ML18184A148
| ML18184A148 | |
| Person / Time | |
|---|---|
| Issue date: | 06/19/2018 |
| From: | Derek Widmayer Advisory Committee on Reactor Safeguards |
| To: | |
| Widmayer D | |
| References | |
| NRC-3782 | |
| Download: ML18184A148 (238) | |
Text
Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION
Title:
Advisory Committee on Reactor Safeguards Future Plant Designs Subcommittee Docket Number: (n/a)
Location: Rockville, Maryland Date: Tuesday, June 19, 2018 Work Order No.: NRC-3782 Pages 1-238 NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers 1323 Rhode Island Avenue, N.W.
Washington, D.C. 20005 (202) 234-4433
1 1
2 3
4 DISCLAIMER 5
6 7 UNITED STATES NUCLEAR REGULATORY COMMISSIONS 8 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 9
10 11 The contents of this transcript of the 12 proceeding of the United States Nuclear Regulatory 13 Commission Advisory Committee on Reactor Safeguards, 14 as reported herein, is a record of the discussions 15 recorded at the meeting.
16 17 This transcript has not been reviewed, 18 corrected, and edited, and it may contain 19 inaccuracies.
20 21 22 23 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com
1 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3 + + + + +
4 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 5 (ACRS) 6 + + + + +
7 FUTURE PLANT DESIGNS SUBCOMMITTEE 8 + + + + +
9 TUESDAY 10 JUNE 19, 2018 11 + + + + +
12 ROCKVILLE, MARYLAND 13 + + + + +
14 The Subcommittee met at the Nuclear 15 Regulatory Commission, Two White Flint North, Room 16 T2B1, 11545 Rockville Pike, at 8:30 a.m., Dennis Bley, 17 Chairman, presiding.
18 COMMITTEE MEMBERS:
19 DENNIS C. BLEY, Chairman 20 RONALD G. BALLINGER, Member 21 CHARLES H. BROWN, Member 22 MARGARET CHU, Member 23 MICHAEL CORRADINI, Member*
24 JOSE MARCH-LEUBA, Member 25 HAROLD RAY, Member*
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
2 1 JOY L. REMPE, Member 2 PETER C. RICCARDELLA, Member*
3 GORDON R. SKILLMAN, Member 4 MATTHEW SUNSERI, Member 5
6 DESIGNATED FEDERAL OFFICIAL:
7 CHRISTIANA LUI 8 ALSO PRESENT:
9 AMIR AFZALI, Southern Nuclear Operating 10 Company 11 GEORGE APOSTOLAKIS, LMP 12 AMY CUBBAGE, NRO 13 KARL FLEMING, LMP 14 BILL RECKLEY, NRO 15 JOHN SEGALA, NRO 16 17 *Present via telephone 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
3 1 C-O-N-T-E-N-T-S 2 Opening Remarks 3 by Dr. Dennis Bley, ACRS . . . . . . . . . 4 4 Staff Introduction 5 by John Segala, NRO . . . . . . . . . . . . 9 6 Draft Guidance Paper -- Modernization of 7 Technical Requirements for Licensing of 8 Advanced Non-Light Water Reactors 9 by George Apostolakis/Karl Fleming . . . . 25 10 Committee Reports 11 SSC Safety Classification . . . . . . . . 100 12 Public Comments . . . . . . . . . . . . . . . . 155 13 Adjourn . . . . . . . . . . . . . . . . . . . . 165 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
4 1 P R O C E E D I N G S 2 (8:33 a.m.)
3 CHAIR BLEY: The meeting will come to 4 order. Good morning.
5 MEMBER MARCH-LEUBA: Good morning.
6 CHAIR BLEY: This is a meeting of the 7 Advisory Committee on Reactor Safeguards Subcommittee 8 on Future Plant Designs. I'm Dennis Bley, Chairman of 9 the Subcommittee.
10 The ACRS Members in attendance or shortly 11 to be in attendance are Dr. Joy Rempe, Charlie Brown, 12 I don't know about Walt, Jose March-Leuba, Dick 13 Skillman, Margaret Chu, Matt Sunseri and Ron 14 Ballinger.
15 ACRS Members Michael Corradini, Pete 16 Riccardella and Harold Ray are attending remotely.
17 That's what all of this shenanigan was about.
18 Christiana Lui of the ACRS staff is the 19 Designated Federal Official for this meeting. The 20 purpose of today's meeting is to hear an introduction 21 to the guidance document entitled, Modernization of 22 Technical Requirements for Licensing of Advanced Non-23 Light Water Reactors.
24 This is an information only meeting. So 25 the Subcommittee will gather information at today's NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
5 1 meeting. The Subcommittee is scheduled to review the 2 final version of the guidance document and associated 3 documents at an October 2018 Subcommittee meeting.
4 Just an aside, when you get to it maybe 5 you'll expand, one of your slides talks about an NEI 6 guidance document which I guess follows on from this.
7 But if you can let us know how that's hooked in it 8 would be appreciated.
9 Full Committee is scheduled to address 10 this matter at the December 2018 full Committee 11 meeting. ACRS was established by statute and is 12 governed by the Federal Advisory Committee Act, FACA.
13 That means that the Committee can only speak through 14 its published letter reports.
15 We hold meetings to gather information to 16 support our deliberations. Interested parties who 17 wish to provide comments can contact our offices 18 requesting time after the Federal Register notice of 19 the meeting is published.
20 We have received no requests to make 21 comments at today's meetings. That said, we also set 22 aside time for extemporaneous comments from members of 23 the public attending of listening to our meetings.
24 Written comments are also welcome.
25 The ACRS section of the U.S. NRC public NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
6 1 website provides our charter, bylaws, letter reports 2 and transcripts of all full and Subcommittee meetings 3 including the slides presented at the meeting.
4 Detailed proceedings for conduct at ACRS meetings was 5 previously published for this meeting in the Federal 6 Register October 4, 2017.
7 The meeting is open to public attendance.
8 As mentioned, time has been allotted on the agenda 9 near the end for comments. Today's meeting is being 10 held with a telephone bridge line allowing 11 participation of public over the phone.
12 Also as mentioned a separate 13 teleconference line has been established to allow 14 participation of three ACRS Members remotely. A 15 transcript of today's meeting is being kept.
16 Therefore we request that the meeting 17 participants on either the bridge line or the 18 teleconference line identity themselves each and every 19 time they speak and to speak with sufficient clarity 20 and volume that they can be readily heard.
21 We request that those participants on the 22 public bridge line keep their phones on mute until 23 they are called on to speak during the public comment 24 period. Participants in the meeting room should use 25 the microphones throughout the meeting room.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
7 1 At this time I ask everyone in the room to 2 please silence all their electronic devices. I remind 3 speakers at the front table to turn on your microphone 4 when you speak and turn it off when you're not 5 speaking just to keep the noise on the line down.
6 We will now proceed with the meeting.
7 Before I introduce others, I just wanted to welcome 8 old friends and colleagues back and everyone who is 9 here to talk today and remind ourselves and some of 10 you that the history of some of what we're going to 11 hear, as far as I know, goes back at least to the late 12 1980s with the modular HTGR application and the NRC's 13 review of that application in NUREG 1338.
14 In 2007, NUREG 1860 originally known as 15 the Technology Neutral Framework was published. A lot 16 of interaction with ACRS on that one and a letter from 17 ACRS generally supporting. But a couple of negative 18 comments on it.
19 In 2010, the DOE NGNP white papers were 20 reviewed by the staff and submitted, including one on 21 licensing basis events. There was a staff assessment 22 of those white papers and we wrote a letter on that 23 staff assessment, made some comments.
24 Pretty much the staff agreed with our 25 comments. But then months later when the final NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
8 1 version of their review was published all mention of 2 this process kind of disappeared, ostensibly because 3 the staff was worried about getting ahead of the 4 Commission on this issue.
5 But that's all part of our history. Then 6 there was NUREG 2150 which one of our speakers today 7 is responsible for, the proposed risk management 8 regulatory framework in 2012.
9 In 2018, we had the staff white paper on 10 functional containment which also included some of 11 these ideas. And Reg Guide 1.232, ARDS, the Advanced 12 Reactor Design Criteria published as guidance for 13 developing principle design criteria was out.
14 And on those, that last one there was also 15 quite a bit of ACRS interaction in the letter. And 16 one side report I'm going to ask you folks about when 17 we get to the defense-in-depth area was in 2016 there 18 was a new kind of NUREG called Knowledge Management-19 0009 on the history and observations of defense-in-20 depth which I believe we wrote a letter on that.
21 And I'm not sure I saw that mentioned in 22 the documents I've reviewed. Anyway, just to bring us 23 up. We've had a long interaction. We've been looking 24 forward to today especially through our discussions on 25 the Advanced Reactor Design Criteria earlier this NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
9 1 year.
2 And almost the whole Committee is either 3 here or on the phone today to follow this. So even 4 though it's an off week it's drawn quite a bit of 5 attention. At this time I would like to --
6 MEMBER REMPE: Before you do that, I've 7 been getting emails from Mike and Pete that are timed 8 at 8:35. They were complaining all they could hear 9 was music. It would be good to --
10 CHAIR BLEY: Please check on it.
11 MS. LUI: It's being addressed right now.
12 MEMBER REMPE: Okay, thank you.
13 CHAIR BLEY: If we get more of those 14 please pass it on to Chris so go ahead with that.
15 (Off record comments) 16 CHAIR BLEY: Thank you. At this time I 17 would like to turn, do you have another comment?
18 MEMBER MARCH-LEUBA: She's not getting the 19 emails. You're not connected are you?
20 CHAIR BLEY: We will now proceed with the 21 meeting. I'm going to call on John Segala, Chief of 22 the Advanced Reactor and Policy Branch Office of NRO 23 to make introductory remarks, John.
24 MR. SEGALA: Okay, thank you. I think you 25 had mentioned just now that we had previously briefed NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
10 1 the Subcommittee on the NRC's vision and strategy and 2 our near term implementation action plan for advanced 3 reactors last year.
4 And then we followed that up with 5 presentations on our Reg Guide 1.232 on principle 6 design criteria for advanced reactors and then 7 functional containment. Industry has been working on 8 this licensing modernization project developing an NEI 9 document which they plan to request the NRC 10 endorsement in a regulatory guide.
11 We are in the process of starting to 12 develop that draft regulatory guide. And so we're 13 looking forward to the Committee's feedback today that 14 we can use to help direct and, you know, incorporate 15 into the development of the draft guide.
16 CHAIR BLEY: So you envision this being 17 something like 0209. You'll endorse either fully or 18 partially --
19 MR. SEGALA: Yes.
20 CHAIR BLEY: -- this paper when it finally 21 comes out. And that's the one we expect will be an 22 NEI document.
23 MR. SEGALA: Yes.
24 CHAIR BLEY: Okay, thanks. And by October 25 when you come back will your review be complete? Is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
11 1 that the expectation?
2 MR. RECKLEY: This is Bill Reckley. At 3 least well enough to support a draft guide.
4 CHAIR BLEY: Okay, thanks. And if you're 5 finished I guess we'll turn it over to Amir Afzali, 6 Afzali, sorry. I can't get all my consonants out.
7 MR. AFZALI: Good enough.
8 CHAIR BLEY: Amir.
9 MR. AFZALI: Good morning, Amir Afzali 10 from Southern Company. I'm the project lead for 11 licensing modernization project. It's a pleasure and 12 we are very excited here to discuss our activities 13 with you.
14 My job was to provide some type of a 15 context on where we are and then pass on all the 16 difficult challenges to George and Karl. But you guys 17 did an excellent job of putting a context on why you 18 are here and what we are doing at this phase.
19 So I'm going to potentially repeat some of 20 the issues that was discussed already or explained 21 already. Next page please. Do I do it or what? How 22 does this work?
23 CHAIR BLEY: Before you pass it on let me, 24 I've just seen this email stream they were talking 25 about, if our Members on the line are there please say NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
12 1 something so we can hear if you can talk to us.
2 MEMBER RAY: This is Harold.
3 CHAIR BLEY: Thank you, Harold.
4 MEMBER RICCARDELLA: This is Pete.
5 CHAIR BLEY: Okay, and, Mike.
6 MEMBER CORRADINI: And Corradini is here.
7 CHAIR BLEY: Okay. We've got you all. So 8 if you want to talk at any time feel free and we 9 should be able to pick you up, thanks.
10 MEMBER CORRADINI: We'll go on mute on our 11 side. We'll keep it to a minimum.
12 CHAIR BLEY: Thank you.
13 MEMBER RICCARDELLA: The same.
14 CHAIR BLEY: Karl or George --
15 MR. AFZALI: Shall I continue? Yes, again 16 here I would do a little bit of introduction then we 17 get some high level comments on our proposal, 18 feasability of our proposal from Dr. George 19 Apostolakis who is one of the Advisory Members of our 20 project.
21 Our other two advisors are former 22 Commissioner Dick Meserve and former Commissioner Jeff 23 Merrifield. And then Karl will, who is leading a team 24 of experts will discuss about the technical basis, our 25 proposal and our technical basis for those proposals.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
13 1 Next page please.
2 So the project fundamentally is trying to 3 develop a foundation for an integrated licensing for 4 advanced reactors. So again, this is not the entire 5 licensing framework. It is the foundation for 6 creating that licensing framework.
7 It's trying to integrate three most 8 important aspects, fundamental and important aspects 9 of licensing which is, which are licensing basis event 10 selection, SSC classification and defense-in-depth 11 evaluation adequacy determination.
12 So those are a fundamental part of 13 starting your licensing process and finishing the 14 licensing. So beginning and the end of your licensing 15 process.
16 MEMBER REMPE: So, excuse me, but just to 17 make sure it's clear in my brain. You're going to do 18 this for folks that are trying to pursue a Part 50 or 19 a 52. You're not trying to limit it to one or the 20 other, right?
21 MR. AFZALI: That's correct. So if you 22 look at the Part 52 or Part 50 the fundamental 23 technical requirements are the same. The processes to 24 go from A to Z is different.
25 But the fundamental technical requirements NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
14 1 are the same. So other than maybe Part 52 requires 2 PRA and Part 50 at this point may not require it.
3 That's debatable. But fundamentally that's a 4 technical requirements.
5 MEMBER REMPE: But when we get to later 6 discussions there's something you're going to need if 7 you're going to tie it to dose if you go with a 52, 8 something associated with some sort of site. Back in 9 the MHTGR days there was EPRI document that we had 10 used to try and tie it to dose.
11 And I didn't see this in the discussion I 12 read. But we can get into that later I assume during 13 Karl's presentation.
14 MR. AFZALI: That's true.
15 MEMBER SKILLMAN: Amir, I feel compelled 16 to follow on Dr. Rempe's question. Was there a bias 17 or a sense or preference that for the newer non-light 18 water reactors Part 50 would be the better path versus 19 Part 52?
20 MR. AFZALI: You're asking me a very 21 difficult question because it's totally a business 22 decision in my opinion. So it just, personal opinion 23 is not LMP's opinion, it is not Southern Company's 24 opinion it's just personal opinion.
25 I would say that a lot of the utilities as NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
15 1 a whole, the people who run it would prefer a Part 52 2 approach. The developers would prefer a Part 50 3 approach just an extremely general statement and 4 personal opinion. So I don't --
5 MEMBER SKILLMAN: I appreciate your 6 candor. Most of us have dealt with both sides of that 7 and understand benefits and challenges associated with 8 one side or the other.
9 But I would think there might in fact be 10 a preference if you're going to have salt cold or lead 11 cold or something that's not a light water reactor 12 given what we've learned and how to do a Part 50 13 versus a Part 52. So it was just a curiosity question 14 and thank you.
15 MR. AFZALI: Thank you. So the last 16 bullet on that slide is very important. We're talking 17 about multiple technology and multiple designs within 18 each technology.
19 Okay. So we are not talking about one 20 technology. We are talking about multiple 21 technologies and multiple designs within each 22 technology.
23 This systematic approach allows us to have 24 technical requirements which are, which provide a 25 balance to all these technologies as a systematic NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
16 1 range is repeatable and doesn't bias one design or one 2 technology.
3 So it, other potential approaches may, if 4 they're not systematic enough, may create differences 5 making different designs simply because of the ad hoc 6 nature. So we're trying to create a coherent path to 7 efficient.
8 But the coherency is the important part of 9 our objective.
10 CHAIR BLEY: Amir, before you go on. Two 11 things. One is a question and one is more of a 12 comment. And I guess I would ask the other presenters 13 this.
14 Your second bullet talked about integrate 15 new advances in risk-informed performance based 16 methods. If there are new advances since 1860 and the 17 DOE white papers maybe you can point those out as we 18 go along. I would find that very interesting.
19 On this last one the coherent path, I 20 recall during the trial application of 1860 they took 21 it back to an existing PWR and one of the people who 22 had been involved in that PWR, the PRA did a lot of 23 that work.
24 But there was a lot of effort in keeping 25 the analysis coherent when you start trying to apply NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
17 1 the criteria that will be talked about later deciding 2 licensing basis events you can break up into smaller 3 and smaller pieces which makes the frequency go down 4 and down and down so you can meet any criteria.
5 So how do you keep this whole thing 6 coherent and group it in such a way that it's a 7 meaningful application? And I would be happy to wait 8 for that until we get to the appropriate parts of the 9 talk today.
10 MR. FLEMING: I will be happy to answer.
11 CHAIR BLEY: Okay. And I know Karl has 12 done that before in other applications. So we would 13 be very interested in hearing that.
14 MEMBER MARCH-LEUBA: I wanted to remind 15 you that you need to turn the phone, the microphone on 16 otherwise you won't be on the record. John Stetkar 17 left me in charge of microphones.
18 MR. FLEMING: I will address your 19 question, Dennis. Thank you.
20 MR. AFZALI: So, as I said at the 21 beginning of this discussion we're going to provide 22 the presentation that familiarizes you with our 23 processes. The three processes we are going to be 24 discussing are selection and evaluation of licensing 25 basis events, SSC classification and defense-in-depth NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
18 1 adequacy determination.
2 You can see I have highlighted the first, 3 we have highlighted the first and the third bullets on 4 there. We believe a very robust conversation around 5 those two topics are going to be fundamental in taking 6 us to the next steps.
7 So although SSC classification is 8 extremely important it is a combination of the first 9 and the third bullet kind of and they are all 10 integrated. So you cannot say which one is most 11 important.
12 But I think the area where we need the 13 most feedback on are the first and the second bullet, 14 okay. Next slide. So just a summary of where we are 15 right now.
16 We are trying to develop a set of 17 procedures which are for the developers and for the 18 NRC endorsement. So when we started it is the how's 19 and the what's.
20 We believe the NRC staff are not in a 21 position to endorse the how's because I think that 22 would be prescriptive on our developers. But the 23 what's are what we are trying to establish.
24 So our white papers include the how's and 25 the what's. The guidance document only includes the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
19 1 what's. So if you are a developer you need to read 2 the white papers from the endorsement point of view as 3 the guidance document.
4 That guidance document is going to be 5 finally issued as an NEI document. I believe, correct 6 me if I'm wrong, 18-04 is the number which is assigned 7 to that particular document. And that would be 8 endorsed through a Reg Guide as previously said by 9 Bill and others.
10 So again, the staff review has been 11 provided in the white papers. We have had at least 12 two new things where we have discussed the 13 requirements on the guidance document, the working 14 sessions.
15 We have kind of modeled our activities 16 similar to what was done for Fukushima response 17 activities where there was a lot of public meetings 18 discussing comments from each side and trying to get 19 the document to the end point as quickly as we can.
20 And the final document, as I said, is a 21 risk-informed performance guidance document.
22 CHAIR BLEY: Before you go on, I 23 understand you want it to be a what document. But 24 you're going to need some at least examples which will 25 to some extent be showing possible how's to do this.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
20 1 Are they going to be part of this guidance 2 or do you envision that as a separate document?
3 MR. AFZALI: So we're going to, I'm going 4 to defer that at a later date to people who understand 5 that juggling the regulatory implications of NRC 6 endorsing a document. Do they endorse a document it 7 gets appendices or not for example or just a document?
8 CHAIR BLEY: It's been done both ways in 9 the past.
10 MR. AFZALI: Yes, so I don't have a full 11 answer to that question yet. But we will make both of 12 them available for the developers and the NRC.
13 CHAIR BLEY: Are you expecting the 14 examples to be in place by October or is that too 15 optimistic?
16 MR. AFZALI: Too optimistic.
17 MEMBER CORRADINI: So this is Corradini.
18 I'm kind of just jumping on with Dennis. I think an 19 example is very important because I found it very hard 20 to follow the what's and then go back to the four 21 white papers to understand the how's without some sort 22 of examples.
23 Maybe I'm just too much of a pragmatic or 24 empirical engineer. But I do think that's very 25 important. And I was going to ask a question. Is the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
21 1 example that you're developing a non-LWR or are you 2 going to go back to the 1860 approach and use an LWR 3 which is more of a known quantity?
4 MR. AFZALI: I'll start and then you 5 finish it. So we are doing, we have done examples.
6 We are doing multiple tabletop exercises. We have 7 completed one already.
8 With that said, I was just talking about 9 how we are going to present that information and when 10 not whether we are going to do it or not. So with 11 that said, Karl, did you want to add something?
12 MR. FLEMING: Yes. I wanted to clarify 13 that we started with the four white papers built on 14 the NGNP white papers and bringing into account more 15 recent information including ACRS comments and so 16 forth.
17 And then we went to the guidance document 18 to abstract down the things that we thought were 19 appropriate for an NRC endorsement. The white papers 20 do have examples.
21 In the PRA and the LBE paper we have 22 examples from the MHTGR and the PRISM in terms of how 23 one works from the probabilistically-derived licensing 24 basis events and deriving, you know, design basis 25 accidents.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
22 1 We also have in the SSE paper in the 2 appendix how the MHTGR came up with functional design 3 criteria safety classification and so forth. The 4 MHTGR applied the steps in the first three white 5 papers with some things that were not available at 6 that time with some exceptions.
7 But they, all the way out to safety 8 classification in the LBEs. So those example are in 9 the white papers and it's our intent that as we 10 finalize the guidance document we will go back and 11 make any adjustments to the white papers to make sure 12 that the white papers are in synch with the guidance 13 document.
14 And as Amir points out, we do plan to do 15 some tabletop exercises to expand the capability, 16 expand the inventory of examples and those are in the 17 planning stages and we'll incorporate those as they're 18 available into these documents.
19 MEMBER CORRADINI: Thank you.
20 MR. AFZALI: So I'm going to finish up 21 with just one statement and then turn it to Dr.
22 Apostolakis. And that statement is that I do realize 23 that there are not many developers coming in and 24 asking the NRC to review their application.
25 But and that sometimes creates this sense NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
23 1 of there are no applications, there is no sense of 2 urgency. Currently there are many, many developers 3 that invested a significant amount of their dollars 4 and the government is doing the same in developing 5 those designs.
6 Developing those designs without a, this 7 is my opinion, developing those designs without a 8 method that allows them to decide on these fundamental 9 questions are going to potentially cause future 10 challenges as they are putting the applications in 11 front of the staff.
12 So if there's no sense of urgency perhaps 13 from the application being in front of the NRC point 14 of view there is a sense of urgency from developers 15 developing a design which ultimately is going to be 16 acceptable to our regulators.
17 With that said, I'm going to turn it over 18 to Dr. Apostolakis.
19 CHAIR BLEY: Before you go ahead there is 20 one other area you speak of and you'll get to it here 21 later. One other use for this work and that's to come 22 up with or support the development of principle design 23 criteria.
24 We've had a year or so of going over where 25 the NRC staff has been maneuvering here and at least NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
24 1 to me as we went through it I think their focus 2 changed from one of having advanced reactor and HTGR 3 and sodium cooled reactor design criteria that would 4 be in a final design criteria to developing their Reg 5 Guide to help people come up with their own design 6 specific criteria.
7 So I had two things I'm interested in.
8 The one I'm pretty sure you'll get to. The other one 9 I hope you'll expand on. One is how you use this 10 process to help support development of those principle 11 design criteria.
12 And the other is if we have Advanced 13 Reactor Design Criteria in the Reg Guide that people 14 can look at to help them, how do we ensure that the 15 principle design criteria are complete, that they 16 don't just look for the ones that have already been 17 identified in a more generic way when they're looking 18 on their own specific plan?
19 If that's something you would rather wait 20 until you get to it that's fine.
21 MR. RECKLEY: I'll address that.
22 MEMBER REMPE: As you address it is the 23 vision that you would avoid any exemptions? I mean is 24 that the vision and is that what you're trying to 25 achieve with this process?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
25 1 MR. AFZALI: The desirable outcome would 2 be getting an application in front of the staff which 3 does not require any exemptions. However, because we 4 haven't actually exercised the activity it's very hard 5 to say that's going to be the case or not.
6 So desired outcome, yes. Whether we 7 believe that's a practical solution I would not be 8 able to make a comment on that. With that, Dr.
9 Apostolakis.
10 DR. APOSTOLAKIS: Excellent, okay. This 11 is something I haven't done before, by the way being 12 on this side of the table.
13 CHAIR BLEY: You know, there are a couple 14 --
15 DR. APOSTOLAKIS: I know.
16 CHAIR BLEY: There are a couple of Members 17 who had they known you were coming this month would 18 have delayed their retirement I'm sure.
19 DR. APOSTOLAKIS: Okay. I'll just make a 20 few fairly high level comments. I have been reviewing 21 the documents and as Amir said I'm a Member of the 22 Advisory Team. We'll come back to that.
23 The first issue I would like to address is 24 the issue of PRA. PRA of course is the foundation or 25 the cornerstone of the approach. And people may raise NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
26 1 questions, you know, how reliable is it and so on.
2 My third bullet is something that I really 3 love. I've noticed over the years that because PRA 4 was the new kid on the block people focused on its 5 limitations, shortcomings and so on and there is an 6 implicit assumption that the existing system is 7 preferred.
8 And I will show you examples that show 9 that the existing system is far from being perfect.
10 And, you know, I've seen IAEA documents and so on.
11 PRA here are the limitations.
12 And there is never a section on the 13 limitations of the traditional system.
14 MEMBER MARCH-LEUBA: Let me be, Dr.
15 Apostolakis, for you and stop you right there. If I 16 understand what we are proposing you are going to do 17 a deterministic analysis, the old analysis to 18 determine, you find your LBEs.
19 And then you're going to weed them out 20 with risk-informed --
21 DR. APOSTOLAKIS: No, that's not my 22 understanding. We start with a PRA defining the 23 licensing basis events.
24 MEMBER MARCH-LEUBA: How can you do a PRA 25 if you don't know what your events are?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
27 1 DR. APOSTOLAKIS: Coming next, next slide.
2 So there is, next slide.
3 MEMBER MARCH-LEUBA: No, no, don't move it 4 yet. Okay, I'll wait for the next slide. You need to 5 show me why your statement makes sense.
6 DR. APOSTOLAKIS: Okay.
7 MEMBER MARCH-LEUBA: Because you made a 8 mention that Charlie wasn't a skeptic. You have found 9 another one.
10 DR. APOSTOLAKIS: And, okay. One 11 interesting development is that there is also a 12 standard from the ASME and ANS on the PRA for advanced 13 non-LWRs and LWRs.
14 I think we're in a situation now that is 15 very similar to what was happening in the early 1970s 16 when the Atomic Energy Commission at that time decided 17 to do a PRA what came to be known as the WASH-1400.
18 Indeed there was no operating experience from, to 19 support the study.
20 They collected failure data for components 21 from all over the world. And of course they used 22 extensively expert judgment especially for human 23 errors.
24 The internal event analysis that WASH-1400 25 did has survived to this day. They did a remarkable NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
28 1 job 45 years or so, 43 years later. For internal 2 events we're still using the same approach.
3 So one of the most influential and 4 consequential PRAs was completed with minimal 5 operating experience at the time. And the next slide 6 we give examples and then we will come to you. Next 7 slide.
8 MEMBER MARCH-LEUBA: I know that I have a 9 clear exception to your statement here. I mean let's 10 go back to the 1970s. Everyone knew, everybody knew 11 that PRA analysis knew that if you survive a large 12 break LOCA for sure you will survive a small break.
13 We don't even need to analyze it. And we 14 are stuck analyzing small breaks after we have 15 operation experience in TMI that the small breaks need 16 to be analyzed.
17 DR. APOSTOLAKIS: No, I disagree. First 18 of all not everybody knew. Everybody assumed that the 19 large break LOCA was a bounding accident. It's not 20 that they knew it.
21 The Reactor Safety Study showed, next, 22 that a small LOCA was a major contributor to it.
23 MEMBER MARCH-LEUBA: Was that performed 24 before or after TMI?
25 DR. APOSTOLAKIS: Before. That was the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
29 1 whole point.
2 MEMBER MARCH-LEUBA: And no action was 3 taken?
4 DR. APOSTOLAKIS: The Commission made a 5 big mistake. Because of the controversy they directed 6 the staff not to use the Reactor Safety Study in any 7 regulatory activities which meant as Norm Rasmussen 8 used to say everybody had 11 blue volumes in their 9 office and nobody read them.
10 Then four years later TMI comes and 11 somebody says, well, gee that looks like it's in the 12 Reactor Safety Study and the study became again legal 13 in the sense that the staff was allowed to use it. So 14 it was a confirmation of what the Reactor Safety Study 15 found.
16 It was four years, five years actually 17 because in 1974 we had the draft report out. So it 18 was the other way around. Transients also the study 19 said are very important.
20 Human errors, it still puzzles me that the 21 traditional system up until that time ignored human 22 performance completely. It was the Reactor Safety 23 Study that said no human errors are important.
24 I still remember when we were doing the 25 Zion/Indian Point PRAs we were in a room in Southern NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
30 1 California and the human error ratio came up and the 2 representative of the utility stood up. He was very 3 agitated and said my operators are trained.
4 They will never do what you say. Nobody 5 will say that today. That Zion/Indian Point PRAs came 6 out four or five years later and showed that 7 earthquakes and fires are among the dominant 8 contributors to risk.
9 And then interestingly enough the 10 Commission in '84 and '88 issued two rules for two 11 reasons. One reason was that there was some operating 12 experience showing that you could have an adverse 13 event and a station blackout but also that PRAs have 14 shown that these were important from the risk 15 perspective.
16 So we had two rules. The reason why I'm 17 saying that is because we can divide the last 40 years 18 or so into two periods. The first years after the 19 Reactor Safety Study and other PRAs like Zion and so 20 on we saw a lot of regulatory activity, new rules, new 21 orders.
22 One of the very first ones was the 23 utilities should make sure that the auxiliary 24 feedwater system is actuated automatically. In some 25 plants it was manually actuated.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
31 1 The Reactor Safety Study said that's not 2 a good idea from the risk perspective. So there was 3 a lot of regulatory activity. So then after the 4 August statement of '95 we started seeing some 5 relaxation of the regulations which creates the wrong 6 impression to many people that PRAs are used to relax 7 the regulations which is absolutely not true.
8 It's because of this separation in time 9 that first when we took care of the regulations and 10 then we started relaxing the regulations. The reactor 11 oversight process provides objectivity through the 12 assessment of the culture of the plant.
13 An example of the extension of the allowed 14 outage times provides flexibility to the licensees.
15 That was the very first risk-informed initially by the 16 way approved by the Agency.
17 And in addition to extending the times it 18 had a major psychological impact. The staff and the 19 industry did not trust each other. So after they 20 allowed, outage time was extended to 14 days I was 21 here talking to the staff wait and see. Those guys on 22 the other side will go to the fourteenth day before 23 they fix anything.
24 Evidence from South Texas Project showed 25 that within five days they had fixed the problems.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
32 1 And that started building trust between staff and the 2 industry that everybody was serious.
3 And of course when it comes to burden 4 reduction and improving safety nothing can compete 5 with risk-informed ISI which I understand all the 6 plants now have implemented.
7 CHAIR BLEY: George, there is one of those 8 examples. There's a lot more examples you could have 9 cited. But there was one that I was personally 10 involved in and interested in.
11 Back to the time of these early PRAs they 12 showed that common cause failure of reactor trip 13 breakers is a lot more likely than people thought 14 because of the way they were forced to disable the 15 shunt trip mechanism on the breakers which makes them 16 much more vulnerable to poor maintenance caused 17 failure.
18 And that was years before the Salem event 19 finally led the staff to let them put the shunt trip 20 back into the reactor trip breaker circuits.
21 MEMBER CORRADINI: So can I make a 22 comment? I think I know where Jose is coming from.
23 But I guess I would argue to Jose, I'm sure he'll 24 respond, is that everybody in their mind does the 25 equivalent of an ill-conceived or inadequate PRA by NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
33 1 even assuming a set of deterministic accidents that we 2 must design against.
3 I think what I heard George is saying that 4 the WASH-1400 regularized the process of engineering 5 thinking so that you came up with insights that you 6 would not have by making assumptions of what are the 7 limits. That's what I hear from George's discussion.
8 MEMBER MARCH-LEUBA: But what I see in 9 reality is that you come out with a list of accidents 10 and then you weed them out based on PRA results.
11 MEMBER CORRADINI: But you don't 12 necessarily, I guess I would disagree, Jose.
13 MEMBER MARCH-LEUBA: You don't?
14 MEMBER CORRADINI: I read the guidance and 15 the starting point is you start off with a list of 16 LBEs. But you don't weed them out. They may grow, 17 they may shrink.
18 They may move relative to each other on 19 terms on frequency and their dose. But you're always 20 iterating on what you think can challenge the system.
21 MEMBER MARCH-LEUBA: Right, which is what 22 they call developing a list of LBEs. What can 23 possibly happen?
24 MEMBER CORRADINI: But I guess, let me 25 just fight back. That's what you would call NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
34 1 developing a systemized approach of risk analysis.
2 MEMBER REMPE: And when I was working at 3 GA we used to laugh about a transient plant design 4 because they would add a system so you could weed out 5 an event and have a lower frequency. And so we would 6 constantly be iterating on what we were analyzing.
7 DR. APOSTOLAKIS: So the point is that if 8 you don't want to do this because the PRA is 9 incomplete I would argue the so-called deterministic 10 approach is incomplete too. How are you going to do 11 it?
12 You're going to get a bunch of guys in a 13 room and say well I think this may happen. Okay, 14 let's make it a design basis accident.
15 MEMBER MARCH-LEUBA: I would call it the 16 Murphy Rule. If it can happen it will happen.
17 DR. APOSTOLAKIS: Sorry.
18 MEMBER MARCH-LEUBA: The Murphy Rule. If 19 it can happen, it will happen.
20 DR. APOSTOLAKIS: Well I'm not so sure 21 about that. That rule has not been approved by 22 Congress.
23 MEMBER MARCH-LEUBA: It hasn't been 24 approved by anybody. But it happens in life.
25 DR. APOSTOLAKIS: But a large LOCA has NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
35 1 never happened.
2 MEMBER MARCH-LEUBA: Right.
3 DR. APOSTOLAKIS: And it's still the 4 cornerstone of regulations. It's very expensive. It 5 affects many other things like the containment spray 6 system and so on.
7 That was a lot of research from Lawrence 8 Livermore National Laboratory showing that you will 9 have a leak before a break. The rule is the same. It 10 was the same as it was in the time of Athens and Rome.
11 That's a major problem of the design basis 12 accidents. They don't evolve.
13 MEMBER MARCH-LEUBA: What is the safety 14 implications of having a LOCA?
15 DR. APOSTOLAKIS: What comes to mind is 16 the automatic actuation of the core spray system when 17 there is no need for it. And it has been observed I 18 understand three or four times.
19 And I know the Agency doesn't worry about 20 that. But the cost, I mean these guys --
21 MEMBER MARCH-LEUBA: We're going back to 22 this. You're thinking cost. You're not thinking 23 safety. So maybe you are --
24 DR. APOSTOLAKIS: Your statement to be 25 said to me that I don't care about safety. My whole NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
36 1 damn career was on safety. I'm sorry. I know a 2 presenter to the ACRS is not supposed to react that 3 way.
4 But, you know, don't tell me I don't care 5 about safety. So the Commission's PRA policy 6 statement in 1995 says a probabilistic approach to 7 regulation enhances and extends the traditional 8 approach by allowing the consideration of a broader 9 set of potential challenges.
10 In this case the potential challenges are 11 the LBEs. Okay, instead of going straight to the 12 design basis accidents we start with a broad set of 13 licensing basis events and then derive the DBAs from 14 there, next.
15 MEMBER SKILLMAN: George.
16 DR. APOSTOLAKIS: Yes.
17 MEMBER SKILLMAN: Let me make, I think 18 you're selling yourself short.
19 DR. APOSTOLAKIS: Okay.
20 MEMBER SKILLMAN: I am in agreement with 21 what you're doing here. But I think what you haven't 22 communicated and what needs to be communicated right 23 up front is that this is driven by a recognition of 24 what are quantitative health objectives and other very 25 high level safety objectives.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
37 1 And if those are held as the, if you will, 2 the guiding limits or the guidelines that develop 3 through a systematic approach how the plant should 4 behave, whatever the plant design might be, then I 5 think the argument that you're projecting really hangs 6 together and it challenges Dr. March-Leuba's 7 challenge.
8 What you haven't said is what this is, is 9 a rigorous systematic approach to make sure that the 10 key health objectives and the key safety objectives 11 are thoroughly implemented.
12 DR. APOSTOLAKIS: Yes.
13 MEMBER SKILLMAN: To me that's the 14 touchstone for this. That's what makes all of this 15 work together. And as long as those safety limits or 16 health objectives are clearly etched and agreed to 17 then a lot of people in diverse places can say I'll 18 argue about how we get there, but I agree where we 19 need to get to.
20 And if we stay focused on that approach 21 then reasonable men and women can say, yes, this is a 22 thorough way for everybody to approach those 23 objectives pretty much the same way. So I think --
24 DR. APOSTOLAKIS: That's a good way of 25 presenting it, yes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
38 1 MEMBER SKILLMAN: I think that's the glue 2 that really makes this come together.
3 DR. APOSTOLAKIS: Yes. The regulatory 4 requirements QHOs that we meet those, yes. Thank you.
5 MEMBER MARCH-LEUBA: So I know Dennis 6 thinks we're wasting time, but we're not. We need to 7 have a discussion and an argument about this approach 8 because otherwise what are we doing here.
9 I would agree with what he said, this 10 method makes a lot of sense in theory. In practice I 11 have never seen a PRA analysis that was complete or 12 accurate --
13 DR. APOSTOLAKIS: Or?
14 MEMBER MARCH-LEUBA: Or accurate. If you 15 look at the basic frequency data that we use for that 16 PRA analysis it's awful. Let me give you an example 17 and I'm not going to make it long but this is 18 interesting.
19 I have four cars in my garage in 20 Knoxville. There are three Toyotas and one Saab 21 convertible. If you get the average of the three car 22 maintenance record they are pretty good.
23 But applying the Toyota liability to my 24 Saab is lunacy I can assure you. So whenever we put 25 the simple data we get the average of the valves out NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
39 1 there and we give it a failure probability of so and 2 so.
3 And now you're using a very specific, in 4 these new plants a very specific, very complex valve 5 that has never been used before and you're saying it's 6 going to behave the same way as these other ones that 7 have been running for 100 years.
8 So my basic problem is I agree 100 percent 9 with what you guys are saying, this implementation.
10 You need to have complete and accurate PRA and I've 11 never seen any.
12 MEMBER CORRADINI: But, Jose, I don't 13 understand why is your engineering judgment any more 14 complete or accurate if that's what you're saying?
15 MEMBER MARCH-LEUBA: What I am saying is 16 you should not be weeding, and you know the event I'm 17 thinking about, the one I've been complaining about.
18 You should not use incomplete PRAs, in my opinion, to 19 weed out particular events.
20 DR. APOSTOLAKIS: This is not the 21 objective of this. And don't forget what we say has 22 to be approved by the staff. So the staff can use 23 arguments like that if they wish if it is appropriate 24 to say, no, here we think you should do that and there 25 would be a debate.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
40 1 So this is not a risk-based approach.
2 It's risk-informed and there is a lot of traditional 3 deterministic. This is a misnomer. Traditional 4 methods that one can apply.
5 But I think the fundamental question is if 6 you are demanding PRA to be accurate and what's the 7 other one --
8 MEMBER MARCH-LEUBA: Complete.
9 DR. APOSTOLAKIS: Complete, I am arguing 10 that the traditional system which is based on judgment 11 is neither accurate nor complete. So you take the two 12 together and you try to come up with something that is 13 fairly complete.
14 That's the whole point. I don't want to 15 start a discussion, well it's not up to me. But I 16 think a discussion on the limitations of PRA without 17 talking about the other system is misdirected.
18 MEMBER MARCH-LEUBA: Let's improve the 19 system. I'm all for improving the other system. Let 20 me put some other numbers on the table.
21 DR. APOSTOLAKIS: Yes.
22 MEMBER MARCH-LEUBA: You have told me the 23 problem, melting fuel reactors on the same day 24 breaching containment, leaking such an amount of 25 hydrogen that would make Hindenburg be proud and then NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
41 1 blowing up the top of the reactor like tv would have 2 been 10-18. Yes, that's what it would say.
3 DR. APOSTOLAKIS: Well is a ridiculous 4 number anyway.
5 MEMBER MARCH-LEUBA: 10-6 times three.
6 That's what you would have said had I asked you ten 7 years ago.
8 DR. APOSTOLAKIS: I don't know what I 9 would have said. I don't know how to answer that.
10 MEMBER RAY: George, George, this is 11 Harold. Would you speak to the quantification of 12 uncertainty in the context that you're talking about?
13 To me that's always been the biggest challenge.
14 DR. APOSTOLAKIS: Well I think Karl will 15 address all this stuff.
16 MR. FLEMING: I just want to make a few 17 comments. No competent PRA engineer would have come 18 up with a 10 to the minus, you know, low number for 19 what happened at Fukushima.
20 In fact there was evidence available even 21 in the country of Japan that indicated the likelihood 22 of a very, very large tsunami was in the 10-2, 10-3 23 range. And given --
24 MEMBER MARCH-LEUBA: What was the 25 published number for that much frequency?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
42 1 MR. FLEMING: Given the fact that the 2 switchgear and the diesel generators were located in 3 the most prone area of the plant the core damage, you 4 know, the conditional core damage probability by any 5 competent engineer would have been close to one. So 6 it wasn't 10-18 event.
7 MEMBER MARCH-LEUBA: And what was the 8 published number?
9 DR. APOSTOLAKIS: And that was an abuse of 10 PRA. And that's why we have a competent staff that 11 would never let it fly. This is not what Amir and 12 Karl and decide here. We have a regulatory staff.
13 MEMBER MARCH-LEUBA: This is my complaint.
14 The objective you didn't like complete. I call it 15 uncertainties of omission. What did you forget to 16 take into account in your analysis?
17 DR. APOSTOLAKIS: Well and again, this is 18 a limitation of the state of knowledge, not a 19 limitation of PRA. If you are going to omit something 20 you're going to omit it in the traditional system too 21 because if you knew about it you would put it in the 22 PRA.
23 MEMBER MARCH-LEUBA: I think I put my 24 concerns on the record and I will continue to do so.
25 DR. APOSTOLAKIS: Okay. Now I was very NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
43 1 glad to see these especially the first line develop an 2 Agency wide process and organizational tools to expand 3 the system of the use of quantitative risk assessment.
4 This is what we tried to do with NUREG 2150 four years 5 ago.
6 What we're doing is consistent with the 7 staff's recommendation.
8 CHAIR BLEY: By the way, we're on Slide 9 Number 11 for people listening in.
10 DR. APOSTOLAKIS: Yes, Slide 11. The 11 Chairman mentioned NUREG 1860 that the Committee 12 reviewed it and so on. And it's interesting to recall 13 that the Committee at that time agreed that the idea 14 of licensing-basis events is a good one.
15 And then the second bullet that if you do 16 that it reduces the risk that licensing-basis 17 requirements will divert attention from events of real 18 safety significance. Joy, you want, you raised your 19 hand now?
20 MEMBER REMPE: No.
21 DR. APOSTOLAKIS: Okay, next.
22 MEMBER REMPE: It goes really high when I 23 raise it and I just interrupt usually.
24 DR. APOSTOLAKIS: As Amir mentioned the 25 former Chairman Meserve, Commission Merrifield I are NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
44 1 members of the Advisory Group and we wrote a letter in 2 February to Mr. Kuczynski, the CEO of Southern 3 Nuclear. And you have the letter.
4 These are excerpts. Now of course we said 5 that we are not in a position to comment on the 6 technological survey.
7 CHAIR BLEY: Will you say something more 8 about that? I read your whole letter. I liked, you 9 told a good story and then you get at the end and said 10 we're not in a position to comment on the technical 11 adequacy. What did you mean?
12 DR. APOSTOLAKIS: Well we are not. We are 13 not because remember this is me, Merrifield and 14 Meserve. And you go down to the technical details.
15 It was not just possible.
16 By my reviewing the documents and I go 17 into technical details. But the group could not do 18 that.
19 MEMBER CORRADINI: So, George, I 20 understood that to mean that you're in favor of the 21 spirit of it but the details have yet to be analyzed 22 completely?
23 DR. APOSTOLAKIS: This is the group's 24 position, yes.
25 MEMBER CORRADINI: Okay.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
45 1 DR. APOSTOLAKIS: Okay. And then we 2 believe that this, that guidance can be developed 3 based on these documents that can be endorsed by the 4 NRC. And the last one I think.
5 And defense-in-depth, this Committee wrote 6 a letter back in 1999 stating that one of the major 7 drawbacks of defense-in-depth, of the principle is 8 that we don't know, there is no guidance how much 9 defense-in-depth is enough.
10 We also praised it that it has worked very 11 well and the plants are safe and so on. But there are 12 some shortcomings. So the methods that Mr. Fleming 13 will present that we claim can be used to decide that 14 the amount of defense-in-depth in a particular design 15 is sufficient are attempting, these methods are 16 attempting to answer that question, how much defense-17 in-depth is enough?
18 So again, you will see the group does not 19 explicitly say this is it. It says we are approving 20 it in principle.
21 But it's interesting in the middle of the 22 second bullet like all risk-informed tools it could 23 also result in the identification of areas where 24 additional requirements are necessary. Okay. And I 25 think I'm done, right.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
46 1 MEMBER REMPE: So now I would like to 2 interrupt. When I look at this document and it says 3 hey, we want to reduce some of the uncertainty with 4 the regulatory process and if we also, what it doesn't 5 sometimes mention is the Commission at one point said 6 the current fleet is safe enough.
7 So if you believe that and you do this 8 whole process in some ways you're going to be 9 identifying new requirements because you're looking at 10 beyond design basis events and having some sort of 11 dose limit they have to meet.
12 But you know, and I think this is where 13 you're coming from, is that the designers if they want 14 to have an economic plan are going and reduce some of 15 the margin. And that document that we were asked to 16 review has a lot of statements about adequate margin 17 is preserved.
18 But I never saw any sort of hard number 19 for what is adequate margin. And I think that if you 20 want to reduce uncertainty in the regulatory process 21 that's going to be a big area of contention on what's 22 adequate margin.
23 And those discussions are going to be 24 extensive when someone goes through them. So maybe 25 the examples will help clarify what that is. But I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
47 1 was kind of wondering what's adequate margin when I 2 was looking at it.
3 And did your Advisory Group discuss that 4 at all and that, that might be a pitfall with this 5 process?
6 DR. APOSTOLAKIS: No, no.
7 CHAIR BLEY: I do have one discomfort with 8 your first bullet up here, George. Significantly 9 reduce risks associated with many of the advanced 10 reactor designs.
11 Jose cited roughly numbers that came from 12 PRAs that weren't plant specific with plant specific 13 data and plant specific external events analysis. We 14 don't have any plant specific PRAs with plant specific 15 operating experience and plant specific external 16 events analysis for any of the advanced reactor 17 designs.
18 So the conclusion that the risks are 19 significantly reduced with them is a little troubling 20 for me and I wonder how you come to it. I get the 21 concept why you're pushing this way.
22 DR. APOSTOLAKIS: Well, maybe that could 23 have been phrased better, you're right. But there is 24 a general feeling that the new advanced reactors 25 employing passive safety systems and so on, as the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
48 1 Commission has said, are expected to be safer.
2 CHAIR BLEY: They are expected to be. But 3 we haven't really placed one at a site and looked at 4 all the external evidence that could affect that.
5 DR. APOSTOLAKIS: That's correct. That 6 statement should have been stated better.
7 MEMBER CORRADINI: So if I might follow 8 up. So you don't think the limits of experience with 9 Fermi 1 and Fort Saint Vrain gives us no insight, 10 Dennis or George?
11 I'm trying to understand what is necessary 12 to give us confidence in the statement changing it 13 from expected to at least is shown by some analysis 14 and limited experience.
15 CHAIR BLEY: Well for me, Mike, since you 16 addressed it partly to me, it's really the external 17 events side of it at a specific site. The local 18 specific cooling water systems it should have.
19 And that's the main thing. And those 20 plants you cited didn't really have full scope PRAs 21 done. So we don't really have the risks.
22 They ran for their time, did okay, had, 23 some of them had quite a few different kinds of 24 problems but nothing that resulted in a big accident 25 of any sort. But they didn't have that long a run NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
49 1 either.
2 So I don't gain a lot of confidence from 3 that. But it's really putting them at a specific site 4 and looking for all of the things that could happen 5 there.
6 And for some of them that run on very low 7 thermal hydraulic margins the effects of aging or 8 fouling of different sorts haven't been included into 9 the design cert PRAs that have been done. So that's 10 another area that leaves me not, it's just not proved 11 in any sense yet for me.
12 MEMBER CORRADINI: Okay. If I might, I 13 guess I want to hear George's view on this. But the 14 way I interpreted what you said, Dennis, is these are 15 not technology specific issues.
16 These are almost technology neutral issues 17 relative to siting, external hazards, ultimate heat 18 sink needs. Am I understanding correctly?
19 CHAIR BLEY: The issues are general. The 20 effects are very plant specific.
21 MEMBER CORRADINI: Right, okay. You're 22 right. George, I'm curious what we can you learn from 23 the past limited experience or is yet to be proven by 24 some sort of demonstration?
25 DR. APOSTOLAKIS: No. Past experience for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
50 1 sure will be included in the guidance and the, 2 whatever other documents are produced. But I think 3 it's also important to bear in mind this is a 4 technology neutral approach.
5 Once you have a specific technology it 6 will go down and develop regulatory guides or whatever 7 else to have technology specific requirements. So I 8 have no doubt and in fact already Karl in the white 9 papers is using examples from MHTGR and so on.
10 So all this experience will be 11 incorporated in whatever documents are produced.
12 Anyway, my final, I'm sorry --
13 MR. FLEMING: I was going to say given I'm 14 sort of a glass is half full type of personality, when 15 I read George's first bullet what I interpreted that 16 is that given the expectations for enhanced safety 17 coming out of the Commission's policy statement if 18 that's true then this, then you know there's a concern 19 that you could end up with enhanced conservatism.
20 That's what I read it to be.
21 DR. APOSTOLAKIS: I think it could have 22 been stated.
23 CHAIR BLEY: I couldn't object to that.
24 It was pointed out that I might be getting tense about 25 the time and I was. But we're almost halfway through NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
51 1 the real slides you're using not counting the extra 2 ones you have at the back and we're not quite halfway 3 through the time you had. So I think we're on track.
4 There's a lot of detail we would love to 5 dig into. And I think we should have probably 6 scheduled this for a longer meeting but we didn't.
7 Maybe the October meeting will be a full day or 8 something because we'll have a lot of detail to go 9 through then.
10 But somebody over here started to say 11 something.
12 MEMBER BROWN: I just have held my tongue 13 since George has announced I'm the resident skeptic 14 for the most part. And I would just let you know that 15 I am not the skeptic relative to the use of PRA to 16 identify those scenarios that you need to protect 17 against.
18 But it's a very, very useful even the 19 iconic naval nuclear program embraced the PRA approach 20 many, many, many years ago to and not at the 21 quantitative side but at the qualitative side of 22 identifying cut sets and other type things that would 23 identify and help you with that.
24 My difficulty is when we apply this now to 25 two things. I keep, my area is obviously I&C, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
52 1 instrumentation and control. And the latest drive on 2 that is to now create a risk-informed performance-3 based regulatory framework.
4 I'm still trying to figure out what a 5 risk-informed framework is for designing and building 6 I&C and what do you mean by performance-based. To me 7 you build it and if it trips the plant when you want 8 it to then you've got a performance-based system.
9 Putting that aside talking, somebody 10 brought up the notion of enhanced safety. Not just 11 say they are significantly reduced risks. As I've 12 looked at each of the ones when we've been in these 13 advanced reactor designs I still have not seen a 14 really good layout of why is a sodium fast reactor 15 necessarily enhanced safe.
16 I'm familiar with the earlier sodium plant 17 that was built and operated as a submarine which was 18 unfortunately thrown over the side and sunk before we 19 applied it to any other ones. I'm just looking for 20 how do we say they are enhanced safety?
21 Where is that list of enhanced things that 22 make them better and we know we can design to ensure 23 those enhancements are there? It's been very 24 difficult for me to grasp that. That's all I have.
25 MR. AFZALI: I apologize. I just want to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
53 1 make sure we are not representing any of the 2 developers here. So I think that this is a very 3 important conversation.
4 But I think to be fair to our developers 5 and have a full conversation which is meaningful 6 conversation we need a representative of the 7 developer.
8 MEMBER BROWN: I don't disagree with that.
9 I'm just, I'm looking forward. I'm trying to go 10 forward and if we're going to go down this path and 11 use it we just somehow we need to get over that bridge 12 as well, that's all.
13 DR. APOSTOLAKIS: I think if you have a 14 specific design and you implement this approach with 15 LBEs one measure of these reactors being safer is the 16 distance between the end point of the accident 17 sequences and the regulatory requirement.
18 That's a margin, which by the way in 1860 19 as I recall when they applied it to a PWR it did not 20 pass the criteria, the existing reactor. And you 21 probably remember Dr. Powers for years saying that we 22 don't know that the existing fleet meets the QHOs 23 because we're not doing level three PRAs.
24 Now you're first comment again, there are 25 certain things that are not in the PRA. Digital I&C, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
54 1 safety culture, you can argue, you know, about the 2 culture affects the data. But basically it's not in 3 the PRA.
4 And that's why you need a risk-informed 5 approach where you scratch your head and say now what 6 can go wrong here. What is happening? And you, if 7 it's a deterministic approach like I&C you do that.
8 If risk analysis can provide some insight 9 you do that. So it's a combination. So I don't know 10 what they're going to do with the risk informing I&C.
11 I think it's a very tough job.
12 MEMBER BROWN: I would agree with that.
13 But that's, if you look at the Commission direction 14 they were tasked with development.
15 DR. APOSTOLAKIS: Actually they showed 16 risk informed --
17 MEMBER BROWN: We're going to talk about 18 that tomorrow.
19 DR. APOSTOLAKIS: -- security should be 20 risk informed.
21 MEMBER BROWN: That's all I had.
22 DR. APOSTOLAKIS: Of the I&C. Security is 23 really, well I'm not on the Commission anymore. It's 24 terrible. It imposes such unnecessary burden on the 25 licensees that, well the problem is we don't know how NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
55 1 to do it.
2 MR. AFZALI: Just before comments start, 3 I'm glad I brought Dr. Apostolakis to talk about it.
4 DR. APOSTOLAKIS: I retain my calm 5 especially when I address Jose.
6 MR. AFZALI: Before we go to the --
7 DR. APOSTOLAKIS: I'm sorry I blew up 8 earlier, okay.
9 MEMBER MARCH-LEUBA: I'm used to it. Most 10 people go 14 to 1.
11 DR. APOSTOLAKIS: Yes, but you told me I 12 don't care about safety.
13 MEMBER MARCH-LEUBA: But, no, I did not 14 say that. I said that the consequences of applying 15 this will reduce safety. But to probably acceptable 16 levels.
17 But this is a good discussion to have and 18 it has to be had not by itself the nuclear industry.
19 But it has to be done in a competitive, adversarial 20 relationship so people bring up what's wrong.
21 And if we don't have those arguments and 22 you have those discussions we will not come back with 23 a good product.
24 DR. APOSTOLAKIS: That's true.
25 MR. AFZALI: I just want to add something NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
56 1 here that our objective is make sure that we achieve 2 adequate safety. But adequate is nobody defined it, 3 right.
4 So we are trying to at least get us closer 5 to that conversation so we know what adequate is. The 6 second thing is fundamentally I am surprised there are 7 people who push very hard on operating experience.
8 They ignore operating experience.
9 So what operating experience tells you 10 over the years risk-informed performance-based have 11 improved safety. Forget about the costs. I'm not 12 talking about the costs.
13 I'm saying the operating experience that 14 everybody rely on says risk informed performance-based 15 improved.
16 MEMBER MARCH-LEUBA: It has in the past.
17 MR. AFZALI: And fundamentally ignoring 18 that proposition, that experience it's kind of, 19 doesn't seem really justified without any, I know you 20 have a different opinion. I'm just --
21 MEMBER MARCH-LEUBA: No, I want to say 22 that I have an equal opinion. I agree the risk-23 informed analysis have created significant increases 24 in safety. Unfortunately that's being used as an 25 excuse to reduce the safety of the future reactors.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
57 1 DR. APOSTOLAKIS: No, no.
2 CHAIR BLEY: Let's go on.
3 DR. APOSTOLAKIS: There are too many 4 stakeholders in this. You being one, the staff being 5 another one, the industry, the international 6 community.
7 If there is any of abuse anywhere it will 8 be pointed out. It's not that three of us decide and 9 go this way. I mean everything comes before this 10 Committee and the Commission and NRR, NRO.
11 There are too many, at 10-19 that you 12 mentioned earlier we would be dead on arrival. Nobody 13 would believe that.
14 MEMBER MARCH-LEUBA: For a single reactor.
15 DR. APOSTOLAKIS: I think Karl has a lot 16 to present and we are taking a lot of his time.
17 CHAIR BLEY: Let's go ahead. Karl, let's 18 do your first set of slides and then we'll take our 19 break. I think this could be the one that bogs us 20 down even more.
21 MR. FLEMING: Okay, thanks. The first 22 thing I want to say as an overview is that as we try 23 to convey a description of our proposed approach here 24 we're not taking anything away from the traditional 25 engineering and judgmental processes that have gone NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
58 1 into licensing.
2 We're expanding the role of PRA to inform 3 the judgments that we're trying to make. I also want 4 to pick up on a thought here is that if I can 5 characterize the traditional approach to describing 6 DBAs, design-basis accidents as a set of judgments to 7 come up with limiting, bounding scenarios for the 8 Chapter 15 analysis that would be appropriate to, you 9 know, form the basis for the licensing process and 10 form a basis for safety related SSEs.
11 Well one says bounding, okay, I have a 12 bounding event. The next question ought to be 13 bounding with respect to what? What was considered 14 when you declared the large break LOCA bounding?
15 And what we're trying to do with LBE 16 aspects, LB selection aspects of this approach is that 17 we're trying to make use of one of the great 18 capabilities of PRA in providing a systematic process 19 for enumerating a very large set of scenarios from 20 which to select the bounding event.
21 So that's what we're trying to do. We're 22 trying to supplant the judgment that this is a 23 bounding event without maybe skipping the step of 24 figuring out what are the possibilities before you 25 consider it bounding.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
59 1 The second part is that in the PRA world 2 we like to start with a realistic assessment of the 3 behavior of the plant on these events. And that's as 4 opposed to necessarily a conservative analysis.
5 And that's a big burden in effect because 6 to do a realistic assessment it requires you to 7 capture and understand all the phenomena going on, you 8 know, in the scenario.
9 And it's interesting to note, I would like 10 to point out one of the things we learned about the 11 large break LOCA is that after all the work was done 12 on that bounding event on the large break LOCA it took 13 an event at Barseback to point out there was an 14 important phenomenon that was not in fact bounded by 15 the traditional deterministic safety analysis, the 16 sump plugging issue and so forth.
17 So in the PRA we try to capture the 18 evidence that we have including the evidence from the 19 other reactors. In this slide this is, this slide in 20 a snapshot tries to identify the key points of our 21 risk-informed approach that we're trying to introduce.
22 And not to take anything away from the 23 deterministic method but to help the deterministic 24 judgments form a complete risk-informed process. And 25 when we say risk-informed we don't mean risk-based.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
60 1 We're just trying to use what the 2 capabilities of PRA are to inform the decisions. Also 3 supplemented with deterministic approaches such as our 4 approach to defense-in-depth which we hope to spend 5 some time on.
6 The way the process works and this is, 7 we're not weeding anything out, we use a design 8 specific PRA to define what we call AOOs, anticipated 9 operations occurrences, DBEs and BDBEs.
10 Now to answer one of Dennis' questions, 11 when you put these LBEs together we group the 12 sequences that are modeled in the PRA based on 13 similarity of challenge to the plant and initiating 14 event type challenges, similarity of a plant response 15 to the events.
16 And if we happen to have a release 17 similarity of the mechanistic source terms. So it's 18 a grouping of event sequences. And the grouping 19 process is intended to eliminate the abuses by 20 subdividing sequences and putting them further and 21 further down the sequences.
22 So that's the first process. Now we, the 23 AOOs, DBEs, BDBEs are evaluated against a frequency 24 consequence target which I'll get to in the next 25 couple of slides.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
61 1 It's not a design requirement, I'm sorry, 2 it's not a regulatory requirement. But it's a design 3 objective to control the risks across the whole 4 accident spectrum or the event spectrum from the 5 anticipated events that are high frequency events, low 6 consequence events to the lower frequency beyond 7 design basis events.
8 MEMBER CORRADINI: Karl, can I break in?
9 Maybe you're going to get to it later and you can 10 postpone this.
11 I'm still trying to understand how you 12 logically bundle individual sequences that then move 13 it, I'll just put it in graphable terms, moves it in 14 the y axis to higher frequencies or in the x axis to 15 higher or lower doses. I'm still unclear about the 16 logic of bundling them.
17 MR. FLEMING: Well I guess the --
18 MEMBER CORRADINI: And if you're going to 19 do this later I'll wait.
20 MR. FLEMING: Okay, well I think I tried 21 to address that earlier but maybe didn't communicate 22 well. We group together similar sequences. When the 23 sequences come out of the PRA we group them based on 24 the, you know, the similarity of the plant response.
25 For example, you know, if you have a, you NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
62 1 may have a turbine trip-induced transient and you may 2 have a feedwater reduction transient and a loss of 3 condenser vacuum transient that all basically create 4 the same set of sequences.
5 So we would organize those and we would 6 sum those up especially if they have the same end 7 state and the same plant response. So we don't permit 8 you to subdivide sequences that basically look 9 identical with different flavors of initiating events 10 for example. So we don't --
11 MEMBER CORRADINI: So what you're saying 12 is it's not just necessarily the initiator. I just 13 don't lump all the station blackouts together. I lump 14 or I bundle things relative to how they involve the 15 accident analysis.
16 MR. FLEMING: Absolutely, absolutely. So 17 we don't, we just don't subdivide based on --
18 MEMBER CORRADINI: I'm sorry, I didn't 19 mean to interrupt you. I'm sorry.
20 MR. FLEMING: Yes, we don't permit the 21 abuses of just subdividing the same sequencing to 22 smaller sequences. And there is a lot of reasons for 23 doing this because we never standardized the level of 24 detail of a PRA.
25 Some people develop more detailed entries NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
63 1 and some people have more simplified entries. That's 2 never been standardized. So I think the white papers 3 tried to describe the criteria for if, the whole plant 4 response has to be similar for it to be given to the 5 same LBE.
6 MEMBER CORRADINI: So let me ask my 7 question this way and then I'll stop because I'm sure 8 we're behind. Dennis is watching the clock. So I 9 think I understand what you're saying in the y axis 10 and how I would bundle them with similar initiators or 11 damage basis.
12 But you're also, the way I heard you say 13 that is how source term, whatever the source term is 14 that would be released to the environment you're 15 looking to have a common source term or a range of 16 source terms? I'm more interested in --
17 MR. FLEMING: A common source term. If it 18 has a different source term we would break it out as 19 a separate LBE. And that would indicate it was a 20 different plant response.
21 CHAIR BLEY: And that implies you're 22 developing essentially scenario specific mechanistic 23 source terms?
24 MR. FLEMING: Absolutely, absolutely.
25 MEMBER REMPE: But in that grouping NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
64 1 there's probably some that are larger and typically 2 people pick the bounding one for that grouping for the 3 source term, right?
4 MR. FLEMING: Well we try to avoid putting 5 dissimilar sequences with different source terms.
6 MEMBER REMPE: But you might, from 7 experience you might have, I don't know, four or five 8 different ones and you'll say well this one is the 9 bounding one, you would pick that. Just to clarify 10 instead of saying the characteristic one.
11 MR. FLEMING: Yes, if we decide to put 12 somewhat dissimilar sequences in the same LBE we would 13 be use the bounding one.
14 MEMBER MARCH-LEUBA: Mike, you always 15 interrupt me in trying to say, explain what I said.
16 Let me explain to you what he's saying.
17 If I take a large break LOCA and now I 18 subdivide it into two events large break LOCA at the 19 left side of the plant and large break LOCA of the 20 right side of the plant. Suddenly the frequency of 21 large break LOCA is half.
22 MR. FLEMING: No, in our approach we would 23 combine those --
24 MEMBER MARCH-LEUBA: I mean exactly. So 25 to prevent that subdivision of left side and right NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
65 1 side of the plant which makes absolutely no sense, you 2 combine them to get the maximum frequency possible.
3 Now you have to do it right. We're giving you the, 4 but if you took that out you guys do it right.
5 MR. FLEMING: Yes, that's the intent is we 6 don't subdivide arbitrarily just to get lower 7 frequencies. If we have different, we only subdivide 8 to get different consequences.
9 DR. APOSTOLAKIS: That has been an issue 10 for a long time since Faulkner published his curve 11 what is a sequence.
12 MR. FLEMING: Now the other thing that is 13 a little bit different from the way PRAs have done for 14 light water reactors is that the frequency consequence 15 target and I'm going to show you what the target is in 16 a second.
17 The frequency consequence target is used 18 to look at the frequencies and consequences of 19 individual LBEs. And that's because of the 20 application where you want to select from these LBEs 21 the design basis accidents.
22 And we have a set of rules that we use so 23 that people would end up with a consistent and 24 reproducible set of DBAs given the same input set of 25 LBEs.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
66 1 MEMBER CORRADINI: So may I break in one 2 last time because something you mentioned in all of 3 this that I want to get clear. You said that to do 4 this you need a mechanistic source term.
5 MR. FLEMING: Yes.
6 MEMBER CORRADINI: So is the alternative 7 source term in current light water reactors 8 mechanistic?
9 MR. FLEMING: Well I don't know that much 10 about the alternative LWR source term. But our non-11 light water reactor PRA standard has requirements for 12 what's mechanistic.
13 So the answer to your question what is 14 mechanistic, it's to meet the requirements in the 15 standard. And it's also to suggest that we're not 16 just going to arbitrarily use the equivalent of a TID-17 14844 source term.
18 We're going to try to make it scenario 19 specific and capture the mechanisms that are important 20 to calculating the releases.
21 MEMBER CORRADINI: So because, so here's 22 where my question comes and I'm still on the x axis.
23 MR. FLEMING: Right.
24 MEMBER CORRADINI: I can't deal with the 25 y axis in my mind. I will leave it to you guys that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
67 1 understand the PRAs. But a mechanistic source term 2 implies analysis of experiments.
3 Experiments imply I've done fuels testing.
4 That tells me there's a whole range of evaluation of 5 or validation of fuels testing that either is going to 6 be done or has been done that can arrive at a 7 mechanistic source term since after decades with the 8 light water reactor I'm still sitting with an 9 alternative source term that doesn't strike me as 10 mechanistic.
11 MR. FLEMING: Well the goal is to have a 12 mechanistic source term that captures the evidence 13 that we have to back up the mechanistic source term.
14 And that means it's done with uncertainty treatment.
15 MEMBER CORRADINI: I'm just, where I'm 16 coming from is I think the burden of proof for a 17 mechanistic source term is much higher than the 18 alternative source term we currently have and/or 19 you're going to have a very wide range of uncertainty.
20 MR. FLEMING: Yes. Well the approach does 21 not prescribe a reduced level of uncertainty 22 mechanistic source term as you're describing. But it 23 does call for a reasonable capture of the state of 24 knowledge about what's behind the mechanistic source 25 term.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
68 1 We were showing some examples to the staff 2 in a little training session we had yesterday. And if 3 you look at how the MHTGR exercise was done in some of 4 their mechanistic source terms the uncertainty was 5 three to four orders of magnitude.
6 MEMBER CORRADINI: Okay, all right.
7 MR. FLEMING: But in those cases it didn't 8 matter because they were still five decades away from 9 the frequency consequence curve. So you can have a 10 very, very large uncertainty. But it doesn't 11 necessarily matter.
12 MEMBER CORRADINI: Okay, thank you for 13 helping me there.
14 MR. FLEMING: Now we also --
15 MEMBER REMPE: Also when calculating the 16 source term is the intent that whether it's an AOO or 17 a DBE they should always do it for 30 days --
18 MR. FLEMING: Yes.
19 MEMBER REMPE: -- a day or, 30 days for 20 all of them?
21 MR. FLEMING: It's a 30 day EAB dose 22 calculated EAB for 30 days.
23 MEMBER REMPE: Okay. And we'll get to --
24 MR. FLEMING: TEDE, TEDE dose.
25 MEMBER REMPE: Later, but some of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
69 1 regulatory requirements like 10 CFR 20 has a two hour 2 limit. You've picked an annual 100 millirem limit.
3 The two hour limit could be more restrictive.
4 So if you do this process are you going to 5 meet all the regulatory requirements because again, 6 you may not have covered all of it?
7 MR. FLEMING: Again, the purpose of the 8 frequency consequence target is to evaluate the risk 9 significance of LBEs. That's all it's for. It's not 10 to meet regulatory requirements.
11 We'll still have to meet the regulatory 12 requirements that the various regulations require. So 13 if there is a two hour dose calculation for something.
14 So the frequency consequence target is simply to look 15 at the risk significance of individual LBEs.
16 And we wanted to have a uniform 17 consequence metric so that we could have a consistent 18 comparison, we could compare AOOs, DBEs and BDBEs on 19 the same graph. It's not a regulatory requirement 20 application.
21 This is something that we throw in, in 22 addition as a tool to help us select the design basis 23 events. Now as we get down --
24 MEMBER REMPE: I think it would be good to 25 clarify that in your document that even though they NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
70 1 may be well below this boundary on your plot it 2 doesn't guarantee you've met all regulatory 3 requirements --
4 MR. FLEMING: It's not intended to.
5 MEMBER REMPE: Yes, it wasn't obvious to 6 me because again, back in the old days of MHTGR we 7 took the two hour limit and made that the AOO boundary 8 instead of the annual limit. So you've got a much 9 less restrictive boundary.
10 MR. FLEMING: Yes. One thing we clarified 11 in this project, I think there was some confusion in 12 the NGNP project because in some documents the curve 13 was described as regulatory criteria, top level 14 regulatory criteria.
15 And early in this project we came to an 16 understanding with the staff that, no, this is a 17 design target. And we used the term target to be 18 explicit as not use the word requirement.
19 The regulatory requirements are still 20 expected to be met, whatever they are. As we look at 21 the DBEs and what we call high consequence BDBEs to 22 determine what we call the required safety functions, 23 and that's the term for our approach here and we're 24 working with the staff to avoid some of the 25 terminology issues that we have because sometimes we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
71 1 use similar words or different words to mean the same 2 thing.
3 CHAIR BLEY: Karl --
4 MR. FLEMING: The required safety 5 functions are the functions that we have to, we 6 determine are necessary to keep the DBEs and the high 7 consequence BDBEs inside the frequency consequence 8 target.
9 Those are required safety functions and 10 those are, that's a tool we use to come up with our 11 safety related SSEs.
12 CHAIR BLEY: Karl, you slid past one 13 bullet there at least for me.
14 MR. FLEMING: Yes, the collective thing.
15 CHAIR BLEY: Yes, the collective one. If, 16 as I read the material you're essentially, you've got 17 a PRA. You have a risk curve from the PRA. And it 18 looks like for the collective risk you're picking 19 specific points off of the risk curve essentially and 20 comparing them to a specific criteria.
21 MR. FLEMING: Right.
22 CHAIR BLEY: There's no overall look at 23 the risk curve to look at a collective acceptability.
24 MR. FLEMING: Well for all of our 25 collective criteria and I think we may have a slide on NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
72 1 that. But for the collective criteria --
2 CHAIR BLEY: I'll be happy to wait for 3 that if you have a slide on it.
4 MR. FLEMING: Yes, we do integrate against 5 all the LBEs. But we have one particular criteria 6 that's intended to capture the risks of the AOOs.
7 And we use the 10 CFR 20 limit now in the 8 terms of an aggregated measure. And then we use the 9 two, the QHOs from the safety goals which are really 10 only going to be exercised in a significant way for 11 lower frequency.
12 CHAIR BLEY: So the collective are some 13 kind of mean value against a criteria?
14 MR. FLEMING: Yes, that's right.
15 DR. APOSTOLAKIS: Can I make a comment?
16 MR. FLEMING: Or exceedance frequency, go 17 ahead.
18 DR. APOSTOLAKIS: I think we need to 19 clarify what the F-C curve means. Certainly 50.34 is 20 a regulatory requirement as I understand it, right.
21 So if a sequence goes to the right of the curve it is 22 unacceptable.
23 The designer has to do something about it.
24 If it's to the left of the curve Joy is right. That 25 doesn't mean it's acceptable. There may be other NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
73 1 things that have to be satisfied.
2 So I think that understanding is important 3 here. To the right unacceptable, to the left we'll 4 look at it again and work up other criteria. I mean 5 if you've got --
6 CHAIR BLEY: That makes good sense to me.
7 But I will point out on your slide you kind of talk 8 about that way to the left it's decreasing risk. In 9 the reports though that area to the left is called 10 risk insignificant which I suspect you might be 11 modifying I hope in the future.
12 DR. APOSTOLAKIS: Yes, yes. Originally 13 the staff objected and I think they were right. We 14 agreed to saying that if it's above it's unacceptable, 15 if it's below it's acceptable. And then we changed 16 the technology.
17 CHAIR BLEY: Risk insignificant bothered 18 me for anything less than 10-4.
19 DR. APOSTOLAKIS: No, you're right.
20 MR. FLEMING: Well, okay. One thing I 21 want to clarify though is that when this came up we 22 had a long discussion about this with the staff 23 yesterday.
24 Again, I'll go back and repeat. The 25 purpose to the frequency consequence curve is to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
74 1 evaluate the risk significance of individual LBEs 2 period.
3 Now we have taken to come up with the 4 anchor points on that curve and I guess we should 5 maybe move on here and not come back, come up with the 6 anchor points on this curve we have interpreted front 7 regulations on things like annual dose limits from 10 8 CFR 20, PAG level doses for triggering off site 9 responses, 10 CFR 50.34 for design basis type of 10 considerations and the QHOs.
11 While we've used those limits with an 12 interpretation of assigning them to a frequency which 13 is similar to what was done in the previous curves we, 14 although there is, my point is where there is 15 regulatory limits and safety goal objectives used to 16 derive this curve, again the only purpose of our curve 17 is to evaluate the risk significance of individual 18 LBEs.
19 When we go back and select the DBAs what 20 we do is after we figure out these required safety 21 functions that these are the ones that I need to keep 22 the DBEs inside the frequency consequence curve and if 23 I have high consequence BDBEs, i.e. BDBEs with more 24 than 25 gram doses I have to figure out what functions 25 do I have to fulfill to make sure those sequence, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
75 1 those points don't migrate outside the consequence 2 chart.
3 And those are declared as required safety 4 functions. And then what I do is I go back to my DBEs 5 and I see what SSEs were available to support each of 6 my required safety functions for each of the DBEs.
7 And normally the designer will then have 8 choices. They can decide, there may be different 9 choices he may have as to which SSEs he wants to 10 declare safety related so that I have at least one 11 safety related SSE to cover each required safety 12 function for all the DBEs and the high consequence 13 BDBEs.
14 Then after the designer selects those we 15 construct a DBA working to be consistent with the way 16 the Chapter 15 analysis is currently done. We force 17 failure of all the non-safety related SSEs that 18 perform these required safety functions so we can come 19 up with an event set that has the same characteristics 20 in the current Chapter 15 analysis.
21 We don't use the frequency to select the 22 DBAs, only indirectly as they come through the DBEs.
23 And in some cases these DBAs that we select have been 24 screened out of the PRA.
25 We show an example in the MHTGR where some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
76 1 of their DBAs were assessed in the PRA at less than 2 10-8. But that result comes from this prescriptive 3 approach of basically forcing failure of all the non-4 safety related SSEs on the incoming DBEs.
5 So once we get into Chapter 15 we're, we 6 don't, we're off the frequency consequence chart. So 7 now we use the rules of 50.34 and we come up with 8 conservative dose calculations to meet 25 gram.
9 CHAIR BLEY: Karl, I read two different 10 things. And one makes sense is you use the DBEs to 11 help define the DBAs.
12 MR. FLEMING: Yes.
13 CHAIR BLEY: Somewhere else it said that 14 all of the DBEs become DBAs. Is that, did I misread 15 or is that the intent?
16 MR. FLEMING: They're all considered in 17 formulating. But what happens is you'll find and we 18 showed examples yesterday, the MHTGR ended up with 11 19 DBEs.
20 Now they had an underlying PRA with 21 hundreds and hundreds of sequences. And they bend 22 them down by the time they have been down the DBE 23 range they end up with only 11.
24 And those map to eight DBAs because you'll 25 have several different DBEs that only vary by what the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
77 1 -- point. And when you force the safety related SSEs 2 to fail they sort like bully and reduce, if you will 3 down to the smaller set.
4 So that's what, that was the example for 5 the MHTGR.
6 MEMBER CORRADINI: So can I ask a 7 question, Karl, because I, this is the other part that 8 I was trying to understand. So once you choose the 9 HGTR numbers that you said, once you have the 11 that 10 broke down to the eight you then from a design 11 standpoint decide which of the systems you want to 12 declare as safety systems that will maintain their 13 functionality.
14 The rest are assumed to have failed. And 15 therefore because of a mechanistic source term these 16 groupings will move to the right. Again, you said 17 you're not using the F-C curve.
18 But I'm still trying to think of it this 19 way. It moves to the right so the source term 20 increases, changes. Is that correct?
21 MR. FLEMING: No, well when you go from 22 DBEs to DBAs there will be one DBA to match up with 23 each of the DBEs. They will have the same mechanistic 24 source term.
25 But when we got into Chapter 15 we're NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
78 1 going to do a conservative analysis versus a realistic 2 with uncertainty analysis that's in the PRA.
3 MEMBER CORRADINI: Okay. What does that 4 exactly mean? You're going to fail all the non-safety 5 related systems the designer has chosen to be non-6 safety and you're going to do something to a 7 mechanistic source term calculation that bounds it?
8 MEMBER REMPE: In the grouping there 9 probably was one sequence that only had the safety 10 related equipment working and that source term is what 11 they used back from when I was a lot younger I 12 remember.
13 If I can answer for you, sir. I have a 14 question though.
15 MEMBER CORRADINI: Well wait a minute.
16 Before you answer for him, is she saying it correctly?
17 I want to understand the procedure of deriving the 18 source terms for the DBAs.
19 I understand what Joy is saying. But is 20 that what's happening?
21 MR. FLEMING: The source term, the only 22 difference between the source term of the DBE and the 23 DBA it maps into, the only thing that's different is 24 the ground rules for conservatism in the calculation.
25 So --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
79 1 MEMBER CORRADINI: But when you say that 2 --
3 MR. FLEMING: It's the same mechanistic 4 source term.
5 MEMBER CORRADINI: -- it's more than just 6 choosing the safety system?
7 MEMBER REMPE: Your response indicated 8 that they had to do more conservative calculations and 9 that's not I think what's done.
10 MR. FLEMING: Okay, let me back up. When 11 you, when the designer is confronted with choices to 12 select their safety related SSEs he has to pick among 13 those SSEs that are available on all the DBEs.
14 So you don't, so when you map a DBE to a 15 DBA you basically have the same source term. It's 16 just that when we do the PRA we do realistic with 17 uncertainty treatment on the source term and when we 18 do the Chapter 15 analysis we do an acceptably 19 conservative analysis of the source term. And --
20 MEMBER CORRADINI: Okay, but --
21 MEMBER REMPE: What would change? What 22 assumptions?
23 MR. FLEMING: The only thing that would 24 change is what systems were assumed to be available.
25 MEMBER REMPE: Right, so it's one of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
80 1 sequences that you had in that group.
2 MR. FLEMING: Yes, that's right.
3 MEMBER REMPE: So I think my answer is 4 correct, Mike. But I have a question that pertains to 5 one of your backups. You mentioned the word about the 6 fundamental safety functions.
7 And you have back in one of your back up 8 slides control heat generation, control heat removal 9 and retain radionuclides. Why do you have nothing in 10 there about control reactivity and the ability to shut 11 the reactor down because that was always, it's in one 12 of your white papers?
13 That was always one of the fundamental 14 safety functions back at GA and --
15 MR. FLEMING: Control reactivity is 16 considered to be part of control heat generation.
17 MEMBER REMPE: We've had a lot of 18 discussions about being able to shut down the reactor 19 and having diverse systems. I would argue with you 20 that you might have different evaluations about ATWS 21 and other things that if you don't have that as one of 22 your fundamental safety functions.
23 Again, we've talked about this with 24 having, meeting the GDCs and the ability to have two 25 diverse site shut down systems.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
81 1 MR. FLEMING: Well in our framework 2 control room reactivity is considered to be part of 3 control heat generation. Right, that's how we --
4 MEMBER REMPE: But you could have just 5 gone with retain, control radiation release because 6 frankly heat generation and reactivity affect how much 7 radiation is released. So why did you pick two 8 instead of three then, I guess?
9 MR. FLEMING: You know, all of these 10 hierarchies of safety functions, I mean there is a 11 hierarchy. I mean in principle the fundamental safety 12 function you could define in just one.
13 MEMBER REMPE: Right.
14 MR. FLEMING: Which is control, you know, 15 control radionuclide inventory. Control heat 16 generation and control heat removal are considered to 17 be necessary for that. So they're not really 18 separate, you know, they're not really separate and 19 independent.
20 And what, the only reason why we refer to 21 those as fundamental is that's what the IAEA calls 22 them, okay. And in our approach what we identify is 23 that each reactor has to figure out what their safety 24 functions are and what their required safety functions 25 are.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
82 1 And we just observe that they're all going 2 to be somehow related to those three functions.
3 MEMBER REMPE: When I think of trying to 4 help designers of what to consider in selecting their 5 sequences in the PRA I would really like to see 6 controlling reactivity as something that they should 7 think of just like heat removal and radiation release.
8 And so I'm not sure that's a good thing to 9 not just explicitly state --
10 MR. FLEMING: Okay, that's good feedback.
11 MEMBER REMPE: -- in this higher level 12 document. I know it's in the white papers because you 13 have the example with the MHTGR and there were several 14 fundamental safety functions there. So please think 15 about that.
16 MR. FLEMING: That's good feedback.
17 That's good feedback. Thank you very much.
18 MEMBER MARCH-LEUBA: So what do you say by 19 weeding down the DBAs into smaller number of DBAs?
20 Why not run all of the DBAs with the fail of non-21 safety related SSEs and make sure the rest is okay?
22 MR. FLEMING: That's what we do. That's 23 what I tried to describe. But the last phrase without 24 crediting the non-safety related SSEs that is the 25 DBAs. It's the same thing.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
83 1 MEMBER MARCH-LEUBA: No, but you do it for 2 all DBEs.
3 MR. FLEMING: But some of the DBEs have 4 non-safety systems working in them and that's what 5 they collapse. They don't account for any of the non-6 safety systems. They're all in the PRA.
7 MEMBER MARCH-LEUBA: So you --
8 MR. FLEMING: But they're not in the 9 deterministic analysis.
10 MEMBER MARCH-LEUBA: Design basis events 11 which is complete or not complete one can argue. And 12 then you run it with everything working. Say the 13 worst are these eight.
14 And then you run those eight with the non-15 safety related things fail.
16 MR. FLEMING: Right.
17 MEMBER MARCH-LEUBA: But maybe before you 18 are doing the procedure with the non-safety related 19 fail are now bad. Why not run all of them? What do 20 you save by weeding them?
21 MR. FLEMING: I'm not exactly sure if I 22 follow how, what you're describing as different from 23 what we're --
24 MEMBER CORRADINI: I think all he's saying 25 is you dropped three. Those three could be done. At NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
84 1 least that's what I thought he meant.
2 MEMBER MARCH-LEUBA: From 11 to eight.
3 MR. FLEMING: Yes, the, those sequences 4 would come out with exactly the same source term. The 5 only thing that's going to be, because in addition to 6 these, to the safety related SSEs working there's also 7 another non-safety related system operational which 8 sort of takes the demand off.
9 There wouldn't be any difference in the 10 consequences. I mean --
11 CHAIR BLEY: I think there's a little 12 confusion. Let me try saying it and tell me if I'm 13 wrong.
14 MR. FLEMING: Okay, thanks.
15 CHAIR BLEY: Say there are four DBEs, one 16 of those because it's in the PRA has all of the non-17 safety systems already failed. Some of the others, 18 some of the non-safety systems are working.
19 Once you fail all the non-safety systems 20 all four of those are the same sequence.
21 MEMBER MARCH-LEUBA: Is that how it is 22 collapsed?
23 MR. FLEMING: Yes.
24 MEMBER MARCH-LEUBA: Is that the only way 25 it's collapsed?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
85 1 MR. FLEMING: The only way it's collapsed.
2 DR. APOSTOLAKIS: And if necessary you 3 look at all of them.
4 MEMBER MARCH-LEUBA: That's why I'm saying 5 why not look at all of them?
6 DR. APOSTOLAKIS: It doesn't say you have 7 to do it. It just happened in MHTGR and --
8 MEMBER MARCH-LEUBA: Analysis is cheap.
9 Let's just look at all of them.
10 MR. FLEMING: Well we are going to look at 11 all of them.
12 DR. APOSTOLAKIS: If necessary, yes.
13 MR. FLEMING: In fact the analysis that we 14 do over on the PRA side is going to have a lot more 15 information because it's going to have the realistic 16 assessment with the full quantification of 17 uncertainty.
18 MEMBER MARCH-LEUBA: You cannot do a PRA 19 analysis unless you've run TRACE to tell you what the 20 sequence was. So you start with an analysis.
21 MR. FLEMING: Yes, we are running the 22 consequences of all the LBEs. And we're rerunning 23 them in Chapter 15 using different ground rules that 24 adhere to the regulations.
25 MEMBER MARCH-LEUBA: Not for all the DBEs.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
86 1 You're running only for a subset.
2 MR. FLEMING: Well this is our way of 3 replacing the judgment that, you know, we consider the 4 large break LOCA, the bounding event for Chapter 15.
5 This is the process where we are using the events 6 above 10-4 for plant year as an engine and a 7 reproducible process to create a set of DBAs that are 8 reproducible for each plant.
9 We're trying to do something that will 10 create a consistent level of treatment for different 11 types of technologies, different types of designs.
12 MEMBER MARCH-LEUBA: Analysis is cheap.
13 I don't see why not do it comprehensive.
14 MR. FLEMING: Yes, if necessary it would 15 be done.
16 MEMBER MARCH-LEUBA: It's easier to run 17 them then you don't have to use judgment.
18 MR. FLEMING: Okay, appreciate your 19 comment. The frequency consequence curve a couple of 20 comments. Do we want to stop for a break? Let me 21 finish this slide and we can go on.
22 The frequency consequence curve, we have 23 a couple of anchor points. We've made some 24 adjustments to the version that was in the NGNP 25 project which is the starting point.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
87 1 We had some issues with the staircase 2 character of some of the other attempts to do this 3 because the staircase shapes somehow permit a higher 4 risk as you go down in frequency which didn't seem to 5 make much sense.
6 There's some language up in the SRP, 7 standard review plan, Chapter 15 that talks about 8 different frequency, different levels of frequencies 9 of AOOs.
10 And there's a statement there that in 11 principle the dose limits for postulated accidents and 12 the AOOs could be equivalent and by following an iso-13 risk contour what we decided to do is sort of connect 14 the dots on our points between the AOO region and the 15 BDBE region rather than having these various staircase 16 aspects to what NGNP had.
17 We moved over the dose limits for the 18 AOOs, the lower frequency AOOs based on an 19 interpretation of something that we found in the 20 standard review plan which talks about lower frequency 21 AOOs could have higher doses than 10 CRF 20 which was 22 the limit used in the NGNP project.
23 And we moved the AOO limit over a decade 24 from 100 millirem to one rem. But that's basically 25 the frequency consequence chart and how it came up.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
88 1 And then for looking at individual LBEs 2 using sort of some precedent set in the PRA standards 3 where if you go into the PRA standards to define event 4 sequences that are risk significant there's a one 5 percent criteria used in the standards for saying if 6 you have a, in an LWR case if you have a sequence that 7 contributes more than one percent of the core damage 8 frequency that's considered to be a risk significant 9 sequence.
10 So using that kind of idea and precedent 11 we set an uncertainty bar, well not an uncertainty bar 12 but a zone of risk significance below the frequency 13 consequence target by two decades down in frequency.
14 And we also took, we took a look at is there a level 15 of dose that we could consider to be not significant 16 in terms of consequence.
17 And we're suggesting that a 2.5 millirem 18 limit might be a reasonable threshold for not worrying 19 about very, very low doses because that would be about 20 ten percent of the background radiation that this 21 person that's taking the 30 day dose is going to get 22 during the 30 days.
23 So does it make sense to worry about 24 fractions of background radiation over a 30 day 25 period? So now the reason for setting this is, you NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
89 1 know, we, LBEs that are defined as risk significant, 2 they're going to get extra scrutiny and attention in 3 the defense-in-depth evaluation we can talk about 4 after the break.
5 And also we're going to come up with SSE 6 risk significance criteria that are tied to these LBE 7 criteria. So in risk significance we're going in two 8 dimensions rather than one. Yes.
9 MEMBER CORRADINI: Okay. I just want to 10 make sure I understood. So on the, on Slide 17 the y 11 axis you've increased by two decades I understand.
12 And the logic of changing the dose by, I'm trying to 13 estimate what it was, but it's more than a factor of 14 ten, less than a factor of 100 was based on what?
15 MR. FLEMING: What I was referring to is 16 the, you know, this cliff edge that we have at one rem 17 and 10-2 where you have this vertical line on the 18 previous consequence chart in the AOO region. That's 19 based on an interpretation of something that we found 20 in the standard review plan that basically says that 21 doses can be higher than 10 CFR 20 for AOOs.
22 For lower frequency AOOs as long as you 23 don't upset off site, you don't impact off site 24 activities. So we interpreted that criteria to be one 25 rem.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
90 1 CHAIR BLEY: I'm going to interrupt for a 2 second and just --
3 MR. FLEMING: The NGNP had a, they had a 4 vertical cliff one decade over at 100 millirem. So 5 that's what I was trying to say.
6 CHAIR BLEY: I'm going to interrupt for a 7 minute. The details of this at least to me smell like 8 something that in the end the staff is going to have 9 to either agree or suggest you do something different.
10 I think in October, because I think we're 11 going to run out of time today, really understanding 12 where the staff falls on this and what you expect to 13 have in your requirements is something we would want 14 to delve into in quite a bit of detail with you. I'm 15 going to, I just wanted to get that out there for them 16 for the next time around.
17 MEMBER RICCARDELLA: Dennis, this is Pete 18 Riccardella.
19 CHAIR BLEY: Just a minute, Pete. I'd 20 like you to finish this set, the next three slides 21 before we take a break. But do them real quickly. Go 22 ahead, Pete.
23 MEMBER RICCARDELLA: Yes, just would you 24 repeat what you said about the 2.5, the vertical line 25 at 2.5 millirem. I just didn't understand it.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
91 1 MR. FLEMING: Yes. So we wanted to see 2 whether we could come up with a criterion for an 3 insignificant dose from the, in the context of the 4 uses of our frequency consequence curve in evaluating 5 LBEs.
6 And the 2.5 millirem is basically, if you 7 go into the NRC website and get the average U.S. value 8 for background radiation and divide that down into 9 what you would get in 30 days it turns out to be 25 10 millirem.
11 So we said let's take ten percent of that 12 25 millirem and say if I have a dose less than 2.5 13 millirem let's just agree that those are not, we 14 shouldn't worry about the risk significance of the LBE 15 if it's, has that low dose.
16 MEMBER RICCARDELLA: Why a factor of ten?
17 Why not 25 millirem?
18 MR. FLEMING: It's a judgment. It's just 19 an engineering judgment.
20 MEMBER RICCARDELLA: Okay.
21 MEMBER REMPE: So this is where I would 22 like you to discuss why you didn't do something for 23 the Part 52 folks to give them guidance about the site 24 parameters that should be picked because we've had a 25 design certification that came in that was valid for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
92 1 one site in the U.S. that we're still evaluating.
2 And to me what was done years ago in the 3 MHTGR where they had this EPRI document that looked at 4 all the sites around and came up with a bounding, I've 5 forgotten it's been many years ago, like 90 percent of 6 the sites have recommended the characteristics, why 7 not go ahead and say to use this design approach use 8 those assumptions for the site?
9 MR. FLEMING: Yes. When we get into 10 things like external events and things like that we 11 do, you know, we do recommend that you define a site 12 parameter envelope that you want to, based on your 13 business case for your reactor where you want to have 14 it licensed and operational for.
15 And then you use that information from a 16 site parameter envelope to do site specific factors.
17 MEMBER REMPE: But even dose requires some 18 assumption about weather too.
19 MR. FLEMING: That's right. If you, one 20 option that you can use if you haven't even selected 21 a site parameter envelope is you can use regulatory 22 guide weather assumptions with this approach to come 23 up with a dose.
24 MEMBER REMPE: It's just silent in the 25 document about what should be done. Why not give some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
93 1 guidance to people?
2 MR. FLEMING: Yes, we have, we've gotten 3 some feedback from the staff that we need to say more 4 about how you do these dose calculations. That's a 5 good comment.
6 MR. AFZALI: Sorry, if I may request I do 7 understand October the staff are going to come with 8 their own portfolio. But between now and then any 9 feedback you have on our approach would be --
10 CHAIR BLEY: We only give the ACRS 11 opinions through our letters after a full Committee 12 meeting. You hear individual comments here and that's 13 the way you should take as individual comments. Thank 14 you.
15 MR. FLEMING: The, this Slide 18 has the 16 cumulative risk targets. And this is to recognize 17 that we have to look at the summation of the risk 18 across all the LBEs.
19 And basically what we're using is the two 20 QHOs for the safety goals the latent cancer and the 21 early fatality safety goal metrics. And so we're just 22 adopting those as goals.
23 And when you do these calculations these 24 are going to be exercised primarily by the lower 25 frequency higher consequence events. And to capture NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
94 1 the higher frequency events and to also recognize that 2 we've used, we've done, we've stepped outside the real 3 meaning of 10 CFR 20.
4 10 CRF 20 is an annualized limit that's 5 summed up over all the releases that happen in a year.
6 So in this case we're capturing maybe the better 7 interpretation of the intent of 10 CFR 20, not to look 8 at individual events like we do in the previous 9 consequence chart but the sum of the over all the 10 events.
11 So we basically do an exceedance frequency 12 curve to make sure the total frequency of exceeding 13 100 millirem does not exceed once per plant year. And 14 then it ends up taking care of the AO overage.
15 So those are the cumulative risk metrics 16 that we use. And then the couple, just a couple of 17 thoughts about the PRA, we just have one slide on 18 that.
19 We recommend early introduction of the PRA 20 into the design process although it's not required.
21 There is flexibility in the approach when the designer 22 wants to do this.
23 But the earlier he gets the PRA in the 24 design process he can help the designers incorporate 25 risk insights in the design long before he, the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
95 1 designer even starts talking to the regulatory bodies 2 about the licensing process.
3 But when it's, the earlier you introduce 4 the PRA in the design process the more simplified the 5 PRA is in terms of scope and level of detail. You 6 make, you know, sensible judgments to incorporate 7 detail and scope in terms of hazards until you have 8 the design and site information that makes it 9 meaningful to do those.
10 So you start probably with internal events 11 and full power operation. And then as more 12 information becomes available you expand to other 13 operational states.
14 You bring in fires and floods when you 15 have general site layout information, cable tray 16 layouts and things like that. You introduce seismic 17 and external hazards when you have a seismic design 18 when you have decided what your seismic design is and 19 you have enough structural and planning information to 20 support a meaningful external hazards PRA.
21 So I want to get back, you know, this PRA 22 process is not a one shot deal. We don't like figure 23 out at an early stage what the DBAs are and then go 24 away and everything goes away.
25 This is, one of the sort of challenges to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
96 1 applying this process is that in each stage of design 2 and site and licensing development you're building on 3 your PRA, you're updating it and you're going back and 4 revisiting your previous risk-informed decisions that 5 are influenced by the PRA.
6 So this thing that's happening today with 7 the operating plants where plants are coming in with 8 applications for license amendments and based in part 9 on some PRA inputs, than those decisions are subject 10 to review and revision as the PRA unfolds even in the 11 site operation.
12 MEMBER MARCH-LEUBA: You're saying that 13 the plant manager is going to let you run a PRA that 14 would challenge his license, that may challenge his 15 license? I can assure you, George should be the one 16 talking about that.
17 DR. APOSTOLAKIS: Definitely they will do 18 it.
19 MEMBER MARCH-LEUBA: If they're doing it 20 generally by Part 21 they need to do it.
21 DR. APOSTOLAKIS: When they come here they 22 will not challenge that site. Dennis and I saw it in 23 the early days where the PRAs that were published and 24 were reviewed by the NRC didn't show that there were 25 any problems.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
97 1 But I remember in internal meetings the 2 fire risk in one of the plants was really unacceptable 3 and the utility decided to do something about it.
4 That was not published.
5 That's a problem because when I was an 6 academic we were looking for PRAs and examples of 7 this. You can't get those. Those are internal. But 8 they never come before you and say, I might have a 9 problem here.
10 CHAIR BLEY: Well you might see --
11 DR. APOSTOLAKIS: But you guys may 12 identify the problem.
13 CHAIR BLEY: -- license amendments. It's 14 happened.
15 DR. APOSTOLAKIS: It could. But in 16 general these things are done internally.
17 MEMBER MARCH-LEUBA: I thought Part 21 18 prevented you from doing that?
19 MR. FLEMING: Now Dennis earlier was 20 asking a question about advances and maybe things that 21 might be different than when this kind of thing came 22 up before. One, back in 2006 the ASME Board of 23 Nuclear Codes and Standards decided that we needed a 24 PRA standard for non-light water reactors.
25 So I was chair of that group. And we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
98 1 issued a trial use standard in 2013. And to capture 2 the input from all the non-light water reactor 3 developers that were involved in that project team it 4 included representatives from the HGTR family, the 5 liquid metal fast reactor family and just recently 6 getting into, getting some involvement from the molten 7 salt reactor family.
8 The, everybody wanted to make sure that 9 they captured the risk of multi module scenarios. So 10 the PRA standard and the PRA is intended to capture 11 event sequences that may involve one or more reactor 12 modules.
13 And there may even be some risk 14 significant non-core sources that are also included 15 depending on what the characteristics of the reactor 16 are. So multi module is sort of captured explicitly 17 in this approach.
18 And they're included in the LBEs. We 19 don't expect to have any in the DBA region. But one 20 of the motivations for introducing the PRA early in 21 the design is to let the designer be aware of the 22 possibility of multi module scenarios and make sure he 23 has design features built into the plant to ensure 24 that they don't become risk significant in the end.
25 The other thing, the final comment on this NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
99 1 slide is that we're not really, we're not advocating 2 a risk based approach. We recognize the limitations 3 and uncertainties associated with the PRA.
4 The PRA doesn't quantify everything. It's 5 never complete. But it's intended to capture the 6 current state of knowledge about events and reactor 7 safety characteristics.
8 And so we augment the insights from the 9 PRA with traditional deterministic methods and most of 10 that is captured in this framework for the defense-in-11 depth framework which we hope to have some time today 12 to cover.
13 CHAIR BLEY: Thanks, Karl. We are 14 certainly running out of time. We're going to take a 15 15 minute break and come back at ten until. I'm going 16 to give you 25 minutes to finish up the next two 17 sections.
18 So think about what you really want to 19 talk about and what you're going to pass on because 20 you, at least for me I would put like ten minutes on 21 the next section and 15 on the one after.
22 MR. FLEMING: Okay.
23 CHAIR BLEY: Either way, when the time 24 runs out we're going to go to the staff. So pick your 25 most important ones. So at this time we'll recess NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
100 1 until ten until. Thanks.
2 (Whereupon, the foregoing matter went off 3 the record at 10:33 a.m. and went back on the record 4 at 10:49 a.m.)
5 CHAIR BLEY: We are back in session.
6 We're going under the SSC safety classification, and 7 Karl is on Slide 24, for the people listening in.
8 MR. FLEMING: Right. And my goal in time 9 management space is to have abbreviated talk on the 10 safety classification. Just talk about, a little bit 11 what's the same or different than what came out of the 12 NGNP process so we have ample time to get well into 13 the defense in depth discussion.
14 The NGNP process came up with three safety 15 classes for SSCs. That was safety-related class which 16 we've already covered how we get those. Those are 17 tied into the process for, you know, putting together 18 our design basis accidents based on several different 19 options.
20 The middle category is called NSRST in our 21 framework, it's non-safety-related with special 22 treatment. And with those we want to pick up other 23 risk-significant SSCs as well as SSCs that might 24 perform an important defense in depth role which make 25 up the safety significant SSCs. So our use of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
101 1 term risk-significant and safety significant is pretty 2 much in alignment with the way these terms have been 3 used in the 50.69 world.
4 However, the criteria we use for risk 5 significance is by definition has to be different 6 because we have to capture more of a technology 7 neutral way to do that. The risk-significant SSCs are 8 basically defined in terms of those SSCs that are 9 responsible for producing risk significant LBEs, and 10 there's a lot of detail provided in the paper on how 11 to get those.
12 Basically, if an SSC performs a function 13 that's necessary to keep one or more LBEs inside the 14 frequency consequence curve, beyond those already 15 picked up in the safety related SSCs, those are also 16 considered to be risk-significant.
17 And also, if we look at the accident 18 sequences where an SSC has failed, if those sequences 19 comprise more than one percent of one of our 20 cumulative risk metrics, than that's another way for 21 an SSC to be called risk-significant.
22 And then later on I'm going to talk about 23 additional considerations from the defense-in-depth 24 criteria that can produce additional SSCs that may not 25 be risk significant but may provide an important NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
102 1 defense-in-depth role. And then we have the rest of 2 the SSCs modeled in the PSA and then of course outside 3 the universe of possibilities we have all the SSCs in 4 the plant.
5 And this diagram was inspired by some 6 interaction we had with the staff to try to get an 7 idea of, you know, how you put these things together.
8 And the way this is put together, it's intended to be 9 each oval inside is expected to be a sub-set of the 10 outer ovals.
11 Well, a feature of this approach is that 12 by virtue of the process we used to select the safety 13 related SSCs, they're all risk-significant because 14 they're all needed to perform a function to keep one 15 or more LBEs inside the risk consequence chart. There 16 may be others up in the AOO region that have that same 17 characteristic, but all the safety related --
18 So we don't have a risk category, what's 19 it called, RISC-2 that in the 50.69 world where you 20 have safety related, but not safety significant. We 21 don't have those. They're all risk-significant and, 22 therefore, safety significant.
23 One of the things that we've done that 24 goes beyond what NGNP did is try to provide more 25 guidance on how we migrate from safety classes to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
103 1 special treatment requirements. And for the safety 2 related -- and I'm going to get back to a question 3 that came up about principle design criteria.
4 For the safety related SSCs, we come up 5 with what we call functional design criteria. And 6 these are reactor design specific criteria that are 7 tied to the safety related SSCs along the LBEs that 8 they're participating in.
9 So they're very, very design specific and 10 you could have two different, say, molten salt 11 reactors come in and they might have different 12 functional design criteria because the designers may 13 have selected a different package of SSCs to become 14 safety related.
15 Karl, and on your PDC question, on your 16 principle design criteria question, one of our ideas 17 here is that when an applicant puts to forth an 18 application for an advanced non-light water reactor, 19 he'll have available to him what comes out of the ARDC 20 world, which are not really design specific.
21 They may be family of design specific and 22 the functional design criteria could be part of the 23 principle design criteria depending on when it's 24 proposed by the applicant.
25 CHAIR BLEY: Makes sense. Back to your NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
104 1 little Venn diagram. Can you give me examples that 2 help me understand how you can be safety significant, 3 but not risk-significant?
4 MR. FLEMING: Yes. So one of our criteria 5 which I'll get to in the defense in depth session is 6 that, you know, it comes from the literature on 7 defense in depth where you don't want to have over-8 reliance on a single design feature to perform 9 something that's really in important role.
10 So these required safety functions that we 11 define are really, really important and it guides our 12 definition of safe related SSCs and design basis 13 accidents.
14 Those SSCs are very, very important, but 15 they perform the function. It's so important, we 16 don't want to rely on a single SSC to perform an 17 important function. So there may be, for example on 18 the MHTGR, it may select a passive RCCS as a safety 19 related SSC for heat removal.
20 We may define a role for maybe one of the 21 active shutdown cooling systems that have blowers and 22 you may have defined that to provided one of these, 23 not for one, criteria from defense in depth. And it 24 may not necessarily be risk-significant and, you know, 25 may or may not be risk-significant depending on how NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
105 1 the numbers come out.
2 So it's conceivable that there may be a 3 defense in depth goal not to rely on a single element 4 of design. So you have to declare one more thing to 5 meet that criteria to be, meet the defense in depth 6 criteria.
7 CHAIR BLEY: I'll have to study that a 8 little and think more about it. I guess one thing it 9 would do is it would keep some potentially really 10 risk-significant SSCs from having a high risk 11 achievement work, because even if they fail they have 12 this other back-up that isn't showing up as risk-13 significant.
14 MR. FLEMING: You're right, well, I'll 15 think about that some more. But that is one example.
16 MEMBER SKILLMAN: Excuse me. This is 17 where I've been waiting to make my comment since we 18 started this morning. It seems to me that the SSC 19 discussion is really subordinate to the overall risk 20 informed base guidance document.
21 And in that document, which is Version M 22 on Page 36, these are the words that I bring to the 23 record. The designer then selects one specific 24 combination of available SSCs to perform each required 25 safety function that covers all the DBEs in high NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
106 1 consequence BDBEs. These specific SSCs are classified 2 as SR, safety related, and Task 5 Alpha and are the 3 only ones credited in the Chapter 15 Safety Analysis 4 of the DBAs.
5 MR. FLEMING: Right.
6 MEMBER SKILLMAN: So with that on the 7 record, let me try to pull on your comment, maybe they 8 are two different salt reactors. One salt reactor 9 designer has chosen one set of equipment to perform a 10 specific set of functions, whereas another might 11 choose another set. Both are acceptable because they 12 fulfil the functions.
13 MR. FLEMING: Right.
14 MEMBER SKILLMAN: But under this category 15 then, now I'm moving into your SSC document, in your 16 SSC document the author of this document used the word 17 credited three times. And two occasions it's credited 18 to Chapter 15 and one time it's credited to 50.34a, 19 excuse me, 50.34, 50.34.
20 MR. FLEMING: Right.
21 MEMBER SKILLMAN: The word relied on is 22 used nine times. And it appears to me, reading all my 23 instances, relied on almost means credited in the 24 Chapter 15 world. And safety-related or safety is 25 used 126 times, non-safety related is used 40, hence, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
107 1 86 others are safety-related.
2 Here's my point, at this very early stage 3 of this very fine fabric that's been woven here, now's 4 the time to get these terms and these words adjusted.
5 And I make that comment to the advisory group and I 6 make that comment to the staff.
7 MR. FLEMING: Right.
8 MEMBER SKILLMAN: I was one the people 9 that was back in 1969, '70 and '71 trying to connect 10 Chapter 15 to the equipment that would be credited and 11 the equipment that would not be credited. It's 12 important to get these terms correct now because the 13 heartache we went back, that we invested back in those 14 early years only became clear once REG Guides 126, 129 15 and now 1.201 would lay out the implementers actually 16 do this.
17 But there's a catch to this, when you 18 finally get your license, whatever it might be, it had 19 good old 50.59, you can make the change. The way the 20 text reads in 50.59 of what is a licensing change is 21 very significantly impacted by the interpretation of 22 those words.
23 Whether it was credited, whether it was 24 relied on, whether it's safety-related, we still have 25 good old 50.55a, important to safety. Then there are NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
108 1 the others who say well it's not important to safety.
2 And all of those words mean something.
3 Now's the time to get the glossary 4 adjusted so that we're all on the same page relative 5 to what those terms mean. It will save gobs to the 6 implementers. Now's the time to do it.
7 MR. FLEMING: That's really good input and 8 we had recognized that problem and it's a work in 9 process. One of the, your term, action items that our 10 team has is to propose a glossary and clean up the 11 terminology, eliminating synonyms where unnecessary 12 and causing confusion.
13 If we mean something different, or well, 14 if we mean the same thing as what's in the current 15 regulatory documents, we use the appropriate term. So 16 that's good input. We recognize we're not there yet 17 and, you know, it's a work in process.
18 MEMBER SKILLMAN: Thank you. Okay.
19 MR. FLEMING: It's been a struggle for us, 20 frankly, to get -- because there's a certain way these 21 terms are used in the PRA world even for light-water 22 reactors and a different way they're used in license 23 process and we recognize that's work to be done.
24 MEMBER SKILLMAN: Thank you. Thank you.
25 MR. FLEMING: One final set of comments on NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
109 1 the safety classification, what we did in the current 2 work on the LMP, we tried to move the ball down the 3 field a little bit in terms of how do we go from 4 safety classification to, you know, additional 5 requirements.
6 And the way our process works is that for 7 safety-related SSCs we have to come up with these 8 functional design criteria that may find some 9 usefulness in formulating the principle design 10 criteria, depending on how the application is put 11 together. As well as lower level design criteria that 12 are specific to the SSCs that are performing the 13 safety-related functions.
14 But for all of the safety significant SSCs 15 in both the safety-related and non safety-related SCC 16 categories, we start by setting reliability and 17 capability performance requirements for all those SSCs 18 that are tied to what reliability was assessed in the 19 PRA and how much deviation from those reliability 20 assessments would influence the decision making as far 21 as targets.
22 And we talk about these as targets, as 23 design targets, they're not necessarily requirements.
24 We don't mean that they would regulatory requirements, 25 but the idea is that we want to make sure that all the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
110 1 special treatment that comes into play here somehow 2 connects the dots to how reliable the equipment has to 3 be and how capable it has to be in terms of margins 4 and performance in order to keep the risk informed 5 process hanging together, to keep LBEs from migrating 6 from region to another, from getting too close to the 7 frequency consequence curve, to getting too close to 8 our cumulative risk matrix and so forth.
9 So that's the starting point for special 10 treatment requirements. And then the rest of the 11 decisions on what special treatment that any safety 12 significant SSC gets really is determined by an 13 integrated decision panel that's the same panel that's 14 going to do the defense in depth evaluation.
15 So what goes into selecting the special 16 treatment is, you know, what's necessary for 17 reliability and capability performance, and then how 18 much additional treatments do we need to basically do 19 things like manage uncertainties and address 20 limitations in the PRA and other considerations that 21 come from our defense in depth criteria. And that's 22 what informs the derivation of special treatment.
23 So we don't specify a specific list of 24 special treatment. That's something that really has 25 to be done on a design specific basis. It's very NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
111 1 difficult to postulate a technology inclusive set of 2 requirements that would be appropriate for any given 3 reactor.
4 MEMBER REMPE: So, I actually pulled NEI 5 00-04 to try and understand about this integrated 6 decision panel and has it been used? I mean, there's 7 a lot of nice words there of the breadth of the 8 experts that should be involved and they get a lot of 9 credit for making decisions like determining adequate 10 margin is in your paper that you provided to us.
11 But has this panel ever been created and 12 used for a 50.69 process. And I assume the panel is 13 a paid panel of experts that's very familiar with the 14 plant.
15 And when you have them with these paper 16 designs, I'm not sure that they're be able to afford 17 that level of experts and maintain them throughout 18 this design process. And so, I'm just kind of 19 wondering, was that discussed much with the staff and 20 can you give me more details about what you expect?
21 And maybe it ought to be documented in this report.
22 CHAIR BLEY: And try to make it concise 23 because we're running out of time.
24 MR. AFZALI: So the panel for X Energy, 25 for example, they did a tabletop exercise, the panel NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
112 1 consisted of external analyst, like solar nuclear 2 development and the designers of that technology.
3 So I would not see this as a -- if I may 4 choose a different venue to come up with a expert 5 panel that they think qualify to make decisions. But 6 that panel will not be substantially different than 7 what we would use for a deterministic evaluation when 8 we are trying to identify licensing basis event.
9 So we have had discussions, we do 10 understand that the outcome of a lack of operating 11 experience, so we have to consider that as part of our 12 conversations. But we have not been prescriptive in 13 nature of what that panel should look like.
14 I take your point that you need to, 15 perhaps it would be advisable to have some guidance on 16 the expert panel formation, the group that is going to 17 be there, what expertise should be in there. We have 18 all those conversations are part of 50.69, license 19 amendment and we came up with a guidelines for that.
20 Maybe we look at those guidelines and see how can we 21 transfer that to a process like this.
22 MEMBER REMPE: So see NEI 00-04 does have 23 a lot of good detail, I just am not sure that the 24 start-up companies are going to be able to adhere to 25 that level of detail.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
113 1 MR. AFZALI: I agree. I'm just wondering 2 what those start-up companies would do identify LBEs 3 if they don't use this process. They still to need to 4 form something and you have to have those 5 conversations to make sure we don't increase anybody's 6 burden, but they want to a have good realistic 7 approach. I do agree. We will look at that option.
8 DR. APOSTOLAKIS: Maybe the Regulatory 9 Guide can give some advice how to do this because it 10 comes from the staff.
11 CHAIR BLEY: Karl, you're down to the last 12 five or ten minutes and you're on Slide 27.
13 MR. FLEMING: Right. So let's go on to 14 the defense in depth approach. We picked up where the 15 NGNP process left off in terms of the DID approach.
16 In our framework, what this approach leads to is that 17 we decided to build on the integrated decision panel 18 approach that was in 50.69.
19 It not only addressed the things that are 20 addressed in 50.69 having to do with safety 21 classification, but to give them the responsibility to 22 put together a documented, from the designers point of 23 view and the developers point of view, a documented 24 basis for the designer's evaluation of defense in 25 depth adequacy.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
114 1 This document would be dated periodically 2 as you go through the different design and licensing 3 phases. At earlier phases, I'm sure, there will be 4 open items that haven't been resolved because not 5 enough design information and analysis information has 6 been brought in. And it's a type of document that the 7 NRC staff could audit to see what the point of view is 8 from the developer on terms of defense in depth 9 adequacy.
10 What inspired this earlier development of 11 defense in depth was remembering the Exelon PBMR 12 discussions with the NRC, there was, somebody made a 13 comment that while this plant doesn't have a leak-14 tight containment and therefore it's not a defense in 15 depth.
16 And back in those days we recognized that 17 we need to find a better way to talk about defense in 18 depth that is more meaningful for different kinds of 19 reactor technologies that have out-catered the safety 20 case to different elements of design.
21 So we picked up on what NGNP came up with 22 on terms of this triangle diagram which the bottom two 23 elements in the triangle is where the defense in depth 24 exists, except there's some physical aspects in 25 defense in depth which show up in the physical part of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
115 1 the design, the plant capability to defense in depth.
2 And there's the programmatic elements that 3 are put on to ensure that the design intent that was 4 in the plan was actually realized when we construct 5 the plant and as it's operated throughout the plant 6 lifetime to give you confidence that the plant 7 capability is actually delivered.
8 And the risk informed evaluation, means 9 that we had deterministic analyses and probabilistic 10 analyses that are going to be going on in parallel 11 with the design evolution that would lead to feedback 12 mechanisms to enhance the plant capabilities as needed 13 and the programmatic capabilities as needed.
14 We came up with a set of attributes and, 15 also, in building on one of the earlier questions, we 16 did review Mary Drouin's fine review and bibliography 17 on defense in depth, and we picked upon that the idea 18 of talking about layers in defense.
19 And we thought that was an advantage in 20 this approach because the whole concept of barriers 21 looks quite a bit different in some of these reactor 22 designs. So we thought that layers of defense versus 23 levels of defense made a little bit of sense here.
24 We adapted from the IAEA, which really 25 still talks about levels of defense, we adopted a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
116 1 diagram that we modified a little bit here to make it 2 more meaningful for beyond design, I'm sorry, for 3 advanced non light-water reactors to talk about the 4 layers of defense. And in this framework, this is 5 really in a great alignment with the way we organize 6 the definition of LBEs.
7 So some of our LBEs are arrested. In 8 fact, some LBEs don't happen because we've done a good 9 job preventing initiating events and we go into 10 different layers and we terminate the LBEs at 11 different layers, depending on the response of the 12 plant and the levels of diversity and redundancy that 13 have delivered that process.
14 So, we've adopted this type of a diagram.
15 The defense and the IDP Panel then goes back and takes 16 a look at what's coming out of the PRA in terms of the 17 LBEs and takes a broader look at the LBEs in terms of 18 identifying what wasn't really analyzed in the PRA, 19 what are the limitations of the PRA and those type of 20 things and factors that in to some recommendations on 21 are there things that could be done to enhance the 22 physical defense or the plant capability defense in 23 depth or the programmatic defense in depth to improve 24 the confidence that the safety case is going to be 25 realized.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
117 1 We also, then, think it's appropriate for 2 this same panel to take the lead on coming up with a 3 specific package of special treatment requirements 4 that are appropriate to not only achieve the 5 reliability and capability objectives or targets, but 6 also to provide greater assurance and a greater degree 7 of confidence that things that may not be adequately 8 resolved in the PRA are addressed.
9 So this is all outside the PRA process, 10 but it also does a critical look at what's coming out 11 of the PRA to make sure that things that are not done 12 very well or are resolved very well in the PRA are 13 given some consideration. We have a set of attributes 14 for each of the parts of our triangle on the plant 15 capability defense in depth--
16 CHAIR BLEY: We're on Slide 30, by the 17 way.
18 MR. FLEMING: Yes. We've gone on to Slide 19 30. We've defined a set of attributes that, and some 20 things that are considered in the evaluation for each 21 attribute to be able to take a look at these defense 22 in depth characteristics.
23 And one of the things that, in terms of 24 reliability, we recognized that we're migrating away 25 from a reactor technologies that are relying on NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
118 1 primarily active redundant safety systems to more 2 passive, utilization of passive and inherent safety 3 functions and so forth.
4 So, we've broadened the idea of looking at 5 redundancy and diversity to considering a combination 6 of inherent characteristics that may be responsible 7 for part of the safety function passive SSCs as well 8 as active SSCs that are performing those functions.
9 That's sort of the focus of that evaluation.
10 These are little bit out of order. I'm 11 going to skip to Slide 32. We have in the 12 programmatic area, we focus on what do we have to do 13 to assure that there is sufficient quality and 14 reliability in our safety significant SSCs in being 15 able to deliver these performance targets.
16 And the IDEP basically takes the lead in 17 setting what these performance targets should be, 18 based on looking at what's coming out of the PRA, 19 considering how far off the reliability might be from 20 what was assumed or assessed in the PRA, as well as 21 the capabilities to mitigate the accident. So they 22 set the performance requirements and then they do a 23 lot of critical look at the uncertainty treatment.
24 The PRA's going to do its best to quantify 25 the range of uncertainty and the frequency and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
119 1 consequence assessments that are associated with the 2 current state of the art of PRA as well as leaving all 3 the requirements in the PRA standard.
4 It's just going to look beyond that to 5 look at other unknowns that just aren't jumping out 6 and addressed, the same kind of thought process that 7 the NRC Staff would normally use to take a look at a 8 license application. But the idea is to build up a 9 set of attributes and evaluation criteria that can be 10 audited by the Staff.
11 Now, back to one of George's suggestions 12 here. We're trying to give some kind of an idea on 13 some evaluation criteria that can be used to 14 established when is enough, enough. And one of the 15 tools that we came up with to look at that is this 16 table that basically is used to examine all the LBEs 17 that are coming out of the PRA in terms of the layers 18 of defense, organized by layers of defense.
19 And we have both quantitative and 20 qualitative criteria for each layer of defense, well 21 for most of them. For the first layer we just have a 22 qualitative semi-quantitative target to make sure that 23 we keep the frequencies of upset transients under 24 control.
25 Then for each of the layers of defense, we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
120 1 need to make sure we keep the LBEs in the right 2 frequency range because it effects the decisions we're 3 going to make, which class that they're assigned to.
4 And we also want to meet some qualitative 5 criteria that minimize the frequency of challenges to 6 our safety-related SSCs, but we also want to adopt the 7 no undue reliance on the single element of design to 8 perform an important safety objective.
9 So, this is where we might get a safety 10 significant SSC added into the mix that may not 11 necessarily be risk-significant. And we also have 12 criteria that go cross the entire frequency 13 consequence spectrum that, again, brings up the not 14 relying on the single element design.
15 So, this was one of the things that went 16 beyond what NGNP came up with that's intended to help 17 reach a conclusion by the integrated decision panel on 18 when they believe there is sufficient defense in 19 depth.
20 There are judgments made as to when it 21 makes sense to begin this process and it sort of 22 depends on the stage of the PRA, stage of design 23 development and so forth.
24 But at some point in the early stages of 25 design, there'll be a baseline evaluation which will NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
121 1 create a document. And this baseline evaluation and 2 document will be updated periodically as the different 3 stages of design and licensing evolve.
4 What they will come to conclusion on, this 5 integrated panel, is they'll come to a resolution of 6 is the plant capability DID considered to be adequate?
7 Are the criteria on this previous Table, that would be 8 5.2 in the document, are satisfied?
9 Review of the LBEs completed with 10 satisfactory result with critical review especially on 11 the risk-significant LBEs? Is the programmatic DID 12 deemed to be adequate?
13 What are the performance targets that have 14 been set for the reliability, capability of all the 15 safety significant SSCs? Are the sources of 16 uncertainty in selecting and evaluating the LBEs 17 identified and have they been adequately addressed in 18 these protective strategies?
19 And then, finally, the panel, the very, 20 very, important outcome is what special treatment 21 requirement should be selected for each of the safety-22 related and non-safety-related SSCs.
23 So, that's the process. It's done by an 24 integrated panel and it'll create a report that will 25 be part of the defense in depth evaluation and can be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
122 1 reviewed or audited by the NRC Staff at any stage of 2 the licensing process.
3 CHAIR BLEY: Karl, thanks. I think we're 4 going to hold any questions here to the end. Can you 5 gentlemen stay until the end of the meeting?
6 MR. FLEMING: Yes.
7 CHAIR BLEY: Okay, good, because I want to 8 talk about what comes next and what we might want to 9 revisit. But at this time I think we want to move to 10 the Staff and hear what they have to say. We'll 11 probably go a little fast, but they say they don't 12 have much to say so what do we do? They will in 13 October for sure. Thank you.
14 MR. FLEMING: Thank you.
15 MR. RECKLEY: Okay, thank you. My name is 16 Bill Reckley with the Staff, and John Segala is here 17 as well. We're going to talk about our plans going 18 forward starting with what you've seen in terms of 19 Revision M of the licensing modernization what we 20 expect to get in an NEI Guidance.
21 And then our development of the associated 22 draft Guide of 1353 Guidance for Technology-Inclusive, 23 Risk-Informed, Performance-Based Approach to Inform 24 the Content of Applications for Licenses, 25 Certifications and Approvals.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
123 1 So you'll start to hear, I think, a little 2 bit of a difference in emphasis in that the REG Guide 3 is intended to -- the rule to which the REG Guide 4 applies is 50.34, 52.79 on the content applications.
5 And so, what Karl was describing in terms 6 of how the design was done, obviously there's a close 7 correlation between how the design is done and what 8 gets put into a licensing application, but our focus 9 is going to be more on the license application.
10 I'll go quickly through the background, a 11 lot of this has been mentioned. Proposals very similar 12 to this one had been brought to the Commission and 13 even to the ACRS. As Dr. Bley mentioned, NUREG 18-60, 14 some similarities. NGNP, obviously, even more 15 similarities.
16 In terms of recent activities, John 17 mentioned earlier, we had come to the ACRS with the 18 vision in strategy and implementation action plans and 19 then more recently with REG Guide 1.232 in the 20 functional containment paper. All of those related to 21 this overall concept of how will non-light-water 22 reactors do design and then from our point of view, 23 make applications.
24 One of the things coming up are additional 25 visits to the ACRS. One of the things that I try to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
124 1 keep in mind is kind of an integrated or holistic 2 approach. And we often talk about safety and whether 3 this design is more safe, less safe. I prefer to talk 4 about not more or less, but how they get to where they 5 are.
6 For large light water reactors, you cannot 7 overlook that part of the safety case for large light-8 water reactors is the ability to move the people out 9 of the way. And so you get emergency planning in 10 terms of the mitigating strategies, it's an important 11 mitigating strategy.
12 As we move forward on advance reactor 13 design, the commission policy statement on advance 14 reactors is, let's put a little less emphasis on 15 moving people out of the way or having operators have 16 to act quickly or increase thermal margins so reactor 17 protection systems don't have to act in a matter of 18 seconds or things go badly fairly quickly if they 19 don't.
20 Increase the thermal capacity of the 21 system we were talking about earlier, that's graphite 22 and sodium. The advantage those coolants bring, they 23 have disadvantages, I'm not trying to make a case one 24 sided here. But one advantage is that they burn is 25 the larger thermal capacities in terms in comparison NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
125 1 to water. And so all of those things keep in mind 2 that are holistic or integrated approach.
3 So, we're here today to line up the 4 discussion of licensing modernization. You have had 5 more recent discussions on the advance reactor design 6 criteria. We talked to you about functional 7 containment, which kind of crosses between both 8 prevention and mitigation. In August --
9 CHAIR BLEY: Is there a paper up at the 10 Commission?
11 MR. RECKLEY: Almost. Shortly. You'll 12 hear in August the proposed emergency planning rule 13 for small modular reactors and non-light-water 14 reactors or other new technologies. So we're trying 15 to keep this holistic thing in mind as we go forward 16 and that's part of the challenge.
17 Dr. Apostolakis mentions security. We are 18 in advance reactor space and small modular reactor 19 space, even bringing some security in with some 20 potential consequence based security measures.
21 So getting back to my point about content 22 applications. This isn't even all, this isn't all 23 inclusive, but basically if you take the 19 Chapters 24 of the FSAR and you take some other aspects or parts 25 of applications as they come in, this is a partial NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
126 1 picture of everything that a designer or a license 2 applicant needs to bring.
3 For large light-water reactors, this 4 puzzle has largely been defined. It's been modified 5 and clarified over the years, but we're at a fairly 6 stable point where this picture is defined for large 7 light-water reactors.
8 For non-light-water reactors and even for 9 small modular reactors as you get into the new scale 10 designs, we'll start to see focuses change like much 11 of the guidance that's prepared, and I'll just pick 12 one out, Chapter 8 on electric power.
13 That's been largely defined because for 14 large light-water reactors using, at least for the 15 operating fleet, using active components to provide 16 the cooling, you need power. And so all of the, or 17 much of the guidance related to off-site and on-site 18 back-up power's been developed with that particular 19 model in mind, the importance of diesel generators and 20 so forth.
21 As you switch over to small modular 22 reactors and non-light-water reactors with the larger 23 thermal capacities, the importance or the need for 24 immediate back-up power or reliance on active systems 25 is diminished. And so, it changes, I'm not going to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
127 1 say it's zero, but depending on the plant design, it's 2 going to change.
3 So, what we need is a way in order to 4 redefine the puzzle for non-light-water reactors. And 5 in addition to that particular challenge, each non-6 light-water technology and each design within a 7 technology can be different. So it's going to be very 8 difficult to simply say we're going to replace.
9 People talk about replacing NUREG 800 or 10 Reg Guide 1.206 with something for non-light-water 11 reactors. That would just be taking a prescriptive 12 approach for large light-water reactors and saying 13 we're going to repeat that for however many designs 14 might face us. So, we're really looking at more of a 15 methodology.
16 CHAIR BLEY: Bill, before you go on. Both 17 Dr. Apostolakis and you brought up security. I 18 vaguely remember, and it's probably something that we 19 saw. I thought for somewhere in this new reactor area 20 there was or was to be a paper on integration of 21 safety and security.
22 We had a briefing from NSIR a year ago, a 23 year and a half ago on vulnerability assessments which 24 are, at least they're comprehensive kinds of analysis 25 on PRAs. Is there any of that that is getting NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
128 1 factored into the process you guys are working on and 2 is that something that will come here?
3 MR. RECKLEY: We are trying, at least on 4 the security basically gets broken down into theft and 5 diversion and then sabotage. So I'll limit the 6 discussion to sabotage because it can more closely 7 align with just external events and so forth.
8 CHAIR BLEY: Exactly.
9 MR. RECKLEY: So, in the area of sabotage, 10 NEI submitted a paper on a consequence based 11 approached basically saying how can a sabotage event 12 lead to a core damage in a potential off-site release 13 and what's the timing of those?
14 And that is going to be in a paper going 15 up the Commission about the same time as functional 16 containment where we ask the Commission to undertake 17 a rulemaking to define that.
18 So, it's somewhat related to what Joe 19 Rivers came and talked to you about in terms of the 20 integration. But that was also, I think, including 21 the operating fleet where --
22 CHAIR BLEY: Well, it was at that time.
23 MR. RECKLEY: Yes. We can be a little 24 more focused on just non-lights and small modular 25 light-water. So the methodology I mentioned in how to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
129 1 put these puzzles together is what we see coming out 2 of the licensing modernization. So I took liberty 3 with the NGNP graphic on defense in depth and kind of 4 say how we can see this coming about.
5 So you have the LMP focus, which is really 6 on the defense in depth, the Chapter 15 traditional 7 deterministic design basis accidents and Chapter 19, 8 which is the probabilistic risk assessment insights 9 and some other related discussions and you take that 10 as being the assessments that are done.
11 And you're going to use those assessments 12 to sharpen really what's the heart of a nuclear 13 design, which is Chapters 4, 5 and 6, the reactor, the 14 reactor coolant system and engineered safeguards and 15 retention of fission products.
16 And this is iterative. One of the things 17 I think got mentioned in the earlier discussion, but 18 is just imperative that people keep in mind is this is 19 not a linear process.
20 This is iterative both by the designers, 21 well iterative by the designers. Hopefully, by the 22 time they're making an application they've gone 23 through all the iterations such that they've made 24 their choices.
25 But, you have not only the plant NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
130 1 capability defense in depth arguments that Karl 2 mentioned, but the programmatic. And we really are 3 looking to see how all of these fit together because, 4 again, you need to take this holistic view.
5 A particular designer, when they're 6 looking at a reactor coolant option, they can try to 7 design away any concerns or they can use a combination 8 of design and programmatic controls to provide the 9 same level of comfort or assurance that that system is 10 going to work.
11 So this isn't really any different than 12 what evolved for the large light-water reactors. But 13 that evolved over a long period of time and what we're 14 trying to do is make sure that when the applications 15 come in, we're already well along on that process.
16 So all of this has to be thought about by 17 the designer by the time they're coming in for a 18 design. And I bring up the importance of the 19 programmatic and physical defense in depth measures 20 just because, again, especially for non-light-water 21 reactors, my personal view, the problematic controls 22 are going to be imperative to make up for some of the 23 lack of operating experience or sparseness of 24 operating experience.
25 And so, when they come in they really have NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
131 1 an option of how much testing they've done to say I 2 don't need to do any additional surveillances once I 3 move over into operations or as part of a business 4 risk, I've done a little less testing ahead of time 5 and I'm going to do online surveillances to make sure 6 it behaves the way I thought. And, again, the 7 business risk is if it doesn't, oops.
8 But, from a public safety point of view, 9 the two can be made equivalent. From the business 10 case, maybe not so much. But that's the advantage of 11 being on the NRC side of the table is we don't have to 12 worry about that aspect.
13 MEMBER CORRADINI: Bill, can I clarify 14 what you just said a different way. So if a designer, 15 vendor, owner all together want to do this, they could 16 take a prototype route and do a power ascension to 17 prove on the device as it's coming up in power -- but 18 they take a business risk. But this would be 19 potentially just an acceptable way of going through 20 it.
21 MR. RECKLEY: From a public health and 22 safety point, yes.
23 MEMBER CORRADINI: Okay, that's what I 24 thought you were getting at.
25 MR. RECKLEY: Okay.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
132 1 MS. CUBBAGE: This is Amy Cubbage. I just 2 want to interject that we're not necessarily seeing 3 would have to be a prototype. If the Staff can make 4 a reasonable assurance based on the information 5 available, plus the programs that are included in the 6 license, then that would be a potentially a non-7 prototype situation.
8 MR. RECKLEY: Yes. It would be a prototype 9 or it could be just a combination of technical 10 specifications and in-service inspection and all of 11 the things that go into monitoring a plant once it 12 gets built.
13 We talked about support systems, what I 14 call support system, instrumentation, electrical 15 power, auxiliary systems, power conversion, all of 16 those things. The level of detail in the FSAR would 17 be informed by and determined by what the actual 18 safety functions and the risk insights that come out 19 of the LMP process to inform that part of the 20 application.
21 And, again, it'll be different for 22 different technologies, it'll be different for 23 different designs within a technology.
24 CHAIR BLEY: Bill, looking ahead, I don't 25 see any slides that go through the comments you made NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
133 1 so far. We've seen some of those. Mostly they seem 2 to me they're kind of detailed questions, some open 3 ended questions and some we think we might prefer some 4 other approaches.
5 Do you consider those well developed at 6 this time or where do you think this is heading? By 7 October you expect to have a document.
8 MR. RECKLEY: By October we expect to have 9 a document. So, let me just skip to Slide 7 real 10 quick. Noting again the demarcations and the cutoff 11 remain under discussion and that's one of the things.
12 Now, when I say we're going to be back in 13 October, that should be a bit of a hit that I don't 14 think we're night and day apart. But whether things 15 shift by a half a decade or something within the 16 curve, that's one of the points we'll discuss and 17 we'll either agree and the industry will change or if 18 we think strongly enough about it, we would take an 19 exception in the REG Guide.
20 One of the things that I wanted to mention 21 because I bring up the cutoff or the lower bound, 22 we're also coordinating this with other activities.
23 For example, NuScale has come in with a report on 24 trying to define credible in order to inform their 25 source term used for various regulatory questions.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
134 1 We obviously see the relationship between 2 that activity and something like a lower bound for the 3 beyond design basis events here. Not to say the 4 answer will be the same, but we need to make sure that 5 we're at least cognizant of simpler regulatory 6 decisions being made in two different areas, if you 7 will, the light-water small modulars and the non-8 lights.
9 And just to reemphasize, and this is the 10 Staff's bullet that they put into the Draft M, the 11 target values shown on the graph are not acceptable, 12 unacceptable, they're a frame of reference. The way 13 I like to think about it, it's a frame of reference.
14 So, you can take safety functions, you can 15 take structure systems and components and you can 16 start to do the assumptions on whether they fail and 17 then that tells you where you're moving on the graph.
18 And the target line, I generally agree 19 with the observation that if you're on the right side 20 of that line or above, you're probably want to give a 21 lot of thought before coming in with an application if 22 you've got things on the other side of the graph.
23 But on the left and lower side, really 24 it's just providing a frame of reference so you can 25 see, hey I need to make that safety function or that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
135 1 SSC much more reliable. Maybe I need to add redundancy 2 or diversity in order to make sure that the 3 reliability of that function is there. Or maybe I 4 need to change and add another wall because I'm going 5 to try to address it from the consequence side of the 6 equation.
7 But the designer is using this to make 8 design decisions and obviously the expense comes in.
9 Is it more cheap or is it cheaper to add diversity and 10 redundancy to improve reliability or is it cheaper to 11 add a wall to lower the consequence? So from the 12 overall perspective of the Staff, either one of those 13 might provide adequate protection, so it's up to the 14 designer to propose.
15 MEMBER REMPE: So I still would bring up 16 what I brought up this morning that you may not be 17 able to meet some regulatory criteria that have two 18 hour2.083333e-4 days <br />0.005 hours <br />2.97619e-5 weeks <br />6.849e-6 months <br /> limits on those values and their just making 19 decision on the design.
20 MR. RECKLEY: And as Karl emphasized then 21 and I'll emphasize as well, this does not mean you 22 meet all, this is populating Chapters 15 and 19. All 23 right? You still have the equivalent of Appendix I, 24 that's for light-water reactors, but even without that 25 you have --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
136 1 MEMBER REMPE: There's a 10 CFR 20 2 requirement 3 (Simultaneous speaking.)
4 MEMBER REMPE: -- why not put that on this 5 plat if you're making design decisions.
6 MR. RECKLEY: Well, again, it's a frame of 7 reference for evaluating events and how much you move.
8 It's not saying that if all of my Chapter 15 and 19 9 events fall a particular way that I've met all of the 10 other requirements like are in Chapters 11 and 12 or 11 in Appendix I or even in the EPA effluent 12 requirements, all of those things still apply.
13 Much of those, including Part 20, and this 14 is one of the age old questions of bringing even Part 15 20 into it, Part 20 is really meant for normal 16 effluence and that's -- so it's kind of artificial to 17 even bring in to assess events.
18 But you had to bring something in and this 19 might be, again, the Staff's not willing to say yet, 20 but you needed to bring something in as a frame of 21 reference at the lower doses and Part 20 has been a 22 proposal since NGNP to use.
23 MEMBER REMPE: And then, of course, if 24 they're designing it and they want to have an economic 25 plant, they're going to want to reduce margins. And NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
137 1 so, I would again emphasize what's an acceptable 2 margin, because that's going to be a discussion in 3 having this Panel making the decision as indicated by 4 one of the slides we just saw.
5 MR. RECKLEY: Right.
6 MEMBER REMPE: It's open ended.
7 MR. RECKLEY: It is, and again, from 8 looking at it from an event standpoint, we can do 9 this. From the other side of the business equation, 10 to say that a designer is going to go out and say I've 11 done this on the cheap and I'm going to release as 12 much radioactive material as the regulator would 13 possibly allow me to do and I want to put it next to 14 your house, you know, that's another aspect of the 15 business case that might not work out. So, all of 16 that would remain to be seen.
17 So, going to what the Staff has as some, 18 a few remaining items to work out between now and 19 September. The F-C target figure, we would like to 20 come to an agreement and the ball is in the Staff's 21 court on that, if we're going to propose anything 22 different that the Staff has to propose.
23 I mentioned the lower range, we are 24 looking at the decisions that are being made in other 25 areas so that we can be consistent. There's some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
138 1 remaining questions on the consequence analysis and 2 the ACRS numbers that we're asking some questions of 3 that as well in terms of the source term and how it's 4 used. Consideration of uncertainty is obviously, 5 that's a big area.
6 The role of the non-light-water reactor 7 PRA standard? We think we have that worked out. In 8 large part the NRC is on board of saying we plan to 9 review and hopefully endorse that standard as it's 10 finished.
11 But one of the things that we have to look 12 at as we're building this particular construct is 13 standards sometimes take a long time to develop too.
14 And so, our process if we're building in a dependency 15 on a future standard, we just have to work out exactly 16 the ramifications of that. The terminology, we agree 17 that it's imperative. We really want to get that 18 right from the start.
19 We've had some feedback on making sure 20 that this system has the flexibility for smaller and 21 simpler designs. The number of designs keeps growing.
22 Some of them are relatively simple. And so, we just, 23 again, we're not saying we see an issue, but we just 24 need to review it and make sure that it would be 25 beneficial for them.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
139 1 External hazards we think needs to be 2 clarified how it's written. I don't think we have an 3 issue, but it just needs to be a little more clear in 4 the write-up.
5 And then interface with other 6 requirements, such as emergency planning. Again, 7 going back, we would like an integrated approach and 8 so we would like to know how all these pieces fit 9 together.
10 And there are areas where the report 11 mentions how it might be used, for example, to show 12 that the doses are less than a certain amount and, 13 therefore, that might help support the discussion on 14 emergency planning zones.
15 We just need to make sure that gets clear.
16 If not in the guidance, then that's an area we would 17 pick up in the REG Guide. And then, although this is 18 largely a design process, we really are curious how it 19 moves into the operations phase so that we have a 20 continuity. That's an area that for the operating 21 fleet I think is not worked out optimally, I can say 22 that.
23 One item I want to add quickly, and this 24 will go to Dr. Bley's suggestion that we make this a 25 full day in October, is there's really another product NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
140 1 that the Staff envisions which is a related Commission 2 paper in that some of the things like the frequency 3 consequence curve and once we agree on the 4 demarcations, we think it's likely that that might be 5 a policy matter.
6 If it's not a commissioned policy 7 decision, it's at least something the Commission is 8 going to be interested in. And so we really see three 9 products being discussed in October, the Industry 10 Guide, the related Regulatory Guide, draft Regulatory 11 Guide and a related Commission Paper in which we bring 12 to the Commission's attention anything that we think 13 they need to be informed of. And, personally, I think 14 there will be some of those to cross over into 15 something the Commission will want to vote on.
16 CHAIR BLEY: Back in the days of 18-60, 17 there was almost rulemaking.
18 MR. RECKLEY: Yes.
19 CHAIR BLEY: There was almost an 20 application project with a new design.
21 MR. RECKLEY: Yes.
22 CHAIR BLEY: And there was a Commission 23 Issues Paper that went up on a lot of these things.
24 MR. RECKLEY: Right.
25 CHAIR BLEY: Wasn't that included at that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
141 1 time? Did the Commission respond?
2 MR. RECKLEY: To some of the issues --
3 CHAIR BLEY: To the F-C curve, in 4 particular.
5 MR. RECKLEY: I don't think for the F-C.
6 CHAIR BLEY: It wasn't in there? Okay.
7 MR. RECKLEY: Because that --
8 CHAIR BLEY: It didn't go back --
9 MR. RECKLEY: That was the 18-60 and the 10 advanced notice of proposed rule making and so I 11 think, my memory is the Commission would have said go 12 ahead and propose, but they didn't accept it.
13 CHAIR BLEY: There was an issue paper 14 separate from the AMPR.
15 MR. RECKLEY: Right. Right, beforehand.
16 CHAIR BLEY: But you don't think this was 17 in there. Okay. Or they didn't respond to it.
18 MR. RECKLEY: Or they didn't respond.
19 CHAIR BLEY: Okay.
20 MEMBER REMPE: So in the document they 21 gave us they refer to the criteria that you had for 22 multi-modular risk and they said that those had almost 23 been approved. What is the status on that?
24 MR. RECKLEY: That's in interim Staff 25 Guidance document. If we can, I forget the number, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
142 1 28? I'm making that up.
2 MEMBER REMPE: But you had issues, but --
3 (Simultaneous speaking.)
4 MR. RECKLEY: For light-water small 5 modular reactors we usually --
6 MEMBER REMPE: Just the light-water ones.
7 MR. RECKLEY: Well, that was the focus.
8 We'll have to look and this is part of the activity 9 that we'll need to do to make sure that it fits, but 10 that was issued with reactors like NuScale in mind 11 which has 12 reactors to a plant.
12 MEMBER REMPE: But you will be evaluating 13 if applicable as part of what we'll see by October?
14 MR. RECKLEY: Yes.
15 MEMBER REMPE: Okay.
16 MR. RECKLEY: And then last bullet on here 17 is there was a more recent SECY paper 18-60, is that 18 coincidence, I guess, SECY 18-60. That is the result 19 of the more recent activities in considerations of a 20 NRC transformation team. One of the recommendations 21 in there, for example, was that we go ahead and pursue 22 a Part 53, the technology neutral rule.
23 MEMBER CORRADINI: So can I interject, 24 Bill? What is your view of that? That seems to be a 25 needless years long approach.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
143 1 MR. RECKLEY: I would say the Commission 2 is going to tell me what my view of that is.
3 MEMBER CORRADINI: I mean, Dennis brought 4 it up in 18-60 when that was going to be there, but 5 that would be ten years ago and probably five of it 6 would have been taken up and using it. I'm very 7 concerned about this pushing it down the road.
8 MR. RECKLEY: I will just chime in that 9 what we're doing now, if we were directed to do a 10 technology neutral rule Part 53, what we envision is 11 taking what we're doing now as the heart of it. And 12 so, it would not be -- it would be kind of like a 13 relay race where we would say okay, now let's take 14 what we've agreed to, maybe we've issued this draft 15 Guide.
16 I think Amir mentioned the tabletops that 17 we'll be doing, some before and some after the 18 issuance of the draft Guide, we'll get some lessons 19 from that. As we go forward that will put us in a 20 better place to do a rule making, I think, if we're 21 directed to do that.
22 And then lastly, just the schedule we've 23 talked about. October 30th is the next subcommittee 24 meeting. I think I would agree that a full day would 25 probably be needed especially if we're throwing in a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
144 1 paper. It is a draft Guide and so in terms of ACRS, 2 that might give you a little flexibility in that you 3 would, as you did for the Advance Reactor Design 4 criteria, you get kind of two bites at that apple.
5 Really, from my point of view, at the 6 draft REG Guide you would just be looking for fatal 7 flaws but not, we could revisit at the issuance of the 8 final Guide more specific areas. But the SECY and the 9 policy issues would also be including some 10 recommendations that you would probably want to write 11 a letter on.
12 CHAIR BLEY: I think so. I mean all of 13 this is really significant. And giving it some time 14 is important.
15 MR. RECKLEY: Right. And we can work out 16 between now and then if there's any other 17 interactions. I don't know if we have time for 18 another meeting, but certainly we can share things 19 with you between now and the end of September.
20 Although, that's not that far away.
21 And so, after the full committee meeting 22 in December, we would plan to issue the draft REG 23 Guide later that month. Our goal is to try to do it 24 by the end of the year. The SECY, we would complete 25 in early 2019. And then the final REG Guide is a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
145 1 little juggling here, because it's going to be 2 informed by the Commission, the ACRS, the tabletops 3 and any number of other things that will be going on 4 at the time. So, with that --
5 CHAIR BLEY: At this time, I want to talk 6 a little bit about that October meeting and if all of 7 you or one of you wants to come up, or just Amir? At 8 least from my point of view, if the OMP group could 9 come back, I think it would be extremely useful, 10 especially somebody to talk about details.
11 And the ones that jump at me as I went 12 through the white papers and the main document, I 13 think if whoever did that pulled out the flow charts 14 from all of those, that would be a good place to walk 15 us through the whole process on each of them.
16 And then we could go into other details, 17 and we would want back-up material to support that.
18 But we've kind of glossed the surface and I think we 19 need some time to really dig in.
20 So if you can support that, I think it 21 would be useful if the Staff presents it from their 22 review, you don't get your voice in here and I think 23 this is a significant kind of feel, I think that would 24 be useful.
25 MR. AFZALI: We would be happy to do so.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
146 1 CHAIR BLEY: Okay. And all of you will 2 work with Derek Widmayer on our Staff to try to focus 3 all together. The date you have up there, September 4 28th is really important.
5 I mean, really important, because this 6 stuff is significant, and I don't know if you're going 7 to have any revisions by that time. We need them a 8 month ahead and if, I don't want to slip your 9 schedule, but if we don't get them a month ahead given 10 the significance, I think we ought to slip.
11 MR. RECKLEY: That's actually why I put 12 that on there. That's a forcing function for us to 13 have everything done, because October 30th sounds, 14 that gives us an extra month. But we really don't 15 have that.
16 CHAIR BLEY: For us, you don't have a 17 month. We need a month to look at it.
18 MR. RECKLEY: Right. And so that --
19 (Simultaneous speaking.)
20 MR. RECKLEY: -- really puts us at the end 21 of September and that's the reason for putting that on 22 there.
23 CHAIR BLEY: Amir, do you expect 24 substantial changes in drafts that we would see 25 beginning of October, end of September?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
147 1 MR. AFZALI: I do not expect significant 2 changes. Our conversation with the Staff may result 3 in some numerical changes here or there, but in terms 4 of our approach and the contents.
5 CHAIR BLEY: Yes. I think most of us are 6 going to need to study those white papers some more to 7 really get comfortable with the detail in there.
8 MR. FLEMING: If I might just add a 9 comment to Amir's is that one of the things that we 10 are going to try to work on is the terminology. We're 11 working on putting together a glossary and we want to 12 get an agreement with the NRC Staff on the glossary.
13 And then we need to do a better job of, you know, 14 getting the language right for the right words, 15 crediting and, you know --
16 CHAIR BLEY: Maybe we should be talking 17 about March or April, because somehow glossary is one 18 of the hardest things to agree on.
19 MR. FLEMING: We won't claim we'll resolve 20 the issues, but we're going to work on it. Okay?
21 CHAIR BLEY: I'm sorry, Amir?
22 MR. AFZALI: As you have spent time a 23 little bit on is that this margin question that has a 24 significant conversation about user margins, if you 25 see the margins. So we want to provide some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
148 1 clarification of how the margins are presented and how 2 those could be used as part of decision making.
3 CHAIR BLEY: I think that's going to be 4 very important. People will be interested -- Dick, 5 was it you? I'm trying to think. Please go ahead.
6 MEMBER SKILLMAN: Yes. Thank you, Dennis.
7 My question is for Bill. And my question is on the 8 topic of terminology, how wide a swath do you see you 9 will be taking?
10 And here's the reason for my question. If 11 I look at the wording in 50-59 and I look at other 12 documents that men and women that work the plant sites 13 use, that's where the terminology issue becomes 14 critical. And it becomes critical in operability 15 determinations, it becomes critical in requests for 16 exigent expect changes.
17 And I know this is a long way down the 18 line, but since were in the embryonic stage of the new 19 reactor designs, now's the time to make sure that even 20 though subordinate or other important licensing 21 regulatory documents line up with the vernacular 22 that's being used in the new design.
23 And so I know that that will take time, 24 but as we use these new words or get aligned on the 25 terminology it needs to be aligned not only in the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
149 1 design documents, but in the implementing documents in 2 the regulations.
3 MR. RECKLEY: No. I agree. And that's, 4 again, when I had the bullet that says how does this 5 carry into operations that's --
6 MEMBER SKILLMAN: Thank you, Bill. That's 7 my point.
8 MR. RECKLEY: And Mike was asking, 9 although I'll take his question as being asking my 10 personal opinion, that's one of the reasons why I 11 don't dismiss Part 53 as quickly, is because even 12 within this discussion that you've had here, you've 13 seen differences between how the very similar terms 14 are used in licensing modernization as they are in 15 comparison how they're used for the operating fleet.
16 And that will be a constant source of issue.
17 Whereas, if we started somewhat with a 18 clean slate and said these reactors are on Part 53, 19 they go their own terminology. Don't confuse it with 20 the 50 year history of the operating fleet, there's 21 some advantage. I know as Mike was saying, there's 22 disadvantages to the rule making, but there's also 23 potential advantages.
24 MR. SEGALA: And we've also had a meeting 25 on LMP a couple weeks ago where we talked about the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
150 1 glossary and the Staff provided feedback on certain 2 terms that we identified that weren't being used 3 consistently. And they started flagging them and then 4 they counted, similar to what you had done, they 5 counted the number of times all those words were used 6 in the documents.
7 MR. RECKLEY: They even used the term 8 anti-glossary. There's a glossary and there's a anti-9 glossary on words to avoid using.
10 CHAIR BLEY: I had one other item I wanted 11 to mention, I didn't raise it when Karl was talking.
12 On the process for selecting and evaluating licensing 13 basis events, and Jose touched on this. As you said, 14 it's a iterative process. You have go round and round 15 and process many times.
16 But there are parts of the PRA, they 17 cannot be really complete until you've got all of your 18 abnormal procedures, your emergency procedures, you've 19 got a crew and a trained crew and, you know, it's 20 pretty far on the design. It's after, at least under 21 the current processes, it's after the licenses have 22 been issued.
23 I think you need to think about more the 24 possibility that in that process new licensing basis 25 events might come into your list from this finalizing NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
151 1 this PRA as everything gets more and more complete 2 both from the HRA and there might be more external 3 things that evolve. And think of what kind of process 4 you can have so you don't need a licensing amendment.
5 Something built into the program before 6 you load fuel that would allow incrementally adding 7 some SSC, safety-related SSCs and LBEs as such that 8 we're not caught in a spot that bringing this thing to 9 conclusion before start-up put you in a spot, you need 10 a new license. And I hope you've thought that through 11 by October.
12 MR. FLEMING: Yes, that's a very, very 13 good question and a very valid concern. Our thought 14 on that question is that the parts of the PRA that 15 lead to the definition of safety-related SSCs in 16 design basis accidents, we have some confidence that 17 that will be stable throughout the process.
18 That, sure, there'll be changes to our 19 PRA, they'll be maybe new LBEs that show up, 20 probabilistic LBEs that will show up, but the 21 robustness that we need in the PRA for selecting LBEs 22 is just, you know, according to our frequency criteria 23 that we have here is ten to minus four per plant year 24 and there's a little bit better handle on stability of 25 the LBEs up in that range than there would be down the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
152 1 BDBE region.
2 CHAIR BLEY: I certainly believe that, 3 but, you know, some confidence is different than 4 complete confidence and the real world can always 5 surprise us. So I think any thought about it ahead of 6 time and having a way out if this should arise is 7 helpful. I think I heard Harold trying to get in.
8 Harold?
9 MEMBER RAY: Yes, Dennis, before you went 10 on to something else, I just want to say what's 11 related in my mind here is I would like us to be able 12 to touch on a future, is you now got our eyes focused 13 on different categories of SSCs, you just mentioned 14 safety-related, but we know how to, on what basis we 15 take confidence in our assumptions about the safety-16 related SSCs in terms of tech specs and all the 17 assurances and so on.
18 I would like to have, at some point, talk 19 about how we're -- life of a plant, how we're going to 20 have confidence in our assumptions about the other 21 SSCs, which all play a role in the new analysis 22 approach that taking.
23 So I just wanted to lay that on the table 24 and say what is it, is it just experience with these 25 things that are going to the be basis of our NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
153 1 assumptions or are we going to have a meet-in-school 2 kind of application that applies to them throughout 3 their life? What is it that we're going to do that's 4 comparable to what we already have high confidence in, 5 in terms of our understanding for safety-related SSCs?
6 MR. RECKLEY: Right.
7 CHAIR BLEY: Okay. Thanks, Harold.
8 MR. RECKLEY: Yes, this is Bill. Karl 9 mentioned before the reliability and capability 10 aspects that'll go into the special treatment 11 definitions and it's a good question. How do we make 12 those fit once we move over into operations?
13 And, again, it would depend somewhat if we 14 have to try to force fit it into something like the 15 maintenance rule because we're sticking with existing 16 regulations or if we think we might move into a new 17 regulation. That be a little actually, a little 18 easier to define how we did that. In the absence of 19 a rule, we could always do it as part of the license 20 or as part of a design certification.
21 CHAIR BLEY: Okay. Yes.
22 MR. FLEMING: Also, I wanted to make a 23 sort of a parallel analysis of if one followed a 24 traditional ad hoc process you come up with your 25 licensing basis and just go through some of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
154 1 history that George was mentioning. We had a couple 2 of ATWS events that happened and we had to overlay on 3 the licensing basis some new requirements to deal with 4 ATWS because we saw it in experience, station 5 blackouts occurred at different plants and so forth.
6 So both processes, the risk informed and 7 the traditional deterministic basis, have to face when 8 new evidence arises that challenge the judgments that 9 went into it. So it's not like this is only going to 10 be happening with PRA. And it is true that the PRA is 11 a, it's a state of knowledge animal and the state of 12 knowledge is not frozen in time. And we'll have to 13 have process for dealing with new insights.
14 CHAIR BLEY: Well before I go for public 15 comments, is there anything else from Member of the 16 Committee?
17 MEMBER MARCH-LEUBA: Are me going to have 18 round table.
19 CHAIR BLEY: Yes. After the public formats.
20 MEMBER MARCH-LEUBA: I'll wait for the 21 round table.
22 CHAIR BLEY: Mr. Brown, Theron Brown, if 23 you can open the phone line, we'll go there. Is there 24 anybody in the room who would like to make a comment?
25 If so, please come to the microphone and tell us your NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
155 1 name and who you're with.
2 (No audible response.)
3 CHAIR BLEY: I hear no noise. I'm not 4 sure the phone lines open, but if there's anybody on 5 the phone line, the public's line --
6 MR. BROWN: It's open.
7 CHAIR BLEY: -- who would like to make a 8 comment, please identify yourself and give us your 9 comment.
10 (No audible response.)
11 CHAIR BLEY: I guess we have none, so at 12 this time I would like to go around to all the 13 Members. I'm going to do it in reverse alphabetical 14 order just for a change. Matt Sunseri.
15 MR. SUNSERI: Thank you, Dennis. Let me 16 start of by saying I appreciate the challenge or 17 skepticism that's been expressed by my colleagues here 18 today. I think we all have the common goal of wanting 19 to achieve reasonable assurance of adequate protection 20 and that's a pretty clear common goal across the 21 board.
22 So I find it always interesting when the 23 Staff, you know, request the Commission to direct the 24 Staff to come up with some kind of risk informed 25 performance based licensing initiative. So when that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
156 1 happens it creates an expectation that such a process 2 will be provided. Okay?
3 So, here comes along license modernization 4 with an approach to satisfy, in part, part of this 5 risk informed performance based process, which in my 6 mind can lead to a situation of group think, right?
7 So the Staff has an expectation. Here's a way that we 8 can satisfy this expectation. Everybody gets on board 9 and moves forward.
10 So our role is to be somewhat skeptical, 11 I mean, by design. We're here to challenge and you 12 should want that challenge and you need that challenge 13 because the last thing that you want, Staff or 14 industry or ACRS for that matter, is to end up with a 15 process that is less that is, I mean, it's a knowledge 16 that the existing process is not perfect. The new 17 process will not be perfect as well. But the 18 challenge here is that the new process is untried 19 also.
20 So at the end of the day we want to make 21 sure that we have, if we decide to move to something 22 new, that it is as effective as the old of providing 23 reasonable assurance of adequate protection. And so 24 today's presentation, at least in my mind, was 25 designed to stop increasing our confidence level in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
157 1 the new thing so that at the end of the day, we all 2 have confidence that we are going to achieve 3 reasonable assurance of adequate protection.
4 So, I appreciate that. Keep in mind that 5 when we ask our provoking questions, we're not 6 intentionally trying to, you know, antagonize you, but 7 sometimes it comes out that way. All right?
8 But that's just, you know, trying to get 9 at the heart of the issues, the hard issues that's 10 going to make this better at the end of the day. So 11 I appreciate all the presentations and comments that 12 have been made. Thank you.
13 CHAIR BLEY: Thank you. Mr. Skillman.
14 MEMBER SKILLMAN: Thank you, Mr. Chair.
15 First of all, thank you to the Southern Team, to the 16 NRC team and to all of those who have brought this 17 process and the oversight team or the consultant team 18 who brought this to this stage of maturity.
19 I found going through the underlying 20 documents logical, coherent, the language was clear to 21 me and I can draw on my some years of experience to 22 identify where I think slight changes will make a huge 23 increase in effectiveness and reduction of ambiguity.
24 So, I commend the individuals who have been involved 25 here for having done a very thorough job, for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
158 1 providing a logical process that I think will bring 2 consistency to new non-light-water reactor designs.
3 I also find that some of the lessons that 4 might come out of this are very applicable to any new 5 actions that you take on the light-water reactors. I 6 think there's benefit on both sides of the equation.
7 But overall, thank you very much for a great 8 comprehensive piece of work. Thank you.
9 CHAIR BLEY: Thank you, Dick. Dr.
10 Riccardella.
11 MEMBER RICCARDELLA: Yes. Thank you, 12 Dennis. I guess I have a concern about what I asked 13 about earlier, the two and a half millirem cutoff. It 14 bothers me a little. I mean, why do we need to 15 consider an event significant that the maximum 16 consequence is only a factor of ten below normal 17 background radiation. I mean, that just seems 18 unreasonably low to me.
19 I wonder, as an industry, why we keep 20 doing stuff like this to ourselves. I just wonder.
21 The response was well it's an engineering judgment.
22 I just wonder if maybe a little more consideration of 23 that engineering judgment. That's all.
24 CHAIR BLEY: Thank you. Dr. Rempe.
25 MEMBER REMPE: Well, I appreciated NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
159 1 everyone's presentations and I guess my questions 2 today were trying to point at areas where I think 3 additional clarification is needed because I would 4 like to have confidence in this process. And so I 5 hope, I mean I can repeat them, but I think I don't 6 need to waste your time on that. But I hope you'll 7 look at the rap and you'll address the comments I 8 made.
9 CHAIR BLEY: Thank you, Joy. Mr. Ray.
10 MEMBER RAY: I don't have anything more at 11 this time, Dennis, other than to say that we do need 12 to be focused and give ourselves time to work with the 13 Staff and others on it as we move forward.
14 CHAIR BLEY: Thank you. Dr. March-Leuba.
15 MEMBER MARCH-LEUBA: Thank you, Dennis.
16 First, Matt is much more polite and political than I 17 am so I wanted to say to what he said, me too.
18 Exactly, I share in his thoughts.
19 With that said, I make a little impression 20 that I don't like that approach. I like the F-C 21 target approach very much. It's logical, I didn't 22 even call it rational. My problem is with 23 implementation. And my problem with implementation, 24 especially on the Y axis, calculating the frequency of 25 loss events.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
160 1 I would like for the Staff to insure that 2 it is bolded, underlined, italics, that the quality of 3 the PRA calculation must be insured to an extremely 4 high level. Not a level we are used to. If you are 5 going to use it for this purpose, this is a different 6 PRA, it has to be better.
7 Let me give you some samples. All right?
8 I took some numbers from one of your reports and I'm 9 looking at an event tree. It says the probability of 10 losing force cooling on the ML, ten percent. The 11 probability of force cooling via SUSD failing, ten 12 percent. The probability of pump -- system failure, 13 ten percent.
14 What those numbers tells me is that the 15 input data that went into this calculation was at best 16 an expert elicitation. We don't really know what the 17 probability those failures have. And all the numbers 18 are like this, probability of failure, two to the 19 minus five. Probability, I mean, they're round 20 numbers which we don't really know how good they are.
21 Once you process them you get numbers with 22 insignificant digits and you start then to start to 23 believe them. And what's worse, we always ask for an 24 uncertainty. And whereas the CSAU method which we all 25 know at this side of the table, maybe you guys are not NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
161 1 familiar, Washington thermal hydraulic calls because 2 we know what the friction coefficient is, we just know 3 it is a plus minus ten percent area. But we know what 4 the friction coefficient is.
5 The probability of force cooling on ML 6 being ten percent is just an expert number. So, there 7 has to be a lot of review. If we are going to use 8 this frequency access on a definite curve, the F-C 9 target, we need to have confidence on that frequency 10 that we have there.
11 Second point, other than the quality of 12 the PRA, we need to make sure we review the 13 completeness of the PRA. What did we forget? The 14 industry are experts and you are going to be running 15 together with a bunch of experts, but I know for a 16 fact, we review a reactor now I found a serious 17 problem with a scram system that nobody even knew 18 about it. And you have to get a big body of expertise 19 to make sure you do not forget anything.
20 And that is my major bone with using 21 probabilistic risk analysis for serious consequence 22 term. Every time we have had a severe accident, it 23 was because there was an event we did not analyze or 24 we chose not to analyze. In that sentence, I mean, we 25 need to go see the completeness of the PRA and also NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
162 1 the completeness of the DBE list.
2 I'm looking through here, and this is a 3 gas reactor, I don't see any event that leads to 4 ingress of air for this particular event. I don't 5 mind the probability is ten to the minus 43, I like to 6 have a sequence that leads to ingress of air into the 7 core accounted for, and tell me the probably is ten to 8 the minus three, forty three. But it was completely 9 ignored. Okay? Calm down.
10 I don't see it here on this small break, 11 the one that I'm looking at. Okay. So, what we have 12 to make sure is that the list of DBEs is thoroughly 13 reviewed and it has to be an adversarial, someone who 14 doesn't believing in it, like me. Okay? That's it.
15 And we did decide on time for question?
16 Maybe we can follow-up after, unless it is something 17 factual. Right? So with that, I get, leaving room 18 for size, the F-C target curve is really good. It's 19 a good concept, it's logical and rational.
20 Implementation is my problem.
21 CHAIR BLEY: Thank you. Professor 22 Corradini?
23 (No audible response.)
24 CHAIR BLEY: Professor Corradini? Going 25 once. He must have -- he's retired again.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
163 1 MEMBER RICCARDELLA: Yes, there was an 2 email from him that said he had to leave at noon.
3 CHAIR BLEY: Okay.
4 MEMBER RICCARDELLA: He sent you some 5 comments, Dennis.
6 CHAIR BLEY: Oh, good. I haven't seen 7 them yet. Just a second. Dr. Chu? Thank you.
8 Professor Ballinger?
9 MEMBER BALLINGER: I guess I would like to 10 echo Pete's caution because I hope we don't Rockwell 11 ourselves in the sense that we paint ourselves into a 12 corner using 2.5 millirems. I just don't understand 13 that.
14 Also, if we were to never build a future 15 plant, no advance system or anything, I think the 16 output of this is kind of a stealth back door method 17 to feedback on the existing system, which I think is 18 a very good thing. So the exercise itself I think 19 will serve a very good purpose. Thank you.
20 CHAIR BLEY: Mr. Brown.
21 MEMBER BROWN: Obviously you don't know 22 about alphabetizing. Alphabetically done. I'll spit 23 that out since Ballinger does not come before Brown.
24 I'll withstand and bypass the embarrassing --
25 (Simultaneous speaking.)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
164 1 CHAIR BLEY: But Charlie's earlier than 2 Ron so that's okay.
3 MEMBER BROWN: That's okay. Anyway, I 4 don't have anything else to say. I just couldn't pass 5 up the opportunity to add some interesting humor back 6 into this very stealthy, complicated conversation.
7 CHAIR BLEY: Certainly, I appreciate your 8 comments. I would like to echo thanks to everybody 9 from the LMP and from the Staff. Most of the things 10 I wanted to say I've already said, except, and this is 11 close what Jose said, I'll just say it a different 12 way.
13 We need, and especially when we think 14 about the defense in depth, we need a way to force the 15 next people, I'm sure you do it great, but the next 16 people who do this, to really make sure they are both 17 honest and very thorough in identifying the sources of 18 uncertainties because those are essential to doing the 19 PRA right and to having a good basis for the residual, 20 I'll call it defense in depth.
21 For the defense in depth that we had 22 beyond what we've already designed into the plan. And 23 it needs to be a process that really makes people 24 think and reexamine their basis. It's the kind of 25 thing that PIRTs are supposed to do, but they don't.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
165 1 People doing them don't always do it. And you really 2 need to categorize where the holes are in your 3 knowledge as you do this stuff. And I hope you talk 4 about that come October.
5 The other thing I would say is we might be 6 -- I think the arguments for the multi-unit plants 7 might be overly optimistic. And I think you need to 8 make sure you have a way to deal with those and I 9 think you do. Should you not be able to make sure 10 they don't come up into your top list because there 11 could be a design where there is no way around that.
12 And the idea that you can always get them down could 13 be good.
14 The area that we might be overly 15 optimistic is in the ability to define mechanistic 16 source terms that at a level that we really believe 17 them and have considered all the uncertainties there 18 and have enough experimental evidence that could back 19 up what we're doing. Other than that, I think it's 20 well on its way and I really look forward to the next 21 round on this material.
22 Thanks to one and all, we are adjourned.
23 (Whereupon, the foregoing matter went off 24 the record at 12:22 p.m.)
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
Licensing Modernization Project Guidance Document Introduction Amir Afzali- LMP Project Technical Lead Southern Company Dr. George Apostolakis- Member of the Advisory Team Karl Fleming- LMP Senior Technical Lead ACRS Future Plant Designs Subcommittee June 19, 2018
Meeting Purpose and Agenda
Purpose:
To introduce the Licensing Modernization Projects (LMP) proposals and the basis for the proposals Agenda:
- Introduction
- (Amir Afzali)
- High Level Comments on the Proposals Feasibility
- (Dr. George Apostolakis)
- Detailed Description of the Current Proposals and Their Technical Basis
- (Karl Fleming) 6/19/2018 Southern Company 2
Principle LMP Objectives
- Create foundation for an integrated approach to licensing modernization embracing three highly interdependent risk-informed and performance-based topics
- Integrate new advances in RIPB methods and applications that can be used in a technology-inclusive manner for advanced reactor design and licensing
- Reflect the culmination of methods and practices available today to operationalize technology-inclusive RIPB practices recognized across decades of policy and incremental progress
- Pave a coherent path to efficient and effective licensing of advanced reactors 6/19/2018 Southern Company 3
Presentation Objectives
- Familiarize the ACRS subcommittee with key aspects of the LMP TI-RIPB approach as described in the draft integrated guidance document
- Review LMP processes for:
- Selection and evaluation of licensing basis events
- Safety classification and performance-based requirements for SSCs.
- As part of this structured SSC classification process, establish better RIPB-focused special treatment that includes input from DID adequacy evaluation process
- Defense-in-depth adequacy determination
- Mechanism to apply RIPB practices to determining DID adequacy and design and programmatic sufficiency (i.e.,
When is enough, enough?)
6/19/2018 Southern Company 4
Document Development Review Approach
- Discrete topic papers
- Start with NGNP as point of departure LBE Industry NRC
- Adjust to make tech inclusive Paper Review Review
- Reflect changes since NGNP
- Reflect LL from NTTF PRA Industry NRC
- NRC staff review Paper Review Review - Feedback on each white paper
- Comments factored into SSC content extracted for Industry NRC Classn incorporation into RIPB Review Review guidance document Paper
- Final RIPB guidance submitted for DID Industry NRC NRC endorsement Prior Paper Review Review NGNP work Extraction of guidance-related content TI-RIPB NRC Guidance Endorsement NEI 18-04 6/19/2018 Southern Company 5
Quantitative Risk-Informed Decision Making
- LMP proposals present a formal and transparent risk-informed and performance-based process for making key licensing decisions
- Very often, criticisms are focused on PRA without discussing the shortcomings of the traditional deterministic system.
- ASME/ANS RA-S-1.4-2013, Probabilistic Risk Assessment Standard for Advanced Non-LWR Nuclear Power Plants, 2013.
6/19/2018 Southern Company 6
Historical Perspective
- No operating experience was available to support WASH-1400
- Failure data were collected from around the world.
- Expert judgment was used, especially for human errors.
- The WASH-1400 methods and insights regarding internal events have largely stood the test of time.
One of the most consequential PRAs was completed with minimal operating experience 6/19/2018 Southern Company 7
Insights from PRAs
- The deterministic Maximum Credible Accident was a LBLOCA +
LOOP + one active single failure.
- WASH-1400 (1975)
- Major risk contributors: Small LOCAs and transients
- Human errors and support systems are important to risk
- Zion/Indian Point PRAs (1981-1982)
- Earthquakes and fires are among the dominant risk contributors
- Early Rules
- Anticipated Transients without Scram (ATWS), 10 CFR 50.62, (1984)
- Station Blackout (SBO), 10 CFR 50.63, (1988)
- Objectivity, Flexibility, and Burden Reduction
- ROP; AOT extension; RI-ISI A systematic risk assessment improves safety while reducing unnecessary burden 6/19/2018 Southern Company 8
The Commissions 1995 PRA Policy Statement
- The Commissions 1995 PRA Policy Statement:
- A probabilistic approach to regulation enhances and extends this traditional, deterministic approach, by:
(1) Allowing consideration of a broader set of potential challenges to safety, (2) Providing a logical means for prioritizing these challenges based on risk significance, and (3) Allowing consideration of a broader set of resources to defend against these challenges.
- Part 52 and the anticipated Part 50 requirement demand PRA models to be developed.
The proposed LMP approach is consistent with the this Policy Statement and builds on the current NRC licensing requirements.
6/19/2018 Southern Company 9
Achieving Modern Risk-Informed Regulation SECY-18-0060, May 23, 2018
- The staff recommends that the Commission direct the staff to:
- develop an agency wide process and organizational tools to expand the systematic use of qualitative and quantitative risk and safety insights; thereby, enabling staff to scale the scope of review and the level of detail needed in licensing to make a finding of reasonable assurance of adequate protection of public health and safety, beginning with licensing reviews for reactors.
- develop a performance-based, technology-inclusive regulation as an alternative approach for licensing for non-light-water reactors The staffs recommendations are consistent with the LMP proposed framework.
6/19/2018 Southern Company 10
ACRS Letter on the Technology-Neutral Framework (September 26, 2007)
- We concur with the staff that a set of licensing-basis events (LBEs) is needed as part of the licensing basis to structure the interactions between the staff and the applicant and to focus the conduct of mechanistic analyses.
- Identifying the LBEs by using the probabilistic risk assessment (PRA) reduces the risk that licensing-basis requirements will divert attention from events of real safety significance.
- The use of a frequency-consequence (F-C) curve is an appropriate way to establish a range of regulatory requirements to limit radiation exposure to the public.
6/19/2018 Southern Company 11
LMP Advisory Group Position on LMP
- Letter to S. Kuczynski by Apostolakis, Merrifield, and Meserve (2/20/2018)
- Although we are not in a position to comment on the technical adequacy of the reports generated by the LMP, we enthusiastically endorse the effort. In particular, we believe that the focus on a systematic and predictable process for early resolution of fundamental technical issues in the licensing of advanced reactors can reduce uncertainty in the development of a design.
- We believe that the work lays the foundation for guidance that can be endorsed by the NRC and we encourage the continuation of the effort.
6/19/2018 Southern Company 12
LMP Advisory Group Position on DID (12/11/2017)
- Given the significantly reduced risks associated with many of the advanced reactor designs as compared to LWRs, a non-risk-informed application of DID could result in excess conservatism in safety reviews.
- Moreover, while we believe that adopting a TI-RIPB DID framework will generally reduce excess conservatism and provide a more effective means to limit unnecessary regulatory burden, like all risk-informed tools, it could also result in the identification of areas where additional requirements are necessary.
- The great value of the LMP DID proposal is that it is a first step toward converting what is currently an ambiguous DID philosophy to a concrete DID process.
6/19/2018 Southern Company 13
Selection And Evaluation Of LBEs 6/19/2018 Southern Company 14
Selection and Evaluation of LBEs
- AOOs, DBEs, and BDBEs are defined in terms of event sequence families from a reactor design-specific PRA
- Individually for risk significance using a Frequency-Consequence (F-C) chart against a F-C Target
- Collectively by comparing the total integrated risk against a set of cumulative risk targets
- DBEs and high consequence BDBEs are evaluated to define Required Safety Functions (RSFs) necessary to meet F-C Target
- Designer selects Safety Related SSCs to perform required safety functions among those available on all DBEs
- DBAs are derived from DBEs by assuming failure of all non-safety related SSCs and evaluated conservatively vs.
10CFR50.34 6/19/2018 Southern Company 15
F-C Target 6/19/2018 Southern Company 16
LBE Risk-Significance Criteria 6/19/2018 Southern Company 17
LBE Cumulative Risk Targets
- The total frequency of exceeding an offsite boundary dose of 100 mrem shall not exceed 1/plant-year to ensure that the annual exposure limits in 10 CFR 20 are not exceeded.
- The average individual risk of early fatality within the area 1 mile of the EAB shall not exceed 5x10-7/plant-year to ensure that the NRC Safety Goal Quantitative Health Objective (QHO) for early fatality risk is met
- The average individual risk of latent cancer fatalities within the area 10 miles of the EAB shall not exceed 2x10-6/plant-year to ensure that the NRC safety goal QHO for latent cancer fatality risk is met.
6/19/2018 Southern Company 18
PRA Development
- Early introduction of PRA into design process facilitates risk-informing design decisions
- Scope and level of detail consistent with scope and level of detail of design and site information and fit for purpose in RIPB decisions
- PRA event-sequences include those involving single and multiple reactor modules and risk significant non-reactor sources
- Limitations and uncertainties associated with PRA addressed in the evaluation of defense-in-depth adequacy 6/19/2018 Southern Company 19
SSC Safety Classification And Performance Requirements 6/19/2018 Southern Company 20
SSC Approach Highlights
- Proposes criteria for SSC risk significance based on absolute risk metrics
- Incorporates concepts from 10 CFR 50.69 and NEI 04 in the context of a forward fit process
- Includes SSC requirements to address single and multi-module risks
- Expands on guidance for deriving performance requirements beyond those in NGNP SSC white paper 6/19/2018 Southern Company 21
LMP Proposed SSC Safety Categories
- Safety-Related (SR):
- SSCs selected by the designer to perform required safety functions to mitigate the consequences of DBEs to within the F-C target, and to mitigate DBAs to meet the dose limits of 10 CFR 50.34 using conservative assumptions.
- SSCs selected by the designer to perform required safety functions to prevent the frequency of BDBEs with consequences greater than 10 CFR 50.34 dose limits from increasing into the DBE region and beyond the F-C target.
- Non-Safety-Related with Special Treatment (NSRST):
- Non-safety related SSCs relied on to perform risk significant functions. Risk significant SSCs are those that perform functions that keep LBEs from exceeding the F-C target, or make significant contributions to the cumulative risk metrics selected for evaluating the total risk from all analyzed LBEs.
- Non-safety related SSCs relied on to perform functions requiring special treatment for DID adequacy.
- Non-Safety-Related with No Special Treatment (NST):
- All other SSCs.
6/19/2018 Southern Company 22
SSC Risk Significance
- A prevention or mitigation function of the SSC is necessary to meet the design objective of keeping all LBEs within the F-C target.
- The LBE is considered within the F-C target when a point defined by the upper 95%-tile uncertainty of the LBE frequency and dose estimates are within the F-C target.
- The SSC makes a significant contribution to one of the cumulative risk metrics used for evaluating the risk significance of LBEs.
- A significant contribution to each cumulative risk metric limit is satisfied when total frequency of all LBEs with failure of the SSC exceeds 1% of the cumulative risk metric limit. The cumulative risk metrics and limits include:
- The total frequency of exceeding of a site boundary dose of 100 mrem <
1/plant-year (10 CFR 20)
- The average individual risk of early fatality within 1 mile of the Exclusion Area Boundary (EAB) < 5x10 -7/ plant-year (QHO)
- The average individual risk of latent cancer fatalities within 10 miles of the EAB shall not exceed 2x10-6/plant-year (QHO) 23 Southern Company 6/19/2018
SSC Hierarchy Risk- Safety Safety Significant Significant PRA Modeled Related SSCs SSCs SSCs SSCs All Plant SSCs 6/19/2018 Southern Company 24
Derivation of Special Treatment Requirements
- SR SSCs
- Functional Design Criteria derived from required safety functions
- Lower level design criteria derived from SRDC
- SR and NSRST SSCs
- SSC reliability and capability performance targets
- Focus on prevention and mitigation functions from LBEs
- Integrated decision making process to derive specific special treatment requirements
- Reflects concepts from 10 CFR 50.69 and NEI-00-04 from existing reactors from a forward fit perspective
- Reflects Commissions expectations for risk-informed and performance based regulation from SRM to SECY 98-0144 6/19/2018 Southern Company 25
Defense In Depth Adequacy Evaluation 6/19/2018 Southern Company 26
LMP DID Adequacy Approach
- Builds on NGNP DID approach also reflected in ANS-53.1
- Evaluation of DID adequacy is both risk-informed and performance-based.
- The layers of defense and attributes of the NRC and IAEA DID frameworks are more visibly represented.
- DID attributes for plant capability and programmatic DID have been enhanced for consistency with the measures defined in the LMP Guidance Document
- This process is used to evaluate each LBE and to identify the DID attributes that have been incorporated into the design to prevent and mitigate accident sequences and to ensure that they reflect adequate SSC reliability and capability.
- Those LBEs with the highest levels of risk significance are given greater attention in the evaluation process.
- The practicality of compensatory actions for DID purposes are considered in the context of the individual LBE risk significance and in a cumulative manner across all LBEs Southern Company 6/19/2018 27
DID Adequacy Framework 6/19/2018 Southern Company 28
Layers of Defense Adapted from IAEA 6/19/2018 Southern Company 29
Plant Capability Defense-In-Depth Attributes Attribute Evaluation Focus PRA Documentation of Initiating Event Selection and Event Sequence Modeling Initiating Event and Accident Insights from reactor operating experience, Sequence Completeness system engineering evaluations, expert judgment Multiple Layers of Defense Extent of Layer Functional Independence Layers of Defense Functional Barriers Physical Barriers Inherent Reactor Features that contribute to performing safety functions Passive and Active SSCs performing Functional Reliability safety functions Redundant Functional Capabilities Diverse Functional Capabilities SSCs performing prevention functions SSCs performing mitigation functions Prevention and Mitigation Balance No Single Layer /Feature Exclusively Relied Upon 6/19/2018 Southern Company 30
RIPB Decision-Making Attributes Attribute Evaluation Focus What can go wrong?
Use of Risk Triplet Beyond PRA How likely is it?
What are the consequences?
Plant Simulation and Modeling of LBEs Knowledge Level State of Knowledge Margin to PB Targets and Limits Magnitude and Sources of Uncertainty Management Uncertainties Implementation Practicality and Action Refinement Effectiveness Cost/Risk/Benefit Considerations 6/19/2018 Southern Company 31
Programmatic DID Attributes Attribute Evaluation Focus Performance targets for SSC reliability and capability Design, manufacturing, construction, Quality / Reliability O&M features, or special treatment sufficient to meet performance targets Compensation for human errors Compensation for mechanical errors Compensation for Uncertainties Compensation for unknowns (performance variability)
Compensation for unknowns (knowledge uncertainty)
Off-Site Response Emergency response capability 6/19/2018 Southern Company 32
Guidelines for Establishing Adequacy of Plant Capability Defense-in-Depth Layer Guideline Overall Guidelines Layer[a]
Quantitative Qualitative Quantitative Qualitative
- 1) Prevent off-normal operation Maintain frequency of plant transients within designed cycles; meet and AOOs owner requirements for plant reliability and availability[b]
- 2) Control abnormal operation, Maintain frequency of all DBEs Minimize frequency of challenges detect failures, and prevent No single
< 10-2/ plant-year to safety-related SSCs DBEs design or
- 3) Control DBEs within the Meet F-C operational No single design or operational Target for all feature,[c] no analyzed design basis Maintain frequency of all BDBEs feature[c] relied upon to meet LBEs and matter how conditions and prevent < 10-4/ plant-year quantitative objective for all DBEs cumulative risk robust, is BDBEs metric targets exclusively
- 4) Control severe plant with sufficient[d] relied upon to conditions, mitigate margins satisfy the five consequences of BDBEs No single barrier[c] or plant feature Maintain individual risks from all layers of relied upon to limit releases in
- 5) Deploy adequate offsite LBEs < QHOs with sufficient[d] defense achieving quantitative objectives protective actions and prevent margins for all BDBEs adverse impact on public health and safety Notes:
[a] The plant design and operational features and protective strategies employed to support each layer should be functionally independent
[b] Non-regulatory owner requirements for plant reliability and availability and design targets for transient cycles should limit the frequency of initiating events and transients and thereby contribute to the protective strategies for this layer of DID. Quantitative and qualitative targets for these parameters are design specific.
[c] This criterion implies no excessive reliance on programmatic activities or human actions and that at least two independent means are provided to meet this objective.
[d] The level of margins between the LBE risks and the QHOs provides objective evidence of the plant capabilities for DID. Sufficiency will be decided by the IDP.
Southern Company 6/19/2018 33
DID Adequacy Evaluation Process
- DID Baseline Evaluation documented by Integrated Decision Panel (IPD) and updated during each design/licensing phase
- Defense-in-depth is deemed by IDP as adequate when:
- Plant capability DID is deemed to be adequate.
- Plant capability DID guidelines in Table 5-2 are satisfied.
- Review of LBEs is completed with satisfactory results.
- Programmatic DID is deemed to be adequate.
- Performance targets for SSC reliability and capability are established.
- Sources of uncertainty in selection and evaluation of LBE risks are identified.
- Special treatment for all SR and NSRST SSCs is sufficient.
Southern Company 6/19/2018 34
Questions?
6/19/2018 Southern Company 35
BACK-UP SLIDES Southern Company 6/19/2018 36
LMP RIPB Framework 6/19/2018 Southern Company 37
Systematic, Wholistic, Technology Inclusive Licensing is a Must Risk-Informed, Performance-Based Method is the Logical Solution
Licensing Basis Events (LBEs)
- LBEs are defined broadly to include all the events used to support the safety aspects of the design and to meet licensing requirements. They cover a comprehensive spectrum of events from normal operation to rare, off-normal events.
- Categories defined as Normal Operations (NO),
Anticipated Operational Occurrences (AOO), Design Basis Events (DBE), Beyond Design Basis Events (BDBE) and Design Basis Accidents (DBA)
- LBE definitions generally consistent with NGNP white papers
- Limited differences with NRC definitions to create consistency with LMP process 6/19/2018 Southern Company 39
LBE Categories Anticipated Operational Occurrences (AOOs). AOOs encompass planned and anticipated events whose frequencies exceed 10-2/plant-year where a plant may be comprised of one or more reactor modules. The radiological doses from AOOs are required to meet normal operation public dose requirements. AOOs are utilized to set operating limits for normal operation modes and states.
Design Basis Events (DBEs). DBEs encompass unplanned off-normal events not expected in the plants lifetime whose frequencies are in the range of 10-4 to 10-2/plant-year, but which might occur in the lifetimes of a fleet of plants. DBEs are the basis for the design, construction, and operation of the structures, systems, and components (SSCs) during accidents and are used to provide input to the definition of design basis accidents (DBAs).
Beyond Design Basis Events (BDBEs). BDBEs which are rare off-normal events whose frequencies range from 5x10-7/plant-year to 10-4/plant-year. BDBEs are evaluated to ensure that they do not pose an unacceptable risk to the public.
Design Basis Accidents (DBAs). The DBAs for Chapter 15, Accident Analyses, of the license application are prescriptively derived from the DBEs by assuming that only SSCs classified as safety-related are available to mitigate the consequences. The public consequences of DBAs are based on mechanistic source terms and evaluated using conservative or best estimate approaches with appropriate accounting for uncertainties.
6/19/2018 Southern Company 40
Frequency-Consequence (F-C)Target
- Purpose is to evaluate risk significance of individual LBEs and to help define the RSFs
- Addressed Staircase feature
- F-C Target anchor points based on:
- 10 CFR 20 annual dose limits and iso-risk concept
- SRP Chapter 15.0 insights on dose limits for lower frequency AOOs
- 10 CFR 50.34 dose limits for DBAs (and DBEs)
- QHOs for prompt fatality individual risk 6/19/2018 Southern Company 41
- 2. Design Top Level Design Requirements for energy Development production, investment protection, public and Analysis and worker safety, and defense-in-depth Selection 3. PRA Development/
Update and 4. Identify/Revise 7a. Evaluate LBEs Against Freq.-
Evaluation List of AOOs, Consequence DBEs, and BDBEs 5a. Identify Target Required Safety Functions Process 6. Select DBAs including Design Basis External 5b. Select Safety-Related SSCs 7b. Evaluate Integrated Plant Risk vs. QHOs and Events 10 CFR 20 7d. Perform 7c. Evaluate Risk Deterministic Significance of Safety Analysis vs. LBEs and SSCs 10 CFR 50.34 including Barriers 7e. RI-PB Evaluation of Defense-in-Depth LBE Evaluations Input to RIPB Decisions:
No Yes - SSC safety classification
- 9. Proceed to Next 8. Design/ - SSC design criteria
- 10. Final Stage of Design LBE Development - SSC performance requirements List of LBEs Development Complete? - Siting criteria
- Emergency planning
- Defense-in-Depth adequacy 6/19/2018 Southern Company 42
Systems Engineering Inputs Select Risk Metrics for Identify/Characterize Risk-Informed Plant Design Concept Radionuclide Sources Performance-Based Decisions Define Radionuclide Barriers and Supporting Structures Plant Functional Analysis Define Reactor Specific Fundamental Safety Functions Safety Functions
- Control heat generation Protecting Each Barrier Design and PRA
- Control heat removal
- Retain radionuclides Identify SSCs and Development Operator Actions Plant/Systems Engineering Supporting Each Safety Function Interfaces Process Hazards Identify Failure Modes of Each Barrier and SSC Analysis (HAZOPs) Providing Safety Function Failure Modes and Effects Analysis (FMEA)
Identify Challenges to Plant Operating Preventing Barrier and Modes and States SSC failure modes Exhaustive Enumeration of Reactor Specific Initiating Events Plant Transient Analysis Plant Response to Events Building Blocks for Reactor and Event Sequences Specific PRA Model Development Accident Analyses 6/19/2018 Southern Company 43
- 1. Identify SSC functions
- 2. Identify and evaluate SSC capabilities and LMP SSC Safety Classification Evaluation in prevention and programs to support mitigation of LBEs defense-in-depth
- Safety-Significant functions include those classified as risk-
- 3. Determine required and safety-significant*
Approach significant or required for defense-in-depth functions Special Treatment for Safety-Significant Functions 6a. Define SR SSC 7a. Define SR SSC 4a. SSC selected** to YES 5a. Classify SSC as reliability and capability functional design criteria, meet Required
- Only those SSCs selected by Safety- Related (SR) requirements to perform and special treatment Safety Function?
designer to perform functions required safety functions requirements required to keep DBEs and high consequence BDBEs inside the F-C No target are classified as SR, All other SSCs not so selected are considered in Boxes 4b and 4c for classification 5b. Classify SSC as Non- 6b. Define NSRST SSC YES 7b. Define NSRST SSC as NSRST or NST. 4b. Non-SR SSC function is Safety-Related with reliability and capability special treatment risk significant? Special Treatment requirements to perform requirements (NSRST) safety-significant functions No 4c. Non-SR SSC Functions required for YES defense-in-depth adequacy?
5c. Classify SSC as Non- 6c. Define NST SSC No 7c. Definee non-Safety-Related with No reliability and capability regulatory NST SSC Special Treatment requirements to meet design requirements (NST) user requirements 6/19/2018 Southern Company 44
LMP Proposed SSC Safety Categories SSCs Including Radionuclide Barriers Safety Related (SR) Non-Safety Related Non-safety Related SSCs SSCs with Special SSCs with No Special Treatment (NSRST) Treatment (NST)
SSCs selected for required safety SSCs performing non-safety SSCs performing risk functions to mitigate DBEs within significant functions significant functions F-C Target*
SSCs selected for required safety SSCs performing functions functions to prevent high required consequence BDBEs from entering for defense-in-depth DBE region beyond F-C target
- SR SSCs are relied on during DBAs to meet 10 CFR 50.34 dose limits using Non-Risk Risk Significant conservative assumptions Significant SSCs SSCs 6/19/2018 Southern Company 45
Comparison of LMP and 10 CFR 50.69 SSC Safety Categories 6/19/2018 Southern Company 46
SSC Classification Summary
- LMP retains the NGNP SSC safety categories of SR, NSRST, and NST
- All safety significant SSCs classified as SR or NSRST
- All SR SSCs are classified as risk significant
- NSRST SSCs include other risk significant SSCs and SSCs requiring some special treatment for DID adequacy
- Specific special treatment for capabilities and reliabilities in the prevention and mitigation of accidents
- Special treatment defined via integrated decision panel using forward fit 10 CFR 50.69 process 6/19/2018 Southern Company 47
DID Evaluation Baseline Summary Concept Qualitative Evaluation of Plant Capability DID Functional Physical Prevention No Single LBE IE Series Multiple Margin and Functional Feature Name Protective Adequacy Mitigation Reliability Relied Measures Balance Upon Normal Operation AOOs DBEs BDBEs DBAs Evaluation Summary - Qualitative Evaluation of Programmatic DID Compensation for Uncertainties Offsite Quality/Reliability:
Response
LBE IE Series Design, Human Mechanic Emergency Name Manufacturing, Unknowns Errors al Failures Response Construction, O&M Capability Normal Operation AOOs DBEs BDBEs DBAs 6/19/2018 Southern Company 48
DID Concept from NUREG/KM-0009 6/19/2018 Southern Company 49
Roles of SSC Capability and Reliability in Prevention and Mitigation of Accidents Plant features Defense-in-Plant prevent SSC1 Prevents SSC2 Limits LBE End State Depth Layers Frequency Dose Distrubance Inititating Fuel Damage? Release?
Challenged [1]
event?
Disturbance controlled with N/A Layer 1 fd 0 Yes no plant trip fd fdp0 LBE-1 F-C Target Frequency ------ >
p0 Yes 1 No fuel damage or release Layer 2 fdp0 0 fdp0p1 LBE-2 No Fuel damage w/ limited p1 Yes 2 Layer 3 fdp0p1 dlow release No fdp0p1p2 LBE-3 p2 Fuel Damage w/ un-3 Layers 4 and 5 fdp0p1p2 dhigh No mitigated release Consequence ------->
[1] See Figure 2-4 for definition of defense-in-depth layers 0 dlow dhigh SSC LBEs Function SSC Performance Attribute for Special Treatment Plant N/A Prevent initiating event Reliability of plant features preventing initiating event 1 Mitigate initiating event Capability to prevent fuel damage SSC1 2 Prevent fuel damage Reliability of mitigation function 3 Help prevent large release Reliability of mitigation function 2 Mitigate fuel damage Capability to limit release from fuel damage SSC2 3 Prevent unmitigated release Reliability of mitigation function 6/19/2018 Southern Company 50
Defense In Depth Adequacy Basic Structure Plant Capability DID Plant Functional Capability DIDThis capability is introduced through systems and features designed to prevent occurrence of undesired LBEs or mitigate the consequences of such events.
Plant Physical Capability DIDThis capability is introduced through SSC robustness and physical barriers to limit the consequences of a hazard.
Programmatic DID Programmatic DID is used to address uncertainties when evaluating plant capability DID and is used where programmatic protective strategies are defined. It is used to incorporate special treatment during design, manufacturing, constructing, operating, maintaining, testing, and inspecting of the plant and the associated processes to ensure there is reasonable assurance that the predicted performance can be achieved throughout the lifetime of the plant. The use of performance-based measures, where practical, to monitor plant parameters and equipment performance that have a direct connection to risk management and equipment and human reliability are considered essential.
6/19/2018 Southern Company 51
Defense In Depth Adequacy Basic Structure Risk-Informed Evaluation of DID This element provides a systematic, holistic, integrated, and transparent process for examining the DID adequacy achieved by the combination of plant capability and programmatic elements. This evaluation is performed by a risk-informed integrated decision-making (RIDM) process to assess and establish whether DID is sufficient to enable consideration of different alternatives for achieving commensurate safety levels at reduced burdens.
The outcome of the RIDM process also establishes a DID baseline for managing risk throughout the plant lifecycle.
Southern Company 6/19/2018 52
Integrated Process Iterate as required
- 1. Establish initial design capabilities
A
- 2. Establish F-C Target Based 11. Perform safety on TLRC and analysis of DBAs Evaluation of DID QHOs A
- 3. Define
- Tasks are not necessarily Risk Significant SSCs
- 4. Define scope 13. Identify NSRST of PRA for current SSCs sequential design phase Other SSCs needed for DID Adequacy
- Tasks can begin early in the 14. Define and A
conceptual design process and 5. Perform PRA evaluate FDC for SR SSCs Color Key mature with the design evolution A Deterministic
- 15. Evaluate uncertainties and DID adequacy evaluation are margins DBE, or BDBE Probabilistic completed when the design A
- 16. Specify
- 7. Evaluate LBE ST requirements baseline for the license risks vs. F-C for SR and NSRST Target SSCs Risk-Informed application is submitted A A
- Programmatic confirmation of
- 8. Evaluate 17. Confirm plant risks vs Programmatic Acronymns Cumulative Risk DID adequacy performance and sustained DID Targets F-C Frequency Consequence DID Defense-in-Depth FDC Functional Design Criteria continues for life of the plant.
LBE Licensing Basis Events NSRST Non-Safety Related with ST
- 18. DID adequacy
- 9. Identify DID SSC Structure, System, Component established; Document/ ST Special Treatment layers challenged Update DID Baseline SR Safety Related by each LBE evaluation TLRC Top Level Regulatory Criteria QHO Quantitative Health Objectives 6/19/2018 Southern Company 53
Hon. George Apostolakis Hon. Jeffrey S. Merrifield Hon. Richard A. Meserve December 11, 2017 Stephen Kuczynski President and CEO Southern Nuclear Operating Company 42 Inverness Center Parkway Birmingham, AL 35242
Subject:
Risk-Informed and Performance-Based Evaluation of Defense-in-Depth Adequacy
Dear Mr. Kuczynski:
As you know, we act as advisors in the effort to develop a modernized framework for licensing advanced nuclear reactors, otherwise referred to as the Licensing Modernization Project (LMP). In our view, the LMP effort, which is intended to develop a technology-inclusive, risk-informed, performance-based (TI-RIPB) process for licensing the next generation of non-light water reactors, is making substantial progress and represents the direct application of the risk-informed principles that the NRC has pursued for the last 20 years.
Over the last six months, industry representatives have been actively engaged with the staff of the NRC to discuss the LMP effort and develop a framework that will be acceptable for future consideration by the Commission. A key element of the LMP is the development of a framework and associated guidelines for establishing, evaluating, confirming, and documenting the adequacy of defense-in-depth (DID) for advanced non-light water reactor (non-LWR) technologies. This effort is focusing on non-LWR technologies because of the need for transparent and principled guidance for the advancement of the multiplicity of novel designs that are being developed. There are many precedents that serve to provide guidance for the application of DID to LWRs, but that same experience does not exist for non-LWR designs.
The concept of defense-in-depth has been a cornerstone of NRC decision making for decades. It serves as an approach to designing and operating a nuclear facility that prevents and mitigates accidents that release radiation or hazardous materialswith the intention of creating multiple independent and redundant layers of defense to compensate for potential human and mechanical failures so that no single layer, no matter how robust, is exclusively relied upon.
4824-8459-2216.v1
Stephen Kuczynski December 11, 2017 Page 2 Defense in depth includes the use of access controls, physical barriers, redundant and diverse key safety functions, and emergency response measures.1 Throughout its history, DID has served the NRC, the industry and the public well by providing layers of protection to prevent and mitigate accidents. LWRs typically have a significant source term, residual decay heat, and complex designs; DID provides the robustness to ensure the reasonable assurance of adequate protection of public health and safety. That said, the DID philosophy has not been without criticism. Some have found it overly conservative and prescriptive, while others have claimed it is too subject to inconsistent application by different members of the NRC staff. Perhaps the most severe criticism is that there is no guidance as to how to implement the DID philosophy and, in particular, to determine how much DID is sufficient. Excessive application of DID can lead to unnecessary regulatory burden. On the other hand, too little DID could lead to an inappropriate increase in risk. The risk-informed initiatives that have been promulgated in the last two decades have attempted to remove unnecessary burden while ensuring adequate protection of the public. The LMP effort seeks to bring this philosophy to the application of DID to advanced non-LWR reactors.
Given the significantly reduced risks associated with many of the advanced reactor designs as compared to LWRs, a non-risk-informed application of DID could result in excess conservatism in safety reviews. That is, it could result in unduly prescriptive regulatory requirements well beyond what is needed to provide reasonable assurance of adequate protection. Moreover, while we believe that adopting a TI-RIPB DID framework will generally reduce excess conservatism and provide a more effective means to limit unnecessary regulatory burden, like all risk-informed tools, it could also result in the identification of areas where additional requirements are necessary.
We agree that DID should continue to play an important part in the NRCs licensing framework.
That said, the development of the LMP for advanced reactors provides an opportunity for the Agency and its stakeholders not only to make DID more risk informed and performance based, but to do so in a manner that is implementable and repeatable. The framework developed by the LMP utilizes existing U.S. and international definitions and philosophies of DID and builds on a DID framework developed in the U.S. Department of Energy Next Generation Nuclear Plant project. While we are not in a position to comment on the specific technical features of the LMP DID proposal, we enthusiastically endorse the effort. We believe it can serve as a positive launching point for establishing a more standardized and principled approach for the application 1
https://www.nrc.gov/reading-rm/basic-ref/glossary/defense-in-depth.html www.pillsburylaw.com 4824-8459-2216.v1
Stephen Kuczynski December 11, 2017 Page 3 of DID. The great value of the LMP DID proposal is that it is a first step toward converting what is currently an ambiguous DID philosophy to a concrete DID process.
We encourage you and the Southern team to continue work on this most important matter.
Sincerely, George Apostolakis Jeffrey S. Merrifield Richard A. Meserve Commissioner, U.S. NRC Commissioner, U.S. NRC Chairman, U.S. NRC (2010-2014) (1998-2007) (1999-2003) www.pillsburylaw.com 4824-8459-2216.v1
Hon. George Apostolakis Hon. Jeffrey S. Merrifield Hon Richard A. Meserve February 20, 2018 Stephen Kuczynski President and CEO Southern Nuclear Operating Company 42 Inverness Center Parkway Birmingham, AL 35242 Re: Licensing Modernization Project
Dear Mr. Kuczynski:
As you know, we serve as advisers to the Licensing Modernization Project (LMP), an effort to develop a modernized framework for the licensing of advanced nuclear reactors. This effort is intended to deal with some of the technical issues associated with the licensing of advanced reactors through the application of a technology-inclusive, risk-informed, and performance-based approach (TI-RIPB). To date, the project has developed several technical reports.1 We are writing to provide our assessment of the project.
Although we are not in a position to comment on the technical adequacy of the reports generated by the LMP, we enthusiastically endorse the effort. In particular, we believe that the focus on a systematic and predictable process for early resolution of fundamental technical issues in the licensing of advanced reactors can reduce uncertainty in the development of a design. It thereby can enable a vendor to avoid the substantial expenditures (hundreds of millions of dollars) required under the current process to develop a detailed design without knowledge of the criteria it must satisfy. We believe that the work lays the foundation for guidance that can be endorsed by the NRC and we encourage the continuation of the effort.
It is helpful to view the LMP in the context of the evolution in the licensing process.
The Atomic Energy Commission (AEC) first licensed commercial reactors in the 1960s at a time during which knowledge about nuclear reactors was limited. Licensing focused on assuring that the reactors could survive certain design basis events, such as a loss-of-coolant accident, by establishing both the need for and the specifications for safety systems. The intention was to select a suite of Design Basis Accidents (DBAs) that encompassed some of the possible events that could disrupt the operation of a nuclear reactor. As the utilities and their regulator gained greater experience, as well as a better understanding of the gaps in licensing, the regulatory 1
The technical papers cover PRA development for licensing basis event selection, approach to PRA for RIPB risk management applications, the SSC safety classification and performance requirements approach, and defense-in-depth adequacy. We have previously commented on the defense-in-depth paper.
Stephen Kuczynski February 20, 2018 Page 2 system evolved to encompass a broader suite of circumstances and requirements than had originally been contemplated. When the AEC first licensed these reactors, they applied deterministic analytical techniques and imposed prescriptive requirements that were frequently based on judgment; probabilistic techniques were unknown at the time the basic regulatory structure was established.2 Beginning in the early 1990s, the nuclear industry began to utilize Probabilistic Risk Assessments (PRAs) to assess and manage nuclear units. These PRAs treated the plant as an integrated system and enabled utilities to identify accident sequences and assess their frequencies. Eventually, PRAs were used to evaluate the risk associated with design or operational changes in an existing plant and later in providing a risk-informed means for changes in a plants licensing basis (e.g., Regulatory Guide 1.174). PRAs had a profound impact on the understanding of the utilities and the NRC about reactor safety. For example, the PRAs identified the risk significance of human errors and accidents not included initially as DBAs, such as anticipated transients without scram and station blackout. The result is that the current regulatory structure at the NRC consists of deterministic requirements with an overlay of probabilistic elements; this framework provides an extensive and complicated set of largely prescriptive requirements.
Because most of the commercial nuclear units were large light water reactors (LWRs),
the NRCs regulatory requirements were tailored to the risks that LWRs presented. Moreover, because the requirements are largely prescriptive, they provide a large measure of assurance and predictability to vendors and utilities of the specific regulatory requirements that must be satisfied. However, many of the requirements in the existing regulatory system are not necessarily appropriate or relevant for advanced reactors. Such reactors, many of which use coolants other than water, may present entirely different risks from LWRs and, indeed, some of the LWR requirements make no sense in application to these designs. For example, requirements for emergency core cooling in LWRs that result from the behavior associated with rapid depressurization of high-pressure water are inapplicable to reactors operating at or near atmospheric pressure, such as those cooled by liquid metal or molten salt. Moreover, there are accident sequences associated with some advanced reactors, such as those involving sodium fires, that are not part of the LWR regulatory framework. Current regulations are based almost exclusively on low-enriched (< 5%) uranium oxide fuel in zirconium alloy cladding, whereas a number of advanced reactors have a variety of other fuel and/or cladding materials and, in some cases, propose to utilize high-assay low-enriched uranium with enrichments up to 19.7%.
The LMP is intended to close gaps in several foundational areas where the current regulations and regulatory guides are either silent or provide inadequate guidance for non-LWR designs. Because the state of knowledge regarding the operation of nuclear power plants has advanced considerably over the period in which the current regulatory structure was developed, the NRC has an opportunity to use new understandings and methods as the foundation for a 2
Probabilistic Risk Assessments for nuclear power plants had their origins in the WASH-1400, or Reactor Safety Study that was authorized by the Atomic Energy Commission and published in October of 1975.
2
Stephen Kuczynski February 20, 2018 Page 3 revised regulatory system and avoid the serious challenges that attend an effort to adapt regulatory requirements developed for LWRs to reactors of very different types. This is the focus of the LMP.
One of the serious challenges in applying the existing regulatory requirements to advanced reactors arises from the uncertainty that is associated with identifying appropriate design basis events and developing the associated requirements to meet them (such as the requirements that must be satisfied by structures, systems and components). The vendors are required to provide costly design details at a point in the regulatory review process when it is uncertain whether the NRC would eventually find the design to be acceptable.
Although the NRC staff is prepared to undertake substantial pre-application consultation with a vendor and to provide guidance, the current regulatory paradigm results in substantial uncertainty about regulatory design requirements until late in the design review process. The process creates regulatory uncertainty for both the vendor and the regulator because the fundamental considerations that should guide design decisions are not well defined and require a large measure of judgment. The result is a process that is both unpredictable and costly and requires a vendor to undertake substantial expenditures (hundreds of millions of dollars) to develop a detailed design without confidence that it will be satisfactory to the NRC.
The LMP seeks to address this situation through the development of a systematic and principled process to define the critical regulatory elements. The effort involves a blend of deterministic and probabilistic inputs in a structured and logical process to refine the foundations for licensing. The LMP is intended to establish these regulatory elements at the earliest possible stages in the evolution of a design and thereby enable the NRC and the vendor to engage in a predictable and cost-effective licensing review. The development of the LMP builds on an LBE white paper for DOEs Next Generation Nuclear Plant (NGNP) that has been reviewed by the NRC staff and the Advisory Committee on Reactor Safeguards. It has appropriately involved extensive consultation across the nuclear industry and with the NRC staff - an effort that is still underway. The aim is the eventual development of guidance documents that the NRC can utilize to review advanced reactor designs.
Reactor developers are currently using PRA as a design tool for new plants to ensure that the risks are acceptably low. The proposed LMP licensing process starts with the identification of a set of Licensing Basis Events (LBEs) using probabilistic methods and relevant regulatory requirements. This approach for selecting LBEs is designed to ensure that an appropriate set of limiting accident sequences for each reactor technology are reflected in the selection of DBAs and that the full set of LBEs define the risk-significant accident sequences. The use of PRA to identify the broad set of LBEs and the resulting structured selection of the DBAs are consistent with the Commissions 1995 Policy Statement:
A probabilistic approach to regulation enhances and extends this traditional, deterministic approach, by: (1) Allowing consideration of a broader set of potential challenges to safety, (2) Providing a logical means for prioritizing these challenges based 3
Stephen Kuczynski February 20, 2018 Page 4 on risk significance, and (3) Allowing consideration of a broader set of resources to defend against these challenges.
In addition to being risk informed, the proposed approach is performance based because it uses quantitative risk metrics to evaluate the risk significance of accident sequences and leads to the formulation of performance requirements on the capability and reliability of systems, structures and components to prevent and mitigate accidents. It, thus, avoids the overly prescriptive nature of current requirements. Since it focuses on performance, the proposed approach is technology inclusive and, therefore, very valuable to the designers of the various reactor designs under development.
In addition to the long-term benefits of the proposed licensing approach, we are pleased to note that there are short-term benefits as well. A recent Draft White Paper issued by the NRC staff3 states that the performance criteria for what are termed functional containment design features are tied to radionuclide release limits for various event categories. Recognizing that an integrated approach is needed, the staff, as a starting point, proposes to use the LMP structure of LBE identification. The staff states further: The structure is sufficiently defined to show the categories and how related acceptance criteria would be derived along with additional consideration of deterministic methods to address uncertainties and ensure sufficient defense in depth.
In order for this project to be successful, it will be essential to develop guidance that is acceptable to the NRC. Given that the proposed process is new and represents significant change from the current licensing process, we anticipate that the development of satisfactory guidance will not be easy and that modifications in early applications may be necessary.
However, we believe that guidance based on the essence of the proposed LMP approach is endorsable by the NRC. Moreover, the full and successful implementation of the process may eventually require modifications of licensing procedures (not included as part of the LMP), as well as the development of further guidance to fill in gaps that extend beyond the current technical papers. The LMP should be viewed as an important start to necessary change and we would expect that there would be significant interactions between the LMP team, the NRC staff and interested industry and external stakeholders to develop and implement the guidance needed to implement this important effort.
We have met, individually, with a number of stakeholders regarding the LMP effort, including some individuals at the NRC, and we believe that this effort is worth pursuing and could provide a very beneficial framework not only for the NRC, but also for the utilities and companies that seek to develop advanced reactor technologies. Success in this effort will require significant and ongoing engagement with stakeholders both within and outside the NRC.
3 Draft White Paper Functional Containment Performance Criteria, November 2017 Draft - Released to Support Public Discussions.
4
Stephen Kuczynski February 20, 2018 Page 5 We appreciate the opportunity to participate in this important project. Please feel free to contact us if you seek any further advice.
Sincerely, George Apostolakis Jeffrey S. Merrifield Richard A. Meserve Commissioner, U.S. NRC Commissioner, U.S. NRC Chairman, U.S. NRC (2010-2014) (1998-2007) (1999-2003) 5
ACRS Future Plant Designs Subcommittee Draft Regulatory Guide (DG) 1353 Guidance for a Technology-Inclusive, Risk-Informed, Performance-Based Approach to Inform the Content of Applications for Licenses, Certifications, and Approvals for Non-Light Water Reactors June 19, 2018 1
Background
- Advanced Reactor Policy Statement
- Commission Papers (e.g., SECY-93-0047)
- Next Generation Nuclear Plant (NGNP)
- Recent Activities
- Vision and Strategy
- Implementation Action Plans
- Regulatory Guide 1.232 (Principal Design Criteria)
- Functional Containment Performance Criteria 2
Integrated Approach to Design and Licensing Functional Containment Proposed EP Rule ARDC, LMP 3
Content of Applications
- General Description of the Plant
- Emergency Planning
- Site Characteristics
- Security
- Design of SSCs and Equipment
- Staffing
- Reactor
- Mitigating Strategies
- Reactor Coolant and Connecting Systems
- Aircraft Impact Assessment
- Engineered Safety Features
- Environmental Report
- Instrumentation and Controls
- Financial
- Electric Power
- Inspections, Tests, Analyses, and
- Auxiliary Systems Acceptance Criteria
- Steam and Power Conversion System
- Insurance
- Radioactive Waste Management
- Fuel Cycle
- Radiation Protection
- Other (design or technology specific)
- Conduct of Operations
- Verification Programs
- Transient and Accident Analyses
- Technical Specifications
- Quality Assurance and Reliability Assurance
- Human Factors Engineering
- Probabilistic Risk Assessment/Severe Accident Evaluation 4
Informing Content of Applications (scope and level of detail)
Assessment: 15, 19 DID: 20*
LMP focus PRA: 19 Deterministic: 15 Key: 4, 5, 6 Programs:
12, 13, 14, Support: 7, 8, 9 16, 17, 18 Site/Design: 2, 3 Power Conv 10 Other: 11
Consolidated Guidance
- Licensing Basis Events
- Deterministic
- Safety Classification
- Function and Risk Considerations
- Safety Related
- Special Treatment
- Defense in Depth Assessment
- Structures, Systems and Components
- Programmatic 6
Identifying Measures to Prevent and Mitigate Events Note that demarcations and cut off remain under discussion with plans to finalize by Sept 2018 The target values shown in the figure should not be considered as a demarcation of acceptable and unacceptable results. The targets provide a general reference to assess events, SSCs, and programmatic controls in terms of sensitivities and available margins.
7
Remaining Discussion Items
- F-C Target Figure
- Demarcations (including lower range for BDBEs)
- Consequence Analyses
- Consideration of Uncertainties
- ASME/ANS Non-LWR PRA Standard
- Ensure Consistent Terminology
- Flexibility for smaller, simpler designs
- Clarify approach for external hazards
- Interface with requirements for areas such as emergency planning, operations, etc.
8
Related Commission Paper
- Details of Approach
- Technology Inclusive, Risk Informed, Performance Based
- Event categories and related demarcations on the F-C Target figure ( including lower range for BDBEs)
- Expected relationships with NRC requirements
- Possible relationship to recommendations made by NRC Transformation Team (SECY-18-0060) 9
Future ACRS Interactions Tentative Timeline June 19 ACRS SC Meeting September 28 Draft LMP Guidance, draft RG, draft SECY to ACRS October 30 ACRS SC Meeting December 6 ACRS FC Meeting Mid-December Issue draft RG for comment Early 2019 Complete SECY TBD-2019 Final Regulatory Guide 10