penained to the NRC Safety Research Program. 32

4 j .. 1

ACRS Comments on T/H Codes The NRC should assure that its major codes are maintained so that they do not lose their value in the international exchange of information.

j In regard to plans to choose between TRAC and RELAP for future development, the

!                        NRC should assure that support is available for licensing audit calculations regardless of t.h.e choice made.

The NRC should focus effort on making code results less dependent on the skill and

[ judgment of the user. i l We believe that more emphasis should be placed on development of the data bank. We

also believe that the goal of the development of an accurate, fast, and user- friendly

code would be best served by a near-term ( 2 years) focus on this effort.

l

^

We believe that the thermal-hydraulic code development is now a mature enough technology so that the competition for development of a fast-running code, now

                        -involving three national laboratories (BNL, INEL, LANL), should be ended as soon as possible, and that resources should be concentrated on one effort.

] 2 i e " Review and Evaluation of the Nucl ar Rerulatory Comminion Safety Ra<aarch

                                                                                 ~

I Prorram for Fiscal Years 1984 and 1985 - A Reoon to the Coneress of the United i States of America". dated February 1983 (NUREG-0963). l Special attention should be given to developing an effective analytical suppon effon for the B&W test program. A comprehensive program of separate-effects tests and 2 associated analyses is a necessary adjunct to this test program. j The focus of the code assessment effort should be shifted more towards assessment of j plant transients with a corresponding reduction in the large-break LOCA assessment { program.

;                         NRC should take an active and aggressive role in planning and coordinating the Plant j                          Analyzer Program to assure its usefulness to NRC.

e " Review and Evaluation of the Nuctaar Reentatory Commi== ton Safety Baeaarch i' Prorram for Fiscal Year 1983 - A Recon to the Conrress of the United States of

                                                                                    ~

America". dated February 1982 (NUR$G-0864) l NRC should continue to suppon the Code Improvement and Maintenance Program. Also, it should consider coordinating the LWR and Advanced Reactors code  : development programs. The development and assessment of TRAC and RELAP-5 codes should be continued. ) I 33 i i

ACRS Comments on T/H Codes The NRC has initiated steps to develop fast running systems codes. This effon should be broadened and accelerated. NRC should undenake a more limited and carefully monitored code assessment program to provide guidance before embarking on a massive assessment program. e "Riview and Evaluntion of the NuA**r Reenistory Commi== ion Safety Ba=aarch Prorram for Fiscal Year 1982 - A Reoort to the Coneress of the United Scarae of

                                                                   ~                                         ;

America". dated February 1981 (NUREG0771) The NRC proposes to complete in FY 1982 best estimate codes for PWR and BWR systems. These codes will be adaptations of TRAC. The NRC appears to believe that TRAC has been adequately developed and ==-==aA so that these effons will be meaningful. We believe that TRAC has been inadequately developed and ==ascad in spite of the large effon that has been expended. We strongly suppon the RELAP-5 effort. The version of TRAC which has just been developed by NRC for BWRs is totally l inadequate. Among other deficiencies, we especially note that the present version of this code has no treatment of the upper plenum. We also note that the NRC has not yet sponsored a BWR version of RELAP-5. We believe that this omission is inexcusable and recommend the prompt development of such a code. It has been noted that the NRC code assessment is inadequate, panicularly in the case  ! of TRAC. In spite of a large effort, TRAC is not yet a code that is adequately i developed and ==-==A In some respects RELAP-5 has indicated greater promise with  ; a smaller effort than TRAC hu received. We nevenheless recommend that both TRAC and RELAP-5 be continued since the NRC has already made a large investment in TRAC and this code may eventually give a 2- or 3-dimensional calculational l capability. We recommend that the NRC drastically reduce, or even terminare, the effort going l into the development of component codes. Funher, the tendency toward a multiplicity of such codes should not continue. These effons should be replaced with attaining increased code capabilities for small-break LOCAs and other reactor transients with RELAP-5 and TRAC. The program of code assessment is so arranged that it is difficult to incorporate changes and improvements into the codes as research and development proceeds. Such a bureaucratic approach is clearly undesirable. We also believe that the code assessment matrix is much too heavily skewed toward concerns with large-break LOCAs. The code assessment matrix should be immediately revised to reflect the need for emphasis on small-break LOCAs and operational transients. Before embarking on a massive 34

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

ACRS Comments on l T/H Codes  ; i l assessment program, the NRC should undertake a limited and carefully monitored code , assessment program on a modest scale.  ; The effective development of RELAP-5 shows the value of an integral facility, even one l of modest scale like 4=i-=le. The close association with observation in such facilities I contributes to better understanding of what is needed to make a code an effective tool l for~the description of reactor transients. ) As a general comment on the code development program, we approve the development  ! of the so-called " advanced codes" like TRAC, but we consider it unfortunate that the  ! NRC does not yet have a full and adequate audit calculation capability. This essential l need of the NRC has not been met even though great attention has been given to these advanced codes. We question whether the WRAP code is being pursued effectively so as to fulfill the NRC audit needs in a timely manner. e " Review and Evaln= tion of the NnA-=r Reen1= tory Co==i== inn Safety b=aarch Prorram for Fiscal Year 1981 - A Reoort to the Conness of the United States of America". Antad February 1980 (NUREG-0657) , The objective of the code development program is the development, for predictive , purposes, of computer codes for the quantitative analyses of reactor transients and accidents. While the objective of the program has not yet been achieved, the program has made some progress.  ; 1 The principal computer code of choice, TRAC, suffers so far from incomplete knowledge of some of the necessary physical parameters. It should be pointed out that a fairly complete description of the possible physical situations in a reactor transient is required for the microscopic description used in TRAC. This microscopic description leads to long running times and thereby limits a rapid survey of the many possible transients. While an effort is under way to develop a fast running version of TRAC, the ACRS believes the RELAP 5 computer code, which is already somewhat faster than TRAC, also should be developed to provide a second fast running code. The ACRS believes that the program is progressing reasonably well in view of the difficulty of the task. The ACRS supports the code development program. The ACRS recommends the continued development of RELAP-5 as another general code of potential value. The ACRS recommends also that a strong program be initiated for the development of methodology and techniques that would facilitate the , implementation of more sophisticated reactor simulators, not necessarily limited to real- l time analysis. This would enable a more detailed understanding of the course of events l in complex transients that include multiple failures and operator intervention. The { l 35 l l

ACRS Comments on T/H Codes ' ACRS believes that the proposed budget is adequate to include these developments without additional funds. o "Commente on the NRC Safety Ra=> arch Prorram Badent". Tuly 1979 (NUREG-0603) .I p The budget request, for the code development program is $15.2 million which has been reduced to $13.2 million by the Budget Review Group (BRG). The entire reduction was in TRAC Au*= ment and Applications which the BRG indicated contained some duplication in TRAC application funds in RES and NRR. The ACRS cannot comment on this point except to say that TRAC Assessment and Applications are imponant activities, e "1978 Review and Evaln=rion of the Nuclear Reentatory Comminion Safety R-rch Procram". December 1978 (NUREG-0496) The ACRS is aware of the criticisms implying that code developers continue to tune their models to fit experimental results, and thus a code may fit particular experiments but still not be suitable for final application to a nuclear power reactor. ACRS members and consultants have reviewed these matters. The present two-step system of code development and independent code ascessment, with the requirements of

• prediction of test results, has been developed in an attempt to address this concern. It should be noted that an independent code assessment has not yet been completed, and thus special attention needs to be focused on this matter in future reviews. Each major test, where appropriate, should be preceded by blind predictions from best estimate codes; blind predictions should also be made with the advanced code, TRAC, for the 3D-2D international test series.

The NRC has planned an approach that would characterize best estimate code un-cenainties and would expose any inherent code errors or bias. It is envisioned that a peak clad temperature probability surface could be obtained from usignment of probabilities for various combinations of plant conditions. This is an important area for continued ACRS involvement. e " Review and Evaluation of the Nue1**r Reen1= tory Comm*= ion Safety F-arch Prorram - A Reoort to the Congress of the United %eae of America". Dae*=har 1977 (NUREG-0392) During the period of review by the ACRS, the NRR Staff recognized that their needs for Best Estimate and Evaluation Model versions of the computer codes were not being met by their Technical Assistance contracts. A formal recpest was made to RES to reorient some of the LOCA analysis computer programs. Tse ACRS hu not reviewed the response to this request, and will include this item in its continuing reviews. 36

b 4 1 . l

, ACRS Comments on l T/H Codes  ;

I The ACRS identified the following matters requiring additional attention: e Inclusion of representatives from the reactor vendors as consultants to the  ! review groups for advanced code development. j e- More thorough planning for code verification for justifying scaling of  ; experiments.  ! l t l 4

                                                                                                                          )

8 l l l l 1 37 i i

1 1 l l l l l ITEM B.3: ! RISK-INFORMED, PERFORMANCE-BASED REGULATION l AND RELATED MATTERS (DR. APOSTOLAKIS) i i i l J l

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

)

!.                                                                                                                              i i                                                                                                                                i

! i j ITEM B.3: RISK-INFORMED. PERFORMANCE-BASED REGULATION AND  ! ! RFT ATED MA'ITERS l During the May 24,1996 meeting with the Commissioners, the ACRS discussed the use of  ; i Individual Plant Examinations (IPEs) in the regulatory process, the Probabilistic Risk i l Assessment (PRA) framework document, pilot applications, and next steps to expand the use l { of PRA in the regulatory decision-making process. The ACRS discussed the status of its j reviews and summarized the points highlighted in the ACRS reports to the Commission dated i ^ March 8 and April 23,1996.  ! l 1' In the May 15, 1996, Staff Requirements Memorandum (SRM), resulting from the April 4, l 1996, staff briefing to the Commission on the PRA Implementation Plan, the Comnussion , uked that, i I

                         "The staff should prepare a policy paper, with recommendations, addressing the                         l resolution of the four emerging policy issues identified in its March 26,1996,                         i memorandum regarding: 1) the role of pedormance-based regulation in the PRA Implementation Plan; 2) plant-specific application of safety goals; 3) risk neutral                    ,

versus increases in risk; and 4) changes to risk-informed inservice testing (IST) i and inservice inspection (ISI) requirements,.."

                                                                                                                                ]

The ACRS heard presentations by and held discussions with the staff regarding these matters ) on August 8-10, 1996. During this meeting, the Committee also discussed the pilot i applications for risk-informed, performance-based regulation. The ACRS Subcommittee on PRA also met with the staff on July 18 and August 7,1996, regarding these matters. The ACRS provided a repon to the Commission on August 15, 1996. In that repon, the Committee made several comments and recommendations, including the following: e The Commission's Safety Goals and subsidiary objectives can and should be used to derive guidelines for plant-specitic applications. It is, however, impractical to rely exclusively on Quantitative Health Objectives (QHOs) for routine use on an individual plant basis. Criteria based on core damage frequency (CDF) and large, early release frequency (LERF) focus more sharply on safety issues and can provide assurance that QHOs are met. The ACRS recommended that the staff develop guidance for handling situations in which high values of CDF occur for shon periods of time. e Accounting for uncertainties is a difficult issue. There are models and formal methods to account explicitly for a large number of uncertainties. However, other uncertainties , are unquantifiable at this time. l l For the characteristics or phenomena which are modeled, the ACRS recommended that i the uncertainties be explicitly quantified and propagated through the PRA. The ACRS i also recommended that the use of mean values of distributions be used for comparison with goals and criteria. i 38 J

        - ,          _      .. c ,-    , ,              ,                , - - -                   - .-      ,..      - - -

• The ACRS agrees with the staff and industry that some increases in risk should be permitted in some situations. The ACRS believes that it is the overall impact on plant risk that is important, and related changes should be handled as a package. The ACRS also recommended that such changes be consistent with the current philosophy of risk 1

management; i.e., that the " bottom-line" numbers should not be the only input to the decisionmaking process, and other concepts such as defense-in-depth must be maintained, e The ACRS agrees with the staff that, where practical, performance-based strategies should be included in the implementation and monitoring steps of the risk-informed decisionmaking process. The pilot programs may provide a more concrete definition

and development of performance-based strategies.

4 In general, the ACRS supports the staff's recommended options regarding these matters. The Executive Director for Operations responded to the ACRS in a letter dated September 6,1996, stating that the insights and recommendations of the ACRS will be reflected in the staff response to the Commission. In the June 11, 1996, Staff Requirements Memorandum, resulting from the ACRS meeting with the Commission on May 24,1996, the Commission requested the ACRS to, ll

                  " continue to interact with the staff in the development of regulatory guidance and a standard review plan for the performance and review of PRAs. In particular, the ACRS should advise the staff regarding appropriate methods for judging the acceptability or unacceptability of assumptions and models to be used when performing PRAs."

The Committee heard presentations by and held discussions with the staff regarding these 4 matters during the ACRS meeting on December 5-7,1996. The ACRS Subcommittee on PRA l also discussed these matters with the staff on October 31, November 1,21, and 22,1996.  ! During these meetings, the ACRS reviewed the following documents: j e Draft Regulatory Guide DG-1061, "An Approach for Using Probabilistic Risk  ; Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Current i Licensing Basis," dated November 8,1996 e Draft SRP Chapter 19, Revision 7, "Use of Probabilistic Risk Assessment in Nant-Specific, Risk-Informed Decisionmaking: General Guidance," dated November 8,1996 e Draft Regulatory Guide DG-1062, "An Approach for Plant-Specific, Risk-Informed Decisionmaking: Inservice Testing" e Draft SRP section 3.9.7, Revision 1, " Standard Review Plan for the Review of Risk- i Informed Inservice Testing Applications," dated November 15,1996 2 1 39

• Draft Regulatory Guide DG-1065, Revision 5, "An Approach for Plant-Specific, Risk-Informed Decisionmaking: Technical Specifications," dated November 11,1996
• Draft SRP Section 16.X, Revision 9, " Risk-Informed Decisionmaking: Technical Specifications," dated November 15,1996
• Draft Regulatory Guide DG-1064, Revision 3, "An Approach for Plant-Specific, Risk-Informed Decisionmaking: Graded Quality Assurance," dated November 15,1996 The ACRS plans to review Draft NUREG-1602, " Standards for Probabilistic Risk Analyses (PRAs) to Support Risk-Informed Decisionmaking," dated October 11,1996, during a future meeting. The ACRS plans to prepare a report to the Commission regarding the overall approach and direction of the SRP sections and associated regulatory guides during a future meeting. The Committee also plans to continue its discussions with the staff regarding technical issues, within the individual guidance documents, during the public comment period.

Attachments:

• Staff Requirements Memorandum dated May 15,1996, from John C. Hoyle, Office of SECY, to John T. Larkins, ACRS Executive Director (pp.41-42)
• Staff Requirements Memorandum dated June 11,1996, from John C. Hoyle, Office of SECY, to John T. Larkins, ACRS Executive Director (pp. 43-44)
• Report dated August 15, 1996, from T. S. Kress, ACRS Chairman, to Shirley Ann Jackson, NRC Chairman,

Subject:

" Risk-Informed, Performance-Based Regulation and Related Matters" (pp. 45-48)

• Letter dated September 6,1996, from James M. Taylor, Executive Director for Operations, to T. S. Kress, ACRS Chairman,

Subject:

" Risk-Informed, Performance-Based Regulation and Related Matters" (p. 49) l 40 l

i l j. a $ . IN RESPONSE, PLEASE 1 REFER TO: M960404A ' May 15, 1996 , .e - ' I MENORANDUM To: James M. Taylor i, Executive Director for operations ' FRON: John C. Hoyle, Secretary /s/

SUBJECT:

STAFF REQUIREMENTS - BRIEFING ON PRA  ; 1 IMPLEMENTATION PLAN, 10:00 A.M., THURSDAY, i APRIL 4, 1996, CONNISSIONERS' CONFERENCE ROON, ONE WHITE FLINT NORTH, ROCKVILLE, 1 MARYLAND (OPEN TO PUBLIC ATTENDANCE)  ! The Commission was briefed by the NRC staff on the PRA l implementation plan. l The Commission requested that the staff provide the following supplemental information.  :

1. The staff should prepare a policy paper, with recommendations, addressing the resolution of the four emerging policy issues identified in its March 26, 1996 memorandum regarding: 1) the role of performance-based ,

reggiation in the PRA Implementation Plan; 2) plant-specific application of safety goals; 3) risk neutral vs increases in risk; and 4) changes to risk-informed inservice testing (IST) and inservice inspection (ISI) requirements, for Commission decision prior to the staff's issuance o.f any final safety evaluation, position, or guidance. However, the Commission does not wish to inhibit any dialog with the public or industry on the resolution of these issues. The

bases as well as pros and cons of alternatives should be included.

2. The staff should provide an update on the implementation and use of subsidiary safety goal objectives including any plans for their use in Standard Review Plans (SRPs) and Regulatory Guides (RGs).

3. The staff should clarify how it intends to address uncertainty in the implementation of risk-informed and performance-based regulation.

(EDO) (SECY Suspense: 9/20/96) The IPE reviews and pilot applications should be reviewed for any lessons that may be applicable to the development and use of SRPs and RGs. 41

1 s. , l , The use of expert judgment, such as is being applied in the i high-level waste area, should be considered as a source of useful guidance to-the expert panels in the maintenance rule. The staff is requested to keep the commission informed of significant policy issues relevant to risk-informed initiatives. I cc: Chairman Jackson commissioner Rogers l Commissioner Dicus OGC OCA OIG Office Directors, Regions, ACRS, ACNW, ASLBP (via E-Mail) PDR - Advance DCS - P1-24 l 42

l

        /          'o,,

UNITED STATES

     !"               g              NUCLEAR REGULATORY COMMISSION
     ,,               j                        w AswiNoTON, D.C. 20555
         %,       j[                                                        IN RESPONSE, PLEASE
            *****                                June 11, 1996              REFER TO: M960524 OFFCE OF THE SECRETARY MEMORANDUM TO:               John T. Larkins, Executive Director Advi ry Co. ittee on Reactor Safeguards FROM:                        Joh      . oyl , Secretary

SUBJECT:

STAFF REQUIREMENTS - MEETING WITH ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS), FRIDAY, MAY 24, 1996, COMMISSIONERS' CONFERENCE ROOM, ONE WHITE FLINT N, ORTH, ROCKVILLE, MARYLAND (OPEN TO PUBLIC ATTENDANCE) The Commission was briefed by the ACRS on the following topics:

1. Use of IPEs in the regulatory process, PRA framework document, pilot applications, and the next steps to expand the use of PRA in the regulatory decision-making process

2. Fire protection issues, including fire PRA models and PRA-based scoping analysis of degraded fire barriers

3. Proposed final revisions to 10 CFR Parts 50 and 100

4. Status of ACRS review of Regulatory Guidance documents related to digital instrumentation and control systems

5. Status of ACRS review of standard plant designs:

e ABWR and System 80+ design certification rules e AP600 design e Test and analysis programs associated with the AP600 and SBWR designs

6. Conformance of operating plants with NRC safety goals The Commission requested ACRS input on GE's recent submittal containing safety significant design changes prior to the Commission briefing on the design certification rulemakings scheduled for late August.

(ACRS) (SECY Suspense: 8/22/96) The ACRS should continue their review efforts, as noted in the meeting, in the digital instrumentation and controls area. 43

7 The ACRS should continue to interact with the staff in the development of regulatory guidance and a standard review plan for the performance and review of PRAs. In particular, the ACRS should advise the staff regarding appropriate methods for judging the acceptability and unacceptability of assumptions and models to be used when performing PRAs. The AciS should provide recommendations on how the Commission's safety goals and safety goal policy should be revised to make them acceptable for use on a plant-specific basis. cc: Chairman Jackson Commissioner Rogers Commissioner Dicus EDO OGC OCA OIG Office Directors, Regions, ACRS, ACNW, ASLBP (via E-Mail) PDR - Advance DCS - P1-24 i k J i 1 f

44 i

l 1

  .       - . . . . . . - - . - -                   . - -.. - - - .                 ..      . - - _ _ . ~           -.-..- _ _
          >                                                                                                                         r
    ~
                     /              o                                  UNITED STATES NUCLEAR REGULATORY COMMISSION

j. '

                    /                 t
    ,              {0               p
                                      ,I          ADVISORY COMMITTEE ON REACTOR SAFEGUARDS WASHINGTON, D. C 20666 l
                                                                                                                                    +

Y g l August 1b, 1996 l i I i  ! The Honorable Shirley Ann Jackson Chairman U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001 l

SUBJECT:

RISK-INFORMED, PERFORMANCE-BASED REGULATION AND RELATED MATTERS During the 433rd meeting of the Advisory Committee on Reactor l Safeguards, August 8-10, 1996, we discussed the issues identified  : in the Staff Requirements Memorandum dated May 15, 1996. We also  ! discussed the pilot applications for risk-informed, performance- J based regulation. Our Subcommittee on- Probabilistic Risk l Assessment (PRA) met with representatives of the NRC staff and the ~! nuclear industry on July 18 and August 7, 1996. We also had the- i benefit of the documents referenced. j i The staff presentations dealt only with the development of i

                   .--       guidelines from the Commission's safety goals to be used as an                                       .l element of the evaluation of licensee-initiated changes to licensing commitments. All of our comments address the application                                      j of risk-informed regulation in that context. At a later time, we will discuss the larger question of the application of the safety goals on a plant-specific basis.

CONCLUSIONS Insue 1: Should the Consission's safety goals and subsidiary objectives be referenced or used to derive guidelines for plant-specitic applications and, if so, how? We believe the safety goals and subsidiary objectives can and should be used to derive guidelines for plant-specific applications. It is, however, impractical to rely exclusively on the Quantitative Health Objectives (QHos) for routine use on an i individual plant basis. Criteria based on core damage frequency , (CDF) and large, early release frequency (LERF) focus more sharply 1 on safety issues and can provide assurance that the QHos are met. They should be used in developing detailed guidelines. * \ 45

4 i i 2 Issue 2: How are uncertainties to be accounted for? This is a difficult issue. There are models and formal methcds to 1 account explicitly for a large number of uncertainties. However, other uncertainties are unquantifiable. The staff proposes to explore a number of options, such as establishing margins in the acceptance guidelines, placing more importance on defense-in-depth, l and others, to deal with such uncertainties. Such approaches seem appropriate, although much work remains to be done. Isnue 3: Should requested changes to the current licensing basis I be risk-neutral or should increases be permitted? We agree with the staff and industry that increases in risk should i be permitted in some situations. Acceptance guidelines expressed in terms of the proposed change in risk and the current risk l estimates should have three regions: a region in which some j increase in risk is acceptable, one in which it is unacceptable, and one in which further analysis and evaluation would be required. Issue 4: How should performance-based regulation be implemented in the context of risk-informed regulation? i We agree with the staff that, where practical, performance-based strategies should be included in the implementation and monitoring step of the risk-informed decision-making process. The pilot l programs may provide an opportunity for a more concrete definition and development of performance-based strategies. DISCUSSION Issue 1 i Even though a CDF could be derived from the QHos that could be greater than 10-3 per reactor-year, the current subsidiary goal of 10-' per reactor-year should be maintained and should be stated as a fundamental safety goal, along with the QHO. Accident sequences that have a high probability of leading to severe consequences could be controlled by the QHos, but a more workable measure would be a subsidiary goal on the LERF. The definition of the latter i needs to be improved. Whether the LERF should be a fixed value or derived from the QHos, which would allow the LERF goal to include site-specific characteristics, needs to be investigat " We recommend that the staff develop guidance for handling situations in which high values of the CDF occur for short periods of time (for example, 10-2 per reactor-year for a day). j i i 46 l 1 l

   -                                          3 Issue 2 In accounting for uncertainties, it is important to distinguish between those plant characteristics or phenomena that are modeled in the PRA and those that are not modeled (e.g., the actual layout of components and organizational factors) .          For those that are modeled', parameter and model uncertainties should be explicitly quantified and propagated through the PRA.                      The resulting distributions should be an input to the decision-making process along with other qualitative input.

Mean values of distributions should, in general, be used for comparison with goals or criteria, although the sensitivity of the mean value to the high tail of a distribution should not be overlooked. For very broad distributions, such as those that typically result when significant model uncartainty is present, reliance on the mean values may not be appropriate and a more detailed investigation of the reasons for this large uncertainty should be undertaken. This could possibly lead to decisions to conduct additional research or to take other measures. Accounting for uncertainty in the case of plant characteristics or phenomena that are not currently modeled at all is much more difficult. The staff proposes to explore a number of options, such as establishing margins in the acceptance guidelines, placing more importance on defense-in-depth, and others. We agree and encourage the staff to actively pursue the resolution of this issue. Issue 3 The concept of a "three-region" approach is consistent with the Electric Power Research Institute's PSA Applications Guide (PSAAG), although the boundaries of the regions used in the PSAAG are not necessarily the ones that the staff will adopt. The staff has raised the issue of how " packaged" requests are to be handled. Packaging is the process by which risk trade-offs can be accomplished. It is a significant benefit of risk-informed regulation. We believe that it is the overall impact on plant risk that is important, and related changes should be handled as a package. Such changes should be consistent with the current philosophy of risk management; i.e. , that the "botton-line" numbers should not be the only input to the decision-making process, and other concepts such as defense-in-depth must be maintained. 47 1 _

k l .

                                                                                                                   ]

4 4 i a

,              We will continue to monitor the progress of the staff on these

! issues. I

Sincerely,

                                                            ,f3 /W j                                                                             T. S. Kress                           j

. Chairman ' ) r t

References:

1. Staff Requirements Memorandum dated May 15, 1996, from John C. i Hoyle, Secretary, NRC, to James M. Taylor, Executive Director

for Operations, NRC, regarding Briefing on PRA Implementation 1

Plan on April 4, 1996  !

2. Memorandum dated June 20, 1996, from James M. Taylor, j Executive Director for Operations, NRC, to the Commission, i

Subject:

Status Update of the Agency-Wide Implementation Plan i for Probabilistic Risk Assessment (PRA) (from March 1,1996 to l May 31, 1996)

3. Electric Power Research Institute, EPRI TR-105396, Final Report dated August 1995, "PSA Applications Guide" 1

J 1 i 1 1 9 4 J 1 1 48 l

pa my

      .,         k                            UNITED STATES

[ j t NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. -m1

      %*****/                                September 6, 1996 Dr. Thomas S. Kress, Chairman Advisory Comittee on Reactor Safeguards U.S. Nuclear Regulatory Comission Washington, D.C. 20555

SUBJECT:

RISK-INFORMED, PERFORMANCE-BASED REGULATION AND RELATED MATTERS I

Dear Dr. Kress:

Your letter of August 15, 1996, to Chairman Jackson provided the Advisory Committee on Reactor Safeguards (ACRS) views on several key policy issues related to risk-informed, performance-based regulation. These issues included: the use of NRC's Safety Goals in lf ant-specific applications; the treatment of uncertainties; whether to allow risk increases in risk-informed regulation; and the role of performance-based regulation in the context of , risk-informed regulation. These are among the policy issues in the May 15, 1996, Staff Requirements Memorandum (SRM) requesting staff recomendations on l the resolution of the four emerging policy issues. The insights and 1 l recommendations in your letter will be reflected in our response to the SRM, which will be attached to the next update of the Probabilistic Risk Assessment 3 (PRA) Implementation Plan due to the Comission on September 27, 1996. The staff has benefited from the meetings with the ACRS and is pleased that

the ACRS supports the staff's preliminary recomendations on the resolution of

,         these policy issues. We will continue our dialogue with the ACRS, and we          !

appreciate your coments ar.d advice regarding risk-informed, performance-based j regulation. i ] Sincerely,  ! , . - / - s M. Tafor

                                           . E cutive Director for Operations cc:    Chairman Jackson Comissioner Rogers Commissioner Dicus Comissioner Diaz Comissioner McGaffigan SECY 4

49

i i I i d ae i l ITEM B.4: 4 POTENTIAL USE OF IPE/IPEEE RESULTS TO COMPARE i THE RISK OF THE CURRENT POPULATION OF PLANTS wffH THE SAFETY GOALS (DR. KRESS) d i l l I I 4 4 a a

l [. {. ITEM B.4: POTENTIAL USE OF IPE/IPFFF RESULTS TO COMPARE THE RISK OF THE CURRENT POPULATION OF PLANTS WITH THE SAFETY GOALS i l In the December 27,1995 Staff Requirements Memorandum (SRM), resulting from the meeting between the ACRS and the Commiazion on December 8,1995, the Commission requested the j ACRS views on,

                                           ~

l "the extent to which the current spectrum of IPEs can be used in the regulatory process." i During its Febmary and March 1996 meetings, the ACRS dimnW the use of Individual Plant j Framinations (IPEs) in the regulatory process and other related matters with representatives

of the NRC staff and provided a report to the Commission dated March 8,1996. In that j report the Committee made several points, including the following

I e The quality and consistency of the IPEs vary and the impact of assumptions and ! analytical models is difficult to assess. On a case-by. case basis, however, additional and ! extended use of these IPEs is possible. As specific regulatory issues arise, the PRA l Standard Review Plan now being developed by the staff can serve u a template for

judging the quality and acceptability of the individual plant PRA for the proposed

} application. e To achieve consistency, some degree of standardization of PRA models and method will be required. IPEs should be reviewed to identify acceptable and unacceptable assumptions and/or models. Codification of m=ptions and models ought not inhibit the continued development of PRA methods. These activities would be a significant first step toward addrening the Commission's statement in the June 16,1995 SRM, "that more meaningful plant-to-plant or scenario-to-scenario comparisons based on risk could be achieved if PRAs were done on a more standardized, replicable basis." e The NRC could make additional use of the present IPEs (expect those that the staff has  ! found to use unacceptable methods or models) for a limited number of applications (e.g., regulatory analyses and prioritization of generic issues). During its meeting on May 23,1996, the ACRS dien=~l he t potential use of IPE/ Individual Plant Fvsmination of External Events (IPEEE) results to compare the risk of the current population of plants with the Safety Goals. This was in response to a SRM dated September 20,1994, in which the Commincion requested further guidance and insight on determining where the current population of operating plants, both individually and collectively, fall in  ! relation to the Safety Goals. The ACRS provided a report to the Commission dated June 6, i 1996. In that report the Committee made several points, including the following:  ! 1 50 l l

f i e The prompt fatality and latent health effects quantitative safety goals are posed in risk terms. Consequently, to establish the statas of the population of plants with respect to these goals, a full-scope level 3 probabilistic risk assessment (PRA) of acceptable quality for every plant would seem to be required. Such PRAs would need to include all internal and external events (including the low-power and shutdown operations) and would also need to take into consideration the individual site characteristics.

• In almost all cases, the IPEs and IPEEEs are not and were not intended to be full-scope PRAs. For example, a large number of IPEEEs used the Fire Induced Vulnerability Evaluation (FIVE) Methodology to search for potential fire ' vulnerabilities and the Seismic Margins Methodology to search for seismic vulnerabilities, neither of which gives a direct expression of risk. Funhermore, shutdown risk was not a pan of the l IPE/IPEEEs. While most licensees performed some type of level 2 containment analysis, the vast majority did not perform a level 3 offsite consequences analysis.

e The BNL study represents a good attempt to estimate the effects of some of the missing elements in the IPEs/IPEEEs. This study did not attempt to evaluate the risk resulting from seismic and fire events, nor did it attempt to evaluate risk in the shutdown mode. l l e Information is available that arguably would make it possible to bound the effects on  ! risk of elements missing from the IPEs/IPEEEs and to develop an approximate comparison with the safety goals. Such a bound would be of questionable value and . would have very large uncenainties. We do not recommend that this be done. I e The evidence from the BNL study, NUREG-1150, other PRAs, and scoping studies of , shutdown risk indicates that, on average, the population of plants meets the safety goals. A definitive determination of this, however, will only be possible when acceptable, full-scope Level 3 PRAs are available for all the plants. e The required effon to develop full-scope level 3 PRAs cannot be justified for the sole purpose of comparison with the safety goals. Such PRAs, however, will be needed in the long run to move toward a coherent risk-informed regulatory system. The Executive Director for Operations responded to the ACRS comments and i recommendations on the above subject in a letter dated July, 18,1996, stating that,  ;

                                         "The staff agrees with the ACRS conclusion 'that, on average, the population of plants    !

meets the safety goals [i.e., quantitative health objectivesIbut that) a defm' itive deermination of this...will only be pouible when acceptable, full-scope Ievel 3 PRAs  ; are available for all the plants.' The staff also agrees with the ACRS conclusion that i

                                         'the required effon to develop such comprehensive PRAs cannot be justified for the sole purpose of comparison with the safety goals."

l i 51

                      --               ,        rr                  N

, Attachments:

e Staff Requirements Memorandum, dated December 27, 1995, from John C. Hoyle, Office of SECY, to John T. Larkins, ACRS Executive Director (pp. 53-54) e Repon dated March 8,1996, from T.S. Kress, ACRS Chairman, to Shirley. Ann Jackson, NRC Chairman,

Subject:

"Use of Individual Plant Examinations in the Regulatory Process" (pp. 55-58) e     Staff Requirements Memorandum, dated June 16,1995, from Andrew L. Bates, Office of SECY, to the File (pp. 59-60) e     Staff Requirements Memorandum, dated September 20, 1994, from John C. Hoyle, Office of SECY, to John T. Larkins, ACRS Executive Director (p. 61) e     Repon dated June 6,1996, from T.S. Kress, ACRS Chairman, to Shirley Ann Jackson, NRC Chairman, 

Subject:

" Potential Use of IPE/IPEEE Results to Compare the Risk of the Current Populations of Plants with the Safety Goals" (pp. 62-64) e     Letter dated July 18,1996, from James M. Taylor, Executive Director for Operations, to T.S. Kress, ACRS Chairman, 

Subject:

" Potential Use of IPE/IPEEE Results *.o Compare the Risk of the Current Population of Plants with the Safety Goals" (pp. 65-66) 52

4 "f nouq'o UNITED STATES 5 c\^ EIN  % NUCLEAR REGULATORY COMMISSION E WASHINGTON D.C 20555 4 December 27, 1995 IN RESPONSE, PLEASE REFER TO: M951208 OFFICE OF THE SECRETARY REVISED MEMORANDUM TO: John T. Larkins, Executive Director

                   .               Advisory Committee on Reactor Safeguards FROM:

SUBJECT:

6r C. Hoyle, Secretary STAFF REQUIREMENTS - MEETING WITH ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS), 1:30 P.M., FRIDAY, DECEMBER 8, 1995, COMMISSIONERS' CONFERENCE ROOM, ONE WHITE FLINT NORTH, ROCKVILLE, MARYLAND (OPEN TO PUBLIC ATTENDANCE) The Commission met with the ACRS on the following topics:

1. Proposed resolution of Generic Issue 78, " Monitoring of Fatigue Transient Limits for the Reactor Coolant System"

2. The NEI petition for rulemaking to amend 10 CFR 50.48,

                        " Fire Protection"

• l

3. Development of improved nondestructive examination  !

(NDE) techniques

4. National Academy of Sciences / National Research Council study on " Digital Instrumentation and Control Systems in Nuclear Power Plants, Safety and Reliability Issues" i

5. Proposed final revision 1 to Regulatory Guide 1.152,  !

                        " Criteria for Digital Computers in Safety Systems of           I Nuclear Power Plants"                                           l l

6. Status report of IPEs The Commission requested feedback from the ACRS on use of RuleNet during the next meeting with the Commission.

(ACRSEBO) (SECY Suspense: 5/24/96) The Commission also requested the J.CRS views on the extent to which the current spectrum of IPEs can be used in the regulatory process. (ACRS) (SECY Suspense: 2/23/96) l 53

_ - _ . _ . - _ _ _ - _ _ . _ . _ _ _ _ _ . _ _ . . _ . _ _ _ _ - _ _ _ = . . _ _ _ _ _ . . . _ _ _ . . . _ . _ . 1 l j . ) j In addition, the Commission requests ACRS views on the PRA I i framework document, its relationship to the pilot applications l l (SECY-95-280), and the next steps in the process to expand the i use of PRA in the regulatory decision-making process. (ACRSEGO) (SECY Suspense: 4/30/96) l J l i

                                                                                                                                                )

cc: Chairman Jackson l Commissioner Rogers l ! EDO I OGC OCA i OIG l Office Directors, Regions, ACNW, ASLBP (via E-Mail) l PDR - Advance l DCS - P1-24 r i

                                                                                                                                              -i 4

l I 54 i

3 i k ,i f

                                    /        g .

UNITED STATES NUCLEAR REGULATORY COMMISSION l d ADVISORY COMMITTEE ON REACTOR SAFEGUARDS g< WASHINGTON, D. C. 30856 j .

***** 1

! March 8, 1996 l l i I The Honorable Shirley Ann Jackson chairman { ' l U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001 i

Dear Chairman Jackson:

' l

SUBJECT:

USE OF INDIVIDUAL PLANT EXAMINATIONS IN THE REGUIATORY ' l PROCESS During the 428th and 429th meetings of the Advisory Committee on Reactor Safeguards, February 8-10 and March 7-9, 1996, respective- l ly, we discussed the Individual Plant Examination (IPE) review i process and findings with the NRC staff. Our Subcommittee on IPEs ' also met with the staff and its contractors on January 26,1996, to j review this matter. We also had the benefit of the documents referenced. This report is in response to the December 27, 1995 Staff Requirements Memorandum (SRM). In the SRM, the Commission requested'"the ACRS views on the extent , to which the current spectrum of IPEs can be used in the regulatory  ! process." We interpret this request as referring to potential regulatory uses of the IPEs that were not delineated in Generic Letter 88-20, " Individual Plant Examination for Severe Accident Vulnerabilities." This report includes comments on both the Generic Letter goals and the Commission request. Goals of Ganaric h tter 88-20 The purpose of the IPE program, as stated in Generic Letter 88-20, was for each licensee: (1) to develop an appreciation of severe accident behavior (2) to understand the most likely severe accident sequences that could occur at its plant (3) to gain a more quantitative understanding of the overall probabilities of core damage and fission product releases 4 55 w -- - , - - - - - - - - - - - - --

i.. 2 (4) to reduce, if necessary, the overall probabilities of core damage and fission product releases by modifying, where appropriate, hardware and proce-dures that would help prevent or mitigate severe accidents. We notg that the IPEs were to be limited to the examination of internal initiating events and internal floods with the reactor at power and that individual and societal risks were not to be estimated. Other programs deal with external events and shutdown risk. The IPE program has been successful at most utilities in meeting . goal (1) and, to a lesser extent, goals (2) and (3) of the Generic Letter. Goal (4) of the Generic Letter also appears to have been l achieved. We were told that most licensees discovered weaknesses and took corrective actions. In addition, this program has been i beneficial in educating a broader segment of the NRC staff about ) l the issues related to these goals. ' t We were told by the staff that all licensees submitted a Level-1 probabilistic risk assessment (PRA) . Most licensees also submitted , a Level-2 PRA, although some addressed Level-2 phenomena in a l , rudimentary manner. The methods and data sources used by different l l licensees varied widely. In some cases, the choices appeared to be l arbitrary. Some licensees chose to include cosmon-cause failures l only for major components, while others chose to ignore them  ; l completely. 1 i l It is difficult to determine the extent to which the variability in i IPE results for similar classes of plants is due to actual plant differences or to modeling assumptions. Although some of the causes for this variability may be immediately apparent, others are l not. The latter include assumptions made about success criteria, the assumed' dependencies between operator actions, and the level of decomposition in fault-tree analyses. (We note that the fault trees were not requested as part of the IPE submittals.) An example of a potentially significant impact of modeling i differences is the range of core-damage frequencies (CDFs) for BWR l 3/4s that the staff has compiled. This range is from about 10 to l about 10" per reactor-year. Although the staff has stated that l such differences are primarily due to plant differences, this range of results seems unrealistic given the similarity among BWR 3/4s. Use of IPEs in the Reculatory Process i As discussed above, the quality and consistency of the IPEs vary l and the impact of assumptions and analytical models is difficult to 56 l l

]'.* 4 , l .I l 3 i a i assess. On a case-by-case basis, however, additional and extended l 1 use of these IPEs is possible. As specific regulatory issues 1 arise, the PRA Standard Review Plan now being developed by the ! staff can serve as a template for judging the quality and accept-ability of the individual plant PRA fpr the proposed application. l As the agency moves toward risk-informed regulation, there will be i an increasing need for full-scope PRAs that incorporate fire risk, ! external events, other modes of operation, and site-specific j consequences. When requests for risk-informed regulatory action i arise, the NRC staff should make it clear that a relevant PRA j should be used. 1 ! To achieve these goals, especially consistency, some degree of ! standardization will be required. Standardizing PRA models and i methods has been a controversial subject. Proponents argue W t it would create a basis for comparison of PRA results, while opponents fear that it would inhibit methodological developments. We recommend that IPEs be reviewed to identify acceptable and l unacceptable assumptions and/or models. Codification of assump-j tions and models ought not inhibit the continued development of PRA ! methods. These activities would be a significant first step toward ! addressing the Commission's statement in the SRM dated June 16, j 1995, "that more meaningful plant-to-plant or scenario-to-scenario

comparisons based on risk could be achieved if PRAs were done on a j more standardized, replicable basis."

l We believe that the NRC could make additional use of the present j IPEs (except those that the staff has found to use unacceptable l methods or models) for a limited number of applications (e.g.,

regulatory analyses and prioritization of generic issues).

i The staff stated that the CDFs for several PWRs are greater than ! 10 per reactor-year. Several BWRs have CDFs that are very close i to 10- per reactor-year and the conditional containment failure ! probabilities for BWR Mark I containments range from about 0.02 to i about 0.6. Although the PRAs have limitations as discussed above, ! these numbers suggest that an investigation would be warranted to i reassess their validity and to verify that the very low numbers l reported by some other plants reflect actual plant differences. Our conclusion is that the IPE program has met successfully the i objectives of Generic Letter 88-20. This program has developed a j risk awareness, both in the utilities and the NRC, that will ! contribute significantly to efforts to establish a risk-informed 3 and performance-oriented regulatory system. The plant-specific 1 l 2 i j I l 57 I I l

4 f IPEs are an extremely valuable asset that should not be permitted to languish unimproved and unused. Sincerely, S. FE

                                                  .. S.      r.ss Chairman

References:

1. Staff Requirements Memorandum dated June 16, 1995, from Andrew L. Bates, Acting Secretary, NRC, to the File regarding Meeting with ACRS on June 8, 1995

2. Staff Requirements Memorandum dated December 27, 1995, from John C. Hoyle, Secretary, NRC, to John T. Larkins, ACRS regarding Meeting with ACRS on December 8, 1995

3. Generic Letter 88-20, dated November 23, 1988, to All Licens-ees Holding Operating Licenses and Construction Permits for Nuclear Power Reactor Facilities,

Subject:

Individual Plant Examination for Severe Accident vulnerabilities - 10 CFR 550.54(f) 4 4

 \

58

j 'g UNITED STATES ygg, f-NUCLEAR REGULATORY COMMISSION WASHWe3 TON.C A 30086 j

       \..O  ...

IN RESPONSE, PLEASE sacasrany REFER TO: M950608A l June 16, 1995 i MEMORANDUM TO THE FILE 1 j FROM:- Andrew L. Bates, Acting Secretary 4

                                                                          /s/

i

SUBJECT:

STAFF REQUIREMENTS - MEETING WITM ADVISORY j COMMITTEE ON REACTOR SAFEGUARDS (ACRS), 10:00 l A.M., TMURSDAY, JUNE 8, 1995, COMMISSIONERS' CONFERENCE ROOM, ONE WHITE FLINT NORTH, ! ROCKVILLE, MARYIAND (OPEN TO PUBLIC l ATTENDANCE) i ) The Commission met with the Advisory Committee on Reactor Safeguards for discussion of the following topics:

1. Thermal hydraulic issues l 2. Status of Westinghouse AP600 design review

3. Regulatory analysis guidelines

4. Application of risk analysis in rulemaking i l 5. Proposed final rule on technical specifications l 6. Cracking and fatigue in nuclear components

7. Digital instrumentation and control j s. Operating reactors conformance to the safety goals -

l status report. i ] During the discussion on cracking and fatigue in nuclear components, the Commission requested the ACRS to land assistance l to the staff in encouraging industry to undertake research to ! improve NDE techniques with the aim of more accurately detecting i and assessing steam generator tube defects.

During discussion of the last topic, the commission expressed

concern about the lack of consistency among current IPEs and suggested that, modifications to the IPE FRAs may be necessary if they are to be used for other regulatory applications. In

' addition, the Commission noted that more meaningful plant-to-plant or scenario-to-scenario camparisons based on risk could be j achieved if PRAs were done on a more standardised, replicable l basis. There were no requirements identified for staff action, i 1 3 4 59 j

cc: The Chairman ! Constissioner Rogers l Commissioner de Planque l Commissioner Jackson ! OGC oCA i OIG l Office Directors, Regions, ACRS, ACNW, ASLBP (via E-Mail) ' PDR - Advance DCS - P1-24 i I l I 60

f

           ,sk               "

UNITED STATES

        ;j                               NUCLEAR RE2ULATORY COMMISSlHN l'                                                        WAaMuteTow. D.3. massenet

• g% * ** IN RESPONSE, PLEASE l REFER TO: M940908A o

l September 20, 1994 I j

MEMORANDUM TO

John T. Larkins, Executive Director i Advisory Committee on Reactor Safeguards t'

FROM: John C. Hoyle, Acting Secretary /s/ I l i

SUBJECT:

STAFF REQUIREMENTS - PERIODIC MEETING WITH ADVISORY COMMITTEE ON REACTOR SAFEGUARDS l (ACRS), 1:30 P.M., THURSDAY, SEPTEMBER 8, i i 1994, COMMISSIONERS' CONFERENCE ROOM, ONE l l WHITE FLINT NORTH, ROCKVILLE, MARYLAND (OPEN l l TO PUBLIC ATTENDANCE) i The Commission was briefed by the Advisory Counittee on Reactor j Safeguards. ! The Commission requested further guidance and insight on l ' determining where the current population of operating plants, l both individually and collectively, fall in relation to the j i safety goals. 5 l The ACRS committed to provide the Commission with a copy of the trip report which discusses the French move toward performance-l based fire regulations. (Subsequently, on September 12, 1994 the ACRS staff forwarded a copy of the report to the Acting Secretary for distribution to the Commission.) i i The Commission requested that the ACRS continue to monitor the NRC's actions and ensure there are no areas being ignored or overlooked through an error of omission. , l cc: The Chairman Commissioner Rogers Commissioner de Planque EDO OGC OCA OIG Office Directors, Regions, ACNW, ASLEP (via E-Mail) PDR - Advance DCS - P1-24 , 61

                 %,                          UNITED STATES j   .* 8          e NUCLEAR REGULATORY COMMISSION j       $                       ADVISORY COMMITTEE ON REACTCR sAFEOUARDS S                                  wAsHmorom, p. c. asses

..... I l June 6, 1996  !

l l i l

The Honorable Shirley Ann Jackson j i Chairman 1

! U.S. Nuclear Regulatory Commission  ! Washington, D.C. 20555-0001 l

SUBJECT:

POTENTIAL USE OF IPE/IPEEE RESULTS TO COMPARE THE RISK OF THE CURRENT POPULATION OF PLANTS WITH THE SAFETY G0ALS This report is in response to a Staff Requirements Memorandum dated ! September 20,1994, in which the Commission requested further guidance and j insight on determining where the current population of operating plants, 4 both individually and collectively, fall in relation to the safety goals. j Dur intent in developing a response was to examine the Individual Plant i Examinations (IPEs)/ Individual Plant Examinations of External Events i (IPEEEs) results to see if they can be extended so as to compare the risk j of the current population of plants with the safety goals. I During the 431st meeting of the Advisory Committee on Reactor Safeguards, ! May 23-25, 1996, we completed our discussions on this subject. During the l 418th, February 19g5, and 419th, March 1995 meetings, we heard j presentations by an ACRS Senior Fellow on an approach for estimating the i risk associated with some of the missing or incomplete elements of the 3 IPEs. During our 431st meeting, we reviewed a study by the Brookhaven , National Laboratory (BNL) (performed as part of the IPE Insights Program) j that investigated the use of some of the IPEs to compare the plant risk to the safety goals. We also had the benefit of the documents referenced. The prompt fatality and latent health effects quantitative safety goals

are posed la risk terms. Consequently, to establish the status of the

population of plants with respect to these goals, a full-scope Level 3 l probabilist$c risk assessment (PRA) of 'eceptable quality for every plant 4 would seem to be required. Such PRAs s 'd need to include all internal j and external events (including low-power and shutdown operations) and would also need to take into consideration the individual site i characteristics.

i' In almost all cases, the IPEs and IPEEEs are not and were not intended to be full-scope PRAs. For example, a large number of IPEEEs used the Fire Induced Vulnerability Evaluation (FIVE) Methodology to search for

;        potential fire vulnerabilities and the Seismic Margins Methodology to i         search for seismic vulnerabilities, neither of which gives a direct                                I l

I

                                                                                                           ~

! 62 1

2 expression of risk. Furthermore, shutdown risk was not a part of the IPEs/IPEEEs. While most licensees performed some type of Level 2 containment analysis, the vast majority did not perform a Level 3 offsite consequences analysis. The BNL study represents a good attempt to estimate the effects of some of the missing elements in the IPEs/IPEEEs. This study did not attempt to evaluate the risk resulting from seismic and fire events, nor did it attempt to evaluate risk in the shutdown mode. Information is available that arguably would make it possible to bound the effects on risk of elements missing from the IPEs/IPEEEs and to develop an i approximate comparison with the safety goals. Such a bound would be of i questionable value and would have very large uncertainties. We do not recommend that this be done. The evidence from the BNL study, NUREG-Il50, other PRAs, and scoping studies of shutdown risk indicates that, on average, the population of plants meets the safety goals. A definitive determination of this, however, will only be possible when acceptable, full-scope Level 3 PRAs are available for all the plants. We believe that the required effort to develop such comprehensive PRAs cannot be justified for the sole purpose of comparison with the safety goals. Such PRAs, however, will be needed in the long run to move toward a coherent risk-informed regulatory system. Sincerely, J. N T. S. Kress Chairman

REFERENCES:

l 1. Memorandum dated September 20, 1994, from John C. Hoyle, Acting l Secretary, NRC, to John T. Larkins, Executive Director, ACRS, l

Subject:

Staff Requirements - Periodic Meeting with ACRS, Thursday, l September 8, 1994

2. Richar11 Sherry, ACRS Senior Fellow, "A Simplified Approach to Estimation of Seismic Core Damage Frequencies from a Seismic Margins l Methods Analysis' l

3. U. S. Nuclear Regulatory Commission, NUREG-1150, " Severe Accident Risks: An Assessment for Five U.S. Nuclear Power Plants," Office of Nuclear Regulatory Research, December 1990

4. U. S. Nuclear Regulatory Commission, NUREG-XXXX, " Individual Plant Examination Program: Perspectives on Reactor Safety and Plant Performance," Draft for Comment dated April 1996

5. U.S. Nuclear Regulatory Cosmission, NUREG/CR-6144, ' Evaluation of l Potential Severe Accidents During Low Power and Shutdown Operations

at Surry, Unit 1," Brookhaven National Laboratory, July 1994 63 l

l l - - _

\. l 3

6. U. S. Nuclear Regulatory Commission, NUREG/CR-6143, " Evaluation of Potential Severe Accidents During Low Power and Shutdown Operations at Grand Gulf. Unit 1," Sandia National Laboratories, March .1995 l

l 64

l na g g '. UNffED STATES j .'u

• NUCLEAR RE2ULATORY C2MMISSIZN wAsmorow, e.o. -

July 18, 1996 Dr. T. S. Kress, Chairman Advisory Committee on Reactor Safeguards U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001

Dear Dr. Kress:

l

SUBJECT:

POTENTIAL USE OF IPE/IPEEE RESULTS TO COMPARE THE RISK OF THE CURRENT POPULATION OF PLANTS WITH THE SAFETY GOALS This is in response to the ACRS letter to Chairman Jackson dated June 6, 1996,

regarding potential use of IPE/IPEEE results to compare the risk of the

;       current. population of plants with the Cosmiission's Safety Goals. As discussed during the ACRS meeting, the results presented from the study comparing the      l IPE results with the Safety Goals were preliminary. Since very few IPEs were based on level 3 PRAs, the comparisons to the goals (i.e., the two              '

quantitative health objectives) were made through the use of IPE level I and 2 information and NUREG-1150 level 3 information. Comparisons were also made with two subsidiary objectives, core damage frequency and the conditional probability of early containment failure. No comparisons were made with the ' one-in-one-million frequency of large release subsidiary objective proposed in the Safety Goal Policy Statement. This proposed objective was found several

years ago to be impractical to implement (as discussed with and agreed to by

' the Committee at that time). As noted in your letter the staff agrees and l

      , recognizes that this study does not include all the relevant risk such as the risk from seismic and fire and the risk from other modes of operation (e.g.,

shutdown). This work has progressed since that time and will be documented in i the upcoming IPE Insights report (NUREG-1560) to be published in October 1996 for public comment. The ACRS will be provided a copy of this report. The staff agrees with the ACRS conclusion 'that, on average, the population of plants meets the safety goals (i.e., quantitative health objectives] [but that) a definitive determination of this...will only be possible when acceptable, full-scope Level 3 PRAs are available for all the plants." The staff also agrees with the ACRS conclusion that "the required effort to develop such comprehensive PRAs cannot be justified for the sole purpose of 65

f j , l 1

comparison with the safety goals." The extent to which such PRA's will be

needed in the long run to move toward a coherent risk-informed regulatory system, as suggested by the ACRS, will be a subject of the continuing interactions between the staff and the Committee on risk-informed regulation.

! Sincerely, l

                                                       ..                                         /

l 7.a _ \ s M. T or ecutive tractor for Operations i cc: Chairman Jackson Commissioner Rogers Commissioner Dieus SECY 1 l l l l 66 l

_am, _us. m..,,- 4,,._ Ass- m a-..m-,,x~.ua,.As2--a -p aAm--D$4,.a,As--_2 sAKa,saAa._,in+r--J4& +4-sk Aa.,_2ra,- ,.hn +6 nem-.s_--an.a w nnaW.

                                                                                                                                 .A .-

mm a . r_ l ITEM B.5: USE OF SAFETY GOALS ON A PLANT-SPECIFIC BASIS (DR. KRESS) l I 1 l l

e ITEM B.5: USE OF SAFETY GOATS ON A PLANT-SPECIFIC BASIS i i In an April 23,1996 repon to the Commission on "Probabilistic Risk Assessment Framework, i Pilot Applications, and Next Steps to Expanding the Use of PRA in the Regulatory Decision-

Making Process," the ACRS stated that a restatement of the Commission's safety goal policy 2

is needed, which would allow the use of safety goals on a plant-specific basis. In the June 11, 1996 Staff Requirements Memorandum (SRM), resulting from the meeting between the ACRS and the Commission on May 24,1996, the Commission stated that: i ! "The ACRS should provide recommendations on how the Commission's safety goals and safety goal policy should be revised to make them acceptable for use on a plant-specific basis." The ACRS Subcommittee on Probabilistic Risk Assessment met with representatives of the staff and industry on July 18 and August 7,1996, to gather information on the pilot applications for risk-informed, performance-based regulation. Based on the information gathered and discussions held during the August 8-10,1996 ACRS meeting, the ACRS issued a report " Risk-Informed, Performance-Based Regulation and Related Matters," dated August 15,1996, which included the following paragraph:

                       "We believe the safety goals and subsidiary objectives can and should be used to derive guidelines for plant-specific applications. It is, however, impractical to rely exclusively on the Quantitative Health Objectives (QHOs) for routine use on an individual plant basis. Criteria based on core damage frequency                ,

(CDF) and large, early release frequency (LERF) focus more sharply on safety l issues and can provide assurance that the QHOs are met. They should be used in developing detailed guidelines." The Committee also noted in the report that it would discuss the larger question of the application of the safety goals on a plant-specific buis at a later time.  ! l

                                                                                                                ~

On November 18, 1996, the Committee provided a report to the Commission on " Plant. Specific Application of Safety Goals," which contained several observations and recommendations including the following: e The Safety Goals quantified "how safe is safe enough" for the population of U. S. , plants. For an individual plant, however, the acceptable level of risk is determined by l the concept of " adequate protection," which in the final analysis means compliance I with the body of regulations. Risk-informed analyses would provide a more rational basis for making regulatory decisions regarding plant-specific requests for exemptions , from the rules or for changes to the licensing basis,'snd the acceptability of new , regulations." i 67 i

I l . \

. 3     An evolutionary and pragmatic approach for using Safety Goals on a plant-specific basis  j would be to use the CDF as the primary criterion for evaluating proposed changes         '

along with a qualitative or quantitative evaluation of the possible Level 2 and Level 3 PRA issues raised by these changes. For a quantitative analysis, the following two options are offered:  ;

1) A full-scope Level 2 PRA (with fission product transpon capability). l To use this option, a conservative value for LERF criterion must be determined. This value, along with the CDF criterion, will provide an acceptable basis for decisionmaking. We note that both the NRC staff and the Electric Power Research l Institute, in its, "PSA Application Guide," are proposing the use of LERF as an acceptable criterion, i

2) A full-scope Level 2 PRA (without fission product transport capability).

To use this option, conservative values for early containment failure frequency criteria for different reactor designs must be determined. These values, along with the CDF criterion, will provide an acceptable basis for decisionmaking. e In the longer term, the agency should move beyond the evaluation of risk associated with proposed changes to individual plant licenses and apply the Safety Goals to assess the acceptability of plant-specific risk. This could be done in terms of the QHOs, along with the CDF, or in terms of the CDF and LERF. To use the QHOs directly, it would be necessary to have full-scope Level 3 PRAs. The use of Level 3 PRAs in the future should be encouraged. Attachments: e Staff Requirements Memorandum, dated June 11,1996, from John C. Hoyle, Office of SECY, to John T. Larkins, ACRS Executive Director (pp. 69-70) e Repon dated August 15,1996, from T. S. Kress, ACRS Chairman, to Shirley Ann Jackson, NRC Chairman,

Subject:

Risk-Informed, Risk-Based Regulation and Related Matters (pp. 71-74) e Letter dated November 18,1996, from T. S. Kress, ACRS Chairman, to Shirley Ann Jackson, NRC Chairman,

Subject:

Plant-Specific Application of Safety Goals (pp. 75-77) 68

    ._ -__ ._ _ . _ _ . _ _ _____ _ _ _ _. _ ._. _ a.             _      ,,,   ., y. _ _ _ . . _ _

1 REFER To M960524 ). i Juns 11, 1996 1!

• l MENORANDUN To: John T. Larkins, Executive Director

] Advisory Committee on Reactor Safeguards

FRON: John C. Hoyle, Secretary /s/

I 1

SUBJECT:

STAFF REQUIRENENTS - MEETING NITH ADVISORY l CONNITTEE ON REACTOR SAFEGUARDS (ACRS), l FRIDAY, MAY 24, 1996, CONNISSIONERS' .; . CONFERENCE ROON, ONE WHITE FLINT NORTH, 1 ROCKVILLE, MARYLAND (OPEN TO PUBLIC ] ATTENDANCE) i j The Commission was briefed by the ACRS on the following topics: W i 1. Use of IPEs in the regulatory process, PRA framework

document, pilot applications, and the next steps to expand the use of PRA in the regulatory decision-making

, process i 4 2. Fire protection issues, including fire PRA models and i PRA-based scoping analysis of degraded fire barriers j 3. Proposed final revisions to 10 CFR Parts 50 and 100 i Status of ACRS review of Regulatory Guidance documents 4. ! related to digital instrumentation and control systems j 5. Status of ACRS review of standard plant designs: i S ABWR and System 80+ design certification rules j 8 AP600 design 4 Test and analysis programs associated with the AP600 and SSWR designs

6. Conformance of operating plants with NRC safety goals i The Commission requested ACRS input on GE's recent sukulittal containing safety significant design changes prior to the

! Commission briefing on the design certification rulemakings scheduled for late August. (ACRS) (SECY Suspense: S/22/96) The ACRS should continue their review efforts, as noted in the meeting, in the digital instrumentation and controls area. I 5 e 69 4

1 I , i k The ACRS should continue to interact with the staff in the

development of regulatory guidance and a standard review plan for the performance and review of PRAs. In particular, the ACRS should advise the staff regarding appropriate methods for judging the acceptability and unacceptability of assumptions and models 1

to be used when performing PRAs. The ACRS should provide reconsendations on how the Consission's j safety goals and safety goal policy should be revised to make l them acceptable for use on a plant-specific basis. i

cc: Chairman Jackson

! Commissioner Rogers i Commissioner Dieus ) EM

OGC I

OCA OIG l Office Directors, Regions, ACRS, ACNW, ASLBP (via E-Mail) '. PDR - Advance j DCS - P1-24 i i i I i } I i k 1 ) 70 1 ) 1

j ., l /

                         /      \,,  ,

UNITED STATES NUCLEAR REGULATORY COMMISSION

g g AovisORY COMMITTEE ON REACTOR SAFEGUARDS

! e J WASHINGTON, D. C. 20H6 1

*e .* l l August 15, 1996 3

The Honorable Shirley Ann Jackson Chairman i U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001

Dear chairman Jackson:

{

SUBJECT:

RISK-INFORMED, PERFORMANCE-BASED REGUIATION AND REIATED ! MATTERS l l During the 433rd meeting of the Advisory Committee on Reactor i Safeguards, August 8-10, 1996, we discussed the issues identified ! in the Staff Requirements Memorandum dated May 15, 1996. We also

discussed the pilot applications for risk-informed, performance-

! based regulation. Our Subcommittee on Probabilistic Risk e Assessment (PRA) met with representatives of the NRC staff and the l 3 nuclear industry on July 18 and August 7, 1996. We also had the l benefit of the documents referenced. The staff presentations dealt only with the development of guidelines from the commission's safety goals to be used as an ! element of the evaluation of licensee-initiated changes to i licensing commitments. All of our comments address the application of risk-informed regulation in that context. At a later time, we will discuss the larger question of the application of the safety goals on a plant-specific basis. CCWCLUSIONS Issue 1: Should the consission's safety goals and subsidiary objectives be referenced or used to derive guidelines for plant-specific applications and, ii no, how? We believe the safety goals and subsidiary objectives can and should be used to derive guidelines for plant-specific applications. It is, however, impractical to rely exclusively on the Quantitative Health Objectives (QHOs) for routine use on an individual plant basis. Criteria based on core damage frequency (CDF) and large, early release frequency (LERF) focus more sharply on safety issues and can provide assurance that the QHos are met. They should be used in developing detailed guidelines. 71

l \ 2 Issue 2: Mov are uncertainties to be accounted for? This is a difficult issue. There are models and formal methods to account explicitly for a large number of uncertainties. However, other uncertainties are unquantifiable. The staff proposes to explore a number of options, such as establishing margins in the acceptance guidelines, placing more importance on defense-in-depth, and others, to deal with such uncertainties. Such approaches seen appropriate, although much work remains to be done. Issue 3: should requested changes to the current licensing basis be risk-neutral or should increases be permitted? We agree with the staff and industry that increases in risk should be permitted ~in some situations. Acceptance guidelines expressed in terms of the proposed change in risk and the current risk l estimates should have three regions: a region in which some increase in risk is acceptable, one in which it is unacceptable, and one in which further analysis and evaluation would be required. Issue 4: Mov should performance-based regulation be implemented in the context of risk-informed regulation? We agree with the staff that, where practical, performance-based strategies should be included in the implementation and monitoring step of the risk-informed decision-making process. The pilot programs may provide an opportunity for a more concrete definition and development of performance-based strategies. nzscussrow l Issue 1 Even though a CDF could be derived fron ' the QHOs that could be greater than 10-8 per reactor-year, the current subsidiary goal of 10-' per reactor-year should be maintained and should be stated as a fundamental safety goal, along with the QHO. Accident sequences that have a high probability of leading to severe consequences could be controlled by the QHos, but a more workable measure would be a subsidiary goal on the LERF. The definition of the latter needs to be improved. Whether the LERF should be a fixed value or derived from the QHos, which would allow the LERF goal to include site-specific characteristics, needs to be investigated. , We recommend that the staff develop guidance for handling I situations in which high values of the CDF occur for short periods  ; of time (for example, 10*8 per reactor-year for a day). 72

i 3 i - a

Issue 2 i

i In accounting for uncertainties, it is important to distinguish between those plant characteristics or phenomena that are modeled in the PRA and those that are not medeled (e.g., the actual layout 1 of components and organizational factors) . For those that are ! modeled, parameter and model uncertainties should be explicitly l quantified and propagated through the PRA. The resulting

distributions should be an input to the decision-making process j along with other qualitative input.

! Mean values of distributions should, in general, be used for comparison with goals or criteria, although the sensitivity of the mean value to the high tail of a distribution should not be

overlooked. For very broad distributions, such as those that

! typically result when significant model uncertainty is present, ! reliance on the mean values may not be appropriate and a more

detailed investigation of the reasons for this large uncertainty I should be undertaken. This could possibly lead to decisions to i conduct additional research or to take other measures.

Accounting for uncertainty in the case of plant characteristics or , phenomena that are not currently modeled at all is much more , difficult. The staff proposes to explore a number of options, such I as establishing margins in the acceptance guidelines, placing more j importance on defense-in-depth, and others. We agree and encourage j the staff to actively pursue the resolution of this issue. i { Issue 3 ) l The concept of a "three-region" approach is consistent with the i Electric Power Research Institute's PSA Applications Guide (PSAAG), 4 although the boundaries of the regions used in the PSAAG are not ! necessarily the ones that the staff will adopt. i l The staff has raised the issue of how " packaged" requests are to be handled. Packaging is the process by which risk trade-offs can be accomplished. It is a significant benefit of risk-informed regulation. We believe that it is the overall impact on plant risk that is important, and related changes should be handled as a j package. Such changes shduld be consistent with the current philosophy of risk management; i.e. , that the " bottom-line" numbers - should not be the only input to the decision-making process, and other concepts such as defense-in-depth must be maintained. 1 i l i 73

4 We will continue to monitor the progress of the staff on these issues. Sincerely, l

h. ri

                                                                              ).
,                                          T. S. Kress                        .

i Chairman I i l 4 i

References:

,        1. Staff Requirements Memorandum dated May 15, 1996, from John C.

i Hoyle, Secretary, NRC, to James M. Taylor, Executive Director for Operations, NRC, regarding Briefing on PRA Implementation Plan on April 4, 1996

2. Memorandum dated June 20, 1996, from James M. Taylor, Executive Director for Operations, NRC, to the Commission,

Subject:

Status Update of the Agency-Wide Implementation Plan for Probabilistic Risk Assessment (PRA) (from March 1,1996 to May 31, 1996)

3. Electric Power Research Institute, EPRI TR-105396, Final I

Report dated August 1995, "PSA Applications Guide" l 4 l l i 74 i

                      - ---     .-                m,

g c:aaasuosaseo e NUCLEAR RECULATORY COMMISSION 3 ADVISORY 00thsTTEE ON REACTOR SAFE *UA".DS

                                    ==aanro . o. c. ==.

.C November 18, 1996 The Monorable Shirley Ann Jackson Chairman U.S. Nuclear Regulatory Commission Washington, D.C. 20555-0001

SUBJECT:

PLANT-SPECIFIC APPLICATION OF SAFETY GQh1.5 During the 436th meeting of the Advisory Committee on Beactor Safeguards, November 7-9, 1996, we discussed the a;pplication of Safety Goals on a plant-specific basis. This subrect was alsce discussed at meetings of our Joint Su h ittees on Probabilistic Risk Assessment and Plant operations on July 17-18, 1996, and of our subcommittee on Probabilistic Risk Assessment on August 7, 1996. We also had the benefit of the documents referenced. In a Staff Requirements Memorandum dated June 11, 1996, we were requested to provide recommendations on how the Commission's Safety Goals and Safety Goal Policy should be revised to make them ( acceptable for use on a plant-specific basis. The Safety Goal Policy Statement made it clear that the Quantitative Realth Objectives (QNos) and the subsidiary Core Damage Frequency (CDF) goal were to provide standards for the NRC staff to judge the overall effectiveness of the regulatory system. That is, if the risk posed by the population of plants on the average proved to be less than tt* Safety Goals, then the staff (and presumably the public) would ts m that the regulatory system had functioned appropriately to protect the health and safety of the public. The Safety Goals quantified "how safe is safe enough" for the population of U. S. plants. For an individual plant, however, the mooeptable level of risk is determined by the concept of " adequate ion," which in the final analysis means compliance with the y of regulations. Risk-informed analyses would provide a more rational basis for making regulatory decisions regarding plant-specific requests for exasptions from the rules or for changes to the licensing basis, and the mooeptability of now regulations. In our August 15, 1996 report, we stated: "the safety goals and subsidiary objectives can and should be used to derive guidelines ' for plant-specific applications. It is, however, impractical to rely exclusively on the Quantitative Realth Objectives (Qu0s) for routine use on an individual plant basis. Criteria based on core l 75 4

I I. 2 f

                                                                                     \

i 'j damag3 frequ^ncy (CDF) and large, early release frequency (LERP) i focus more sharply on safety issues and can provide assurance that l , the glos are met." i i In developing plant-specific criteria, it is important to consider ! the regulatory needs in the near future and to ensure that the 1 j process will be evolutionary rather than so revolutionary that it j j might discourage the licensees from using this approach. It ' ] appears to that most of the anticipated licensee requests for changes

their current licensing basis will deal with Inval 1

) probabilistic risk assessment (FRA) issues, e.g., inservice  : ! inspection, extension of allowed outage times. Furthermore, most j licensses have only recently familiarised themselves with Level 1 ! PRA methodology for the narrow regime of power operations. They are just beginning to integrate findings of such Level 1 risk assessments with the safe operation of their plants. Even the NRC l staff is still coming to grips with the implications of Level 1 i risk assessment results for regulation of nuclear plants. Many l licensees do not have Awass to the technologies for facile conduct of full-scope tavel 2 or Level 3 FRAs that treat power operations,

low power / shutdown operations, as well as accidents initiated by

external ever.ts. Commonly accepted standards for such extensive,

' in-depth ana'./fses do not exist. i

f-An evolutionary and pragmatic approach for using Safety Goals on a i I plant-specific basis would be to use the CDF as the primary criterion for evaluating proposed changes along with a qualitative l or quantitative evaluation of the possible Level 2 and Level 3 PRA

issues raised by these changes. For a quantitative analysis, the I

foilowing two options are offered: ! 1) Full-scope Inval 2 FRA (with fission product transport capability). To use this option, a conservative value for a' LERF criterion must l. be determined. This value, along with the CDF criterion, will provide an acceptable basis for decisionmaking. We note that both j the NRC staff and the Electric Power Research Institute, in its,

. "FSA Application Guide," are proposing the use of LERF as an

acceptance criterion.

i l 2) Full-scope Level 2 FRA (without fission product transport j capability). ! To use this option, conservative values for early containment i failure frequency criteria for different reactor designt must be ! determined. These values, along with the CDF criterion, will

provide an acceptable basis for decisionmaking.

1 i In the longer term, we believe the agency should move beyond the evaluation of risk associated with proposed changes to individual plant licenses and apply the safety Goals to assess the ] (f' l 76 1

i !! 3 ) i - acceptability of plant-specific risk. This could be done in terms 1 . e of the 3 0s, along with the CDF, or in teras of the CDF and LERF. l To use the QHos directly, it would be ma==ary to have full-scope Inval 3 FRks. future should be encouraged. We believe that the use of Level 3 FRAs in the I i Sincerely, i

h. 7e T. S. Kress

) Chairman l

Rafarances:

) 1. Staff Requirements Memorandum dated June 11, 1996, from John j Moyle, Secretary, MRC, to John T.1arkins, Executive Director, ! ACRS,

Subject:

Meeting with ACRS, Friday, May 24, 1996 l 2. ACRS report dated August 15, 1996, from T. S. Eress, Chairman,

ACRS, to Shirley Ann Jackson, Chairman, RRC,

Subject:

Risk-1 Informed, Performance-Rased Regulation and Related Matters i 3. Electric Power Research Institute Report TR-105396, "PSA { Application Guide," prepared by ERIN Engineering and Research,

Inc., August 1995 4

t I 2 i I i i 1 1 J l 1 i l 1 i' 77 i i I

A,Ae ,_ 4,j 4 w-. Mn 4-e,--emeL---,...-A e _x ss e 6%, _ s a aAM sr. -n-s--.L--mmm.m---M -.4whea -osa. ns saes- A we e xs z. _.a aammama-m aeses--m- -s eamwa , w gwa l l l ITEM B.6: USE OF RULENET IN THE REGULATORY PROCESS l (DR. SHACK) l l j l l I i I l l

l l ITEM B.6: USE OF RUT _FNET IN THE REGULATORY PROCESS e Use of RUT FNET in the Renulatory Prome In the December 22,1995 Staff Requirements Memorandum (SRM), resulting from the meeting between the ACRS and the Commission on December 8,1995, the Commission requested,

                   " feedback from the ACRS on the use of RuleNet" in the rnlamaing process.

The NRC undertook a project called RuleNet, designed to use statwf-the-art computer technology to maximize communication between the NRC and the public on important nuclear power plant safety issues. The project was intended not only to provide the NRC and l the public with valuable information, but also to test the usefulness of computer-based communication as a tool in the rulemaking process. The Nuclear Energy Institute (NEI) petition for rulemaking to amend 10 CFR 50.48, Fire Protection, was chosen as the pilot RuleNet action. The RuleNet process was broken into three phases. Phase I was the issue identification and discussion phase. Phase II was the alternative identification phase. In this phase, proposed solutions to the challenges and/or issues identified in Phase I were solicited. The analysis of  : alternatives and final comment phase was Phase III. This phase was intended to solicit more concrete proposals on how to deal with the issues or challenges. r During the 430th ACRS meeting, the Committee heard presentations by and held discussions with representatives of NEI and the NRC staff. NEI provided the following comments regarding the RuleNet program: e RuleNet may have merit and can improve communications between the NRC and the public if it is properly managed, is cost effective, and adds value to the regulatory process. The pilot project hu demonstrated that computer technology can be used in soliciting public input on regulatory issues, and in identifying areas of common understanding as well as differing opinions. e NEI supports the use of computer technologies to improve the regulatory process if they are cost-effective and add value. However, NEI believes there are a number of significant questions that need to be resolved in the RuleNet effort. For example, NRC's disposition of communications among participants, qualification of participants, potential influence by special interests, condu t of private versus open caucuses among l participants, and representation by organizations versus individuals, etc., have not been addressed by the RuleNet procedures. Until these questions can be fully addressed,  ! l RuleNet should be considered only as a tool to assist the NRC in achieving better l communication on regulatory issues. 78

o i I

   .                     The NRC staff stated that the use of electronic communication in licensing poses problems          i because the information that used to consist of hard copies, which became pan of the public        i' record, is now transmitted between computers.

The NRC sought to use RuleNet to reach agreement with the public on the statement of the  ; issues associated with the fire protection rulemaking. On August 29,1996, the staff issued SECY-96-188, "RULENET," which provided the staff's evaluation of the RuleNet process.  !

                       . It concluded that the RuleNet pilot program accomplished its two principal objectives, which were to a8d to the NRC's information base about fire protection issues and to serve as a           !

demonstration of the use of electronic communications in the regulatory sphere. In SECY-96-178, " Action Plan to Address Outstanding LSS Issues," the staff advised the Commission of issues penaining to the Licensing Support System (LSS) and its proposed use in the Depanment of Energy's license application for a high-level radioactive waste repository. As described in SECY-96-178, the staff intends to conduct an electronic discussion with the LSS ' Advisory Panel to identify issues related to the future direction of the LSS. This discussion will use many of the tools and processes developed for NRC's RuleNet. The NRC is posting this phase as LSSNET. ACRS OBSERVATIONS: e The use of computer technology can enhance the communications between the public and the nuclear industry and can be cost effective and add value to the regulatory process. However, because a significant ponion of the general public either does not have access to the needed computer technology or is not familiar with its use, other avenues for participating in the rulemaking process will have to be maintained. e Much of the discussion was in the form of open caucuses among the panicipants. It was difficult for the staff to determine whether comments were addressed to the NRC or to the other panicipants in the caucuses, o More guidelines should be provided to the panicipants with regard to proper format for providing comments. e RuleNet was a useful first step in the development of procedures and guidelines for using computer technology in the regulatory process in the future. This development should continue. I Attachments: e RuleNet pages from Internet, including NEI views (pp. 80-82) e Licensing Support System (LSS) Rulemaking Issues (pp. 83-84)  ; I e Letter dated December 1,1995, from William H. Rasin, NEI, to NEI Nuclear Strategic ) Issues Advisory Committee,

Subject:

"NRC RuleNet Program" (pp. 85-86) 79                                          !

_ -_ _ ga,,,,, m,,- 4 i United states

             @ym. Nuclear Regulatory Commission
                                                                          . m ,=

- m .. .

About NRC Rulemaking a 9 NRC Rulemakino Process

          # Petitinnino NRC to Initiste. Modify or Terminate a Rule a

9.NRC Repulsenry Acends Rulemaking and Petitions for Rulemaking on Fedworld i 9'Informatinn about the Rulemakino Bulletin Board System O. News Inn.....minn and cantmets for current Rulemakinac (tableformat) 8 News. Infnnnatinn and contnets for cumnt Rulemakinac (non-tableformat) l lateractive Rulemaking Pilot ! The U.S. Nuclear Regulatory Commission conducted an interactive rulemaki.ng pilot known as "Rulenet." The pilot allowed for interactive rule development through use of World-Wide Web browsers ' and associated software tools. Using fire protection rules. the pilot focused on changes that would make the rules more performanced based and less arescriptive. The public was able to access RuleNet from i i January 5.1996 to February 9.1996. While TuleNet is closed for comment, the description and discussions which occured in each of the three phases ofRuistist is available for browsing. t ruletop.htmllast updated on Tuesday, March.1996 by cihi@nre nv . i l l i 1 i i i 4 l 80 scrt ""

I mnemosses nnpassengomemlopesmm l i i

                                                                           ,.                 m-h.-                        -

i . 5 ($, P_ ((i i,13. L- M '- 4 l f ..

                                             . u. 1 .T ---                            :        WF            i                                    , t,l Choose this for a textual representation of this page.

s l To provide participants with an overview of discussions in RuleNet, news pages were published daily. nese pages included a summary and a detailed news story for each discussion. These News cories have , been maintained in the News archive. 1 -. RuleNet was open from January 5,1996 throught February 9,1996. It proceeded through 3 phases. The

description and discussions which occurred in each of these phases are accessable below. Please feel free i

to bmwse the discussions space. l Phase I Discussions j ne first phase of the process was a " virtual kickoff" in which all participants were able to

commumcate m a simultaneous discussion via computer. His was followed by a 'od of five

! days for any caucuses; for the posting of questions and requests for clarification, ' cted either to l the NRC or to other participants, and for the posting of answers to those questions; and for the j identification of any funher issues to be addressed, or challenges to be met, in the rulemaHT l Phase II Discussions j The second phase of the process comprised approximately 10 days. The NRC solicited proposed i solutions to the challenges and issues identified in the first phase. It was also an opportumty for 1 participants to respond to comments and suggestions made during the first phase. Phase III Discussions During the third phase of the process, the NRC technical staff, acting with the assistance of staff suppl,ied by the contractor. consolidated and synthesized the challenges and the proposed so .utions, usmg them to develop more concrete proposals, which were posted electronically, he i participants were then allowed to respond to the proposals. As before, there was the opportunity l for panicipants to caucus. O References ) Help k Info 7 Icedhatk l 3 News et Updates ( l orum . i llome To eenanannt en shis servace: caamamena@nssellaLyov Technicalp.=eMenas: ar&+fadilasselinLeov Last Updese:Janesty IL I996 (Dheininarri UCRL-MI.!l9788 81 g, out2s608.312

1 annemums . 1, 4 l l i h-eRegulatory-- Nuclear COMMISSION l RuleNet Discussion Group - Contment_On_Rulenet

e

Subject:

NEI Views

Front gew@nel.org Date:Fri,9Feb 199619:04:27 GMT Message-Id: 199602091858.SM28305@nssc.llnl. gov l

4

Name

George Wu l

j ! NEI believes that RuleNet may have. merit in enhancing communications  ; j 1.r;;a the public and the NRC if properly managed, is cost effective, and adds value. We believe that the pilot project has demonstrated that computer technology can be used in soliciting public input on regulatory I issues. and in idenufying areas of common understandmg as well as

!           differing opinions. We also believe that information exchanges on RuleNet
!            can promote a better understandmg at the NRC and among interested individuals ofissues that need to be addressed.

We note. however, that procedural questions remain with how the i communications on RuleNet will be used to supplement information developed ! through traditional rulemaking approaches. For example, NRC's disposition i of commumcations among parncipants, c* of pnvase versus open l caucuses among participants, and what constitutes t se fonnal record, are l some questions that can be identined. Until these questions can be fully i addressed, we believe that RuleNet should be considered only as a tool to j assist the NRC in achieving better communication on regulatory issues. ' We note also that the benefits of RuleNet to rul*==Wna must be detennined based on an assessment of the cost effectiveness of the' program and value added in the regulatory process. both from the wve of managing the ! RuleNet process as well as the resource 'tures of participants. (As the NRC stated "Before the type of electronic exchange being demon =ed

in the RuleNet project became a part of the agency's usual process for the development of rules, it would have to be shown to be cost effective."

60FR57372) Pending this demonstration of cost effectiveness and value added, we believe that the use of computer-based technology to improve ! communications abould not be considered for other regulatory activities of

the NRC. We also believe. based on lessons learned from this

demonstration, that the NRC should publish more detailed guidance on how RuleNet and the end results will be used. The NRC's determination

              +Jmg the cost effectiveness and overall value of RuleNet also should be published in the Federal Register for public review and comment.

- Messages sorted by

I due if thrend if mblect if anthne 1 I tuhiect-shetemet 1

                         -. Next message: gew@nei arc "Re Direct i ieen=* and Public Are*=e" i                         : Prwious message: devonrue1@nnt enmi

• Ret Direct I icen=ee and Puhlie Aeee=="

                          - Next la thread: devonme1@nni enm "Re NET Views" 1
^

l Ref erem es 3 llelp ht info O l eedb.u k l News k Updates 7 l-orum O flome d

Welcome!

LSS Rulemaking Issues l

The Licensing Support System (LSS) concept grew out of the Nuclear Regulatory Commission's concern regarding how best to review the DOE license application for a high-level radioactive waste (HLW) repository. A centralized, electronic database, accessible by all parties appeared to offer the opportunity for significant time savings in conducting the licensing proceedmg for the repository and, simultaneously, for the enhancement of any party's opportunity for effective participation. Plans for the LSS were first initiated in 1986 and were based on computer l technology available in that time frame. It was intended to provide a central, shared, federally funded database of l licensing information beginning in 1995. Budgetary shortfalls, however, and the unanticipated length of time that it  ; ! would take to develop the licensing application for the repository, not only delayed the development of the LSS, but l also resulted in the accumulation of a tremendous amount of potential licensing information, much of which may no I longer be relevant to a licensing proceeding which may not begin until about 2002. In addition, since document I capture may now involve much larger backlogs than originally contemplated, the risk of failing to capture all relevant material in the LSS is substantially larger than originally assumed. Whil.> the development of the LSS remained stalled, the state of technology in document automation and retrieval ovc::nok the technology of 1986 on which the original LSS was to be based. With the widespread and common place use of computers to generate and maintain the documents of a party to the HLW licensing proceeding, the universal availability of the Intemet to tie disparate and geographically dispersed systems together, and the availability of commercially available software applications relevant to LSS functionalities, the centralized LSS envisioned at the time the LSS rule was developed may be obsolete. Consequently, the Commission intends to evaluate how these new technologies can be integrated into the LSS rule while still maintaining the primary functions of the LSS:

1. A mechanism for the discovery of documents before the license apnlication is filed;

2. Electronic transmission of filings by the parties during the proceeumg; l

3. Electronic transmission of orders and decisions related to the proceeding; and

4. Access to an electronic version of the docket. l It is the intent of the NRC staff to focus this rulemaking on how best to address changes in technology in regard to the LSS. There is no intent to re-visit the basic functionalities of the LSS that are reflected in the current 10 CFR Part 2, Subpart J.

To attempt to address these issues, the NRC is posting the following " topics" to guide the discussion during this phase of LSSNet: Topic 1 - What are the costs and benefits of moving from a dedicated, centralized system to a distributed system based on the Intemet? Inpk.2 - How should other improvements in computer technology be incorporated into the LSS? Topic 3 - What provisions of the LSS rule will ne-i to be changed to reflect the incorporation of new technologies? Tonic 4 - How should the backlog of"uncaptured",amd possibly irrelevant, repository-related information be 83

I

;                                  auw casw a                                                                                                                                I

! i

  ,                                 Topic 5 - What would the role of the LSS Administrator be under a distributed system?
    ,                               Topic 6 - How should advice from potential users of the LSS be provided for?

4 l 1

                                                                          ' Ref erent es                                                 llelp A Info             ' feedbc k

! 9 New s A l'pdates s I or um 11ouw References l Nggs l Help & Info l Forum l Feedback l Heat

                                                                     ~

i j To comment on this wavice: c-- = G - = =

• ff=Laov l Technical problems: red-fm/4==="r l'al gov i

Last Update: October 24,1996 l Disclaimer: UCM-MI.11M08 I

} I t  ! i l l 1 1 84

i i ~ l1 ~ () I l i Es(tut triest lasterett l pG 4

                                                                                                                }

i Deosaber1,1995 - f l TO: NEZ Nuclear Striangle Issues Advisory ComeMetet 1 l SUBJRCT: NRC RuisNet Program i l On November 15,1996, the NRC published in the Federal Regisser-(ao Fed. Reg.

57870) a notice d avaitelnlity of a cosaputer based pilot program onBed "RuleNet."

! N intent of RuleNet is to ==44= casamunie=+ian between the NBC and the

                                                   . public on rn1==aMag issues, and it is the Srst time that sneh a program has been implemented by any federal agency. N program is being applied by the NRC speciscally to the 5:e protection regulatory improvement area as a tjilot esort. A copy of the notice is enclosed for your information.

In sunuaary, the RuleNet program will aBow interested individuals to participate in discussions on See protection rata ==Mng issues through the uno doosaputer networks, such as the World Wide Web. % RuleNet program is in progrees with participant registration, and electronic discussions among participants win occur between January 2 and February 9,1996. Procedures for the program are not clearly deined; however, and the NRC indicates that they can be developed and communicated through implementation of the pilot esort. 1 Discussions with the NRC stafindicate that RuleNet was developed by the NRC in i cos$anction with activities related to the National Performance Review, and will be  : observed by other federal agencies for potential adaptation. In aMitian the NRC . expects to apply this program to other rule ==W- actions should the pGot eBart be saaneesful for the issue of Ane protection.  :) NEI supports the use of computer technologies to improve the regulatory process if it is cost eSootive and adds value. However, we believe that there are a number of signiacant questions that need to be resolved in the RuleNet puot osart. Per example, NitC's disposition of coauaunications annong participante, qualianation of participants, potentialinSuence by specialinterests, conduct ofprivate versus open caucuses among participants, and representation by organisations versus individuals, etc., are aosne of the questions that have not been addressed by the 85 ' i m , r... m w. . , = .c : = .. m . => . , m ,. ...

1 i s

y- '

i

         '                                                           Nadest Strategie Iseems Advisory F==ia*==

j December 1,1996 i Pages a

c==w to

! RalsNet procedens. We intend to pursue discussions with the ! alarid!r the RuleNet program. i Panding darineesion etthe prosses, NE! is agistering to participam in RuhNet on l behaNdtbs ladestry. - ~_ Boonnes See @, ray, app. ?.._ Q # ._ - M

                                                                                                                                                                               ~

{ , we do not encourage direct i,.r' t -C- la RuleNel) ! useasses or other ladustry organisationsf newever, we acto that i -- m perumpsass on unwm - -, == --- ..iby others, and we h*h U es-e=T,=. u_ m$ = . . - - - - 2 -

                                                                                                                                                                                             -n ,         so t                                                                                                                                                                                                                          !

t --- -_ , p won--

                                                                                                                  -_         _,a s                                           - -1so w ' ~~;-- - *Marev                    i j                                                                     g,f_ _,,,, , - -- ..a = :. ' =' -i. ; - "" - *= ---* - ;-- ^ " " "

j i j E

                                                                      -~

E ,, ,,n , i 1 M *** Please contentime (303 789 8088) i ,,0eorge Wu (302 73H006) of the NE1 Mshould you hw say W. s l D 1 e i William H.Ramin WHE/GCW/ sad ! Enclosure i C Y k Fire ProtectionWorking Group l Chainaan,NSSS Ownere Groups i 4 I I f I 1 2 86 j i i _ _ _ _ _ - _ - _ _ _ _ _ _ _ . _ _ _ _ - _ _ _ . _ - - _ - - . - . _ _ ,-}}