ML18325A049
| ML18325A049 | |
| Person / Time | |
|---|---|
| Issue date: | 10/30/2018 |
| From: | Derek Widmayer Advisory Committee on Reactor Safeguards |
| To: | |
| Widmayer D | |
| References | |
| NRC-3948 | |
| Download: ML18325A049 (359) | |
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, October 30, 2018 Work Order No.: NRC-3948 Pages 1-359 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 OCTOBER 30, 2018 11 + + + + +
12 ROCKVILLE, MARYLAND 13 + + + + +
14 The Subcommittee met at the Nuclear 15 Regulatory Commission, Three White Flint North, Room 16 1C3 & 1C5, 11601 Landsdown Street, at 8:30 a.m.,
17 Michael L. Corradini, Acting Chairman, presiding.
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
2 1 COMMITTEE MEMBERS:
2 MICHAEL L. CORRADINI, Acting Chairman 3 RONALD G. BALLINGER, Member 4 DENNIS C. BLEY, Chairman*
5 CHARLES H. BROWN, JR., Member 6 VESNA B. DIMITRIJEVIC, Member 7 WALTER KIRCHNER, Member 8 JOSE MARCH-LEUBA, Member 9 JOY L. REMPE, Member 10 GORDON R. SKILLMAN, Member 11 12 DESIGNATED FEDERAL OFFICIAL:
13 GIRIJA SHUKLA 14 15 16 *Present via telephone 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
3 1 A-G-E-N-D-A 2 Opening Remarks . . . . . . . . . . . . . . . . . 4 3 Staff Introduction .. . . . . . . . . . . . . . . 6 4 Licensing Modernization: Technology-Inclusive, 5 Risk-Informed, Performance-based Approach 6 for Non-Light-Water Reactors. . . . . . . . 7 7 Guidance Paper NEI 18-04. . . . . . . . . . . . . 86 8 Draft Regulatory Guide DG-1353 and 9 SECY-18-00XX 233 10 Public Comment . . . . . . . . . . . . . . . . . 262 11 Adjourn.. . . . . . . . . . . . . . . . . . . . 265 12 13 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 8:30 a.m.
3 ACTING CHAIRMAN CORRADINI: Okay, why 4 don't we get started? Good morning. The meeting will 5 come to order.
6 This is a meeting of the Advisory 7 Committee on Reactor Safeguards Subcommittee on Future 8 Plant Designs.
9 My name is Mike Corradini. I'm chairing 10 this meeting for Dennis Bley who is chairman of the 11 Future Plant Designs Subcommittee.
12 ACRS members in attendance are Charles 13 Brown, Ron Ballinger, Jose March-Leuba, Dick Skillman, 14 Walt Kirchner, Joy Rempe and Vesna Dimitrijevic.
15 Dennis Bley as I said is on the 16 teleconference line and he'll let us know if he has 17 questions through one of the members since we're on an 18 open line that is muted.
19 Girija Shukla, the ACRS staff, is the 20 designated federal official for today's meeting.
21 The purpose of today's meeting is to 22 review the working drafts of the NRC staff and NEI 23 guidance documents to implement a technology-inclusive 24 risk-informed performance-based approach for approving 25 non-light-water reactors also known as the licensing 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 modernization project.
2 The subcommittee will gather information, 3 analyze relevant issues and facts and formulate 4 proposed positions and actions as appropriate for 5 consideration by the full committee.
6 It is scheduled the full committee to 7 address this matter in the December full committee 8 meeting.
9 The ACRS was established by statute and is 10 governed by the Federal Advisory Committee Act or 11 FACA. That means that the committee can only speak 12 through its published letter reports.
13 We hold meetings to gather information to 14 support our deliberations. Interested parties who 15 wish to provide comments can contact our offices 16 requesting time after the Federal Register notice of 17 the meeting is published.
18 That said, we also set aside time for 19 extemporaneous comments from members of the public 20 attending or listening to our meetings. Written 21 comments are also welcome.
22 The ACRS section of the U.S. NRC's public 23 website provides our charter, bylaws, letter reports 24 and full transcripts of all our full and subcommittee 25 meetings including all slides presented at 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
6 1 meetings.
2 Detailed proceedings for conduct of the 3 ACRS meetings was previously published in the Federal 4 Register on October 4, 2018. The meeting is open to 5 public attendance and we have received no requests for 6 time to make oral statements. However, time has been 7 allotted in today's agenda in case of extemporaneous 8 comments.
9 Today's meeting is being held in telephone 10 bridge line allowing participation of the public over 11 the phone. A transcript of today's meeting is also 12 being kept.
13 Therefore we request that meeting 14 participants on the bridge line when they are called 15 upon to identify themselves when they speak and to 16 speak with sufficient clarity and volume so they can 17 be readily heard.
18 Participants in the meeting room shall 19 also use the microphones located throughout the 20 meeting room when addressing the subcommittee.
21 I'll note that we have a challenge in our 22 new conference setting so we'll be looking for the 23 presenters if they have experts they need to bring to 24 the mike to come over to the other side and identify 25 themselves.
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'll ask the attendees to 2 please silence all cell phones and other devices that 3 make noises to minimize disruptions.
4 Also I remind the speakers in front of the 5 table to turn on the microphone which is indicated by 6 the illuminated green light when speaking and 7 otherwise turn off the microphone when not speaking.
8 We'll proceed with the meeting and I'll 9 call on John Segala, chief of the Advanced Reactor and 10 Policy Branch of the Office of NRO to make our opening 11 comments. John.
12 MR. SEGALA: Thank you, Dr. Corradini, and 13 the other committee members. We're pleased to be here 14 today to discuss the licensing modernization project.
15 The NRC staff sees this as a key aspect of 16 licensing and risk-informing advanced reactors.
17 I wanted to step back for a moment and 18 just provide some context of where we've been. Back 19 in April 2017 industry submitted the first of four 20 white papers on the licensing modernization project.
21 We reviewed those, provided feedback.
22 They then consolidated those into an NEI document 18-23 04. We presented that to the ACRS committee in June 24 of 2018. We also gave the committee some initial 25 thinking on the development of a regulatory guide 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
8 1 potentially endorse the NEI document.
2 We took the feedback we received from 3 ACRS. We developed a working draft of the regulatory 4 guide and an associated commission paper. So today 5 we're going to be presenting an overview of the NEI 6 document, the commission paper and the draft guide.
7 We're looking for the committee to write 8 us a letter on the commission paper. And again we 9 look forward to the insights and the feedback that the 10 ACRS provides us today. With that I can turn it over 11 to Bill Reckley.
12 MR. RECKLEY: Thank you, John. So the 13 order of the presentation today will be we'll provide 14 a little background to answer one specific request 15 from the ACRS. We're going to spend the first few 16 minutes talking about the enhanced safety focused 17 review approach which is -- Ian Jung will go into.
18 But that's primarily for light water small modular 19 reactors.
20 But we've referenced it in the licensing 21 modernization project discussions as kind of a 22 stepping stone to where we're ending up. So it fits 23 in well with the background.
24 Then I'll talk about the overall non-25 light-water reactor program, just a summary because 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 again we've been to the subcommittee a couple of times 2 in the context of the program and then in the context 3 of the advanced reactor design criteria and the 4 functional containment performance criteria paper.
5 Then I'll give a summary or high-level 6 overview of the technology-inclusive methodology.
7 And then after the break the industry group, NEI, 8 Southern Company and other participants in the 9 industry effort will go over the licensing 10 modernization and in particular the guidance that's 11 now in the draft, the working draft of NEI 18-04.
12 And then we'll close the day this 13 afternoon with a discussion of the specifics of the 14 SECY paper which John mentioned. We'll be asking for 15 a letter on that paper.
16 And the draft regulatory guide and the 17 ACRS can decide on whether they want to weigh in on 18 the draft guide or wait until public comments are 19 received and we move to the next step to finalize the 20 guide.
21 So with that I'll turn it over to Ian.
22 MR. JUNG: Good morning, Chairman and 23 committee members. My name is Ian Jung. I recently 24 took a position as a senior reliability and risk 25 analyst within the same division as John and Bill 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
10 1 working.
2 I've been with the INC as a branch chief 3 for chapter 7. Some of you have heard about design 4 specific review standards for chapter 7 which is 5 somewhat consistent with some of the framework we are 6 talking about from instrumentation control systems.
7 So I bring some background from a technical aspects of 8 it.
9 The reason I'm here is to specifically 10 talk about enhanced safety focused review approach.
11 Some of the members may not have appreciation for some 12 of the background.
13 So, Mr. Ray asked for specifically on this 14 topic. So I want to spend a few minutes on overview 15 of the enhanced safety focused review approach and its 16 potential relationship with the LMP, licensing 17 modernization project.
18 I think there's some relationship and I 19 want to briefly touch upon that.
20 So, this particular approach is a staff's 21 approach for NuScale specific review. The intent was 22 to focus on safety. I'm going to go over that a 23 little bit more.
24 This particular approach is about the 25 tools and strategies for staff to use in defining 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 scope and depth of the review which can have an impact 2 on efficiency and effectiveness of the staff reviews.
3 This is a particular approach is a 4 companion to NUREG-0800 SRP, standard review plan, 5 introduction part 2. I believe the committee has been 6 briefed on this particular topic on SRP update related 7 to small modular reactor reviews.
8 And also I think staff has briefed the 9 committee on design specific review standards and in 10 particular chapter 7 was with the committee several 11 times for mPower as well as NuScale design specific 12 review standards where the whole SRP has been 13 reformatted and restructured to be consistent with the 14 fundamental design principles focus that Mr. Brown has 15 working with us. We have a very positive letter on 16 it.
17 The whole approach is intended to be used 18 during both pre-application and during actual review 19 process. And pre-application and collaboration with 20 potential applicant is critical in success of this 21 particular approach. Next slide.
22 So the overall objective of this enhanced 23 safety focused review approach is to increase 24 effectiveness and efficiency for staff reviews to meet 25 the customer's needs. Also to meet the statutory NRC 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 mission of the regional notional (phonetic) safety 2 finding in an efficient manner.
3 This particular approach is also 4 Commission directed. There are a couple of documents 5 that I just -- I don't want to go over that but bottom 6 line is that Commission told the staff to focus its 7 review and resources on -- to risk-significant SSCs, 8 structures, systems and components and other aspects 9 of the design that contribute most to safety.
10 I think on this topic related to SRP 11 introduction part 2 as well as the design specific 12 review standard and this enhanced safety focused 13 review approach there were presentations to the 14 committee several times. In addition for chapter 7 I 15 think we came to the committee multiple times to deal 16 with the chapter 7 design specific review standards.
17 Next slide.
18 So I just want to highlight there are 19 multiple tools and activities that went on to help the 20 staff with the NuScale review. And one of the review 21 tools that we provided to the staff and had a multiple 22 training and other sessions is this particular tool 23 that providing sort of the table and logic that 24 considers various elements of the -- various elements 25 that could have an impact on the staff review's safety 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 significance or risk significance.
2 This whole A1 A2, B1 B2 approach. A1, A2 3 refers to safety-related risk significant or not risk 4 significant. B1, B2 corresponds to the non-safety 5 portion of that.
6 Of course complying with the specific 7 regulations, how to meet those. Novel nature of the 8 design. NuScale had a multiple areas of novel design.
9 Also related to interaction between safety 10 and non-safety as well as the safety interactions that 11 could have dependencies that could have an impact on 12 safety or risk.
13 Unique licensing approach. The NuScale 14 had some areas where exemptions are made as well as in 15 other areas. Of course the Regulatory Guide 1.174 has 16 elements in safety margin and defense-in-depth.
17 Of course how to deal with the operational 18 programs and additional risk insights to be 19 considered.
20 Through these considerations without 21 dealing some of the issues in a piecemeal I think the 22 intent was to have the staff members consider these 23 various elements in deciding the scope and the depth 24 of the staff reviews. Next slide.
25 This is my final slide. So status 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
14 1 future. This particular approach was applied during 2 the pre-application and during the early reviews.
3 We had various aspects of it that not all 4 place, all disciplines used this approach based on 5 uniqueness of each discipline or the timing of the 6 reviews.
7 We had successes. Chapter 7 is one that 8 I keep referring to. But there are other areas that 9 staff made a conscious decision on considering various 10 elements in deciding the scope and depth of the 11 reviews.
12 The staff is currently developing lessons 13 learned. We expect to have a memo developed to share 14 with the office and in other places.
15 We believe that this particular approach 16 can be used in the future including advanced reactor 17 reviews. We are coordinating with Bill's branch. I 18 think there's more to come.
19 The nexus of that particular approach I'm 20 just discussing with the future licensing 21 modernization project is that -- most of today's 22 discussion is on the framework approach and for 23 industries to use and as an endorsement of it.
24 But I think there's going to be another 25 piece related to what staff can review, what 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
15 1 what scope, efficient and effectiveness of the staff 2 review. So I think standard review plan or this 3 particular enhanced safety-focused review approach I 4 think staff has been discussing how to go about doing 5 that part of the piece.
6 But we are following this particular 7 licensing modernization project very carefully as we 8 deal with the staff portion of the review.
9 Overall I think the underlying concept of 10 the enhanced safety-focused review and -- is 11 consistent with the agency's risk-informed and 12 performance-based approach. That's the end of my 13 presentation. Any questions or comments?
14 ACTING CHAIRMAN CORRADINI: Questions by 15 the committee?
16 MEMBER REMPE: So my understanding --
17 Harold's not here, but my understanding his concern 18 was that some of the required content that has to be 19 submitted would be reduced or would be eliminated from 20 this Reg Guide 1.206 is why he asked us to discuss 21 this at this meeting today. Which is reasonable.
22 But I guess he also was interested in how 23 this would affect how ACRS interacts on such reviews.
24 You can weigh in here, Mike, but with what we saw with 25 NuScale I thought ACRS was pretty much kept onboard.
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 There were interactions with ACRS to make 2 sure we were aware of where you were focusing on. Is 3 that the intent? Like if you're going to say well, 4 certain things don't have to be included as mandatory 5 anymore in the submittals based on your interactions 6 with the designer you would have some way of always 7 coming to ACRS and interacting with us so we 8 understood why certain components would not be 9 required.
10 MR. JUNG: So I think the question is much 11 broader than just enhanced safety-focused review 12 approach. This particular approach is more of a 13 staff's review approach based on what's expected, the 14 information that is expected to be submitted through 15 other vehicles.
16 Regulatory Guide 1.206 is one of those 17 attempts. But I think we recognize that for advanced 18 reactors in particular Regulatory Guide 1.206 update 19 I don't think in my understanding is it does not 20 really -- we didn't create Regulatory 1.206 to be 21 completely up to date associated with the additional 22 approach.
23 But I think I expect that, I mentioned 24 about the standard review plan being updated for the 25 future. I think that's a vehicle that I think 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
17 1 committee has the opportunity to discuss that.
2 ACTING CHAIRMAN CORRADINI: I guess I want 3 to just stay on this slide for a minute to make sure 4 I understand.
5 So to put it in the simplest terms, is the 6 LMP that we're going to hear about throughout the day 7 today a natural outgrowth of what you did for the 8 enhanced safety-focused review for NuScale?
9 MR. RECKLEY: This is Bill. We certainly 10 are taking the lessons learned from that and in 11 previous discussions with both the industry and with 12 the ACRS we talk about bringing the enhanced safety-13 focused review approach forward.
14 One primary difference to keep in mind is 15 that for light waters which is Reg Guide 1.206 and 16 also the SRP this is an overlay of that existing 17 framework to say where additions and maybe 18 subtractions should come in the staff's focus.
19 As we go forward with the non-lights we're 20 going to take some of these concepts like the 21 consideration of operational programs, the focus on 22 safety and so forth, key concepts, but as opposed to 23 overlaying it on that framework we're going to build 24 a framework with those concepts embedded if you will.
25 We're not taking the SRP and scaling 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
18 1 back for non-lights. We're tweaking it for non-2 lights. We're planning on building this framework 3 which you can see the first step in what we're going 4 to talk about today on how to build it basically from 5 the ground up.
6 ACTING CHAIRMAN CORRADINI: Okay. So let 7 me say it back to you so I've got it right. So you're 8 looking at really assembling an SRP that is 9 technology-inclusive.
10 MR. RECKLEY: Right. Which is why it ends 11 up looking more like a methodology than a list. Most 12 of the guidance for light water reactors are lists.
13 There's specific items --
14 ACTING CHAIRMAN CORRADINI: That have to 15 be looked at, that have to be reviewed.
16 MR. RECKLEY: Right. Whereas for what 17 we're going to talk about today since it's technology-18 inclusive it's more of a methodology that any designer 19 for any technology can use to construct an application 20 and then as Ian said we'll have companion guidance for 21 how we're going to do reviews.
22 But it won't look -- our current plans are 23 it won't look so much like a list.
24 ACTING CHAIRMAN CORRADINI: To do things.
25 MR. RECKLEY: 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
19 1 MS. CUBBAGE: I'm seeing Joy's looking a 2 little confused still. This is Amy Cubbage, NRO.
3 So basically in a nutshell the NuScale 4 application was generally developed based on the way 5 applications have always been developed. And then 6 ESFRA was a way for the staff to in certain areas with 7 more or less emphasis.
8 This will develop a different type of 9 application from the bottom up. LMP.
10 MEMBER REMPE: I think I understand. And 11 again, I'm trying to interpret also what Harold 12 conveyed to us. So basically you'll have a process 13 and designer X will come in and they may only need 3 14 of the 10 components that were on the old list.
15 And somewhere the staff will interact with 16 him and concur. And then Harold was concerned how 17 will ACRS fit into this process. And at that point 18 you'll interact with us and we'll say yes, we agree 19 with you, or no, we don't agree with you, you need to 20 include another component. Is that kind of -- are we 21 talking the same thing?
22 ACTING CHAIRMAN CORRADINI: But let me 23 just back up a step because I think Joy said it very 24 well. I think Dennis had some other questions.
25 Dennis.
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 CHAIRMAN BLEY: Yes.
2 ACTING CHAIRMAN CORRADINI: Did you want 3 to ask your question?
4 CHAIRMAN BLEY: Yes, I had a comment and 5 a question. The comment goes back to Joy speaking 6 about what Harold was concerned about.
7 And Reg Guide 1.206 rev 1 states that the 8 technical information that used to be in 1.206 at 9 least the way I read it is going to show up in interim 10 staff guidance or some other form in a while. And I 11 guess the question on that is what's awhile. Is that 12 going to be available about the same time as this reg 13 guide or what are people supposed to do.
14 ACTING CHAIRMAN CORRADINI: There's some 15 background noise on the line so whoever's out there is 16 going to have to mute themselves. Bill, did you get 17 it?
18 MR. RECKLEY: Yes. We are going to 19 continue. We're mixing apples and oranges a little 20 bit as we bring in the non-light discussions and Reg 21 Guide 1.206 update which will continue to be for light 22 water reactors. So just want to keep that. There's 23 two things. They're related but there are separate 24 activities.
25 There are activities underway to provide 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 as Ian mentioned an update and further risk inform the 2 SRP. Those things will take a little while. I don't 3 think they are intended to be companions to the update 4 to Reg Guide 1.206.
5 Ian, do you have any more or maybe John 6 Monninger? No.
7 CHAIRMAN BLEY: I'm a little confused by 8 that because in section B of 1.206 it actually points 9 to the fact that this will be reflected in interim 10 staff guidance by NUREG or some other management 11 document to pick up that technical information that's 12 disappearing from -- so we will leave that on the 13 table if nobody there wants to talk to it.
14 MS. CUBBAGE: So you mean, is that in 15 general or what was that in the context of ESFRA? I 16 think in general, and please, John Monninger, correct 17 me if I'm wrong, I think there is an attempt with the 18 new version of Reg Guide 1.206 to put more of the 19 guidance into the SRP in the future and less in Reg 20 Guide 1.206. But John is coming to the mike.
21 CHAIRMAN BLEY: Okay.
22 MR. MONNINGER: Good morning. This is 23 John Monninger from the staff. I'm the director, 24 Division of Safety Systems, Risk Assessment and 25 Advanced 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
22 1 So I think it's a good discussion and with 2 the staff in approaching the revisions of Reg Guide 3 1.206 and then the SRP they recognize that there was 4 tremendous overlap between the two.
5 So the intent was to the extent possible 6 could you pull out the technical details out of 1.206 7 and put the technical acceptance criteria within the 8 SRP.
9 However, it will take a while to update 10 the SRP so the staff is considering how best to do 11 that and I think that's the concept of how the ISGs 12 were brought into play. 1.206 was meant to be just 13 the format and content of the applications and the 14 real technical criteria the staff is trying to focus 15 that within the SRP.
16 The problem is when we had technical 17 criteria in two different documents when new insights, 18 lessons learned, you know, it was difficult to keep 19 the two documents consistent so the thought was to 20 focus all the criteria within the SRP.
21 I'm not up to speed on the details of the 22 schedule for the ISGs but during lunch we could run it 23 down with the appropriate staff and chit chat in the 24 afternoon session.
25 CHAIRMAN BLEY: I think that's great. 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
23 1 know 1.206 isn't the focus for today, but since we 2 mentioned a few things about it I may -- comment.
3 It seems to me this has gone the wrong way 4 to providing more guidance than just how to put 5 together the application for people who want to come 6 and talk with the staff early and do the kind of 7 things that have been evolving over the last year or 8 two. Seems pretty thorough on that.
9 The other point is although it is for 10 light water reactors right in the second paragraph the 11 staff says they also consider this to apply to other 12 types of power reactors. So I would agree with that.
13 It's kind of slipped off of just being an 14 LWR document, right, even if it's introductory steps.
15 That's about all from me on this, Mike.
16 ACTING CHAIRMAN CORRADINI: So let me try 17 one more time to simplify for me. Maybe everybody 18 else gets it. I'm still -- so it's fair to say there 19 will be a 1.206 prime in some fashion for non-light 20 water reactors and there will be a standard review 21 plan prime.
22 Or will it be just -- because you use the 23 word overlay but I sense it's more than an overlay.
24 It's almost like a buildup from scratch. Which of 25 those two is it? I'm still -- I'm not completely 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 clear.
2 MR. RECKLEY: What we're going to do as we 3 go forward you have NEI 18-04 and Draft Guide 1353 4 which is a first step.
5 We're going to then continue to work with 6 the industry to see what other guidance is needed. If 7 the feedback is more detail is needed on how to 8 construct an application then we'll focus on that. If 9 it is on how to do a particular area within NEI 18-04, 10 maybe one of the analytical discussions that we're 11 going to have later today and the developers think 12 they need more guidance in that area then we'll focus 13 on that.
14 ACTING CHAIRMAN CORRADINI: Okay. That 15 helps. Thank you.
16 MEMBER KIRCHNER: Mike, may I ask a 17 question? So Bill or John or whoever, I know we'll 18 hear about this later today. I'm just a little 19 concerned maybe about reconciling all these different 20 approaches.
21 I'm looking right now on my computer at 10 22 CFR 50.34 and I'm wondering why you wouldn't start 23 there in a technology-inclusive manner and proceed.
24 Because you loop back to that later in your -- in the 25 LMP.
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 So I'm just somewhat concerned about how 2 all these different approaches get reconciled and 3 which ones take precedence in terms of establishing 4 some certainty.
5 You're looking for efficiency and 6 effectiveness in the regulatory process, but I see 7 complexity being built. But maybe I'm not 8 appreciative of how you see this being streamlined 9 when you're done. So maybe it's a discussion for 10 later in the day but just put that marker down.
11 MS. CUBBAGE: Maybe I could just offer --
12 again, this is Amy Cubbage -- that we brought in the 13 ask for discussion at the beginning just specifically 14 to address Member Ray's question relative to the 15 committee's review of Reg Guide 1.206.
16 And really other than the fact that there 17 are some principles in common we're not applying ESFRA 18 in the future for non-LWRs. That's something that was 19 developed for the NuScale review, maybe used again if 20 the opportunity arose, but we see the LMP as really 21 the way we're going for the future for the non-LWRs to 22 develop and bake in the process from the beginning to 23 be risk-informed, performance-based and technology-24 inclusive.
25 It's difficult for going forward with 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
26 1 non-LWR to start with the standard review plan that's 2 largely light water reactor-centric and also isn't 3 even applicable to the non-LWRs. So this process you 4 see in the draft guide is really where we're headed 5 for the non-LWRs and I don't want there to be left any 6 confusion on the ESFRA.
7 MEMBER KIRCHNER: Then let me be specific 8 because you have slides up there that suggest 9 otherwise. What lessons learned have you so far 10 derived from this process and what is being considered 11 in coordination with LMP?
12 MR. JUNG: In terms of lessons learned 13 there's a draft report in there so I don't want to go 14 too far with that. I sort of briefly mentioned that.
15 Because of the timing and uniqueness of 16 the discipline applying I think not everybody, not all 17 the disciplines were able to execute that in a manner 18 that was originally intended.
19 But I think the underlying concept of 20 being able to -- the linkage that I was referring to, 21 there's lessons learned that Bill was also mentioning.
22 There's some elements that are applicable to -- it is 23 a generic because if you look at the definition of 24 risk-informed and performance-based regulation the 25 staff's effort focusing on most safety significance 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
27 1 it. The underlying concepts are the same.
2 But I think as we apply some of these 3 concepts for the future I see some valuable lessons 4 that can be shared in terms of how we approach it.
5 But specific elements of what documents to be 6 submitted and how the specific regulations that under 7 Part 50 specific, those individual regulations, how to 8 deal with that as well as the staff guidance, I think 9 the message is somewhat clear that staff needs to work 10 on and work with industry to come up with something.
11 But I think the underlying safety issues 12 and the elements, we have a great number of staff 13 members who can use the current framework. I think 14 the message from Amy and Bill is that perhaps there 15 are new way of doing business in that regard.
16 MR. MONNINGER: If I could just add two 17 comments on lessons learned. This is John Monninger 18 from the staff.
19 So I think it's -- I think ESFRA was a 20 very important worthwhile effort. I think we really 21 had two big challenges.
22 One is the design of NuScale. It's a 23 light water reactor design, compliance with the 24 current requirements, compliance with the current 25 SRPs. So we tried to come in with an approach 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
28 1 would then almost afterwards sort of dissect that. So 2 that was very difficult to then apply in terms of the 3 design, what to focus on, what should be submitted.
4 As a matter of fact it didn't even impact 5 what was submitted on the NuScale design. So the 6 NuScale design, the ESFRA approach had no impact on 7 the actual submittal.
8 So we had all the existing SRPs, the 9 entire application come in from NuScale and then to 10 tell the staff to focus more heavily on these areas, 11 not to focus as much on these areas.
12 It really represented some internal 13 challenges with how we proceeded.
14 In addition to that the DSRSs that we 15 developed really didn't benefit from a risk-informed 16 approach in development of the DSRSs. Those had to a 17 large extent begin prior to a lot of the ESFRA 18 efforts.
19 The others I think in terms of just change 20 management within the NRC staff. The real ESFRA 21 efforts and focus probably occurred about a year, year 22 and a half prior to the application coming in. So it 23 was difficult in terms of our roll-out and our buy-in 24 on that.
25 Here we're trying to build it in up front 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 well in advance of the applications coming in. The 2 effort is actually being led by and run by the 3 projects, the licensing staff.
4 So the whole issue of -- to me it's two 5 things. It's one in terms of the challenges with the 6 NRC for change management and here we're trying to 7 bake in the process from the ground up.
8 And the other is in terms of the applicant 9 and the design and the material coming in. Have the 10 material coming in and the approach consistent with 11 how we intend to review it. So I think it's two 12 things.
13 A lot of it is change management and 14 execution within the staff and the other is in terms 15 of the actual application of material and expectations 16 on the applicant. They would be the two top lessons 17 learned that I would throw out there.
18 CHAIRMAN BLEY: While John's still up 19 there can I slip something in?
20 ACTING CHAIRMAN CORRADINI: Sure. We will 21 need to move on.
22 CHAIRMAN BLEY: The safety-focused review 23 approach is actually called out in Reg Guide 1.206 rev 24 1. I'm just a little curious, John. And this is not 25 terribly relevant for safety, but your group spent an 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 awful lot of time coming up with this phrase safety-2 focused review. And now suddenly it seems to be 3 replaced by an incomprehensible acronym. I just 4 wonder what led to that. And I'm off.
5 ACTING CHAIRMAN CORRADINI:
6 Incomprehensible acronym. That's what he said.
7 MR. MONNINGER: This is John Monninger.
8 If ESFRA is the incomprehensible acronym. So I think 9 that really represents some internal challenges with 10 change management.
11 Originally the team working on it talked 12 about a risk-informed approach. There are some optics 13 within the agency about risk-informed, risk-based, a 14 reliance upon risk.
15 So really risk and safety, we view it as 16 being one, hand in hand the same thing. However, 17 there are some internal challenges there so we 18 deliberately -- it's the same approach for risk-19 informed performance-based approach but in an attempt 20 to address challenges internally with change 21 management we used the incomprehensible acronym.
22 CHAIRMAN BLEY: Thank you.
23 ACTING CHAIRMAN CORRADINI: Okay, onward.
24 MR. RECKLEY: Okay. And as we go through 25 you can judge to the degree that we've tried 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
31 1 incorporate some of those concepts into what we're 2 doing now for non-light water reactors as we shift 3 over to the primary focus of today.
4 We've been before the subcommittee a 5 couple of times as I mentioned to talk about the 6 overall program, our strategy, and our implementation 7 and action plans.
8 One goal that we've had all along is 9 wherever possible to be technology-inclusive. And 10 that kind of drives a lot of the discussion today as 11 to why we lean towards methodologies.
12 We had the same discussion when we were 13 before you talking about the functional containment 14 performance criteria, that it is more of a 15 methodology. The performance criteria is not a leak 16 rate from a structure, it's a methodology on how well 17 a design using whatever combination of barriers is 18 able to retain the radioactive material.
19 So just a quick summary. The 20 implementation and action plans that we've had from 21 the beginning is divided into six strategies, building 22 the staff's knowledge, developing the tools like 23 computer codes and the ACRS has had a recent meeting 24 on that topic with DOE and the laboratories.
25 Strategy three is to develop a licensing 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 framework. Strategy four is to work with the 2 standards development organizations, ASME, ANS to 3 develop consensus codes and standards.
4 Strategy five is the resolution of policy 5 issues. And again the ACRS has been involved. The 6 proposed rulemaking on emergency planning, SECY 18-7 103, that's going up to the Commission. The 8 functional containment performance criteria, SECY 18-9 96 recently went up to the Commission.
10 Strategy six is communications. And then 11 down at the bottom I have just a couple of points that 12 the staff is trying to remain aware of potential first 13 movers to see if we need to accelerate an activity or 14 change our focus if a particular design or technology 15 is moving ahead of the others.
16 And then a recent topic that's come up in 17 the context of the Defense Authorization Act and 18 elsewhere is micro reactors and the possible 19 development and deployment of those technologies.
20 But the focus today is on the last block 21 under the licensing framework, the licensing 22 modernization project.
23 ACTING CHAIRMAN CORRADINI: Let me ask 24 about the purple circle. This is still an option for 25 the industry.
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 MR. RECKLEY: Yes.
2 ACTING CHAIRMAN CORRADINI: So, not to 3 take us back. If they choose not to use the option 4 they would essentially have to go on a case-by-case 5 exemption under a light water reactor set framework.
6 MR. RECKLEY: Yes, or develop something 7 totally on their own.
8 ACTING CHAIRMAN CORRADINI: Okay, that's 9 what I thought. I wanted to make sure. Thank you.
10 CHAIRMAN BLEY: This is Dennis. Question 11 for Bill. Actually a comment. When this all first 12 started we really pushed for the staff to focus on 13 strategies three and five and I think that's been done 14 pretty well.
15 As you've pointed out all of these pieces 16 are really tied together. Have you heard anything 17 back from the Commission yet on the functional 18 containment paper?
19 MR. RECKLEY: Not yet.
20 CHAIRMAN BLEY: Okay. Because without 21 that I think all of this stuff starts to unravel.
22 MR. RECKLEY: Yes, we agree, and that's 23 why we wanted to send it up first. And what we've 24 explained to anyone who asks is if you have any fixed 25 -- well, I'll get into that in the next slide 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 actually. It's a good one to just lead into.
2 ACTING CHAIRMAN CORRADINI: But before you 3 do. So let's just stay on the policy list because 4 Dennis picked one. What is the status of the others?
5 Or should we -- or is it that the functional 6 containment is probably the leading policy issue that 7 needs to be settled? I see a couple of others there 8 that would concern me to be clear.
9 MR. RECKLEY: Right. So what we're 10 currently working on, on the first one, siting near 11 populated centers. We have guidance and the most 12 restrictive part of the guidance is that we look at 13 population density out to 20 miles. And the guidance 14 is 500 people per square mile out to 20 miles.
15 For the deployment when we talk to DOE or 16 the laboratories or others that's a particular 17 challenge. And so we're looking to see if that is an 18 appropriate factor for smaller reactors or reactors of 19 different technologies.
20 We're currently working on that. We 21 issued a preliminary white paper not with a proposal 22 but just kind of to frame the issue. And we're 23 currently working through our periodic stakeholder 24 meetings to undertake that. And we have a contract 25 with a laboratory to help us evaluate particular 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 possible options.
2 In terms of insurance, the Price-Anderson 3 Act, we the NRC, the agency owes a report to Congress 4 in 2021. We plan to have a section on advanced 5 reactors to say whether we think the current 6 requirements are fine, whether we think they should be 7 changed or whether we think more study would be 8 warranted in terms of what insurance is required for 9 non-light water reactors. So that's an early 10 activity. We have it identified but we've really not 11 done too much yet.
12 Consequence-based security. That SECY 13 paper is identified there, 18-76. That's currently 14 before the Commission where the staff proposed a 15 rulemaking somewhat similar to the EP rulemaking to 16 say we would do a performance or consequence-based 17 approach to security and if certain performance 18 measures could be met requirements such as the number 19 of armed responders might be revised.
20 And then we're always looking to see if 21 there are other policy issues or key technical issues 22 that are identified that we would add to the list.
23 There are others that we didn't list here. We just 24 listed the primary ones that we're currently working 25 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
36 1 ACTING CHAIRMAN CORRADINI: Thank you.
2 MR. RECKLEY: So, one of the goals that 3 the NRC staff has in any case is to try to look at 4 this in an integrated fashion. And as Dr. Bley was 5 mentioning these things are all interrelated. And 6 that makes it difficult because for the light water 7 reactors much of it was put in place in the fifties 8 and sixties and then it was added and revised over the 9 decades in various areas to say what are the events 10 that need to be addressed, what are the controls or 11 barriers to address those threats or events, and what 12 potential measures might be taken to mitigate the 13 consequences if there's a release.
14 So this bow tie diagram was used in the 15 functional containment paper just to kind of lay out.
16 It's got its limitations as an assessment tool 17 perhaps, but it's a good representation of how to 18 consider a number of factors as you're looking at the 19 overall program.
20 So going back to that policy list you can 21 see I've just -- I resist all along trying to put 22 specific things on the blocks in the generic diagram 23 in terms of what are the barriers or controls for 24 different technologies.
25 But just as an example you can put up 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
37 1 and for example EP, emergency planning, the evacuation 2 of people is usually considered the last step, the 3 last mitigation measure that if you have to you 4 reserve the right to move the people out of the way if 5 you're unable to keep the radionuclides from being 6 released.
7 So you see we have an activity in that 8 regard. Insurance and liability and environmental 9 reviews I mentioned as well as siting. That is a key 10 factor not only in terms of things like external 11 events maybe on the prevention side but it's also a 12 measure that's used on the mitigation side. That's 13 why you have population density criteria for example.
14 And then functional containment. The 15 containment function goes beyond just the design basis 16 events, traditional design basis events, and goes into 17 the beyond design basis events if you do have in light 18 water reactors a core damage accident or what we've 19 defined for non-light water reactors the top level 20 event being a plant damage state with the unplanned 21 movement of radionuclides. You need to come up with 22 terminology like that because some reactor designs 23 have a planned movement of radionuclides in the form 24 of molten salt going around the coolant system.
25 So, that is kind of what we were looking 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 at and trying to make sure that as we go forward and 2 look at any particular area that we are also looking 3 at the integrated and how the whole picture fits 4 together.
5 And so for example in the emergency 6 planning proposed rulemaking it points back for non-7 light water reactors to the LMP in terms of where are 8 you going to identify the events that you have to 9 assume in order to assess whether the dose remains 10 less than the protective action guides and perhaps you 11 can justify a smaller emergency planning zone. And 12 I'll get to that in a second.
13 So, whereas that goes beyond the immediate 14 scope of licensing modernization there is a 15 relationship there and the staff is trying to make 16 sure that we remain cognizant of all of these 17 different proposals and that they all fit together in 18 the end to make an integrated approach.
19 As if that figure wasn't complicated and 20 busy --
21 ACTING CHAIRMAN CORRADINI: I congratulate 22 you on the denseness of whatever that is.
23 MR. RECKLEY: So, one of the challenges as 24 you change technologies is the tendency, and we face 25 this all the time, and I do it myself, everybody does 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 it, to start where you're comfortable which is for 2 example we talked earlier of NUREG-800 and so forth.
3 And say okay, we're going to apply it to 4 something different now and how does it change. The 5 more we've looked at that the more we conclude that 6 you're better off to start with First Principles and 7 borrow from NUREG-800 where it's applicable but don't 8 become so wed to it that it actually hampers you going 9 forward.
10 So what this slide which is included in 11 the working draft of the SECY paper is trying to 12 convey is the three fundamental safety functions with 13 the highest level safety function being the retention 14 of fission products or the retention of radioactive 15 materials.
16 And that can be modeled through just 17 basically a set of barriers or controls in saying how 18 well can that barrier attenuate the radioactive 19 materials or the release of radiation or another form 20 of the equation what's the release fraction across 21 each barrier as you go through the process.
22 And one of the things that you'll see is 23 a different reliance on different barriers for the 24 different technologies. And so we thought it was 25 important to start with again high-level First 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 Principle kind of approach.
2 The formula there if you broke this down 3 into a formula is the basic DOE five-factor formula on 4 the retention of radionuclides or the source term for 5 the release from a reactor or non-reactor facility.
6 So, the bottom half of the figure is the 7 other two fundamental safety functions, the heat 8 generation and the heat removal. And it's basically 9 again just trying to represent that you can do that at 10 a high level just by the heat generated from the decay 11 heat or from the core or from whatever source that 12 you're addressing.
13 And then the heat removal through the 14 various paths ultimately out to the ultimate heat 15 sink. So, for passive reactors it does generally look 16 something as simple as this where it's just going from 17 the core to the reactor coolant system or primary 18 system or whatever you want to call the primary system 19 through a building and then to a reactor cavity 20 cooling system or something where it's released to the 21 environment.
22 The failure on the bottom either in heat 23 generation or heat removal such that you have a 24 mismatch is in general what causes the degradation of 25 the barriers in the top level approach. And so 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
41 1 is how these things kind of generally fit together.
2 I know it's an over-simplification but it 3 was just an attempt on our part to try to focus the 4 staff as we developed what is the content in an 5 application and what the staff's going to look at 6 during the review to focus on what's important.
7 You start with the fundamental safety 8 functions as basically being a good place to start.
9 And then as we build through this process as we're 10 going to talk during the day using various analytical 11 tools, probabilistic risk assessment, deterministic 12 assessments and other tools, you're basically looking 13 at how well does a design satisfy these fundamental 14 safety functions.
15 MEMBER REMPE: If you only look at these 16 or what's in this diagram why would you need to worry 17 about having redundant shutdown systems because you 18 could have a low power reactor that stays critical for 19 a long period of time as long as you can remove heat.
20 So you've gotten rid of the general design criteria 21 needing to have redundant shutdown systems, right?
22 MR. RECKLEY: Well, as we go through the 23 process you would have to show that whatever you're 24 relying on provides you the needed confidence.
25 And so if it is small enough and simple 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 enough could one conceive that it be as you said, 2 perhaps. But you would have to have the confidence 3 that that heat path for example couldn't be 4 interrupted, and if it could be interrupted then maybe 5 you need either a diverse redundant and/or diverse 6 function in order to serve that function.
7 And that would come out of all of the 8 assessments we're going to talk about during the day.
9 But if you go down to could it be small 10 enough or simple enough that you didn't need it, I 11 wouldn't rule that out. But you'd have to see and the 12 point would have to be proven that the reliability and 13 the confidence that you have in that single thing 14 would be enough.
15 So going back to the bow tie I tried to 16 represent in general terms what we're going to be 17 talking about today through licensing modernization.
18 And it captures basically this part of the bow tie.
19 The internal plant events, malfunctions, failures of 20 plant equipment, external hazards, the plant systems 21 and operational programs that are there to address 22 those events, and in the beyond design basis category 23 if the technology has a plant damaged state with an 24 unplanned movement of radioactive materials what the 25 plant might include to address that particular 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 scenario.
2 It doesn't -- LMP doesn't feed over into 3 the external responses, things like siting and 4 emergency planning. It doesn't directly address 5 environmental reports. And it doesn't directly 6 address security, radiological sabotage type events.
7 Although in all of those areas you can 8 draw some information from LMP.
9 So again looking at the LMP and how it 10 fits into the regulatory structure. And this came out 11 of our June meeting so I wanted to touch on this a 12 little bit.
13 Within the licensing modernization 14 activity there are specific regulations that are 15 mentioned and credited for how this system -- this 16 methodology would work.
17 Examples of that are quality assurance in 18 the maintenance rule. As you go through the 19 methodology it's going to define the desired 20 reliability of equipment, for example. How do you 21 ensure once you go from the design stage into 22 operations that that reliability is maintained.
23 You'll use something like the maintenance rule or 24 something related to the maintenance rule in order to 25 help provide that confidence.
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 As I mentioned the LMP interfaces although 2 it's not specifically mentioned in the document or 3 addressed specifically it is an interface with other 4 regulatory requirements like siting, emergency 5 planning and environmental reviews.
6 As I mentioned under emergency planning 7 when you say for non-light water reactors how will you 8 define the event by which you'll judge whether you're 9 remaining under one rem to activate the PAGs, the 10 protective action guidelines, that will come out of 11 the events that are identified through the LMP.
12 There are requirements that are beyond LMP 13 in which the LMP doesn't directly interface but which 14 an applicant would have to address. Some of those are 15 just routine effluents, Part 20.
16 If we do an equivalent to Appendix I for 17 non-light water reactors Appendix I and 10 CFR Part 50 18 which address those routine effluents. Worker 19 protections and other Part 20 kind of requirements.
20 As I mentioned security and aircraft 21 impact assessments not directly affected. But 22 designers should be looking at these requirements as 23 they're looking at LMP to see that the overall design 24 is meeting all of these requirements and from their 25 perspective that they meet it in the most efficient 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 way they can.
2 One easy probably example is aircraft 3 impact. If they do LMP and look at all the natural 4 events like perhaps wind is an easy example and decide 5 that a building structure has to be X and they 6 continue along on that assumption all the way until 7 they get further in the design and then they'll say 8 now we're going to do our aircraft impact assessment 9 then they face the potential to say oh, that building 10 should have been thicker, or some combinations of 11 walls should have been different, or maybe we should 12 have given more thought to putting it below grade.
13 So they need to be aware of all of these 14 things as they're doing the design and I think this is 15 the case. We all experience that they're well aware 16 that they need to address all of these things. But I 17 did want to just separate out. LMP doesn't answer 18 every question, it doesn't answer every regulation, 19 that there are others out there that they'll have to 20 address.
21 MEMBER SKILLMAN: Hey Bill, before you 22 change that slide. This is Dick Skillman. This list 23 appears to me to be a list that was constructed or 24 developed by designers.
25 And let me make a contrast. Over the last 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 couple of decades we've watched the regulations 2 change. Give you an example. In 1971-72 Appendix B 3 to 10 CFR Part 50. Later on I think the gold standard 4 was 50.65 the maintenance rule. I mean that was a 5 fundamental change.
6 Industry resisted that like the dickens 7 and it has turned out to be one of the most important 8 changes in regulation at least from my years of 9 experience.
10 But there have been other lessons learned 11 that may not be represented here that come from the 12 operating teams. As I said this appears to be a list 13 developed basically by designers.
14 I'm wondering are there some key lessons 15 learned from the operating side of industry and from 16 the oversight of operations by the NRC that would add 17 to this.
18 Actually, make it better.
19 MR. RECKLEY: I would assume that there 20 are.
21 MEMBER SKILLMAN: I think so too.
22 MR. RECKLEY: Let me clarify that this 23 wasn't intended to be all-inclusive.
24 MEMBER SKILLMAN: This is not a 25 comprehensive list.
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 MR. RECKLEY: Yes. And it wasn't by 2 designers, it was just by me.
3 But the primary reason I wanted to address 4 it was to say that LMP doesn't answer every question.
5 MEMBER SKILLMAN: It's a good place to 6 start. What I'm suggesting is that there is a I don't 7 want to say list. There is a recognition by the 8 operating individuals that yes, you have to design it 9 properly, yes, you have to include design features and 10 functional performance requirements to ensure that the 11 machine does what it's supposed to and that the health 12 and safety of the public are protected.
13 But beyond the if you will design features 14 there are probably some other issues that need to be 15 woven into quote "other requirements" to protect or 16 further enhance the level of safety of new plants 17 whether they're light water plants or they are non-18 light water plants.
19 MR. RECKLEY: I agree with you.
20 MEMBER SKILLMAN: Thank you.
21 MR. RECKLEY: And perhaps when we get into 22 the defense-in-depth discussions and the integrated 23 decision-making panel they can touch on that a little 24 bit later this morning or this afternoon.
25 MEMBER REMPE: So, I actually am glad 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
48 1 hear you say this and I noticed in the difference 2 between whatever version we're looking at, N versus 3 the ones we looked at last summer, they actually added 4 a paragraph explicitly saying hey, just because you 5 meet the top level regulatory criteria that we 6 identified doesn't mean you're going to satisfy all 7 the regulations.
8 That was a concern I had when I read the 9 version last summer. So I'm glad to see both of you 10 guys emphasizing that now.
11 MR. RECKLEY: We thank you. That was 12 directly in response to the question.
13 CHAIRMAN BLEY: This is Dennis Bley. I'm 14 going to follow up on Dick's comment and your 15 response.
16 One place where we've really seen the kind 17 of things Dick's talking about is on newly designed 18 plants. The main control room board and operating 19 procedures linked together through I'll say software 20 but are linked together provide the operators with 21 additional tools to understand things about their 22 plant, or to a large extent based on events that have 23 happened in the past. It kind of fits in that 24 category.
25 MR. RECKLEY: Yes. Again, agreed. One 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
49 1 the things that we do have to keep in mind I think is 2 I don't want to overstate this too much, but at least 3 the operating fleet is operating largely in a nineteen 4 seventies world.
5 And so as we go through things like the 6 man-machine interface that you're mentioning, Dr.
7 Bley, the technology has developed a lot over those 8 decades and I think it's --
9 PARTICIPANT: I'm just joking. I would 10 never suggest such a thing.
11 ACTING CHAIRMAN CORRADINI: I think we 12 have people online that have to mute.
13 MR. RECKLEY: That generated a response 14 anyway. So I think it's fair to say that people 15 designing plants today are looking at the available 16 technology and areas like man-machine interface and so 17 forth.
18 So going forward and getting again to try 19 to lay out a little bit of the high level and then the 20 industry folks are going to talk to you for a couple 21 of hours about the details. And they're also going to 22 go through largely at the suggestion from the June 23 meeting some experience that has been gained through 24 tabletops with different designs.
25 ACTING CHAIRMAN CORRADINI: Did we already 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 do slides 13 and 14 and I just missed it? Just 2 checking.
3 MR. RECKLEY: I'm a little repetitious.
4 So you wouldn't have missed anything anyway.
5 The general approach within the regulatory 6 guide and also the companion SECY paper building off 7 of NEI 18-04 is to divide the framework into licensing 8 basis events, and that gets looked at both from a 9 probabilistic risk assessment viewpoint as well as 10 deterministic viewpoint.
11 The safety classification and performance 12 criteria, how do you define those for structure 13 systems and components. Looking at what function does 14 that SSC play, which ones would be identified as being 15 safety-related and therefore subject to the higher hat 16 in terms of quality assurance.
17 I think probably more importantly to some 18 degree is how do you look at the non-safety-related 19 equipment and determine what special treatment 20 requirements, what are the reliability and 21 capabilities you're crediting for that equipment and 22 how do you assure it once you get into operations.
23 We have that now to some degree through 24 things like regulatory treatment of non-safety 25 systems, RTNSS. And if you go over to 50.69 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
51 1 it.
2 But this is again not an overlay in how 3 can you change a design, or how can you change your 4 operations for a reactor that's already been designed, 5 but from the beginning how can you build in this 6 logic.
7 Going back to the first discussion. In 8 personal opinion, one of the better things about this 9 overall approach again in my view is the marrying of 10 the design and operations better than we traditionally 11 did under Part 50. And there will be a talk later on 12 about looking at the plant capability or the hardware 13 and the companion performance and operational programs 14 that go along with that.
15 And that is included also in the defense-16 in-depth assessment which is the last bullet up here 17 looking again at the programmatic areas, at the 18 hardware and then giving it a good scrub through an 19 integrated decision-making process looking at it 20 through multi-disciplinary going to Dick's point, the 21 operations as well as design to see how it carries 22 forward.
23 MEMBER REMPE: Before you leave that 24 slide. The one thing when I read through this and I 25 think about it, this integrated decision panel 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
52 1 which they do have additional guidance on in NEI 00-2 04.
3 Has the staff ever interacted with such a 4 panel? Especially the way they've placed such 5 emphasis on this panel it's going to be with the 6 design from its inception through licensing. I'm just 7 wondering what kind of issues might crop up with it 8 versus how the regulator and the panel cite their 9 opinions.
10 MR. RECKLEY: We've --
11 MEMBER REMPE: -- it will work.
12 MR. RECKLEY: We've had closely related 13 experience I would say through both 50.69 type reviews 14 and our reviews of PRAs and the peer review process.
15 But maybe I'll just ask the IOU for the 16 industry presenters if they have any other examples 17 where the staff has interacted.
18 I think there's been a couple of close but 19 not exactly from the point of the design where we are 20 now going forward on these non-light waters.
21 MEMBER REMPE: So in past experiences with 22 50.69 how did it work? Was it well documented? Did 23 you like how this multi-experience whatever background 24 panel came up and supported the design?
25 MR. RECKLEY: I might have to take an IOU 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 on that unless Jason?
2 MR. REDD: Good morning, Jason Redd from 3 Southern Nuclear.
4 I believe that we can make some comments 5 on this -- make some comments on this topic in our 6 session coming up soon. Thank you.
7 MEMBER REMPE: Thanks.
8 MR. RECKLEY: That would at least be from 9 the industry side. I'll take an IOU maybe during 10 lunch to see if I can get with NRR. I haven't been 11 personally involved so I can't.
12 MEMBER REMPE: Even with the tabletops I 13 don't think that you've had that interaction yet with 14 the LMP process at all. So I'm real curious on how 15 it's going to work.
16 MR. RECKLEY: So, one of the -- another 17 area is the key considerations as the staff looked at 18 this and developed the draft guide and the SECY paper, 19 the enclosure 1 to the SECY paper and we mention it in 20 passing in the draft guide goes through the evolution 21 of this approach.
22 You can take it probably back further than 23 this if you want but I tend to start with the 24 development of the Advanced Reactor Policy Statement.
25 There was an immediate test of the Advanced Reactor 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 Policy Statement through interactions funded by DOE 2 and the staff looked at various designs including 3 PRISM, modular high-temperature gas reactor and PIUS 4 at the time as well as the CANDU 3.
5 Lessons learned from that in the 6 identification issues was in SECY 93-092.
7 Around this same time the risk-informed 8 performance-based focus with the PRA policy statement, 9 the 1999 Commission white paper on risk-informed 10 performance-based regulation was issued.
11 That obviously related to the things that 12 were going on at the same time. Those efforts were 13 applicable to both light water operating reactors as 14 well as the development of the non-light water reactor 15 technologies.
16 SECY-0347 was a follow-up where we came 17 back to the Commission to propose resolution of some 18 of those policy issues. That ends up being a key 19 paper and I'll talk about it a little more this 20 afternoon.
21 Just as an example of the marrying of the 22 risk-informed approaches and the development of non-23 lights as well as other reactors you had the 24 development and issuance of NUREG-1860 which is a 25 feasibility study for a risk-informed 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
55 1 Throughout all of that you can see that 2 some similarity to the traditional light water reactor 3 licensing structure is maintained, but one of the 4 things as we go through today and as you look at the 5 draft guide and the Commission paper is there are 6 differences and some of those differences are hard to 7 recognize on first blush because the terminology 8 that's used uses some of the same terms but with a 9 different definition.
10 And so just be a little careful as you go 11 forward to say oh, I know how design basis events are 12 analyzed. Design basis events are defined for light 13 water reactors, they're defined for non-light water 14 reactors using this methodology. It's a different 15 definition.
16 Safety-related. The derivation of how 17 something is safety-related is slightly different here 18 than it is in Part 50, Part 1000 for light water 19 reactors.
20 Anticipated operational occurrences. Same 21 term, slightly -- and similar but slightly different 22 definition in this case versus what you may be 23 accustomed to in chapter 15 of the light water 24 reactor.
25 So it's just a caution that whereas 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
56 1 overall structure is similar there are key differences 2 and some of those differences are hard to pick up on 3 in part because the same terms are used with different 4 definitions.
5 ACTING CHAIRMAN CORRADINI: So you'll 6 remind us of this since we're forgetful.
7 MR. RECKLEY: One of the reasons to bring 8 it up now is so when you bring up a question the 9 answer might be careful, this is one of the areas 10 where our definition is different than the Part 50 11 definition.
12 ACTING CHAIRMAN CORRADINI: So let me ask 13 you, maybe you said it, I didn't hear you mention the 14 next generation, the NGNP.
15 So I guess I'm empirical enough that I 16 want an example. So what is it about what we're going 17 to hear that's different than what was proposed for 18 the NGNP?
19 MR. RECKLEY: It is most similar to NGNP 20 and I should have listed it up there. It's on future 21 slides. It is most similar to the approach of NGNP.
22 It's been refined a little bit based on 23 interactions both with the staff and also as the 24 effort was made to ensure it would be technology-25 inclusive it was tweaked 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
57 1 But it closely resembles NGNP. I see Karl 2 Fleming, so Karl, if you want to weigh in.
3 MR. FLEMING: Karl Fleming, LMP project.
4 During my presentation this morning and maybe early 5 this afternoon I will highlight the similarities and 6 differences with NGNP.
7 But Bill is correct, it's primarily the 8 NGNP process with some refinements.
9 ACTING CHAIRMAN CORRADINI: So, if you can 10 hold on a second. Then you'll tell us more, but at 11 this point I personally found reading through this 12 stuff difficult. Maybe it was because it's process 13 and framework.
14 I really think if it's that similar an 15 empirical example would really help. Maybe the 16 industry gets it, but at least me trying to wade 17 through the documents, I kept on asking myself gee, 18 how is this different.
19 Because the frequency consequence curve is 20 1860, the NGNP frequency consequence curve was 1860 21 with attempts to place DBEs and LBEs on it.
22 So I think it would help for the less than 23 completely involved individuals in this to marry those 24 because I just think that would be a nice way of 25 walking through 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
58 1 I really had a hard time in some sense 2 trying to understand the process steps which you're 3 going to go through.
4 MR. FLEMING: Good feedback.
5 MEMBER SKILLMAN: I'd like to weigh in on 6 that just for a second. It seems out in the operating 7 plant world we use a term called error likely 8 situations. This is one. But it's right here in the 9 staff.
10 And it just seems that it might be useful 11 if we're using the same acronym at least mark the 12 unique use of the acronym with a sign or something 13 that communicates this is for the different 14 application so that those who would read would say ah, 15 get it, this is not identical, it's similar, caution.
16 But this really is an error likely 17 situation for those who are trying to digest this 18 information. Thank you.
19 MEMBER BALLINGER: Might we ask for a 20 table that clearly lists the differences?
21 MR. RECKLEY: You can ask.
22 MEMBER BALLINGER: Can we make it a formal 23 request?
24 MS. CUBBAGE: In the back of the NEI 18-04 25 document there is a table that lists a number of terms 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 and in the right column if it says LMP that means it's 2 a definition that came from LMP, and if it's the same 3 definition as elsewhere it says where it came from.
4 MEMBER BALLINGER: I read that but I 5 wasn't sure whether it was complete.
6 MR. RECKLEY: We will take a look and by 7 the full committee we will prepare -- we'll prepare as 8 best we can.
9 I just want to make sure, your request was 10 on terminology or a comparison with NGNP?
11 MEMBER BALLINGER: Terminology.
12 MR. RECKLEY: Okay. Terminology is a 13 little easier.
14 And as Amy pointed out one of the major 15 things that was developed as we went through this was 16 the glossary that's at the back of 18-04.
17 So again, keeping at kind of the high-18 level discussion as I mentioned the methodology 19 consists of the three primary elements, the licensing 20 basis event selection and analysis, the classification 21 of equipment and the derivation of performance 22 requirements in assessing defense-in-depth.
23 I'll say it probably a few times going 24 through the day but the emphasis here is that this is 25 an integrated approach and the staff is looking 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
60 1 these three elements within this methodology and they 2 are like three legs to a stool. They're all 3 complementary and they're all interdependent.
4 And so when we say that this is an 5 approach that's okay for the selection of licensing 6 basis events that goes to that's okay because of the 7 way the defense-in-depth is also addressed within this 8 methodology.
9 Likewise the safety classification and the 10 assessment of the defense-in-depth. These things 11 interplay with each other and we're saying that the 12 three elements fit in this process and work together.
13 You would be challenged just to pick up 14 one of these elements and say I'm going to pick my 15 licensing basis events this way but I'm not going to 16 do safety classification or a defense-in-depth 17 assessment in the same way.
18 Then the next bullet on the slide.
19 Another thing to keep in mind is when it comes to the 20 actual regulatory decisions the criteria are basically 21 the same in this methodology as are in the current 22 rules.
23 The 50.34 25 rem number, that's used in 24 this methodology. The safety goal, the NRC safety 25 goal at the lower end of the curve, that's also within 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 this methodology as one of the aggregate measures 2 that's ultimately used to show the adequacy of a 3 design.
4 The assessments are performed both using 5 risk-informed and deterministic approaches and as was 6 mentioned that includes the engineering judgment that 7 would come from the integrated decision-making 8 process.
9 And the methodology includes a specific 10 element and step for looking at defense-in-depth and 11 how that's provided using both hardware and 12 programmatic controls, and how the programmatic 13 controls are developed to support the defense-in-depth 14 assessments, the uncertainties that might exist in a 15 particular design and so forth.
16 And for me that becomes a very important 17 point to keep in mind as you go forward because one of 18 the questions that often arises for non-light water 19 reactors is how do you address the availability of 20 less operating data, of less operating experience.
21 And one key way that that's done is 22 through this defense-in-depth assessment and really 23 looking at both the plant capabilities or the hardware 24 and what would be appropriate in terms of 25 surveillances and monitoring and reliability targets 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 and all the other things that you can set on the 2 operating side in order to try to address some of 3 those uncertainties.
4 MEMBER REMPE: So on the second bullet in 5 your discussions with industry this does -- especially 6 for -- well, design certification you have to assume 7 some sort of characteristics about the site.
8 In the past, in the MHTGR example they 9 reference that EPRI document that had like some 10 hypothetical site that bounded 85 percent or some 11 fraction of the site.
12 Did you try and push -- I mean, you're 13 going to have a lot of different vendors coming in 14 theoretically with a bunch of different designs. And 15 if they would all just pick the same theoretical site 16 wouldn't that make things easier and did you guys 17 discuss that with NEI?
18 MR. RECKLEY: It might make it easier in 19 some regards for us. The problem that arises is that 20 these technologies to some degree have different 21 potential uses, customers and locations that makes it 22 kind of hard to say we're going to pick a generic 23 envelope if you will.
24 Whereas some designs might be able to say 25 off the bat we don't see Alaska as a potential siting 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 others are being developed specifically for those kind 2 of environments.
3 And so we are basically comfortable 4 leaving it up to the designers to say you know the 5 marketplace that you're trying to pursue. When it 6 comes to picking an external envelope to try to bound 7 where you want to put these it's really up to you to 8 do.
9 MEMBER SKILLMAN: Excuse me, Dennis, this 10 is Dick. Go ahead.
11 CHAIRMAN BLEY: Okay. Bill, you said 12 something earlier that got me curious. You were going 13 into the (telephonic interference) NEI 18-04 and 14 something we put together.
15 In the guidance that you're going to get 16 to this afternoon, it looks as if NRC plans to endorse 17 NEI 18-04 with a few exceptions or clarifications.
18 How -- when you look at NEI 18-04 is that kind of a 19 consensus between the industry -- the NRC, or is it a 20 separate product that you're evaluating later -- new 21 reg guide?
22 MR. RECKLEY: It's a separate document, 23 NEI 18-04 that the industry owns. And they'll be 24 asking for our endorsement via the regulatory guide.
25 At the same time they didn't develop it 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
64 1 a vacuum and we've gone through -- you'll notice it's 2 revision N. I think we've seen at the staff level 3 during interactions three or four of those iterations 4 and provided feedback of our own as well as what the 5 industry has provided.
6 And as Amy just mentioned plus the white 7 papers that preceded it, plus NGNP that preceded that.
8 And so it's their product and they're free to put in 9 what they want. At the same time we've provided 10 feedback in order to if possible minimize the number 11 of exceptions or even clarifications that we might 12 need to add.
13 MS. CUBBAGE: And Dennis, this is Amy 14 Cubbage. If you're specifically getting at the 15 glossary in the back we specifically discussed that 16 with industry at multiple public engagements and we 17 provided input to them on that.
18 CHAIRMAN BLEY: Thanks. I was just trying 19 to generalize.
20 MR. RECKLEY: And by the way, that's not 21 any different than other guidance documents that are 22 developed by the industry and then ultimately endorsed 23 by the NRC.
24 MEMBER SKILLMAN: I'm going to hold up, 25 thanks.
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 MR. RECKLEY: And again I'm going to just 2 touch on these because there will be additional 3 discussions of the actual methodology.
4 I just wanted to put a little staff 5 context and maybe overview to prepare for the 6 subsequent presentations by the developers of NEI 18-7 04.
8 A key aspect of this methodology as well 9 as the NGNP and ANS 53.1 and basically the whole 10 methodology that largely arises from the gas cooled 11 reactor community and is being revised and updated 12 here was the use of the frequency consequence diagram.
13 And one of the things that we would like 14 to emphasize here is what's in the bullet which is an 15 extract right from NEI 18-04 and it's an extract more 16 or less right from the reg guide is that the target 17 figure is a useful tool when you're doing the 18 discussions assessment, when you're doing the safety 19 system classifications, but don't look at it as an 20 acceptance criteria where on one side of that line 21 you're okay and on the other side of the line you're 22 not okay.
23 The other caution --
24 ACTING CHAIRMAN CORRADINI: So can I --
25 with your first caution. But as we at least I thought 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 we said in June if I start approaching the line things 2 become concerning. That's the point of having some 3 line.
4 MR. RECKLEY: That's right.
5 ACTING CHAIRMAN CORRADINI: Okay.
6 MR. RECKLEY: Yes, the closer you are to 7 the line the more concern. And there is a point where 8 it's unacceptable but we're trying not to use this 9 curve that way.
10 As I mentioned earlier ultimately the 11 regulatory decisions are made using basically the same 12 metrics we use now which are the aggregate measure of 13 the NRC safety goal policy statement, the specific 14 assessments that are done against the criteria in 10 15 CFR 50.34, the dose reference values, the 25 rem 16 number.
17 For those designs or projects that are 18 pursuing a reduction in the emergency planning zone, 19 that EPA PAG dose limit is marked on the figure. That 20 might become a reference value that they need to 21 address in the design.
22 But overall the figure is used in the 23 context of identifying risk-significant licensing 24 basis events. It is used in the defense-in-depth 25 assessment and in the safety classification.
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 It's I think familiar to you, the 2 anticipated operational occurrences, the DBEs. Again 3 this is one of those cases where there's a definition 4 difference.
5 DBEs are those event sequences between 10-2 6 and 10-4. In light water reactors a DBE, that 7 terminology is used as a broad category that is used 8 within the definition of safety-related equipment and 9 includes design basis external events, anticipated 10 operational occurrences and design basis accidents or 11 postulated accidents. I forget the exact terminology 12 but they're the same thing. And special events.
13 And then to me what is actually key to 14 this methodology is the inclusion of the beyond design 15 basis events from the beginning and the assessment of 16 those low likelihood events within the methodology.
17 I'll let it go into this afternoon for a little more 18 discussion of that, but again you can contrast that 19 with the existing framework which through being 20 conservative in the assessment of postulated accidents 21 and anticipated operational occurrences was largely 22 trying to address the fact that they didn't address 23 some lower likelihood beyond design basis -- what we 24 now call a beyond design basis event.
25 The last category on the curve there 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
68 1 that DBAs, design basis accidents, are maintained as 2 a category within the licensing basis events. They're 3 done largely the same as is used now within chapter 15 4 of a typical safety analysis report crediting safety-5 related equipment and using analytical methods that 6 are consistent with the guidance that the staff has 7 issued for chapter 15 transient and accident analyses.
8 MEMBER REMPE: So a few weeks ago I was at 9 a meeting and a designer put up a plot that showed the 10 risk of the plant as a function of years based on 11 their increased knowledge. And I would have actually 12 liked to have seen a similar plot that also had the 13 risk of their plant as a function of dollars invested 14 in the design development because it was going up and 15 down.
16 The reason I'm bringing that up now is 17 that it might be good to provide some perspective 18 about what was it Rickover's letter that said a paper 19 reactor is very, very safe and then as you have more 20 knowledge and more information that you find out that 21 it has more issues that you have to address.
22 I just think that some perspective might 23 be useful in your document of what the staff expects 24 or some caveats to the design developers that are 25 coming out with their concepts claiming they're 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
69 1 safe.
2 MR. RECKLEY: We try to do that through 3 the pre-application discussions that we have with them 4 as does EPRI and others that are involved in various 5 exercises. So we try.
6 MEMBER MARCH-LEUBA: Okay, this figure, I 7 don't want to call it -- is merely a mathematical 8 problem. I guess the left side of my brain. And the 9 issue is I may not be using the proper methodology, is 10 segmentation of events.
11 If I take LOCAs and I decide to call it 12 LOCAs that happen at midnight, LOCAs at 1 a.m., LOCAs 13 at 2 a.m. suddenly the frequency of my LOCAs is 24 14 times more.
15 So when you plot only -- by making my 16 events very, very specific I get a lot more events and 17 I don't change the line. See what I'm talking about?
18 There has to be some guidance.
19 ACTING CHAIRMAN CORRADINI: I think what 20 Jose is asking is what Dennis asked in June which is 21 the bundling of these so they're appropriately bundled 22 so that I don't by parsing enough they all get --
23 well, that's what I think you said.
24 MR. RECKLEY: This is a question that's 25 come up. I guess I'll ask Dr. Fleming if he wants 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
70 1 MR. FLEMING: It's a very good comment.
2 The LBE in this process is defined by constructing 3 event sequence families. And it requires you to group 4 together event sequences that have similar initiating 5 event challenge to the plant safety functions and if 6 there is a release mechanistic source term.
7 So you're required to group the sequences 8 that are similar to avoid abuses like subdividing like 9 that.
10 MEMBER MARCH-LEUBA: That has to be very 11 specific in the guidance and should be on the standard 12 review plan. That should be part of the review that 13 they didn't cheat on the generation.
14 And also at the end of the day if I have 15 a house that is downwind from this reactor I don't 16 care what my risk is due to a LOCA I want an internal 17 risk. And I don't know how you add up all these 18 points to give me my risk in my house five miles 19 downstream. This will give you a risk for each 20 particular event. Again I want to know what is my 21 risk.
22 MR. RECKLEY: Right. And there are 23 aggregate measures where you take the whole risks.
24 The summation of the sequences.
25 MEMBER MARCH-LEUBA: But you probably 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
71 1 have eliminated a whole bunch of events to do the 2 aggregate. You will only aggregate the DBEs.
3 MR. RECKLEY: And the beyond design basis 4 events.
5 MEMBER MARCH-LEUBA: You will aggregate 6 everything?
7 MR. RECKLEY: Yes.
8 MR. FLEMING: Yes, if I might amplify.
9 Because one of the applications is to select licensing 10 events for different applications including coming up 11 with our design basis accidents we needed a tool to 12 look at the risk significance of individual LBEs 13 separately.
14 However, we also have three cumulative 15 risk metrics where we accumulate the risk from all the 16 event sequences against the two QHOs from the NRC 17 safety goals. And we also have a metric for the high-18 frequency low consequence events that's based on 19 assuring that 10 CFR 20 is maintained.
20 So we have the aggregate measures and the 21 separate measures.
22 MEMBER KIRCHNER: Well, let me ask a 23 specific question. Which individual risk is the 24 anchoring point down at the bottom right of the -- is 25 this 750 rems which is the large release, or is 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
72 1 early fatality within one mile?
2 MR. RECKLEY: And this is the caution that 3 we bring out both in the reg guide and elsewhere is 4 one of the reasons again that this methodology we 5 think works within the overall construct is that the 6 bottom figures actually don't correlate to actual 7 criteria.
8 For example, the bottom that you'll see is 9 the effective dose over a month whereas the criteria 10 for emergency planning will use a different number, a 11 different time period.
12 The 750 rem roughly correlates maybe to 13 the prompt fatality but we didn't want to argue --
14 MEMBER KIRCHNER: That's much greater than 15 a prompt fatality.
16 MR. RECKLEY: But we didn't want to --
17 MEMBER KIRCHNER: It's not roughly 18 correlating. LE50 is a much lower number.
19 MR. RECKLEY: Yes, it's a couple of 20 hundred. So we knew that as we went in and for the 21 purposes of the methodology. Again, this is why I 22 keep coming back. As an integrated process to look at 23 the methodology we're fine that that number does not 24 actually correlate to the 50.50 prompt fatality 25 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
73 1 MEMBER KIRCHNER: Because you can't 2 reconcile all those it is cleaner to show it as fixed 3 points and a solid line. But like in 10 CFR 50.34 4 there is that footnote that cautions people that we're 5 not intending -- let's see, it's there, the 50.34 dose 6 limit.
7 The intention is not to approach that 25 8 rem exposure. So I just would feel personally, this 9 is just one opinion, if there were some band on this 10 that suggested you don't want to be approaching this 11 line from what would be on the left side of decreasing 12 risk significance. You don't want to be bumping up 13 against this line with your quote unquote "advanced 14 design."
15 The expectation is that you're not going 16 to really come close to this or it's not an advanced 17 design. By policy statement of the Commission.
18 So, the expectations are not to press that 19 envelope. And I don't know visually how best to do 20 that other than putting some kind of hatched area on 21 the lower side of that that kind of suggests. And I 22 know, I guess the designers will go and say oh, I see 23 what they mean, it's now not 25 rem, it's 20.
24 But something that suggests that you're 25 not expecting these designs to push this envelope.
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 MR. RECKLEY: Right. The presentations on 2 18-04 will specifically address one there is a hashed 3 area that's two orders of magnitude lower than this 4 line for looking at what you'd call a risk-significant 5 event.
6 Then as I mentioned earlier most of these 7 designs are going to go for the one rem at the fence 8 objective. That would limit it as well.
9 But I think we'll get into it and if it's 10 not addressed then I'll be back this afternoon. But 11 it's a good comment. Yes. And we tried to address 12 that specifically within the regulatory guide by 13 saying don't look at these points as the acceptance 14 criteria.
15 Dr. Corradini mentioned 1860 earlier.
16 I'll offer a personal opinion. It was a great 17 document but the stair step approach, many of us like 18 the straight lines and that causes you to have some 19 compromises here or there versus having so many 20 different break points as 1860 had.
21 So, but I understand your point and as we 22 get into it if it's not addressed as we go through it 23 we can talk. That would be something we could tweak 24 in the reg guide.
25 So I'm going to just quickly go 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
75 1 the last couple of slides because again all of this is 2 going to get repeated.
3 So the safety classification. This slide 4 just has the definitions which we'll get to as we go 5 through.
6 MEMBER MARCH-LEUBA: I have questions and 7 maybe I need to ask this afternoon. First is 8 language. I don't understand what you say on the 9 first bullet. If I'm reading this correctly the 10 designer selects which SSCs are safety-related or not.
11 And he decides to -- those SSCs that are needed to 12 meet the classification of DBEs must be safety-13 related. Correct?
15 MEMBER MARCH-LEUBA: DBEs. The first 16 couple of sentences, to mitigate the consequences of 17 DBEs --
18 MR. RECKLEY: Within the curve. Yes.
19 MEMBER MARCH-LEUBA: And then it says to 20 mitigate only those DBAs which -- that only rely on 21 SRs. There are other DBAs that can rely on long --
22 I'm talking about language.
23 MR. RECKLEY: Okay. If there's a 24 confusion we can take that as a comment but the intent 25 is that DBAs just much like they do now assume safety-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 related equipment.
2 MEMBER MARCH-LEUBA: All DBAs must have 3 safety. That's not what that sentence says.
4 Now, the most important comment is the 5 second bullet which I think you are trying to address 6 my concern. I detect a circular logic here. Let me 7 give you a simple example.
8 I have a very strong containment and I 9 have an accident that melts the core but nothing comes 10 out of containment. Therefore the frequency 11 consequence is very small, it's way to the left to 12 your line and it's not a safety -- a risk-significant 13 event. Correct?
14 So then I decide that because it's not a 15 risk-significant event I don't need a containment 16 because I don't need to have it safety grade. This is 17 circular logic there.
18 MR. RECKLEY: It's actually what's trying 19 to be addressed here is that if you have something 20 you've placed in the beyond design basis event 21 category as a result of a low frequency and you're 22 relying on a particular barrier to limit the 23 consequence of that event that that is reason to make 24 it safety-related because if you took away that 25 barrier you would move 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
77 1 MEMBER MARCH-LEUBA: It will move up. So 2 what are we gaining --
3 MR. RECKLEY: I might have explained that 4 wrong.
5 MEMBER MARCH-LEUBA: No, no, you did it 6 perfectly. What do you gain by doing it that way?
7 You should have -- make that event a design basis 8 event that you have to analyze and make the SSCs that 9 you rely on safety grade. Where are you baking it, I 10 don't understand.
11 ACTING CHAIRMAN CORRADINI: I think we can 12 come back to this one. Karl can come back to it. You 13 can take it up with Karl.
14 MEMBER MARCH-LEUBA: Okay.
15 MR. RECKLEY: So just the last two. I've 16 addressed this largely, the defense-in-depth 17 assessment and this is going to be talked about by the 18 industry in the context of NEI 18-04. Again stressing 19 that it includes PRA, deterministic assessments, it 20 includes hardware and programmatic controls. And so 21 in our view it's a good tool to apply to a design to 22 make sure that you're addressing the uncertainties and 23 other objectives that we have in this process.
24 Then lastly I did want to touch on that 25 the reg guide that we're preparing is on content 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
78 1 applications. That is the rule that this reg guide is 2 being used for. So we felt it necessary to add a 3 little discussion, more than what's in 18-04 as to how 4 this guidance is used in the development of the scope 5 and level of detail of information that we expect to 6 be in applications.
7 And so I'll get into this this afternoon 8 in a little more detail, but primarily if you look at 9 the fuel, the primary systems and the other primary 10 barriers if you go back to for example that first 11 principle slide what is retaining your radionuclides 12 those kind of barriers would largely need to be 13 described much as they are now because that's where 14 you're going to get how do you get a release. You get 15 a release because you're failing the fuel, you're 16 failing the matrix, you're failing a primary system.
17 So that kind of information would largely 18 be similar.
19 But then as it relates to other systems, 20 ancillary systems we want to focus on what is the role 21 of those systems in supporting again those fundamental 22 safety functions. And from the beginning we know that 23 many of these designs are going to rely less on those 24 ancillary systems, things like ac power and to some 25 degree forced cooling water or other active 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
79 1 And so this process we would hope would 2 build into the beginning that this is how you decide 3 how much information you need to provide on those kind 4 of systems.
5 Likewise whenever you're relying on 6 programmatic controls that needs to be addressed in 7 the application so that you're looking at the same 8 time in concert to hardware and the programmatic 9 controls to provide the needed assurance.
10 I think with that then I'll just set up 11 that the next presentations will be 18-04 that you'll 12 hear about for a couple of hours. And included in 13 that discussion will be some recent example through 14 tabletops that were done with various designs.
15 And then we'll come back, the staff will 16 come back to specifically talk about the draft SECY 17 and the draft reg guide because in the end the ACRS is 18 here to make recommendations or observations on the 19 staff's activities. Those are the two things that we 20 plan to issue and so we will be asking at the December 21 meeting for a letter at least on the SECY paper and at 22 your discretion either on the draft guide or an 23 acknowledgment that you'll get another shot at the 24 guide after we address public comments and the 25 Commission's decisions on the SECY paper.
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 So with that I apologize for being a 2 little late.
3 ACTING CHAIRMAN CORRADINI: No, we're 4 good. Final questions from the members.
5 MEMBER KIRCHNER: Yes. Bill, since you're 6 trying to do this technology-inclusive I would suspect 7 that first movers may not be non-LWRs but LWRs for 8 many of the issues that you're addressing.
9 So in your tabletop exercises have you 10 tried to walk through with an advanced LWR design --
11 I'm not saying NuScale, I'm thinking just an advanced 12 design to just conceptually since as you said you're 13 doing methodology and process. Just to see how it 14 works.
15 MR. RECKLEY: We haven't. If someone were 16 to come forward I guess we could entertain it.
17 The dilemma that you get in and I'll take 18 NuScale as the most recent example. Since they 19 started largely with the existing structure and how 20 they did the design and the arguments it gets a little 21 difficult to then apply this methodology that's 22 intended to be used both during the design process and 23 the license application process.
24 It was done for a large light water 25 reactor but NUREG-1860 includes an appendix where 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
81 1 tried to do this exercise for a large light water 2 reactor and they ran into some of the same problems.
3 And by the way the same problems the staff and 4 industry have faced for the last 30 years on trying to 5 undo a methodology that was based so heavily on that 6 large break LOCA and how it was incorporated both into 7 the design and into the licensing structure.
8 So, the short answer is no, we haven't 9 really entertained it. And the only way we could do 10 it is if somebody came forward and asked for us to do 11 it, a developer. For example, one of the other light 12 water SMR developers.
13 MR. SEGALA: This is John Segala from the 14 staff. But the focus of this, the NEI document and 15 our draft guide is on non-light water reactors. So, 16 when we say technology-inclusive we're referring to 17 the different non-light water reactor designs that are 18 out there versus light water reactors.
19 MR. RECKLEY: That was actually the whole 20 intent of saying technology-inclusive versus the old 21 term of technology neutral.
22 ACTING CHAIRMAN CORRADINI: Walt? Follow-23 up?
24 CHAIRMAN BLEY: This is Dennis. This 25 bothers me a bit. And maybe Karl will talk about 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
82 1 at the next session.
2 I'm not sure what I see that is non-LWR 3 specific about any of this material.
4 MR. RECKLEY: I don't think we're trying 5 to say it could not be used. We're simply saying that 6 the target audience that we're developing this for and 7 the community that's been engaged with us is the non-8 light water community.
9 I don't think we would disagree, and Karl 10 can weigh in later, that these notions would 11 potentially apply to a light water SMR but that's not 12 what we're trying to develop.
13 ACTING CHAIRMAN CORRADINI: Dennis, all 14 right?
15 CHAIRMAN BLEY: Okay.
16 ACTING CHAIRMAN CORRADINI: Okay, why 17 don't we take a break till quarter of.
18 (Whereupon, the above-entitled matter went 19 off the record at 10:28 a.m. and resumed at 10:44 20 a.m.)
21 ACTING CHAIRMAN CORRADINI: Okay, why 22 don't we try to come back together here and start our 23 next session.
24 Which Michael is going to lead us off.
25 Mr. Meier --
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 MR. AFZALI: Actually I'll start us off.
2 ACTING CHAIRMAN CORRADINI: Oh, I'm sorry, 3 Amir. I apologize. I was looking over there by the 4 computer. Go ahead.
5 MR. AFZALI: Good morning. It's a 6 pleasure to be here again. We have based on our last 7 conversation, June conversation, Dr. Bley asked us to 8 come back and have a detailed conversation about the 9 proposal we are making. And we have put a great team 10 together to come and answer your questions.
11 We look forward to your insightful 12 comments. We thought it would be appropriate for our 13 utility representative to say a few words before 14 starting the conversation.
15 To that end I've asked Dr. Meier, a 16 regulatory affairs VP and Mr. Steve Nesbitt, I'm going 17 to read his title, director of nuclear policy and 18 support, to say a few words. So, Dr. Meier.
19 MR. MEIER: Good morning and thank you all 20 for the opportunity to appear before the ACRS Future 21 Plant Designs Subcommittee.
22 Southern Company has 46,000 megawatts of 23 generated capacity and provides clean, safe, reliable 24 and affordable energy to its -- throughout our service 25 territory.
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 What's important to note, our CEO Tom 2 Fanning announced to our generation fleet that we have 3 a goal to be low to no carbon by the year 2050. And 4 he has set some goals in between there.
5 In order to do this we're going to have to 6 focus on technologies that will allow us to reduce 7 these carbon emissions. With nuclear energy, and we 8 have talked about this a lot in the company, is going 9 to play a major role in that.
10 Regulatory modernization, however, is 11 going to be necessary for us to remove any of these 12 unnecessary challenges and reduce inefficiencies in 13 order to make this happen.
14 NEI 18-04 proposals provide a robust 15 systematic and a flexible foundation for modernizing 16 the regulatory requirements for these advanced light 17 water reactors.
18 Given all the variety we have on these 19 non-light water reactor designs being developed by the 20 advanced reactor community it's imperative that we 21 have a good foundation as well as a follow-on 22 regulations made available to the developer community.
23 We are encouraged and we are excited by 24 the cooperation between the NRC, DOE and the industry 25 to take concrete steps toward developing 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
85 1 foundational framework and we look forward to the ACRS 2 suggestions to make the products even better as well 3 as expediting endorsement by the NRC.
4 Finally, I would like to thank the NRC 5 staff, DOE management, our developers and the industry 6 partners for diligently and effectively getting us to 7 where we are today.
8 Again, I want to thank you for your time.
9 MR. NESBIT: Good morning and thanks for 10 the opportunity to appear before the ACRS Future Plant 11 Designs Subcommittee.
12 So why are we here. At the risk of 13 repeating the obvious the current nuclear power 14 reactor regulatory framework dating from the nineteen 15 seventies and even before has proven to be effective 16 although not always efficient in providing adequate 17 protection to public health and safety.
18 This project is about leveraging 19 knowledge, experience and technological advances over 20 the past 50 years to put in place a methodology that 21 will work in the 21st century when applied to the 22 range of innovative and diverse reactor designs many 23 of which bear little resemblance to the light water 24 reactors we've become so adept at operating today.
25 Duke Energy, the nation's second largest 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 nuclear power plant operator, supports the licensing 2 modernization project. The 2017 Duke Energy climate 3 report to shareholders outlines a scenario in which 4 our company would achieve a 72 percent reduction in 5 CO2 emissions by the year 2050 compared to 2010 6 levels.
7 In addition to phasing out coal-fired 8 electricity generation this scenario envisions 9 preserving generation from all 11 currently operating 10 reactors, increasing energy efficiency, expanding 11 renewable generation, expanding energy storage and 12 deploying innovative technologies we refer to as zero 13 emitting load following resources, or ZELFRs.
14 A ZELFR has essentially no carbon 15 emissions, can generate power continuously and can 16 adjust its output to match load.
17 To meet customer needs in this scenario 18 Duke Energy analyses indicate 13 percent of our year 19 2050 generation will need to come from these ZELFR 20 technologies that may not exist today.
21 Nuclear power generation has been a great 22 asset for Duke Energy and its customers in North 23 Carolina and South Carolina and we believe advanced 24 reactors are good candidate ZELFR technologies.
25 There are of course challenges to 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
87 1 deployment of advanced nuclear generation. One of 2 those challenges is the need for a modern, flexible, 3 adaptable regulatory framework.
4 For innovative and diverse nuclear power 5 reactor designs we must have a methodology that 6 continues to provide adequate protection of public 7 health and safety and works in a timely and 8 predictable manner.
9 NEI 18-04 is a key foundation for that 10 regulatory framework about which you have heard 11 already today and you're going to hear more.
12 I've been encouraged by the progress made 13 to date on this endeavor and in particular on the 14 constructive engagement I've seen among industry, 15 national laboratories, the Nuclear Regulatory 16 Commission staff and other stakeholders.
17 And our team looks forward to receiving 18 your observations and insights.
19 MR. AFZALI: Mike, did you want to add 20 anything? Okay. So you heard why we are here. We 21 are excited to demonstrate to you the how part.
22 And we have a team of three who sit at the 23 table and a team of contributors sitting in the 24 audience to answer any detailed question you may have.
25 With that said we're going to leave 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
88 1 bring the real team over to the table.
2 ACTING CHAIRMAN CORRADINI: The technical 3 team versus the leadership team.
4 MR. TSCHLITZ: Good morning. My name is 5 Mike Tschlitz. I'm the senior director of new plants, 6 SMRs and advanced reactors at NEI. Thank you for the 7 opportunity to come before the ACRS and give this 8 presentation.
9 So, one of the objectives of my 10 presentation here today albeit very short is to 11 discuss the importance of this initiative and NEI 18-12 04 and to the overall vision for where the industry 13 needs to head.
14 To paint that picture I'll point to the 15 paper that's on the slide. It's entitled Ensuring the 16 Future of U.S. Nuclear Energy: Creating a Streamlined 17 and Predictable Licensing Pathway to Deployment. It 18 was issued January 23rd this year and cosigned out by 19 NIA, NEI and NIC. Sent to Chairman Svinicki.
20 And it laid out the near term regulatory 21 reforms that the industry saw as being necessary for 22 licensing advanced reactors.
23 And we'll go through all of these but the 24 second bullet there talks about aligning the 25 regulatory framework for advanced reactors with 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
89 1 inherent advanced safety and that's what in part we're 2 trying to accomplish through NEI 18-04.
3 In this paper we also provided a vision 4 for the future with a modernized NRC licensing process 5 where the reviews of advanced reactors become more 6 efficient and timely while continuing to protect 7 public health and safety.
8 The methodology in NEI 18-04 will play a 9 large role in enabling a technology-inclusive risk-10 informed and performance-based approach, a more 11 safety-focused and predictable regulatory review 12 process and ultimately the licensing and deployment of 13 innovative and safe nuclear technologies.
14 MEMBER REMPE: Mike, I had a couple of 15 questions about this slide.
16 First of all, this comment about the trend 17 of increasing costs. And I looked at slide 28. And 18 although it's good for exciting some folks on the Hill 19 I'd suggest that maybe it's incomplete.
20 For example, I believe the APR 1400 if you 21 had that cost might show a difference in trend. And 22 in fact, in general when you already have an operating 23 plant like the system 80 as well as the APR 1400 I 24 think the staff has done things more efficiently.
25 It's sometimes maybe design incompleteness that 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
90 1 leading to increased costs.
2 MR. TSCHLITZ: Sure. The information that 3 we're using was based upon information that was 4 reported to Congress in 2015 over the last 20 years 5 for the reviews and it showed a four time increase in 6 the cost of reviews.
7 That being said the staff deserves some 8 credit. I mean, the NuScale review and the APR 1400 9 reviews are proceeding on schedule. The APR 1400 10 review as you know is basically an uprate of an 11 existing design so it's kind of in a different 12 category. So to say that's a completely new and 13 different design that you can compare apples to apples 14 for the cost would be a challenge.
15 But for NuScale, if you look at the cost 16 of the design reviews for NuScale they're approaching 17 and predicted to be about the same as ESBWR for a 18 design with about one-third of the safety systems.
19 So you're wondering -- and a much lower 20 overall risk profile. So, I think what we're finding 21 is the staff is becoming more timely in its reviews as 22 evidenced by APR 1400 and NuScale review, but the 23 efficiency associated with that we're not seeing. So 24 that's the basis.
25 MEMBER REMPE: Everyone could improve, 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 I'll agree with you, but I just was thinking that that 2 chart is a little incomplete.
3 MR. TSCHLITZ: So if you go to the next 4 slide.
5 MEMBER REMPE: Actually, I have another 6 question too.
7 MR. TSCHLITZ: Okay.
8 MEMBER REMPE: This last bullet, providing 9 additional flexibility for changes during 10 construction. And I'm thinking about what happened 11 with another certified design where some issues were 12 identified and they had to change during construction.
13 And it's expensive to change a certified design under 14 Part 52. What are you thinking about doing here?
15 MR. TSCHLITZ: So if I can go to the next 16 slide I'll refer to a paper here. So the paper in the 17 lower right-hand corner of this slide, Assessment of 18 Licensing Impacts -- I can't even read it myself.
19 MEMBER REMPE: On Construction.
20 MR. TSCHLITZ: On Construction. So it's 21 a paper that we recently issued that looks at the 22 experience with -- it started with the Vogtle and the 23 Summer plants but ended up just looking at the Vogtle 24 3 and 4 constructions about all of the license 25 amendments that had to be issued during construction.
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 And we did a review along with Southern.
2 Southern was very instrumental in us being able to 3 develop the data to support the conclusions in this 4 report.
5 We found that a lot of the changes in the 6 licensing had no safety impact. And they were 7 basically causing additional costs because of the 8 staff that's necessary to be maintained basically 9 around the clock so you don't impact construction when 10 you find an issue that requires some type of 11 disposition that may require an amendment.
12 So the ongoing carrying costs and then the 13 cost of writing amendments and having the NRC review 14 them is not justified from a safety perspective. So 15 there's these additional carrying costs for having the 16 ability to make changes to the licensing basis on an 17 ongoing basis throughout the construction period where 18 the vast majority of the changes had no real 19 connection to safety.
20 And so I would point out that that's a 21 report that you can read and see all the details. We 22 provide some specific examples in there. We look at 23 tier 2 star information. We look at the level of 24 detail that's provided for some of the civil 25 structural part of the licensing basis.
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 We're basically suggesting that there be 2 a reconciliation process during construction that 3 allows a period of time where the construction 4 continue in non-conformance with the licensing basis 5 and allow a period of time for some development and 6 submittal and the NRC review of a change while 7 construction continues.
8 That goes at this. That was not going to 9 be the subject of this talk today.
10 MEMBER REMPE: Sure, I just was curious so 11 thank you. I'll look at the paper.
12 MR. TSCHLITZ: Okay. So on this slide as 13 I noted in the January 23 paper we set priorities for 14 what needs to get done in the near term.
15 The four documents shown on this slide 16 were written over the past nine months and provide 17 recommendations for making regulatory reviews more 18 safety focused and efficient, providing guidance for 19 developing a regulatory engagement plan that supports 20 staged licensing, proposing a process for providing 21 additional flexibility during construction under Part 22 52, and the topic we're here to discuss today, NEI 18-23 04 which provides a technology-inclusive risk-informed 24 performance-based guidance for identifying licensing 25 basis events, SSCs, and determining the adequacy 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
94 1 defense-in-depth.
2 MEMBER KIRCHNER: Can I ask you to go 3 backwards?
4 MR. TSCHLITZ: Certainly.
5 MEMBER KIRCHNER: Since you highlighted it 6 in yellow aligning the regulatory framework for 7 advanced reactors with their inherent enhanced safety.
8 I think I know where you're going with that but it 9 would seem to me that the regulator requires the 10 applicant to demonstrate the inherent enhanced safety.
11 That's not a given going in even though on paper many 12 of the designs look promising.
13 I'm just quibbling with your choice of 14 words. If I were in your shoes I'd want to expedite 15 my way through the safety review with a focus on 16 safety and what's important to safety and risk.
17 This sounds like retooling the regulatory 18 environment because we think these reactors have 19 enhanced safety features yet to be demonstrated. It 20 seems to me that's not your real objective. Your 21 objective is to demonstrate that these reactors are 22 indeed lower risk, they have more margin and therefore 23 -- I'm just struggling with the words there because on 24 paper it's incumbent on the applicant to make that 25 demonstration that they really do provide an enhanced 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 level of safety.
2 MR. TSCHLITZ: So, I agree with what you 3 say. I think what we are trying to communicate there 4 is -- I'll give you two examples.
5 Consequence-based emergency planning. If 6 you make changes to the regulation that allow based 7 upon the consequences associated with events to set 8 the EPZ as appropriate that's aligning the regulatory 9 framework with enhanced safety.
10 Consequence-based security measures.
11 Aligning the security at the site with its enhanced 12 security features. For advanced reactors that's 13 changing the framework.
14 So those are the type of changes I think 15 we were after. And this NEI 18-04 also fits in that 16 category whereas you're looking at a different 17 approach to determining licensing basis events that 18 basically will focus on the most important parts or 19 aspects of the design.
20 MEMBER KIRCHNER: Let me repeat my 21 question to Bill from the last session. Is your 22 document going to be amenable to an LWR based 23 technology?
24 MR. FLEMING: Well, we never intended this 25 to apply to an existing light water reactor. If an 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 advanced non-light water reactor came forward with 2 safety characteristics that were essentially the same 3 as a light water reactor using -- relying on an 4 inventory of coolant, metallic fuel, reactor vessel 5 and so forth and a leak tight containment the process 6 should accommodate such a design approach.
7 We didn't intend it to exclude any 8 technology but we didn't intend it to be applied to 9 light water reactors.
10 MEMBER KIRCHNER: Again, at risk of 11 repeating myself the first movers may likely be LWR 12 designs that will challenge some of the existing 13 policies. And we have such an application for an 14 early site permit before us to look at doing more of 15 a risk-based and performance-based approach to the 16 emergency planning zone as an example. Thank you.
17 MR. TSCHLITZ: So Jason, if you can go to 18 my banner slide. So this slide reflects NEI's near 19 term activities which have been focused on the topics 20 on the four banners shown on this slide.
21 And the risk-informed performance-based 22 technology-inclusive approach of NEI 18-04 has 23 impacted the areas that I've highlighted in red 24 circles that don't really show up that well on the 25 slide here but I'll talk briefly about each one 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
97 1 those.
2 In the area of safety-focused reviews 3 experience over the last two decades with the DC, COL, 4 and ESP applications indicates that the NRC staff has 5 to a large extent remained deterministic in its 6 licensing reviews even though regulation and guidance 7 allow the NRC staff flexibility to adjust its review 8 on the basis of safety significance.
9 Costs of ongoing NRC reviews remain high 10 leading to the conclusion that the advantages of safer 11 designs appear to be of little benefit when trying to 12 reduce regulatory review costs. Future NRC reviews 13 should better utilize risk information in combination 14 with the principles of defense-in-depth and 15 maintenance of safety margins.
16 In the area of risk-informing advanced 17 reactor licensing basis, information included in the 18 licensing basis that doesn't have a connection to the 19 safety basis in the NRC's determination of adequate 20 protection imposes a burden on applicants who have to 21 invest resources to develop the information and pay 22 for the NRC to review unnecessary content.
23 In addition, there are ongoing costs 24 associated with maintaining and evaluating changes to 25 this information over the life of 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
98 1 Content that is not needed to demonstrate 2 compliance with regulations and/or lacks a nexus to 3 subsequent NRC oversight poses a regulatory burden 4 with no benefit to safety. Inclusion of this 5 information during initial certification or licensing 6 is not necessary. These practices increase licensing 7 review costs without a corresponding increase in 8 safety.
9 NEI 18-04 provides the starting point for 10 adjusting the content of applications and the focus of 11 NRC's review based upon safety and risk significance.
12 Reversing the trend. In this area data 13 submitted to Congress in 2015 shows the costs of NRC 14 reviews have increased substantially over time.
15 As I mentioned briefly the NuScale example 16 demonstrates that the projected licensing fees of 17 advanced reactor designs are similar to other large 18 light water reactors.
19 These design certification review costs 20 have been normalized to 2017 dollars and have 21 increased by a factor of approximately four over the 22 last 20 years.
23 This shows that the advantages of safer 24 designs have not resulted in reduction of regulatory 25 review costs.
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 CHAIRMAN BLEY: Mike, this is Dennis Bley.
2 Can you tell me anything about the success or failure 3 of applicants who have challenged the staff that 4 things they're looking at are not important to safety?
5 MR. TSCHLITZ: That's a good question.
6 I'm probably not in the best position to answer that, 7 but I can offer one example.
8 For the NuScale review, the chapter 9 9 auxiliary systems which have no impact on safety or 10 mitigating beyond design basis events. I guess it was 11 earlier on in the review and I'm sure this information 12 has changed as the review has continued and shifted on 13 to chapter 15.
14 But at one point in time 30 percent of the 15 staff's RAIs were focused on chapter 9 issues.
16 Chapter 9 as I said has no real nexus to safety.
17 So I think the vision would be in the 18 future for those types of systems that don't have a 19 direct connection to the safety case there would be a 20 high-level description without a lot of detail in the 21 application. And that should be sufficient for the 22 staff's understanding of the design.
23 So at this point --
24 CHAIRMAN BLEY: That doesn't really get at 25 what I was trying to ask. You showed increases 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
100 1 costs over quite a few years and during that time have 2 the applicants tried to challenge NRC in these areas?
3 Or do they just kind of go along with it?
4 MR. TSCHLITZ: Well, I don't know if I can 5 offer a really good answer to that question because it 6 involves a lot of different applicants over a long 7 period of time.
8 I can say in general that there is a 9 reluctance to challenge the NRC in some of these areas 10 during the course of a review.
11 CHAIRMAN BLEY: I'm not sure that won't 12 continue even with this new framework so something to 13 think about.
14 MR. TSCHLITZ: I think the framework helps 15 focus the discussion. So if you can show things more 16 in black and white as you can on the frequency 17 consequence curves as you'll see when you look at the 18 results of the tabletops it helps focus the discussion 19 I think on the issues.
20 So at this point in my presentation I'm 21 going to make some introductions and invite some 22 people who come to the meeting to support us to come 23 to the mike and introduce themselves and explain their 24 connection to the project.
25 So the first person is Jim Kinsey from 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 Idaho National Lab.
2 MR. KINSEY: Good morning. I just wanted 3 to make a couple of remarks related to -- I know the 4 NGNP project came up earlier in the day.
5 Back during that discussion we developed 6 a process based on inputs from the three modular HTGR 7 developers in the U.S. and also partnered with Entergy 8 at the time to get some insights from an owner-9 operator organization.
10 And the risk-informed performance-based 11 approach that we presented to this subcommittee back 12 at that time was intended to work toward our marching 13 orders of moving gas reactors forward, but it was 14 always envisioned that it could be a technology-15 inclusive process.
16 So our involvement with LMP has been to 17 bring some of that history to bear, provide insights 18 from those previous reviews and as you'll see as we go 19 through the day the current team is much larger, 20 includes NEI, includes tabletops and evaluations from 21 other technology types and includes other owner-22 operators. So it's provided some, as Bill mentioned 23 earlier, some tweaks and refinements to that original 24 process but it's still largely based on the foundation 25 from that previous review. So we appreciate 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
102 1 insights today.
2 MEMBER REMPE: Jim, did the NGNP -- I 3 can't remember. Did it have this integrated decision 4 panel as part of that process?
5 MR. KINSEY: I don't know that it had that 6 discussion in detail. I think that concept was out 7 there, but I think one of the refinements that was the 8 more significant one in the LMP approach is further 9 defining the details of the defense-in-depth strategy 10 and how you go about actually implementing and 11 managing it. That's probably one of the more 12 significant additions that we'll talk about. Other 13 questions?
14 MR. TSCHLITZ: Thanks, Jim. The next 15 person is Ed Wallace, consultant to Southern Company.
16 MR. WALLACE: Good morning. My name is Ed 17 Wallace. I've been involved with advanced reactors 18 since 2001 through the PBMR NGNP technology neutral 19 framework and NuScale activities. And a member of the 20 ANS Standards Board focused on risk-informed 21 performance-based practices within the standards 22 community.
23 Part of my purpose with the consultation 24 to Southern is to bring that experience to bear in the 25 evolution of this process which stems back 35 years or 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 more to the MHTGR days.
2 My role has been to focus on the 3 discussions aspect of it because of the comments that 4 have already been made and the need to provide a 5 practical way to perform that assessment in a 6 consistent manner and consistent with the risk-7 informed performance-based information that's derived 8 in the other activities that you're hearing about.
9 If you have any questions today I'll be 10 glad to answer them. Thank you.
11 MR. TSCHLITZ: Thanks, Ed. The next 12 person is on the phone line. Brandon Waites of 13 Southern Company. And he's going to be speaking on 14 behalf of X-Energy.
15 MR. WAITES: Yes, this is Brandon Waites.
16 I'll just take a quick pulse to make sure everyone can 17 hear me.
18 MR. TSCHLITZ: Brandon, just give us one 19 moment to turn off area mikes so we don't get 20 feedback. Thank you. Brandon, please go ahead.
21 MR. WAITES: Okay, thank you. I really 22 appreciate the opportunity to speak today. My name's 23 Brandon Waites. I'm new projects manager at Southern 24 Company and I wanted to speak just real quickly on 25 some activities we had regarding the LMP 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
104 1 year.
2 Earlier this year the LMP team completed 3 the first demonstration of the LMP process using a 4 real world example with the X-Energy high temperature 5 gas cooled reactor design.
6 And for this I'd like to take a quick 7 minute to mention that the LMP team is grateful to X-8 Energy for their support and allowance and significant 9 support of this demonstration.
10 Just to get quickly to the outcome of the 11 demonstration we concluded in a report that is 12 publicly available that the demonstration was 13 successful and produced several actionable insights 14 both in the area of -- for the LMP process itself and 15 also insights into the X-Energy high temperature gas 16 cooled reactor design.
17 ACTING CHAIRMAN CORRADINI: Brandon, can 18 you give us a reference so the staff can get us a copy 19 of that report? I'd be interested in seeing that.
20 MR. REDD: Michael, this is Jason Redd 21 from Southern. We'll get you that reference 22 momentarily. We've got it available. We'll provide 23 it to a member of the ACRS staff before we leave 24 today.
25 ACTING CHAIRMAN CORRADINI: Thank 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
105 1 Thank you very much.
2 MR. TSCHLITZ: Thanks, Brandon. The next 3 person is Gary Miller from GE-Hitachi.
4 MR. MILLER: Good morning. I'm Gary 5 Miller, manager of PRA at GE-Hitachi. We're 6 responsible for all PRA aspects including design and 7 licensing.
8 We used our PRA of the PRISM sodium fast 9 reactor as a basis for supporting two of the LMP white 10 papers on PRA and LBE selection and also we used it to 11 demonstrate the methodology that we're going to talk 12 about today. I'll be happy to answer any questions 13 you might have.
14 MR. TSCHLITZ: Thanks, Gary. The next 15 person is Steve Krahn from Vanderbilt University.
16 MR. KRAHN: Good morning. I'm Steve 17 Krahn. I head up the nuclear environmental research 18 group at Vanderbilt University where we do risk and 19 hazard assessment on advanced nuclear technology.
20 Specific to the subject of today's meeting we have 21 been involved for the last four and a half years doing 22 hazard and risk assessment of molten salt reactors and 23 two of the outcomes of that research are part of the 24 package that will be briefed this afternoon.
25 MR. TSCHLITZ: Thanks, Steve. The next 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 person is Dave Grabaskas from Argonne National Lab.
2 MR. GRABASKAS: I'm Dave Grabaskas. I'm 3 a principal risk analyst at Argonne National Lab. I'm 4 also the vice chair of the ASME ANS non-light water 5 reactor PRA standard.
6 I was also the Argonne lead for the 7 collaboration with GE to update the PRISM SFR PRA. In 8 advance of issues we see with advanced reactor 9 licensing our research has focused on passive system 10 reliability, mechanistic source term and developing 11 component reliability databases for advanced reactors.
12 And particularly applying them to the NEI 13 framework but also its predecessors too with the NGNP 14 and NUREG-1862. So I'd be happy to answer any 15 questions you have in those areas.
16 MR. TSCHLITZ: Thanks, Dave. And the last 17 person is Jim August from Southern Nuclear.
18 MR. AUGUST: Good morning. My name is Jim 19 August. I'm with Southern Nuclear at Vogtle. I'm 20 very excited to be here.
21 The reason I'm here is in my first post 22 Navy commercial job I started off as a reliability 23 engineer at Fort St. Vrain in 1981 and worked at Fort 24 St. Vrain through about 10 years of operations and did 25 a lot of work trying to resolve technical issues 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
107 1 well as licensing issues that surrounded that high 2 temperature gas reactor prototype commercial plant.
3 As a result of those experiences when the 4 ANS decided to reconstitute and redevelop their 5 standard for safety design of high temperature gas 6 reactors which are now termed modular helium cooled 7 reactors I volunteered to join that committee.
8 From 2004-08 I was a member, 2008 I became 9 the chair and we completed the standard ANS 53.1 which 10 led to a lot of the work we're discussing here which 11 was the safety design standard for modular helium 12 cooled reactors.
13 My motivation for doing that work was 14 largely the experience I gained at Fort St. Vrain 15 which included a significant amount of frustration 16 that related to us continually being judged in what I 17 will call a light water reactor environment. I'm here 18 to answer any questions that you might have.
19 MR. TSCHLITZ: So at this point I'll turn 20 it over to Jason Redd.
21 MR. REDD: Thank you, Mike. Good morning.
22 My name is Jason Redd from Southern Nuclear Operating 23 Company. I'm pleased to be here with you all today, 24 members of the committee.
25 The LMP methodology is ultimately focused 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 on establishing a systematic, coherent framework for 2 establishing a technology-inclusive risk-informed 3 performance-based aspects of the licensing basis.
4 Given the wide variety of non-light water 5 technologies that are proposed on the relatively near 6 horizon a top down path of establishing technology-7 inclusive methods to establish compliance requirements 8 such as the NEI 18-04 document, the advanced reactor 9 design criteria which were released last year after 10 collaboration between NRC staff and the Department of 11 Energy leading to methods for establishing technology 12 specific requirements such as the high temperature gas 13 reactor and sodium fast reactor design criteria 14 contained within the advanced reactor design criteria 15 both leading to reactor design specific design and 16 compliance basis, for example, the principle design 17 criteria is an appropriate and effective pathway.
18 NEI 18-04 guides prospective applicants in 19 answering the following questions. And we're going to 20 come back to these questions again at the end of the 21 presentation so certainly stay tuned through Karl.
22 What are the plan initiating events, the 23 event sequences and accidents that are associated with 24 that particular reactor design, how does the proposed 25 design and its structured systems and components 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 respond to initiating events and event sequences, what 2 are the margins provided by the facility's response.
3 Again we've heard in the Commission's 4 policy statements the margins are of significant 5 interest both to the Commission and to the staff and 6 to the designer and operator community as those 7 margins relate to the prevention and mitigation of 8 radiological releases within prescribed limits for the 9 protection of the public health and safety.
10 And is the philosophy of defense-in-depth 11 adequately reflected in the design and operation of 12 this facility.
13 With these opening remarks I'll now turn 14 it over to our technical lead --
15 MEMBER REMPE: Just a second, I have a 16 question.
17 MR. REDD: Yes.
18 MEMBER REMPE: To make sure I understand 19 because I did find when I looked through the document 20 you're considering low power and shutdown events, 21 you're considering external events. Hazards 22 associated with the spent fuel pool. That should also 23 be considered.
24 And then later on when you get to the 25 tabletop discussions, did they have PRAs 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
110 1 considered all those types of phenomena?
2 MR. REDD: Let me answer the first part.
3 Yes. The LMP process in NEI 18-04 is designed to 4 address all of the radiological sources within the 5 plant whether that's the reactor vessel -- the primary 6 coolant system, spent fuel pool. Obviously some 7 advanced reactor designs also have radionuclide 8 inventory such as off gas holdup vessels and similar 9 or storage tanks.
10 All of those sources of radionuclides that 11 could pose a hazard to the public are included within 12 the LMP process.
13 I'd like to invite Karl to answer the 14 question about whether the demonstrations we've done 15 so far have included those aspects.
16 MR. FLEMING: I'm not sure if we'll get to 17 it this morning but certainly in the early afternoon 18 I'm going to give you a breakdown of all the steps of 19 our process and what was exercised and not exercised 20 in each of the tabletops so far.
21 So in general most of the experiences 22 focused on full power operation so the experience base 23 is limited on some of these other sources. But I'll 24 give you more details on that later.
25 MEMBER REMPE: Thank 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
111 1 CHAIRMAN BLEY: Karl, this is Dennis.
2 Before you get started two things.
3 We did invite you guys back to hear more 4 and more broadly and new material. So as you go 5 forward if you can emphasize the new material and de-6 emphasize the repetitive stuff that would be great.
7 And number two, back in June we had draft 8 Mary, M. Now we have draft November. Has there been 9 any substantive changes that you can tell us about in 10 the guidance since the last time we talked with you?
11 And I'll go offline.
12 MR. FLEMING: In respond to your first 13 question I'll do my best not to repeat things that 14 you've seen before and try to emphasize the new 15 material.
16 I'll invite Jason to comment on revision 17 N versus M.
18 MR. REDD: Good morning. The changes from 19 draft Mike to draft November were primarily the 20 incorporation of comments from this committee in the 21 June time frame.
22 The major changes have been an expansion 23 of the discussion of certain aspects. There was 24 increased discussion especially of how defense-in-25 depth is applied.
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 A lot of clarifications here and there in 2 response to both staff feedback both industry feedback 3 and the committee's June comments.
4 There was no change whatsoever in the 5 underlying philosophy or the methodology. I would 6 characterize these changes as editorial and 7 explanatory.
8 MR. FLEMING: I'd just add one point to 9 what Jason mentioned and that is that each of the 10 revisions has reflected our evolution in being more 11 precise about our terminology.
12 So the use of our terminology in avoidance 13 of synonyms for key terms and cleanup of our glossary 14 has continually been improving along the way.
15 Thank you very much, Jason. If I can 16 start my talk here. The technical presentation that 17 we have outlined has two parts to it. One of them, 18 the first part is to just amplify on some methodology 19 refinements that we made since the NGNP days and to 20 point out some technical items that fill in some of 21 the gaps from Bill Reckley's presentation.
22 And then the second half of our 23 presentation is geared towards the lessons learned 24 from our tabletop pilot applications.
25 On this first slide which outlines 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
113 1 principal focus of this methodology the things that --
2 I just wanted to amplify on some of the things that 3 Bill Reckley has already mentioned.
4 This is an integrated process for license 5 event selection safety classification and defense-in-6 depth. And they're really interrelated in terms of 7 the safety classification refers to functions that are 8 performed by the SSCs on the licensing basis events.
9 The defense-in-depth refers back to both 10 the SSC functions and the LBEs that are participating 11 in preventing and mitigating accidents. And the 12 defense-in-depth aspects have a lot to do with setting 13 the performance requirements for our system structures 14 and components that come out of safety classification.
15 The process leads to a systematic 16 identification of the design basis accidents that will 17 go in chapter 15 using a process that we believe is 18 repeatable, reproducible and so far has produced 19 nothing but sensible and consistent results.
20 Uncertainty is a very major focus of this 21 activity. It's addressed within the state of the art 22 of PRA in terms of estimating frequencies and doses 23 with their associated uncertainties but it identifies 24 sources of uncertainties that are captured and 25 evaluated very carefully in the integrated decision 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 processes associated with establishing defense-in-2 depth adequacy.
3 The evaluation of plant capabilities and 4 programs for defense-in-depth is one of the areas 5 that's extended beyond what was done in the NGNP 6 process.
7 We think this process is risk-informed 8 using Chairman Jackson's original idea behind that 9 term in that it involves a balance of probabilistic 10 and deterministic inputs. It's not risk-based by any 11 shape of the imagination but our rationale for 12 starting with a design-specific PRA that's integrated 13 into the design process is it's a way to enumerate a 14 systematic and exhaustive set of scenarios that we can 15 draw from to build the license application.
16 One area that we have enhanced from the 17 NGNP days, we've tried to emphasize more of the 18 performance-based aspects of the approach.
19 Performance-based includes using plant level metrics 20 for measuring the risk significance of licensing basis 21 events, but also in setting performance requirements 22 for SSCs that are phrased in such a way that can be 23 tracked and monitored throughout the plant operation 24 and lifetime to get adequate assurance that a safety 25 case is being upheld.
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 And the other aspect of this approach is 2 that unlike the light water reactor model for 3 prevention and mitigation which has been largely 4 focused on preventing core damage and mitigating the 5 consequences of core damage, these reactor designs 6 that we're dealing with have many different end 7 states, many different event sequences, different uses 8 of barriers and layers of defense.
9 So finding a general way to talk about 10 prevention and mitigation linked to balancing, 11 preventing and mitigating the releases from 12 radioactive material from the plant.
13 If we go on to the next slide. To clear 14 up some of the discussion earlier on how we come up 15 with our design basis accidents, we start with 16 defining accident families in which we group event 17 sequences according to the similarity of plant 18 challenge initiating event, plant response and if 19 there is a release mechanistic source term.
20 We group them and classify them by 21 frequency into three regions. And from that we 22 evaluate the -- we start with the design basis events 23 and the design basis events region and we look at the 24 design basis event as candidates for 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
116 1 The idea is that we want to have a 2 reasonably complete enumeration of design basis events 3 that challenge the safety case.
4 When we get into this part of the analysis 5 the LBEs that have no consequences are equally 6 important if not more important than the ones that 7 might have a risk significance. The risk significance 8 is part of this but what we want to mine out of this 9 is what are the features in the plant that are 10 responsible for preventing releases from these 11 accident sequences and then what do I have to preserve 12 in my design basis to enforce that result.
13 ACTING CHAIRMAN CORRADINI: So can I -- if 14 this is the wrong time to ask a question you can just 15 hold me off.
16 So is there a standard process in risk 17 assessment that one understands how to bundle these 18 things? Because I know you've said it a number of 19 times and we said it in June, but I don't -- I'm still 20 trying to get a handle on a guidance here that it's 21 perfectly clear what's a good judgment and what's an 22 inappropriate judgment on the bundling. Whether it be 23 based on source term or based on frequency. Or type 24 of initiator. And I can't tell yet.
25 MR. FLEMING: Okay. We handle 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
117 1 through the ANS ASME light water reactor, non-light 2 water reactor PRA standard. There are technical 3 requirements in that standard for defining event 4 sequence families and this is fundamental to analyzing 5 the contributors to risk in that framework.
6 So we actually have -- it's in that 7 standard that was issued for trial use in 2013. David 8 Grabaskas was alluding to one of the pilot studies 9 done to exercise that.
10 ACTING CHAIRMAN CORRADINI: So if I have 11 -- pardon if this is too simple, but if I have a 12 station blackout event as we might have in a light 13 water reactor but with a range of source terms that 14 would all be bundled various station blackout events 15 with various initiators, or would it be more akin to 16 bundling them based on source term?
17 I'm trying to think in my mind that I've 18 got an x-y plot where y is the frequency and x is the 19 source term essentially for all intents and purposes.
20 And I'm trying to understand how you bundle these 21 things if I get a disagreement about how I bundle them 22 based on initiator or source term.
23 MR. FLEMING: Well, first of all we want 24 to bundle them based on source term. If there is a 25 source term we don't want to have dissimilar source 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 terms in the same event sequence family.
2 ACTING CHAIRMAN CORRADINI: That's the 3 first principle.
4 MR. FLEMING: That's one. Then beyond 5 that of all those that have the same mechanistic 6 source term among those we want to identify those that 7 have the same challenge to my safety functions. So 8 what systems were working, what systems weren't 9 working, what functions were fulfilled.
10 So we want to preserve the character of 11 how the safety case was challenged by the event sq.
12 ACTING CHAIRMAN CORRADINI: So did I miss 13 that, or is that written somewhere in 18-04?
14 MR. FLEMING: No, it's not written in 18-15 04. It's referred to in the PRA standard, the non-16 light water reactor PRA standard.
17 ACTING CHAIRMAN CORRADINI: Okay. And 18 that's referred to in 18-04.
19 MR. FLEMING: Yes.
20 ACTING CHAIRMAN CORRADINI: Okay.
21 MEMBER KIRCHNER: Karl, can I interrupt 22 and ask a question? So you're in the early stage of 23 an advanced design. You can probably bound the source 24 term obviously, whatever the core design is. But 25 there are design characteristics, I guess I'm asking 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 an uncertainty question in an indirect way. Without 2 getting into specific designs. Let me see if I can 3 phrase this generically. Technology-inclusive.
4 There are things like reactivity insertion 5 accidents, or there are fuel failure modes that early 6 on can have a large uncertainty associated with them 7 until you've done the actual detailed design or you've 8 done a fuel qualification program or et cetera. So 9 how do you best include uncertainty early on so that 10 you don't get down the road and find that systems that 11 you thought weren't risk significant or weren't 12 safety-related then you get into a backfit situation 13 of revising your design well down the road which 14 obviously would be a nightmare for any advanced 15 concept trying to expedite its way through the system.
16 So how do you deal with that uncertainty 17 early on when you're going through establishing your 18 design basis and other events and then you're going 19 through it, and then you're selecting your safety-20 related systems and such. Then do the DBA analysis.
21 But put the DBA analysis aside.
22 I'm just curious how best in this process 23 you avoid a major redesign, or a major backfit, or a 24 major change in the quality level of systems and 25 components as the design matures.
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 MR. FLEMING: Well, let's see. If you 2 break the LMP process down into its full level of 3 detail it's like an 18-step process. And many of 4 those steps involve evaluations of what you have so 5 far with feedback loops to go back to the beginning 6 when you have to choose the design in order to get a 7 satisfactory result.
8 In the PRA part of this process based on 9 where you are in the design when you apply the PRA 10 standard roughly half the requirements in the PRA 11 standard have to do with uncertainties. Have to do 12 with identifying sources of uncertainty, trying to 13 account for them to the best you can and your 14 estimates of the source term and the frequencies of 15 occurrence.
16 The ones that you cannot handle that way 17 beyond the state of the art to do that then you have 18 to do sensitivity studies. But you have to document 19 all of the sources of uncertainty in the overall 20 process.
21 After the PRA has taken its best shot to 22 deal with these in I'd say PRA space when we get into 23 the defense-in-depth adequacy evaluation the defense-24 in-depth adequacy evaluation looks at these issues of 25 uncertainty, takes a critical look at what was done 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
121 1 the PRA, what was assumed in the PRA, what sources of 2 uncertainty were identified in the PRA and then 3 identifies compensatory measures.
4 And the compensatory measures could range 5 anywhere from changing the design to putting in 6 programs, doing testing, experiments and those types 7 of things.
8 So the process certainly does not shy away 9 from this challenge of uncertainty. And I think the 10 process accommodates it.
11 ACTING CHAIRMAN CORRADINI: So can I 12 follow up Walt's question? So I'm still back to 13 principles of using this because I'm still kind of 14 muddled about this.
15 You said there are three possibilities if 16 you go through your iteration loop. One was to change 17 the design. One was to I'll call it sharpen my 18 pencils and do better analysis. One was to use 19 compensatory measures, some sort of programmatic --
20 MR. FLEMING: Or do testing.
21 ACTING CHAIRMAN CORRADINI: Or testing.
22 MR. FLEMING: Yes.
23 ACTING CHAIRMAN CORRADINI: Okay. So is 24 the principle that if I can do something with low 25 uncertainty and high confidence I would choose 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
122 1 over something with large uncertainty?
2 In other words I might change the design 3 and now I have a hardware fix that solves it with a 4 much smaller band of uncertainty. Is that preferred?
5 MR. FLEMING: Well, that would certainly 6 be taken into account in whatever decision would be 7 made. It's hard to prejudge.
8 ACTING CHAIRMAN CORRADINI: But it's not 9 necessarily preferred.
10 MR. FLEMING: It's hard --
11 ACTING CHAIRMAN CORRADINI: I'm going 12 somewhere with this, but I'm trying to understand it 13 because it strikes me that unless I start off with a 14 relatively sophisticated, or some level of 15 sophistication in the design and the PRA I'm going to 16 have a lot of uncertainty.
17 So the more I can change the design to 18 minimize my uncertainty band the better off I am.
19 MR. FLEMING: Right, but there has to be 20 sort of a cost-benefit part of that decision-making 21 process to figure out what the most -- I'm reluctant 22 to give a one size fits all answer.
23 ACTING CHAIRMAN CORRADINI: I understand.
24 MR. FLEMING: Given the different designs 25 and different stages of design and so forth.
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 MEMBER SKILLMAN: Karl, let me ask a 2 question here. I'm following your discussion with the 3 material that was presented. You're explaining task 4 4.
5 The event sequences modeled and evaluated 6 in the PRA are grouped into accident families each 7 having a similar initiating event, challenge to the 8 plant safety functions, plant response and mechanistic 9 source term if there is a release.
10 MR. FLEMING: Yes.
11 MEMBER SKILLMAN: Now, here's my question.
12 Can the family assignment affect the PRA's conclusion 13 or frequency such that random selection will identify 14 a sequence as an AOO one time and a DBE another time?
15 MR. FLEMING: I don't believe -- I believe 16 if the words that you just read are followed properly 17 that shouldn't result in any different classification 18 randomly. I can't see how that would happen.
19 MEMBER SKILLMAN: So would it be accurate 20 to assume that the family grouping is consistent 21 whether it's done by PRA analyst A or B or D or Q?
22 MR. FLEMING: That's the reason why we 23 develop standards. The whole idea of the standard is 24 to create a reproducible process.
25 MEMBER SKILLMAN: Thank you, Karl.
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 MR. FLEMING: Getting back to the slide, 2 after we define the DBEs in the DBE region the idea is 3 to have a comprehensive set of challenges to my safety 4 case.
5 And I go through a process and I ask 6 myself what are the essential functions that I have to 7 fulfill to keep these design basis events inside my 8 frequency consequence target that if I didn't have 9 these could easily flow outside the target.
10 And that's when I come up with what we 11 call the required safety functions. These are the 12 required safety functions.
13 Now they relate to the fundamental safety 14 functions that Bill mentioned, but each reactor has 15 the opportunity to come up with a specialized set of 16 safety functions that fulfill the fundamental safety 17 functions. So this is what we call the required 18 safety functions.
19 This was the insight that was always in 20 the process but needed better discussion that was 21 fleshed out in the X-Energy pilot demonstration. It 22 led to some substantial enhancements to that part of 23 the process.
24 Then we look at, okay, what SSCs are 25 available and not available during all the DBEs 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
125 1 perform those required safety functions. And that 2 process leads to presenting the designer with a set of 3 options that he can select among those that are 4 available on the design basis events, he can select 5 based on his overall strategies. Another integrated 6 decision process, by the way.
7 He selects the safety-related SSCs that he 8 wants to declare safety-related and then we construct 9 -- from each DBE we construct a DBA where we remove 10 any credit for the performance of any non-safety-11 related SSC and that leads to a set of DBAs.
12 And this process has now been done for 13 three or four different plants and when we get to the 14 end everybody thinks that yes, these make sense for 15 this reactor.
16 ACTING CHAIRMAN CORRADINI: So can you 17 give me an example of a list of required safety 18 functions that are technology-inclusive?
19 MR. FLEMING: No. The point is --
20 ACTING CHAIRMAN CORRADINI: I'm 21 struggling. I'm reading the words. I'm trying to 22 understand. Because you said X-Energy this is 23 something that was illuminated in the tabletop 24 exercise was X-Energy. So I'm thinking there would be 25 some required safety functions that are essentially --
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 I guess to put it a different way you're identifying 2 safety functions that remove vulnerabilities.
3 MR. FLEMING: That's right. The required 4 safety functions will be reactor-specific. So the 5 fundamental safety functions that Bill talked about 6 are generic to all reactors, remove core heat, control 7 reactivity and contain fission products.
8 But then when you develop these for 9 specific reactors, for example, in the high 10 temperature gas cooled reactor family controlled 11 chemical attack always comes up because that's 12 necessary for the fuel integrity. They don't want to 13 have oxidation processes go on.
14 We'll show you what the required safety 15 functions were for GE PRISM this afternoon.
16 ACTING CHAIRMAN CORRADINI: Okay, that's 17 fine. If we're going to get to it later that's fine.
18 Thank you.
19 MEMBER MARCH-LEUBA: The staff 20 presentation had a second bullet what safety functions 21 are required to maintain the beyond design basis 22 events to prevent them from going to design basis 23 event in frequency. I don't see you addressing that.
24 Do you understand my question?
25 They had two bullets on the selection 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
127 1 which structures are safety-related. And the second 2 bullet said if you need a structure to make sure that 3 your beyond design basis event does not increase in 4 frequency and becomes a DBE. Those seem to be 5 addressing that part.
6 MR. FLEMING: We do the safety 7 classification, we may have a beyond design basis 8 event that has a very high consequence above 25 rem.
9 So part of the safety classification process is to 10 prevent those BDBEs to go up into the DBE region. So 11 that's another input to the safety classification.
12 This covers the safety classification that 13 comes from mitigating the DBEs.
14 MEMBER MARCH-LEUBA: So you're proposing 15 to do that only for those beyond design basis events 16 that have high consequence?
17 MR. FLEMING: Yes. The goal is to make 18 sure that if there's some degradation in performance 19 of the safety-related SSC that you don't get outside 20 the consequence target.
21 There's two ways to get outside. One is 22 horizontally and the other is vertically. So that's 23 the reason for that.
24 MEMBER MARCH-LEUBA: But you only do it 25 for the high consequence 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
128 1 MR. FLEMING: For safety classification, 2 yes. For safety classification. Now there's other 3 aspects of the frequency consequence that come into 4 the non-safety-related with special treatment which 5 I'll get to in a second.
6 MEMBER MARCH-LEUBA: Well, we'll talk 7 about this when you have the figure.
8 MR. FLEMING: Go to the next slide, 9 please.
10 MEMBER SKILLMAN: Karl, let me ask a 11 question before you go on. I'm back to my homework.
12 Going to read a sentence to you.
13 Part of the LBE frequency dose evaluation 14 is to ensure that LBEs involving releases from two or 15 more reactor modules do not make a significant 16 contribution to risk and to ensure that measures to 17 manage the risks of multi-module accidents are taken 18 to keep multi-module releases out of the list of DBAs.
19 MR. FLEMING: Those are design objectives.
20 What you're referring to there are design objectives.
21 And since the beginning of this process 22 which started in the MHTGR days and carried up through 23 the NGNP part of this development it's always intended 24 that this is a multi-module application.
25 So rather than worry about the lessons 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
129 1 Fukushima after the fact to worry about what you're 2 going to do about multi-module risk we wanted to get 3 the multi-module treatment built in from the ground 4 floor.
5 So what you were just reading is sort of 6 a statement of a design objective. It's really the 7 motivation for taking on multi-module event sequences 8 is we want to take them on so the designer was aware 9 of them so he can make decisions about sharing 10 equipment.
11 There's benefits to sharing equipment 12 because it provides more backup capability and 13 redundancy, but there's down sides associated with 14 maybe introducing the likelihood of a multi-module 15 event.
16 So by embracing the multi-module 17 considerations in the process we give the designer a 18 tool to manage the risk of multi-module events as part 19 of this design. So that's what that statement is.
20 ACTING CHAIRMAN CORRADINI: Thank you.
21 MR. FLEMING: On the safety classification 22 as Bill mentioned we have three safety classes. The 23 safety-related, the non-safety-related with special 24 treatment and the non-safety-related with no special 25 treatment. Those are the three classes that we 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
130 1 The integrated decision process associated 2 with defense-in-depth has an impact on this because 3 the second category, non-safety-related with special 4 treatment, there's two ways to get in there.
5 One, it's a risk-significant SSC based on 6 some risk significance criteria that are outlined up 7 here, or the SSC performs in a function that's 8 considered necessary for adequate defense-in-depth and 9 that's the result of an integrated decision process 10 that looks at the design, that looks at the 11 redundancy, the diversity, the layers of defense and 12 determines some SSC functions may be critical for 13 adequate defense-in-depth. Those are the two ways to 14 get into NSRST.
15 And that aspect of the classification 16 process is analogous to some aspects in 50.69 although 17 I don't want to say we're using 50.69 but that 50.69 18 also classifies safety significant SSCs as risk-19 significant or defense-in-depth adequacy.
20 ACTING CHAIRMAN CORRADINI: So when it's 21 time maybe in the afternoon I'd be interested in an 22 example about the risk-significant or performed 23 functions necessary for defense-in-depth adequacy.
24 I had a hard time in the document 25 understanding the logic so an example might help 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
131 1 that regard.
2 MR. FLEMING: Yes. We actually have 3 examples this afternoon for GE PRISM we can show you.
4 One of the features of this approach is 5 the use of what we refer to as absolute risk metrics 6 for risk significance rather than relative metrics.
7 What I mean by that is in the traditional 8 light water reactor risk-significant approach for 9 operating reactors you measure the importance of a 10 piece of equipment relative to your baseline result.
11 And if you have a core damage frequency that's one or 12 two orders of magnitude lower that's not reflected in 13 the relative importance of the metric.
14 In the ESBWR application they adopted more 15 of an absolute risk metric approach and we've adopted 16 that here in the sense that we measure risk 17 significance on how close you are to the frequency 18 consequence target as far as licensing basis events 19 are concerned and how far away you are from the 20 cumulative risk targets that we have.
21 So the risk significance is tied to 22 stationary numbers that don't change with your design.
23 And that's a very, very important distinction between 24 what we have for operating light water reactors and 25 the LMP 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
132 1 The risk significance criteria by the way, 2 the numerical risk significance criteria is something 3 that was added since NGNP. NGNP did not come up with 4 SSC risk significance criteria.
5 ACTING CHAIRMAN CORRADINI: So, can I say 6 the second major bullet a different way, or the sub 7 bullet in that.
8 What you're really saying is how close to 9 the line can you get before something, it alarms you.
10 And you're saying you have to get within --
11 MR. FLEMING: One percent, yes.
12 ACTING CHAIRMAN CORRADINI: Either in the 13 frequency or in the dose.
14 MR. FLEMING: Well, 1 percent of the 15 frequency as a function of dose. And I'll show you 16 the chart. It's coming up. I'll show you the chart 17 that shows that.
18 We also screen out doses that are so low 19 that they're a small fraction of background which we 20 talked about in June. Next slide, please.
21 So if we look at the universe of SSCs in 22 the plant we have all the -- the rectangle represents 23 all the SSCs in the plant.
24 The large oval there is what's modeled in 25 the PRA. And the idea there is the PRA safety 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 functions are supposed to capture all the SSCs that 2 participate in either preventing or mitigating the 3 release of radioactive material from any source, 4 radionuclide source. So that's the logic for getting 5 it into the PRA.
6 The safety significant SSCs are those that 7 are risk-significant or they provide an adequacy of 8 defense-in-depth. And therefore the risk-significant 9 is a subset of that.
10 We also have our safety-related SSCs and 11 they're almost always risk-significant but if there's 12 a lot of redundancy in your ability to meet your 13 required safety functions they're not necessarily 14 risk-significant but they're always safety-15 significant. So we refer to that as the Segala-16 Cubbage diagram because it resulted from a long 17 discussion we had with Amy and John about how these 18 things relate.
19 MEMBER KIRCHNER: Since the point was made 20 of excessive review and time and enhanced cost comes 21 into play can you use this to make an argument that I 22 only need, I'll make up a number, 10 chapters out of 23 the standard application versus 18 or whatever we're 24 up to in a typical application and the level of detail 25 that's needed for say auxiliary 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
134 1 If you can argue that they aren't safety-2 significant or they aren't risk-significant then can 3 you propose a means to the staff that they should fall 4 off the table in terms of the review?
5 MR. FLEMING: Right. Well, we haven't 6 gone into great detail on this. The general 7 understanding is that in the license application we 8 would provide substantial information for the staff to 9 review the safety-significant SSCs and their 10 performance. And the ones that are not safety-11 significant would not be described in great detail.
12 I mean, that would be the intent.
13 The motivation going back to the MHTGR, 14 this process really started with the MHTGR application 15 back in the nineteen eighties. And the motivation 16 that General Atomic had to launch this approach is 17 that they wanted to end up with a correct set of 18 safety-related SSCs because that was viewed to be the 19 thing that drove the cost of the facility.
20 They didn't want it to be larger or 21 smaller than necessary, but they wanted to get the 22 right set of SSCs. So that's obviously the motivation 23 is to not spend a lot of time arguing and sending RAIs 24 back and forth on non-safety significant SSCs.
25 And of course within the two categories 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
135 1 safety-related and non-safety-related special 2 treatment the understanding is, the general 3 expectation is that there would be a lot more focus on 4 the safety-related SSCs given their importance and 5 somewhat less level of detail on the non-safety-6 related with special treatment. That's the general 7 understanding.
8 MR. REDD: I would of course add that Bill 9 and Amy will be discussing this topic on application 10 content further this afternoon. But I agree with what 11 Karl said. The ultimate goal is to focus on those 12 most safety-significant aspects that could affect 13 public health and safety.
14 CHAIRMAN BLEY: This is Dennis --
15 MR. FLEMING: I think Dennis had a 16 question.
17 ACTING CHAIRMAN CORRADINI: Yes.
18 CHAIRMAN BLEY: I want to make sure I 19 understand your diagram. Things that are safety-20 significant are either risk-significant or they're 21 needed for defense-in-depth. Is that correct?
22 MR. FLEMING: That's correct.
23 CHAIRMAN BLEY: And in your evaluation of 24 defense-in-depth you're considering the uncertainty in 25 the performance with barriers and other equipment.
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 MR. FLEMING: That's right.
2 CHAIRMAN BLEY: Okay. Go ahead.
3 MR. FLEMING: That's correct. Just a 4 couple of comments about this diagram is that one of 5 the areas that we went a little bit further compared 6 to NGNP is that after we had the safety classification 7 how do we come up with special treatment requirements 8 for each of the categories.
9 The thing that's new here in the LMP 10 process is that we start with both safety-related and 11 non-safety-related special treatment, we start the 12 process by setting performance requirements for 13 reliability and capability.
14 Reliability because if you look at all the 15 special treatment requirements you can sort of get 16 into those two categories. Some of them give you 17 greater assurance of reliability. Some of them give 18 you greater assurance that they've got adequate 19 margins when they perform that they'll get the job 20 done to perform their function.
21 So we set the requirements for reliability 22 and capability. Those requirements are set with input 23 from the integrated decision process for evaluating 24 defense-in-depth. They're looking at the 25 uncertainties. They're looking at the whole package 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 of things.
2 We set those requirements including 3 numerical requirements for reliability and 4 availability and performance requirements. And then 5 the rest of the special treatment flows from that.
6 And for the non-safety-related with 7 special treatment our thought is that in most cases 8 all that should be required is putting into case a 9 monitoring program to monitor the performance of those 10 SSCs against those performance requirements.
11 If there's other special treatments or 12 compensatory measures that are needed the IDP process 13 would identify those whereas in the safety-related one 14 there would be a more extensive set of special 15 treatment requirements.
16 So that's an area where we've gone beyond 17 what's actually in the NGNP documents.
18 ACTING CHAIRMAN CORRADINI: So Dennis 19 actually clarified the one thing about what safety-20 significant SSC is. It's both defense-in-depth and 21 risk-significant together.
22 MR. FLEMING: That's right.
23 ACTING CHAIRMAN CORRADINI: Where is 24 safety -- oh, I see. Safety-related is the smaller of 25 those.
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 MR. FLEMING: And getting to -- I won't go 2 into the details on this. There's a table 5-2 for 3 example in the guidance document that talks about the 4 minimum requirements for plant capability defense-in-5 depth.
6 And one of the principles that's in that 7 table is that for required safety functions and 8 critical elements of your safety case you can't have 9 over-reliance on a single design feature or a single 10 element of your design or a single programmatic 11 measure to assure that that's fulfilled.
12 Where that leads to is the need to have at 13 least a couple of different ways to perform your 14 required safety functions.
15 So all the safety-related SSCs are 16 definitely necessary for defense-in-depth. And in 17 most cases they're also risk-significant because if 18 they don't perform their function you could easily 19 have a point creep outside the frequency consequence 20 target.
21 ACTING CHAIRMAN CORRADINI: Thank you.
22 MR. FLEMING: The final point I wanted to 23 make on this is that this big rectangle, the change 24 left over after you modeled everything in those ovals 25 and everything, there's typically screening done 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 because the PRA model doesn't include all the SSCs in 2 the plant. So there's all kinds of screening 3 assumptions made and screening sometimes based on low 4 frequency or whatever.
5 The integrated decision-making process 6 takes a look at that to say gee, is there some 7 compensatory measure we've got to put in place to make 8 sure that the assumptions to screen that component out 9 of the PRA model is enforced.
10 So that's another example on how this is 11 not a risk-based process. It's -- we get what we can 12 out of the PRA process but then we supplement it with 13 defense-in-depth.
14 MEMBER MARCH-LEUBA: Going back to your 15 previous comment about not over reliance on a single 16 thing. Is that single failure criteria light?
17 MR. FLEMING: It may be, I don't know.
18 You may look at it as single failure heavy because the 19 way it's typically manifested in the examples that 20 we've gone through in the pilot studies is you end up 21 having diverse -- in some cases you may have passive 22 inherent feature to perform a safety function. And 23 maybe the second item that's added to the defense-in-24 depth adequacy is an active system.
25 So it's more likely to result in diversity 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 rather than redundancy. However, redundancy would be 2 one of the tools that you would have to meet your 3 reliability requirements. So after you set your 4 reliability requirements redundancy may be necessary, 5 it may not be, on a case-by-case basis.
6 MEMBER MARCH-LEUBA: So your guidance does 7 not have single failure criteria, yes. It's a 8 guidance.
9 MR. FLEMING: Not as an arbitrary 10 requirement.
11 MR. REDD: How we would address that is 12 again point out that through the PRA process you look 13 at all forms and failure combinations including 14 combinations that are extraordinarily unlikely so you 15 get that same value of looking at a single limiting 16 failure through a much more systematic and 17 comprehensive evaluation through the PRA process.
18 Karl, is that a fair statement?
19 MR. FLEMING: That's a fair statement.
20 ACTING CHAIRMAN CORRADINI: But to get to 21 Jose's point, if it becomes a DBA I still would use a 22 single failure criteria.
23 MR. REDD: No.
24 ACTING CHAIRMAN CORRADINI: No.
25 MR. REDD: No, we do 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
141 1 ACTING CHAIRMAN CORRADINI: Your point is 2 -- then your point really is that the design when 3 going through this exercise you'll see diversity or 4 redundancy beyond safety-related equipment. That's 5 when you said it was -- you called it single failure 6 heavy.
7 MR. FLEMING: Heavy. Yes. In that 8 respect, yes.
9 ACTING CHAIRMAN CORRADINI: So let me do 10 a process question. We're at noon. A natural break 11 point at least if I see it in the slides is after 12 slide 15. Where would you want to break? That's what 13 I guess I wanted to ask you.
14 MR. FLEMING: I think if you give me five 15 minutes and then we'll get to a logical break point.
16 ACTING CHAIRMAN CORRADINI: Okay, thank 17 you.
18 MR. FLEMING: At the June meeting Joy 19 brought up a question about safety margins which we 20 didn't have a chance to really give a good answer for 21 so we prepared this slide specifically for you, Joy.
22 This is summarized in the guidance 23 document. There's a half a page or so text that 24 basically wraps around this.
25 But there's -- the approach to safety 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 margins in the LMP framework there's a plant-level 2 safety margins and those are reflected in the margins 3 between where the frequency consequence points plot 4 against the frequency consequence target.
5 And as we got feedback from the staff in 6 an earlier version of our paper by making this 7 comparison of where our points plot relative to 8 frequency consequence target it's one way to 9 demonstrate enhanced safety margins consistent with 10 the Commission's Advanced Reactor Policy Statement.
11 Then we also have SSC level safety margins 12 and those are set in both the reliability targets that 13 we set as well as the performance targets we set by 14 selecting design codes in order to be able to perform 15 the safety functions with adequate assurance.
16 So we have both the plant-level and SSC-17 level safety margins and we confirm the adequacy of 18 these margins as an important element of the defense-19 in-depth process.
20 If we can go on to the next slide, please, 21 unless there's a -- did that answer your question, 22 Joy?
23 MEMBER REMPE: Maybe I missed it but does 24 it talk about how the defense-in-depth process will do 25 this? I know that there's programmatic and 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
143 1 level defense-in-depth and different things like that, 2 but does it really -- is it going to be something 3 where you kind of have to feel it out with another 4 demo and have this integrated decision panel look at 5 this to really understand how it's going to work?
6 MR. FLEMING: I think that will probably 7 help. The integrated decision-making process -- well, 8 the integrated decision-making process on defense-in-9 depth will measure these margins up in the plant 10 level.
11 They'll say, okay, you put together a 12 table. There's example tables in the guidance 13 document that show how far away -- what are your order 14 of magnitude margins in both the frequency and 15 consequence scale for all your design basis -- I'm 16 sorry, all your LBEs.
17 You do that process and that's an input to 18 say based on a frequency consequence what kind of 19 margin do you have in that. So they do that.
20 And then in the SSC safety margin area the 21 IDP process is actually taking a lead role in setting 22 what the reliability requirements are going to be and 23 what the performance requirements are going to be for 24 all the special treatment. So they have a big 25 influence on what comes out of the special treatment 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 box in the process.
2 MEMBER REMPE: When you said the IDP you 3 mean the panel will help specify this, or do you mean 4 just the process?
5 MR. FLEMING: The process.
6 MEMBER REMPE: The process. And the panel 7 will review it.
8 MR. FLEMING: We like to emphasize the 9 integrated decision process. There's a panel exercise 10 here but it doesn't do all this work. There's a lot 11 of integrated decision process that goes along the 12 way.
13 That's one thing we tried to clarify in 14 the last version of our guidance document is that it 15 initially appeared as that all this important stuff 16 was going to be done when we convened this panel at 17 the end of the process and that was a misleading 18 picture.
19 MEMBER REMPE: I guess I still had that 20 concept. In the MHTGR and NGNP has this process been 21 fully exercised yet?
22 MR. FLEMING: No. Well, in the MHTGR it 23 was embedded, implied in their process but they didn't 24 call it that. But if you go ask Fred Silady and 25 others how they actually put together their design 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
145 1 actually -- it was a joint interaction between the 2 design team, the PRA team, the analysis team. They 3 all got together and made decisions.
4 MEMBER REMPE: -- part of that process and 5 we used to have a transient plant design. But the way 6 the NEI --
7 MR. FLEMING: So it was embedded in the 8 MHTGR. But if you look at the documents IDP doesn't 9 appear in the documents. They didn't make a big deal 10 about it. It was just the natural way to do it.
11 MEMBER REMPE: In this NEI document it 12 implies it's more formalized, and I don't recall it 13 being that formalized back in the MHTGR days.
14 MR. FLEMING: Well, we tried to put more 15 structure in this process.
16 MEMBER REMPE: Did the NGNP have this more 17 structured process implemented?
18 MR. FLEMING: Well, in the NGNP we didn't 19 really do much to apply this to a design.
20 MEMBER REMPE: It's not really been 21 exercised is where I'm kind of going, and I think 22 maybe it may need to be that way more.
23 MR. FLEMING: The one part that has been 24 exercised and we'll tell you about this afternoon is 25 in the GE PRISM tabletop they took a cut at looking 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
146 1 their non-safety-related with special treatment SSCs 2 that are necessary for defense-in-depth.
3 Now, they didn't get into special 4 treatments and performance requirements yet, but 5 putting components into the NSRST box was -- we have 6 an example of that for PRISM and we'll show you the 7 results after lunch.
8 MEMBER REMPE: Thank you.
9 MR. FLEMING: I think given where we are 10 it's probably a good time to stop for lunch.
11 ACTING CHAIRMAN CORRADINI: So we'll pick 12 it up with the penultimate diagram.
13 (Whereupon, the above-entitled matter went 14 off the record at 12:08 p.m. and resumed at 1:14 p.m.)
15 ACTING CHAIRMAN CORRADINI: Okay, why 16 don't we begin. Karl, you stopped on slide 13. You 17 wanted to move on to the most important.
18 MR. FLEMING: Before I make a key point on 19 this slide I wanted to revise and extend my remarks on 20 a couple of items that came up before lunch.
21 With the question of our experience in 22 handling the scope of different hazards through the 23 LMP process I forgot to mention in the MSRE work 24 that's ongoing and Steve Krahn will talk to in a 25 little bit they are looking at the offgas system 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
147 1 addition to the fuel cell system. So we are in fact 2 getting some experience outside the normal core thing.
3 The second point I wanted to make refers 4 back to discussion we had on uncertainty and it also 5 refers to a request that Dennis made that we identify 6 things that have changed in the guidance document.
7 We added a paragraph in the guidance 8 document, I can guide you to the specifics if I can 9 remember it. I can't, but we can get that to you.
10 But we wanted to make an emphasis on the 11 point that the LMP process is designed to be flexible.
12 It can be introduced early in the process and we offer 13 -- we identify some advantages to doing that.
14 Or you can also apply it late in the 15 process in more of a confirmatory mode. The GE PRISM 16 example that we went through is maybe one example of 17 a design that was designed using more traditional 18 approaches and then the process came about later.
19 Kairos is planning to do a demonstration 20 project on their fluoride high temperature salt 21 reactor. And in their case they're going to risk-22 inform qualitatively their safety design approach with 23 a view towards using LMP to confirm the selections 24 that they made rather than to develop them.
25 So the reason I'm bringing this paragraph 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 up is that when we talked about the question about 2 uncertainties I wanted to point out that this question 3 about uncertainties is not a property of the LMP 4 process. It's not a property of even trying to do a 5 PRA. It's the property of our state of knowledge 6 about the machine we're trying to license.
7 And I just wanted to point out that we can 8 see advantages to early introduction of this process 9 we believe will help flesh out what the uncertainties 10 are earlier in the process and hopefully minimize the 11 chance that you end up with costly backfits. That's 12 just a value judgment that I wanted to make.
13 On the current slide, the frequency 14 consequence chart, this was actually alluded to in the 15 earlier morning discussion. We've adopted a set of 16 risk significance criteria for licensing basis events 17 and we're setting those at 1 percent of the frequency 18 all the way down the frequency consequence target.
19 If any part of your uncertainty bands on 20 both the dose and the frequency get inside this zone 21 then we consider it a risk-significant LBE and of 22 course we look at that much more carefully than we 23 would look at other LBEs that are not in that process.
24 So the discussion we had this morning and 25 the concerns about the selection of the 750 rem 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
149 1 which anchors the lowest point in the BDBE region down 2 there, this is something that was carried over from 3 the NGNP process.
4 And I wanted to point out this is a 5 surrogate. This is a dose surrogate, dose to be 6 calculated at a fixed point at the plume center line.
7 This is a surrogate for verifying that 8 you've met the QHOs. The QHOs for early fatalities is 9 the average individual risk in a doughnut-shaped hole, 10 a doughnut-shaped area from the site boundary to one 11 mile beyond the site boundary for early health 12 effects.
13 If the doses at the plume center line, at 14 the EAB happen to be 750 rem the average doses in the 15 doughnut hole are well below the threshold for early 16 health effects.
17 So there was actually some work done to 18 demonstrate -- this is actually a conservative 19 selection, but it's just a surrogate for a more 20 elaborate individual risk calculation away from the 21 site boundary.
22 Because as the doses get one mile beyond 23 the site boundary the dose versus distance profile 24 will dilute the dose quite a bit. Just wanted to make 25 that 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
150 1 MEMBER SKILLMAN: Karl, for those hatched 2 areas where you express the caution in the 1 percent 3 zone is there sufficient guidance to prevent there 4 from being the kind of tortuous discussion that we 5 might have with an item that's more than minor, if you 6 will a gray definition of what it means to be only 7 slightly more than risk-significant but not to the 8 limit and therefore you burn up hundreds of hours 9 barking and arguing over trivia.
10 Is the dotted line sufficiently identified 11 or codified that the designers would know once you 12 cross that line you need to consider more action?
13 MR. FLEMING: First of all, it's very 14 important to emphasize that this is a statement of 15 risk significance. If you look at any PRA result from 16 a light water reactor you'll see the identification of 17 risk-significant event sequences or accident 18 sequences, risk-significant basic events and so forth.
19 This is the tool that we would use based 20 on absolute metrics to say what is a significant risk 21 as far as an LBE is concerned.
22 It just means it's significant. It 23 doesn't mean it's not acceptable. However, when we 24 get events that start to encroach into that zone 25 they're going to get much more focused in the defense-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 in-depth evaluation.
2 So the defense-in-depth evaluation, the 3 integrated decision process used there is going to 4 look very carefully at the results coming out of the 5 PRA, the limitations of the PRA, the screening 6 criteria and so forth.
7 And when you're getting into these risk-8 significant LBEs they're going to drill down and 9 understand what's behind that calculation. First of 10 all, the definition of the LBE in terms of the event 11 sequence families and also the estimation of the 12 frequency and consequence.
13 So the trigger point is not a trigger 14 point of unacceptability, it's more of a trigger point 15 for focusing the resources of the defense-in-depth 16 evaluation. Hope that answers your question.
17 MEMBER SKILLMAN: It does, but to me it 18 raises the issue that we've all dealt with and that is 19 once you set a line or a limit for better or for worse 20 it becomes a discussion item. And depending on the 21 strength of the personalities depends on how much more 22 resource you're going to expand to determine how much 23 further you're going to go.
24 MR. FLEMING: Right.
25 MEMBER SKILLMAN: So unless 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
152 1 definition is very, very clear that it is a guide and 2 not a drop dead consideration then what you have 3 stated seems to make sense.
4 But having dealt with this kind of thing 5 my whole life I know sure as shooting someone's going 6 to say we crossed the line. It's obviously got to be 7 in that bin, not that bin, so it needs more QA, it 8 needs more of this, more of that, more analysis. And 9 the only way you can undo that is to make sure that 10 it's very clear that that line is not a drop dead 11 go/no-go gauge.
12 It's a trigger for greater consideration, 13 but it's not in itself a limit.
14 MR. FLEMING: That's right. It's very 15 important to note that.
16 And also the fact that it's -- these are 17 absolute definitions, i.e., there's a fixed frequency 18 consequence curve and a fixed 1 percent line below 19 that as opposed to looking at significance relative to 20 the baseline result which is the way light water 21 reactors do risk significance.
22 So, there's always something significant 23 in a light water reactor PRA because it's just a 24 relative metric. So we would expect in most of the 25 case studies we've seen we haven't really seen 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
153 1 many examples at all of any LBEs show up in this 2 region, but it's certainly possible.
3 MEMBER SKILLMAN: Thank you, Karl.
4 MR. FLEMING: If we can go on to the next 5 slide. This last of the process slides. The defense-6 in-depth evaluation. We've already talked quite a bit 7 about that so I didn't plan on doing a soup to nuts 8 discussion on that.
9 In the guidance document we break down 10 attributes of defense-in-depth for each of the three 11 yellow cornerstones up here, the plant capability 12 defense-in-depth, the programmatic defense-in-depth 13 and the evaluation of risk-informed performance-based 14 evaluation of adequacy.
15 And so those attributes are used by the 16 integrated decision process and the panel to come up 17 with a baseline defense-in-depth evaluation which is 18 documented and then provides a basis for change 19 management as the design goes through various stages 20 of development, licensing and siting and so forth.
21 So I think we've talked about most of 22 these. I didn't want to spend too much more time on 23 this. This is I think one of the cornerstones of 24 advancing the technology-inclusive approach that came 25 out of the NGNP project.
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 There was a defense-in-depth white paper 2 for NGNP that is consistent with this but I think 3 we've taken -- in sort of a football analogy we think 4 we've advanced the ball down the field on this topic.
5 MEMBER REMPE: So out of curiosity I was 6 trying to think of this when I was reading the 7 material. Is there an emphasis to make sure you have 8 some plant capability as well as programmatic 9 capability defense-in-depth. You should draw from 10 both types of options. Plant capability is the device 11 basically and you do have to have --
12 MR. FLEMING: Well, yes. They're sort of 13 like different kind of animals. One way to look at 14 plant capability defense-in-depth is defense-in-depth 15 on paper. So if I build this plant according to the 16 way it was designed and I implement the safety design 17 approach and there's no changes and whatever, no 18 uncertainties, then it's sort of like an as designed 19 sort of defense-in-depth.
20 What the programmatic defense-in-depth, it 21 does two things. Number one is that it puts in 22 processes to make sure that if you build it according 23 to the design it will be maintained and operated 24 through the life of the plant maintained in that 25 design envelope.
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 And also to address uncertainties in all 2 the decisions that went into putting the features into 3 the plant capability defense-in-depth including things 4 like what do we have to do to assure the reliability 5 and capability of the SSCs that are part of my 6 defense-in-depth, the uncertainties that may have come 7 out of the frequency consequence evaluation in 8 evaluating LBEs.
9 So it's more of a preservation of defense-10 in-depth through all phases of building the plant and 11 operating it, licensing it and managing uncertainties 12 and deviations, temporal deviations in performance.
13 MEMBER REMPE: So your response implies to 14 me you have to have --
15 MR. FLEMING: You have to have both.
16 MEMBER REMPE: Thanks.
17 MR. REDD: Let's ask Ed Wallace can you 18 comment briefly on the balance about whether we have 19 to have -- Dr. Rempe I think your question is do you 20 have to have balance or does the guidance tell you to 21 balance programmatic and --
22 MEMBER REMPE: I just was exploring it.
23 MR. WALLACE: A couple of thoughts here to 24 add to Karl. One is when you look at this equation of 25 sorts these are contributions to reasonable assurance 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 and adequate protection.
2 And so adequate protection is more aligned 3 with how does the plant perform and reasonable 4 assurance is how confident are you about how does the 5 plant perform.
6 And so the programmatic activities start 7 to go to things like tech specs. As long as you stay 8 in the operating box that the design was built for 9 then there's a higher likelihood that you're not going 10 to run into a problem.
11 As long as you monitor the systems to make 12 sure they're not degrading in service then there's a 13 higher likelihood -- that's a special treatment.
14 Higher likelihood that you're not going to get 15 surprised later in life.
16 So there's a series of things. You can 17 include QA. Karl mentioned earlier and I think Dr.
18 Corradini also made a point about early in the design 19 your sources of uncertainty are extremely important 20 when you look at what's going on. There's a lot of 21 unverified assumptions in the design process that get 22 worked through.
23 And one of the answers when you do this 24 may be gee, this phenomena we don't know enough about 25 and it's driving the uncertainty bands around the mean 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 which might tickle if you will to the question that 2 Dick asked the cross hatch zone. We'd say why is that 3 happening.
4 And so the purpose of the defense-in-depth 5 process is systematically to say we ought to be 6 looking at that harder in a different structured 7 manner and we ought to be looking at is it driven by 8 plant capability, is there compensatory measures that 9 you could take that would be more programmatic but not 10 change the plant capability one of which is go run 11 some more tests in your integrated non-nuclear test 12 facility if you have one so that you can sharpen your 13 pencil to use the term that was used earlier about 14 that uncertainty and its significance to your overall 15 plant performance.
16 So you end up with a set of things in both 17 camps and the design process sort of weighs the best 18 way to solve the problem and part of the defense-in-19 depth description is if it's already in concrete your 20 design options are limited and so you may have a bias 21 towards trying to solve the problem programmatically 22 because tearing out concrete is not a good idea if you 23 can avoid it.
24 MR. REDD: Thank you, Ed. Karl, which 25 slide.
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 MR. FLEMING: Let's go to the next slide, 2 please. Now the balance of our presentation is going 3 to focus on lessons we're learning by applying this 4 process to different technologies. And they're 5 summarized here on one slide, everything that has been 6 done or is planning to be done by the spring of 2019 7 to support the processes in the LMP.
8 And I put them in to sort of accident --
9 reactor type families. The high temperature gas 10 cooled reactors, the liquid metal cooled reactors, the 11 molten salt reactors and then we have some other 12 reactor concepts that have different combinations of 13 fuel coolant type arrangements.
14 ACTING CHAIRMAN CORRADINI: So these are 15 like pilot applications of the LMP?
16 MR. FLEMING: Well, yes. Each of these 17 contributes to some element of experience in applying 18 the LMP process. I'm going to show you a matrix 19 coming up that breaks it down into which reactors 20 apply to which steps of the process to give you an 21 idea of where we are today.
22 ACTING CHAIRMAN CORRADINI: But none of 23 the four if you want pilots have exercised the whole 24 process.
25 MR. FLEMING: That's correct.
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 ACTING CHAIRMAN CORRADINI: Is the intent 2 that they will eventually?
3 MR. FLEMING: That's -- I doubt whether 4 every aspect of the process will be demonstrated in 5 the pilots. There's just too much resources.
6 ACTING CHAIRMAN CORRADINI: The reason I'm 7 going there is to the extent that the industry works 8 together and understands it together the better off it 9 is downstream versus a fragmentary understanding.
10 So the thought that you're running -- I 11 keep on using the word pilots. The fact that you're 12 running four of these strikes me as interesting. It 13 would be more interesting if they completed them 14 because then any other particular vendor in a 15 particular type can look back and see an empirical 16 example of how --
17 MR. FLEMING: Right.
18 MR. REDD: And I want to add that as we've 19 progressed through these demonstrations that the 20 amount of detail we've been able to go into and the 21 further through the process we've been able to go has 22 been beneficial. Especially one site, the GE-Hitachi 23 PRISM exercise given that they're by far the most 24 complete design. We were able to exercise a good bit 25 of the process there that we'll discuss further 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
160 1 MR. FLEMING: And when I get to the next 2 slide, Michael, I think I can get a more complete 3 answer to your question.
4 In the high temperature gas cooled reactor 5 family this process was really first started with the 6 MHTGR licensability submittal that was done back in 7 the nineteen eighties.
8 That was done in conjunction with a 9 preliminary safety information document, a PRA, an NRC 10 staff and NRC staff contractor review. This is 11 probably the most complete application of the process 12 although some steps of the process were invented after 13 MHTGR so they weren't able to do it all.
14 ANS 53.1 and Jim August mentioned that he 15 was the chairman of that group that put together that 16 standard, that built upon the methodology in between 17 the Exelon PBMR interaction and the NGNP project is 18 when ANS 53.1 came along.
19 And it basically documents a design 20 process that follows the basic elements of the LMP 21 framework.
22 We recently completed and Brian Waites 23 reported this morning we completed a limited scope 24 demonstration on the XE-100 pebble bed reactor and we 25 have a public domain report that documents that. I'm 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 going to show some example results that we got from 2 that.
3 In the liquid metal -- in the sodium 4 cooled fast reactor family GE PRISM similar to MHTGR 5 submitted a licensability submittal, a preliminary 6 safety information document, a PRA, NRC staff and 7 contractor review and published NUREG-1368.
8 GE-Hitachi has also been actively involved 9 in supporting with many other advanced reactor 10 developers in developing a non-light water reactor PRA 11 standard.
12 We issued a trial use standard in December 13 of 2013 and it was intended to be piloted by a number 14 of projects. There were quite a few different 15 projects that piloted including the Chinese HTRPM that 16 was used to license the pebble bed reactor being 17 designed and just about ready to start up in China.
18 And one of the things that transpired was 19 the Department of Energy granted a project to GE-20 Hitachi to modernize their PRA specifically to pilot 21 the non-light water reactor PRA standard and give 22 feedback to the standard process.
23 And that gave the opportunity to have a 24 very modern PRA project that was done that supported 25 a demonstration project, the tabletop project 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
162 1 we're going to talk about here in a few minutes.
2 In the molten salt reactor area there's an 3 activity underway that Steve Krahn will tell you more 4 about in a few minutes involving using the molten salt 5 reactor experiment as a design and a plant that's 6 already operated and had some service experience to 7 work through the process to support the molten salt 8 family reactors using the LMP process.
9 And they've already published a report 10 where they've started to take a look at licensing 11 basis events for the MSRE using the LMP process 12 starting from basically a blank sheet of paper.
13 And they're also advancing the technology 14 of using HAZOPs, process hazards analyses like HAZOPs 15 and failure modes, effects analysis to build the 16 knowledge base that you would need to start this 17 process. And Steve will amplify on that in a few 18 minutes.
19 We also have planned some demonstration 20 projects, some pilot projects on the Kairos fluoride 21 salt reactor and also the Westinghouse micro reactor, 22 the eVinci heat pipe reactor. And both of those are 23 planned demonstrations to be completed by the spring 24 of 2019.
25 So now I'd like to go on to the next slide 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 and start getting into what they actually did. So in 2 this matrix we've identified this is progress to date.
3 So this is actually what's been completed to date.
4 It doesn't credit what we plan to do for 5 the ones that haven't been finished yet.
6 So this matrix shows it's broken down.
7 There's about 18 different steps of the LMP process 8 and we tried to capture here some of the key steps of 9 the process including developing an internal events 10 PRA, a seismic PRA, a PRA that covers both single and 11 multi-module sequences enough to define the AOOs, DBEs 12 and BDBEs using the accident families coming out of 13 those studies, evaluate the LBEs against the frequency 14 consequence target and the cumulative risk targets, 15 identify the required safety functions that are 16 necessary and sufficient to keep our design basis 17 events inside the target, and have at least example 18 selections of safety-related SSCs that would perform 19 those functions.
20 We also have the process step of 21 developing functional design criteria for the safety-22 related SSCs and safety-related design conditions.
23 That's SSC-level design criteria. Those two steps 24 have only been performed for the MHTGR.
25 This actually involves a completed design.
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 It's difficult to tabletop some of these steps.
2 And we also have looking into the defense-3 in-depth steps we've broken down two parts to that.
4 That's evaluating the plant capability for defense-in-5 depth and we have some limited experience with the GE 6 tabletop on that.
7 And also the rest of the defense-in-depth 8 process including the programmatic defense-in-depth 9 and the application of the integrated decision 10 process. So this is the matrix that shows you where 11 we are today.
12 CHAIRMAN BLEY: This is Dennis. Two 13 things I wanted to ask you about. On all of these 14 cases you were able to compare the LBEs with the FC 15 curves which implies you were able to develop source 16 terms for all of these scenarios. Are those described 17 in something we can look at, how that was done?
18 MR. FLEMING: We'll get into the extent of 19 source term development. The examples we have up 20 here, the MHTGR and the GE PRISM are supported by 21 mechanistic source term analyses to develop the doses.
22 The XE-100 based on its stage of design we just 23 provided estimates of those based on scaling from 24 power level based on results of other studies.
25 And the molten salt reactor experiment 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 I'll let Steve tell us what they plan to do for source 2 terms on that.
3 ACTING CHAIRMAN CORRADINI: So let me just 4 make sure I understand what N/A means.
5 MR. FLEMING: N/A means it wasn't -- that 6 step wasn't available when the MHTGR was done. The 7 MHTGR was done in like 1988 and we hadn't invented 8 these DID steps. We hadn't invented that.
9 ACTING CHAIRMAN CORRADINI: Excuse me. So 10 NGNP is not a potential pilot. This is the 1986 11 MHTGR.
12 MR. FLEMING: The NGNP really didn't 13 involve a pilot. There were several different designs 14 but it wasn't really an actual demonstration like was 15 done for the MHTGR.
16 CHAIRMAN BLEY: They hadn't actually 17 settled on their design.
18 ACTING CHAIRMAN CORRADINI: Dennis, it's 19 hard to understand you. Can you say it a little 20 louder, please?
21 CHAIRMAN BLEY: I said NGNP didn't 22 actually -- had not actually settled on a design.
23 MR. FLEMING: That's right.
24 ACTING CHAIRMAN CORRADINI: I think they 25 didn't settle on their reactor core design, 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
166 1 thought most of the equipment between the two designs 2 were similar outside of the core geometry.
3 I think we have -- I'm looking at -- where 4 did he go. I'm going to drag Jim up to the mike to 5 make him properly characterize this.
6 MR. KINSEY: We had some preliminary 7 information from the three players who were involved 8 with pursuing designs but there was no selection and 9 it didn't get -- those designs didn't progress far 10 enough in their level of detail to be able to apply 11 the process.
12 We used the GA MHTGR design as sort of our 13 surrogate for all of those discussions during NGNP as 14 our poster child.
15 ACTING CHAIRMAN CORRADINI: So the reason 16 I asked the question --
17 CHAIRMAN BLEY: -- my memory.
18 ACTING CHAIRMAN CORRADINI: I'm sorry, 19 Dennis, go ahead. I apologize.
20 CHAIRMAN BLEY: I said what Jim said 21 agrees with my memory.
22 ACTING CHAIRMAN CORRADINI: Okay. But 23 since Jim is there can I make sure I'm clear? In the 24 MHTGR analysis they didn't do the ones that are N/A 25 but in terms of functional -- I'm still back to what NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
167 1 we've already written about and what we have taken a 2 position on in terms of functional containment.
3 These are particularly of interest to me.
4 So I'm kind of curious in terms of defense-in-depth 5 adequacy is the only example this has been exercised 6 is with the PRISM?
7 MR. FLEMING: So far, yes.
8 ACTING CHAIRMAN CORRADINI: Okay, thank 9 you.
10 MR. FLEMING: That's where we are.
11 MEMBER DIMITRIJEVIC: Okay, so my 12 question, you show us the graph with the -- the risk-13 significant, the safety-significant. That was only 14 done on PRISM as much as I can see, right? Is it 15 done? Because I'm not sure that these, like for 16 example the necessary required safety function that's 17 where we can identify systems which will get safety 18 classification.
19 And then I can see the risk was done, the 20 significance was done and if defense-in-depth was done 21 then you have a safety significance there too, right?
22 MR. FLEMING: That's right.
23 MEMBER DIMITRIJEVIC: So will you show us 24 example how many SSCs we have --
25 MR. FLEMING: For PRISM.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
168 1 MEMBER DIMITRIJEVIC: You have that.
2 MR. FLEMING: For PRISM. We'll have that.
3 It's on one of the slides.
4 MEMBER DIMITRIJEVIC: Okay, all right.
5 Excellent.
6 MR. FLEMING: Okay. Shall we move on to 7 the next slide? Some general points that we've 8 learned collectively and many of these lessons are 9 already reflected in draft November or whatever we 10 call it of the guidance document.
11 So we now have experience with at least 12 including some that are sort of still a work in 13 process on the three major families of advanced non-14 light water reactors meaning the high temperature gas, 15 the liquid metal and the molten salt reactors.
16 The feedback we've gotten from the 17 developers and I'm going to ask them to give us this 18 in their own words, but they found the demonstration 19 to be useful. They like the approach and it seems to 20 produce results that they think are reasonable. So 21 that's the feedback we've gotten from them. And we'll 22 get some more specifics on that from the developers in 23 a little bit.
24 They really liked the idea of using 25 absolute metrics for determining risk significance.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
169 1 It really focuses the statement of what's risk-2 significant to what's really important.
3 If we had stuck with the relative 4 significance we might be -- we might have an 5 unnecessary burden.
6 And also I want to point out an important 7 insight. This actually happened from the GE PRISM PRA 8 modernization that was done to pilot the non-light 9 water reactor standard.
10 When we first wrote the risk significance 11 criteria in the non-light water reactor standard we 12 used the light water reactor model. There was a lot 13 of pressure by the Joint Committee on Nuclear Risk 14 Management to keep everything consistent with light 15 water reactors unless it had to be different.
16 So we used the risk significance criteria 17 where we mapped the requirements for CDF-based risk 18 significance to all the release categories.
19 And what GE PRISM found was it just 20 created a mess because they had 40 release categories.
21 So they had 40 sets of Fussell-Veselys and risk 22 achievement worth and so forth.
23 So it was actually feedback from the GE 24 PRISM PRA modernization project that had a great 25 influence on how we do safety significance -- risk NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
170 1 significance in the LMP. So that's already baked into 2 the guidance document.
3 And we're now working on putting those 4 absolute risk metrics for risk significance into the 5 standard. We're working on the next generation of 6 this standard now.
7 So that was a huge -- it has a huge 8 impact.
9 MEMBER SKILLMAN: Karl, by changing from 10 relative to absolute did the conclusions change?
11 MR. FLEMING: Oh, yes. The population of 12 risk-significant SSCs was --
13 MEMBER SKILLMAN: Night and day.
14 MR. FLEMING: Yes, night and day.
15 Absolutely. Huge, huge difference.
16 The other thing we learned and again we 17 learned this, this was in the GE-Hitachi demonstration 18 project. The Venn diagram I showed you that showed 19 all the different safety-significant, risk-20 significant, safety-related SSCs.
21 We used to think that all the safety-22 related SSCs were risk-significant based on our 23 definition and we put that down there. And it was 24 actually the GE-Hitachi PRA people who had experience 25 with the ESBWR and had done a similar approach where NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
171 1 in their example they had 14 different ways to get 2 water in the vessel and therefore whatever safety-3 related SSCs for vessel injection that they had in the 4 ESBWR that was not risk-significant because they had 5 14 backups.
6 So they clarified that Venn diagram for us 7 and got us a better understanding of the relationship 8 between risk-significant, safety-related and safety-9 significant. So that was a good insight.
10 MEMBER KIRCHNER: Just quickly going back 11 to your Venn diagram. How do you square that so to 12 speak with the NRC's definitions for safety 13 classification and performance criteria? Where you 14 have three, safety-related, non-safety-related special 15 treatment and non-safety-related with no special 16 treatment.
17 MR. FLEMING: Yes. For those three 18 categories -- well first of all, having those three 19 categories was something that actually was imported 20 from the NGNP process. The NGNP process also had 21 those three safety classes.
22 But as far as the current reactor is 23 concerned this is similar to the 50.69 process where 24 you're dividing up and you're defining safety-25 significant sequences other than safety-related 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
172 1 on risk significance and defense-in-depth.
2 So we think there's some alignment there 3 with the 50.69 just as far as the safety 4 classification process.
5 MR. REDD: And just to emphasize we're not 6 trying to implement the 50.69 process here. It's just 7 a similar framework looking forward. That's one key 8 takeaway we want to be very clear on.
9 MEMBER KIRCHNER: I'm just thinking in 10 terms of clarification and simplicity are we going 11 with three categories or four categories?
12 MR. FLEMING: We only have three.
13 MEMBER KIRCHNER: Maybe I'm just confused 14 then.
15 MR. FLEMING: Oh yes. Okay, that's right.
16 We do have four, that's right.
17 ACTING CHAIRMAN CORRADINI: I was going to 18 say --
19 MR. FLEMING: That's right. There are 20 four. Yes, I'm sorry, I misspoke. There are four.
21 It's the ones outside. Thank you.
22 MEMBER KIRCHNER: Is there some 23 qualitative understanding with the NRC about how these 24 definitions are going to be used?
25 I mean, where I'm going is so you made 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
173 1 pitch in the beginning that the process requires the 2 applications too much extraneous so to speak 3 information and such.
4 It would seem to me your Venn diagram 5 ought to be first order of basis for cutting out a lot 6 of material that isn't important to risk, right?
7 I guess I'm missing something here.
8 MR. FLEMING: Should we go back to the 9 Venn diagram?
10 MEMBER KIRCHNER: Doesn't this provide you 11 a means for streamlining the application in terms of 12 content and such if you can demonstrate that there are 13 SSCs that are outside the envelope of contributing 14 significantly to risk or whatever terminology you're 15 going to use.
16 It seems to me that for consistency this 17 would be the rationale or basis for then eliminating 18 material or excess material from consideration in the 19 license application.
20 MR. FLEMING: Well, I guess it's our 21 general view that application of the LMP process vis-22 a-vis a conventional ad hoc process should lead to a 23 more streamlined safety analysis report. But we 24 haven't actually gone through chapter by chapter to 25 actually demonstrate that yet.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
174 1 MEMBER KIRCHNER: Thank you.
2 MEMBER REMPE: Could you go back to slide 3 17. In light of all the insights that you've gotten 4 from doing what you've done why not -- are you 5 confident you wouldn't get a lot more insights if you 6 go through and finish the rest of these steps with the 7 PRISM design?
8 Again, I'm still hung up on how this whole 9 defense-in-depth and integrated decision panel is 10 going to work. I know you've said we don't have the 11 resources. Department of Energy has a lot of 12 resources.
13 I'm just wondering is there something 14 important you would glean if you went another step 15 further and did some additional work in this area?
16 MR. FLEMING: I'm sure that we could glean 17 more insights if we did more but really that question 18 is really something that needs to be collectively 19 asked to the developing community.
20 All of these developers are volunteering 21 their services to come and apply this process because 22 they're excited about using it and they're investing 23 their resources to do it.
24 ACTING CHAIRMAN CORRADINI: But I think 25 what Joy is asking is another way of what I was saying NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
175 1 also which is the idea that you guys have gotten 2 together and have again I use the word pilot these 3 four different approaches, or four different classes 4 and running through it. To the extent that you can 5 more completely do it I think it would be better for 6 the community.
7 It will clarify a whole lot of things it 8 strikes me.
9 MEMBER REMPE: Yes, especially since 10 you've spent extra time now to define defense-in-11 depth. That's what I'm thinking of and the integrated 12 whatever, decision panel and all that.
13 I just think that there's some fuzziness 14 in my mind from what I'm reading and maybe it's clear 15 to you. But it seems like if you would step through 16 those additional steps you might learn some important 17 nuggets that ought to be considered. It's something 18 to think about. That's kind of where I'm going.
19 MR. REDD: Certainly. We thank you for 20 the feedback. And again we've got several more 21 demonstration opportunities coming up.
22 MEMBER REMPE: But those designs are less 23 mature and I think you're going to have to use one of 24 the more mature designs to get that useful feedback.
25 Unless somebody really jumps their design ahead 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
176 1 hurry.
2 MR. REDD: Certainly. Thank you.
3 MR. FLEMING: Any experience we get we 4 should be able to benefit from. I certainly wouldn't 5 disagree with that at all.
6 If we can go on to that next slide. The 7 next to last bullet --
8 MR. AFZALI: Sorry. Two of our exercises 9 fundamentally the approach and the methodology hasn't 10 changed. The insight we've gained, we're mostly 11 focused on ease of application. So we have provided 12 additional clarity but the fundamental approach hasn't 13 changed.
14 So we totally agree with the concept of 15 more applications to further improve the process and 16 execution part of it. But fundamental part of our 17 process we have a pretty much confidence that it 18 works. It gives reasonable results. Just wanted to 19 clarify that point.
20 And Karl, I would like you to make sure 21 I'm not misstating our findings as a result of our 22 pilots.
23 MR. FLEMING: Yes, that's a very important 24 point. We're not finding the need to modify the 25 methodology from these tabletops but we are providing NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
177 1 -- we're getting opportunities to provide better 2 guidance on how to most efficiently implement it.
3 That's what we're basically getting out of it.
4 MEMBER REMPE: Has the regulator been 5 involved with these demonstrations at all? Or is it 6 just the industry coming in and working with NEI?
7 MR. FLEMING: NRC staff sat in on one day 8 of the GE PRISM demonstration. The GE PRISM 9 demonstration went over a period of several months and 10 there were lots of interactions and meetings between 11 the LMP team and the GE-Hitachi team that was working 12 on that.
13 But the final day of sort of like 14 presenting the results the NRC staff participated in 15 a one-day meeting.
16 MEMBER REMPE: Were some additional areas 17 for clarification was needed identified because of 18 that interactions with the regulator?
19 MR. FLEMING: I don't recall.
20 MR. REDD: Bill or John Segala if you all 21 would like to provide any input.
22 MEMBER REMPE: No, okay. Thank you.
23 MR. FLEMING: The next to last point I 24 wanted to emphasize is that one thing that's come out 25 in our demonstrations is the importance of thinking NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
178 1 through all steps of the process before making 2 decisions. In other words if you're looking at your 3 frequency consequence charts and you're starting to 4 think about incorporating risk insights before you 5 start thinking about making a change to anything 6 because you don't like the results work your way all 7 the way through the process so you know all the tools 8 that are available to affect the results including 9 what you're going to find out of defense-in-depth.
10 So this gets back to the fact that this is 11 an integrated process and it's not designed to be 12 taken piecemeal and applied partially from one aspect 13 of it.
14 This really -- what Amir was jumping up to 15 say was really the last bullet point on that is that 16 what we're finding is we're finding ways to improve 17 the guidance.
18 In fact, the GE and the X-Energy tabletops 19 that were already completed already are reflected in 20 changes in the guidance document. So we're getting 21 better guidance out of it.
22 The next slide here is one slide we have 23 on the XE-100. This is a pebble bed reactor, a very 24 small pebble bed reactor. It's being designed by X-25 Energy.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
179 1 And this is a good example of what this 2 process would look like at a very early stage of 3 design. The conceptual design of the XE-100 is 4 recently started. It started earlier this year. So 5 they're very, very early in their analysis.
6 And before the LMP project came along they 7 decided to do a very, very high-level limited scope 8 PRA just to help make some design decisions to support 9 some tradeoff studies on how they were going to design 10 their core heat removal systems and also to get some 11 rough idea of what the licensing events were going to 12 look like that they need to worry about in the 13 conceptual design. And that was -- that PRA was done 14 several years ago not necessarily tied to applying the 15 LMP process in general.
16 And it was a high-level PRA. The event 17 trees and event sequence diagrams from that PRA are 18 actually in the PRA white paper that show examples of 19 how you can develop the first building blocks of a 20 high-level PRA at an early stage of design.
21 But they knew what their sources of 22 radioactive material were, they had insights from 23 prior PBMR type PRAs. And what we did in the tabletop 24 exercise was help the XE-100 folks develop rough 25 estimates of what the doses would be because their NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
180 1 mechanistic source term analysis and their tools 2 weren't ready yet to come up with their own estimates.
3 So we provided information from NGNP 4 studies and PBMR studies and MHTGR work and based on 5 power level scaling arguments they were able to come 6 up with rough estimates.
7 They actually do have in the library of 8 work they actually do have uncertainty estimates on 9 these frequency consequences. But in the time 10 available it was hard to get them plotted on the 11 chart.
12 MEMBER MARCH-LEUBA: Karl, we have this 13 example here. I'm a very visual guy. We can go how 14 you select for this particular example the safety-15 related component. Let me see if I understand what 16 you do.
17 You will take the two green points which 18 are the AOOs and say those are going to be DBAs for 19 me. And identify --
20 MR. FLEMING: Let me walk you through the 21 process. We start with the events in the DBE region.
22 MEMBER MARCH-LEUBA: The other ones.
23 MR. FLEMING: Between 10-2 and 10-4 per 24 plant year. Now, when we do that we capture events 25 whose uncertainties straddle the boundary. So we make NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
181 1 the set as large as we can.
2 And then we analyze each LBE one at a time 3 and we ask ourselves what safety functions were being 4 fulfilled that kept the doses as low as they are.
5 And by the way, also shown on here on the 6 y axis are LBEs that have no dose whatsoever. There's 7 no dose at all.
8 So we look at those and ask ourselves what 9 are the required safety functions that are needed to 10 keep those inside the frequency consequence chart.
11 And for this reactor the three they came 12 up with was controlled core heat removal, controlled 13 core heat generation or reactivity control and 14 controlled chemical attack.
15 And those in turn will assure the 16 containment of radionuclides.
17 MEMBER MARCH-LEUBA: And then for your 18 chapter 15 you will run all those points with --
19 MR. FLEMING: Hold up, I'm not quite done 20 yet. So after we've done that we look at all the DBEs 21 that we started with and we identify which SSCs were 22 available or not available to support each of those 23 required safety functions.
24 So we go DBE by DBE. We look at every one 25 singly and then we say can we select -- what are 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
182 1 options to select a single SSC that will perform each 2 required safety function for all the DBEs, that cover 3 all the DBEs that's available in all the DBEs.
4 And normally there's options. For the 5 pebble bed reactor it turned out the options were the 6 reactor cavity cooling system or the heat sinks, the 7 passive heat sinks in the reactor building. Those 8 were the two options.
9 By the way, very similar result is mapped 10 out for the MHTGR in the LBE white paper. So then 11 they decide okay, which of those options do they want 12 to make safety-related.
13 And that's as far as we took that tabletop 14 exercise but that's the process.
15 MEMBER MARCH-LEUBA: I didn't have any 16 problem with that. I want to go beyond ten to the 17 minus four. Now the beyond design basis event. You 18 said earlier that the highest one gives you 20 19 millirems so you say if you want more than 2 and a 20 half rem you would consider it.
21 The beyond design basis events you are not 22 going to evaluate what safety-related functions make 23 them safe?
24 MR. FLEMING: Well, by making the safety-25 related selections that we did make based 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
183 1 evaluating the DBE region it helped to reduce the 2 frequency of some of the BDBEs.
3 MEMBER MARCH-LEUBA: For those points that 4 you -- the purple points you put there on that figure, 5 do they include all of the components, even the ones 6 that are not safety-related?
7 MR. FLEMING: Yes, it includes all of 8 them.
9 MEMBER MARCH-LEUBA: So how do you know 10 that those -- if you have fail a non safety grade 11 component, you fail it, that will move to 2,000. How 12 do you know it doesn't?
13 MR. FLEMING: Well, we're going to pick 14 that up when we do the risk significance evaluation.
15 MEMBER MARCH-LEUBA: See, the thing that 16 makes sense to me is you go through your red and green 17 points, decide what your systems are going to be 18 safety-related, fix them, rerun the BDBEs and those 19 components and none else. And then you know that 20 you're okay. But that's not what you're planning to 21 do.
22 ACTING CHAIRMAN CORRADINI: No, I think 23 you're one step ahead of what he said. Once I 24 identify -- I see it as a number of steps. The first 25 thing you identify your safety systems, safety-related NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
184 1 systems. The next step is you identify your risk-2 significant ones. Then with a defense-in-depth 3 adequacy judgment which still I'm cloudy about you 4 might have an even larger population than risk-5 significant.
6 MR. FLEMING: That's right.
7 ACTING CHAIRMAN CORRADINI: And then when 8 you want to do a DBA analysis you basically assume 9 everything that is not safety-related fails.
10 MEMBER MARCH-LEUBA: No, that's not --
11 MR. FLEMING: -- chapter 15.
12 MEMBER MARCH-LEUBA: No, chapter 15 you 13 don't do BDBEs. You only do the DBEs.
14 (Simultaneous speaking.)
15 MEMBER MARCH-LEUBA: I'm asking him to do 16 chapter 15 for all of them.
17 ACTING CHAIRMAN CORRADINI: Why would you 18 do that? That doesn't make any sense.
19 MEMBER MARCH-LEUBA: Let's take a 20 hypothetical. I have an event that melts my core, 21 breaches the vessel, but containment is intact and 22 nothing comes out. Is the containment a safety grade 23 or not? If you run the calculation and assume that 24 your containment failed that event kills a million 25 people.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
185 1 But if you run it this way you say I don't 2 need a containment, it's not safety grade.
3 MEMBER REMPE: Well, doing the risk 4 assessment sometimes things may fail, some things may 5 not. But if you do a realistic analysis --
6 MEMBER MARCH-LEUBA: Your mike is not on.
7 MEMBER REMPE: But again if you do a 8 realistic assessment for beyond design basis events 9 sometimes things will fail, sometimes things won't but 10 you consider that --
11 CHAIRMAN BLEY: Can't hear.
12 MEMBER REMPE: Sorry.
13 MR. FLEMING: When we determine our 14 required safety functions we're going to assess okay, 15 if I didn't do that function, like if I didn't have a 16 containment would the doses go outside that line. And 17 if they do then that's assurance why we have to --
18 MEMBER MARCH-LEUBA: In all those events 19 I can assure you that the containment atmosphere 20 remain inert meaning air didn't come inside the 21 containment otherwise we will have a fire.
22 And that's because there was a component, 23 a window that remained closed and didn't break.
24 Should that window be safety grade?
25 MR. FLEMING: That should show up 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
186 1 required safety functions. If that window is so 2 important it would be identified --
3 MEMBER MARCH-LEUBA: I can guarantee you 4 it is. I can guarantee you it is. If you flood the 5 containment with air and you have graphite that's a 6 bad scenario, really bad scenario. And you're going 7 to want to make that window safety grade. I can 8 guarantee you that too.
9 MR. FLEMING: Well, without getting into 10 the details the phenomena of graphite oxidation is 11 tracked in the evaluation of the high temperature 12 reactor LBE so graphite oxidation is tracked.
13 If you depressurize the helium pressure 14 boundary in the reactor building you will displace the 15 air from the building --
16 MEMBER MARCH-LEUBA: And then the window 17 will close.
18 MR. FLEMING: And even if you have a 19 vented structure you'll end up with basically a 20 helium-rich --
21 MEMBER MARCH-LEUBA: Because the window 22 closed and did not allow air to come in.
23 MR. FLEMING: But meanwhile, so assume I 24 have a break in the helium pressure boundary if I've 25 lost my coolant. If I cool I'm not going to have any NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
187 1 graphite reaction at all. But if I have the heat up 2 what's going to happen is that my thermal expansion, 3 the helium inside the helium pressure boundary is 4 going to expand and expel helium outside in the 5 reactor building.
6 And later on when the core cools down it 7 will start to bring in a helium-air mixture back into 8 the system. And the graphite phenomena is analyzed as 9 part of the deterministic calculation.
10 So anyway, we're getting into a specific.
11 I think we can --
12 MEMBER MARCH-LEUBA: I didn't want to get 13 into the specific but in my opinion all those purple 14 points should be analyzed only with the safety grade 15 components.
16 ACTING CHAIRMAN CORRADINI: I don't see 17 that. I don't think that was the intent. The intent 18 would be if the -- my interpretation of the process.
19 I could have this wrong.
20 If the uncertainty of the purple -- that's 21 the color. If the uncertainty of the purple points 22 starts essentially encroaching upon 10-4 then I would 23 have to include them. But if I'm sitting down two 24 orders of magnitude lower I see no reason that I would 25 make that assumption as part of the analysis.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
188 1 MEMBER REMPE: And furthermore would you 2 be doing that with the current fleet? We don't do 3 that with severe accidents in the current fleet.
4 You're asking us to take severe accidents and analyze 5 them with only safety -- and it may be even more 6 severe and the frequency would be much lower. So 7 what's the point?
8 MR. FLEMING: Well, the LMP approach when 9 you get to the beyond design basis region there are 10 two points.
11 One is if you happen to have a high 12 consequence BDBE you have to make sure that the 13 reliability of the SSCs that keep it down there are 14 adequate and that gets into a safety classification.
15 It could be safety classified if it's not already 16 there for some other reason.
17 And then the other open question is is it 18 risk-significant or not. If it's risk-significant it 19 could contribute to some of the NSRST.
20 MEMBER MARCH-LEUBA: But you're telling me 21 it's not risk-significant because the non-safety 22 component is working. Let's drop it. I understand.
23 But do you understand. You assume the low safety 24 grade device works.
25 MR. FLEMING: I don't assume it works. 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
189 1 calculate the probability it works, the probability it 2 fails. I look at the consequences when it works and 3 the consequences when it fails.
4 Some LBEs have non-safety-related failing, 5 some have them working. And when we get to the DBE 6 region we're just going to extract out what we're 7 going to call safety-related SSCs and we're going to 8 then calculate the DBA analyses only crediting those 9 SSCs.
10 We're trying to go through a process that 11 gets back into some kind of alignment with the 12 existing licensing process. And I forgot to mention 13 one of the constraints of the LMP framework is to 14 provide a set of licensing events, safety classes and 15 defense-in-depth evaluation that would fit within the 16 current regulatory structure. Because something more 17 than that would require a rulemaking and would take 18 much longer to implement. It wouldn't meet the 19 objectives of the project.
20 MEMBER KIRCHNER: What were the 21 assumptions. I know this is just a specific example, 22 but how was the source term generated?
23 MR. FLEMING: Well, in the -- a lot of 24 this work came from the MHTGR work that was done back 25 in the nineteen eighties. But they had computer codes NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
190 1 and mechanistic source term models that would first --
2 it would basically validate -- it would do the core 3 thermal response for the different accident sequences 4 for steam generator tube rupture events. There would 5 be water ingress and graphite-water reactions to 6 consider.
7 There's a whole mechanistic source term 8 white paper developed for the gas cooled reactors that 9 gets into the physical and chemical processes of 10 those. So that's really what was behind the source 11 terms for most of these.
12 XE-100 hadn't finished their design 13 specific source term calculation so we basically 14 scaled some information.
15 This was a demonstration project. It 16 wasn't a real application. We just wanted to see if 17 it worked and if the design team thought it was a 18 useful way to proceed and the answer was yes.
19 MEMBER DIMITRIJEVIC: Karl, since you're 20 familiar with pebble bed can you give me a couple of 21 examples from each category. Those have sequences, 22 right?
23 MR. FLEMING: The S, M and L are the 24 small, medium and large breaks in the helium pressure 25 boundary.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
191 1 MEMBER DIMITRIJEVIC: Okay, so let's look 2 in small breaks. That's these, right?
3 MR. FLEMING: Yes, it's less than 10 4 millimeters.
5 MEMBER DIMITRIJEVIC: So that small break 6 we have in every category.
7 MR. FLEMING: Yes, that's right.
8 MEMBER DIMITRIJEVIC: So let's say --
9 let's look at this 01, what is that, just small break, 10 nothing else?
11 MR. FLEMING: Nothing else happened.
12 Circulating activity. This had circulating activity.
13 MEMBER DIMITRIJEVIC: And 02 which is in 14 design basis region, what is that?
15 MR. FLEMING: It's probably a loss of 16 forced cooling event combined with a --
17 MEMBER DIMITRIJEVIC: Small LOCA.
18 MR. FLEMING: With a small.
19 MEMBER DIMITRIJEVIC: And then when you 20 have in the beyond design basis events.
21 MR. FLEMING: We have the large breaks 22 which are the L's and then the steam generator 23 scenarios --
24 MEMBER DIMITRIJEVIC: But let's say you 25 have SD03 or something, right. That's a small LOCA NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
192 1 with what now happen.
2 MR. FLEMING: I have to go back to the 3 event trees.
4 MEMBER DIMITRIJEVIC: So I have a little 5 problem that we call these events because they are 6 definitely not events, they are sequences.
7 MR. FLEMING: They're sequences.
8 MEMBER DIMITRIJEVIC: -- small LOCA and a 9 small LOCA is a design basis event. Here the small 10 LOCA --
11 MR. FLEMING: It's an event sequences.
12 They're event sequences. Yes, they're event 13 sequences.
14 MEMBER DIMITRIJEVIC: So why do we call 15 them events. That confused everybody who works on the 16 deterministic side of the thing, you know, that this 17 small LOCA here belong in every category.
18 MR. FLEMING: It would be more accurate to 19 say event sequences. Yes, it would. All the LBEs are 20 event sequences in our approach.
21 ACTING CHAIRMAN CORRADINI: To make sure 22 that we're catching Vesna's point is they're all 23 bundled.
24 MR. FLEMING: They're all grouped.
25 MEMBER DIMITRIJEVIC: They're grouped 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
193 1 different groups.
2 MR. FLEMING: They're grouped in families.
3 MEMBER DIMITRIJEVIC: -- itself which is 4 small LOCA. The event as well.
5 MR. FLEMING: And Vesna, there is a public 6 domain report where you can actually get detailed 7 answers to all the questions, what is this sequence, 8 where is it in the event trees. It's all mapped out 9 in the report, the public domain report.
10 MEMBER DIMITRIJEVIC: Right.
11 MR. FLEMING: I didn't memorize --
12 MEMBER DIMITRIJEVIC: No, no, no, I know.
13 It's a bunch of the sequences. I just want to point 14 out that --
15 MR. FLEMING: They're all event sequences.
16 MEMBER DIMITRIJEVIC: -- looking in 17 combination of systems and events.
18 MR. FLEMING: Absolutely.
19 MEMBER DIMITRIJEVIC: Right. So they're 20 calling this event, it's not an event anymore. It's 21 an actual sequence --
22 MR. FLEMING: They're event sequences.
23 MEMBER DIMITRIJEVIC: It goes beyond -- it 24 will go beyond containment for light water reactor 25 here. I assume there is some building containment.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
194 1 So it is a total different, our understanding how is 2 treated is totally different because if you come to 3 the chapter 15 that doesn't apply. And this will not 4 apply --
5 MR. FLEMING: Fifteen is also sequences.
6 It says large LOCA, loss of offsite power.
7 MEMBER DIMITRIJEVIC: But that's a 8 different assumption they make for large LOCA. For 9 small LOCA they would not have four different 10 categories.
11 MR. FLEMING: Well, in the word 12 definitions we have them. In the guidance document 13 it's very clearly stated that all the LBEs are event 14 sequences. So that's our intent.
15 MEMBER DIMITRIJEVIC: I just wanted to 16 make sure --
17 MR. FLEMING: I'm sorry for the confusion.
18 MEMBER DIMITRIJEVIC: -- all understand 19 that that's what we've got.
20 MEMBER REMPE: So to follow along on that 21 they're really though, they're grouped. Like medium 22 break LOCAs are something bigger than such --
23 something else. It's a group of sequences that are in 24 that event category. Or whatever.
25 MR. FLEMING: The initiating event, 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
195 1 this case there's basically four initiating events 2 that shut off. They're small, medium and large breaks 3 in the helium pressure boundary as initiating events 4 and those are defined by ranges. Ten millimeter, up 5 to 10 millimeter small, 10 to 65 millimeters, the size 6 of a fuel pipe is medium, greater than that's a large.
7 So that's the way we define it.
8 But then every LBE is a family of event 9 sequences as Vesna points out that has the response of 10 the plant all the way out to the source term.
11 MEMBER REMPE: Characterized by that 12 sequence.
13 MR. FLEMING: That's right.
14 MEMBER REMPE: Also, why are there no 15 ATWS?
16 MR. FLEMING: They're in there. They're 17 in there. For the MHTGR ATWS has no adverse 18 consequence. The reactor shuts down on negative 19 temperature coefficient.
20 MEMBER REMPE: And the same is true for 21 this XE-100.
22 MR. FLEMING: Yes. ATWS you don't have 23 like a pressure spike and a challenge to your reactor 24 coolant system. It doesn't in and of itself create 25 any different dose. I mean, we track them, we model NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
196 1 them, we calculate them.
2 MEMBER REMPE: It's just not showing up.
3 Okay.
4 MR. FLEMING: And we put them in different 5 families because we know there's interest to keep 6 track of that. But they don't jump out on a frequency 7 consequence chart at all. For high temperature 8 reactors. And for a lot of these reactors.
9 MEMBER KIRCHNER: So Karl, just walk 10 through if you would, please, what you do when you 11 have things on the cusp or on the margin. Let's just 12 pick one. Steam generator I assume that says steam 13 generator tube rupture 18 sitting right there at 1 14 times 10-4.
15 But it doesn't matter. I'm not asking the 16 specifics of the design or anything. You have 17 something that lies close to that line. What happens 18 next?
19 MR. FLEMING: Well, okay. So we have 20 rules for how we process each of the three regions, 21 AOOs, DBEs and BDBEs.
22 And we also not shown here will address 23 the uncertainties on the frequency and the dose. And 24 when we have a straddle situation or something comes 25 really close to the line we'll evaluate it on both NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
197 1 sides of the line. In other words we're not going to 2 get into one of these gee, if I can multiply by 0.98 3 where can I find a 0.98 factor and get below. We 4 don't play those kinds of silly games if you will.
5 So we consider the uncertainties and if 6 we're close to the boundary we'll evaluate the event 7 as though it's either a DBE or a BDBE and apply the 8 rules. So we're not sensitive to the line in the sand 9 problem.
10 MEMBER KIRCHNER: Three decimal points.
11 MR. FLEMING: Right. May we go on?
12 MEMBER KIRCHNER: But the assumption here 13 is with your mechanistic source term you're not 14 assuming a significant failure in the case of the 15 HTGR.
16 MR. FLEMING: Well, we're trying to model 17 the actual phenomena that would dictate either the 18 retention or release of radioactive material and how 19 much.
20 I'm going to say a few words about the GE 21 demonstration.
22 MEMBER KIRCHNER: Let me belabor this a 23 little bit because it's an important point. This 24 particular example benefits certainly from the amount 25 of effort that was invested in the MHTGR which NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
198 1 included a lot of experimental work. Do you sense 2 that your colleagues understand the challenge that's 3 in front of them for developing mechanistic source 4 terms? Versus making an assumption in the LWR 5 business is significant failure and proceed from 6 there.
7 MR. FLEMING: I think that's a good 8 question for some of our developers.
9 MR. REDD: I think that's an excellent 10 question regarding mechanistic source term. But I 11 think it also puts us on kind of a point Karl brought 12 up earlier that the uncertainties are there regardless 13 of whatever approach we take, but we haven't found --
14 at least from the work we've done we haven't found 15 these uncertainties insurmountable or anything like 16 that.
17 Yes, experiments might need to be done to 18 help inform your decisions but if you cycle through 19 the LMP process and you find that your uncertainties 20 even if they are large but that you're still okay with 21 that range of uncertainties then maybe you can 22 demonstrate that you don't have to have such an 23 extensive experimental program. You could actually 24 use it as a justification for not doing some work to 25 reduce uncertainties.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
199 1 So it depends a little bit on the specific 2 case but I think at least having the LMP structure 3 there provides a way to prioritize uncertainties, 4 especially in an area like mechanistic source term 5 where there could be uncertainties all over the place.
6 Some you might be able to live with and some you might 7 not be able to.
8 MR. REDD: Was that responsive to your 9 question, sir?
10 MEMBER KIRCHNER: Actually I was making a 11 statement, a cautionary statement. Just as a designer 12 in the past I would just submit that when the 13 uncertainties are large you design robustness into the 14 design.
15 I hope that the DID, the defense-in-depth 16 process, that would be a result that would come out 17 that you would go back.
18 And there are cliff effects for all these 19 designs that it's not just uncertainty in the sense 20 how uncertain I am about my calculations. There are 21 real cliff effects for -- I won't enumerate them, but 22 for each of the designs on your table that get you 23 into -- could you get into a situation where you have 24 significant release.
25 MR. FLEMING: Right. And just to confirm NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
200 1 it is something that's looked at very carefully as 2 part of the defense-in-depth adequacy evaluation.
3 MEMBER DIMITRIJEVIC: I have one other 4 concern. Let's say that we apply this to light water 5 reactor existing fleet which we have most information.
6 Let's not even talking advanced light water reactor.
7 Every point will correspond to the release 8 category sequences, right?
9 MR. FLEMING: Well, also as I mentioned 10 earlier we want to capture the no release sequences 11 and understand why we don't have a release. That's 12 fundamental to --
13 MEMBER DIMITRIJEVIC: All right.
14 MR. FLEMING: In the GE PRISM exercise 15 they started out with a PRA that was focused on the 16 traditional reason for doing a PRA finding the risk of 17 severe accidents.
18 MEMBER DIMITRIJEVIC: I understand that.
19 MR. FLEMING: And they had to put more 20 emphasis on the success states to do this process.
21 But anyway.
22 MEMBER DIMITRIJEVIC: But I am more 23 interested in release. So this will correspond to the 24 risk category that is hundreds and hundreds of those 25 sequences. If we separate them like this, 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
201 1 based on initiator and where they belong we may 2 satisfy this all the time. But when you sum them the 3 large release may not satisfy.
4 MR. FLEMING: That's why we have the 5 cumulative -- we keep showing this slide but this is 6 only used to look at the risk significance of 7 individual licensing event families.
8 We also have our cumulative risk targets 9 for the QHOs --
10 MEMBER DIMITRIJEVIC: Those have some --
11 MR. FLEMING: Where we aggregate for three 12 different metrics. One based on a Part 20 to look at 13 high frequency low dose scenarios and the two QHO 14 metrics. So we sum for those.
15 MEMBER DIMITRIJEVIC: So my -- one other 16 comment because we discussed this this morning. This 17 curve it's not practical to apply to existing or large 18 light water reactor. I mean, you know. That's one 19 insight. I don't see what would be point.
20 MR. FLEMING: The purpose of applying this 21 to a light water reactor was to revisit what are the 22 design basis accidents.
23 MEMBER DIMITRIJEVIC: Well, but you're 24 talking design basis accident versus design basis 25 sequences. So we are changing nature 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
202 1 regulation. That's completely different issue. It's 2 a qualitative -- quantitative jump, so --
3 MR. FLEMING: I'll say a few words about 4 the GE demonstration. Gary Miller is here and also 5 David if there's questions on some of these aspects.
6 I'll sort of summarize this.
7 This is an example of an application after 8 already developing a design back in the late nineteen 9 eighties and then taking advantage of this 10 modernization of the PRA project that was sponsored by 11 DOE a few years ago. There's a public domain report 12 on that by the way.
13 One of the features of the PRA that went 14 into this is that they demonstrated the ability to 15 meet our PRA standards requirements for putting 16 together a component reliability database for a new 17 kind of reactor. Also for demonstrating passive 18 component reliability which is really -- it's 19 primarily an uncertainty analysis in the phenomena 20 that are responsible for the passive heat removal 21 features and so forth, and also the mechanistic source 22 term.
23 The PRA standard goes all the way out to 24 consequence analysis, radiological consequence 25 analysis and has separate requirements for mechanistic NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
203 1 source term. So they demonstrated the ability to meet 2 those requirements in their original PRA.
3 As I mentioned earlier the work they did 4 on the PRA modernization really had a big influence on 5 how we defined our risk significance criteria for the 6 LMP process.
7 So more recently since the PRA was 8 completed we did a demonstration that GE-Hitachi 9 performed and this went on for a number of months of 10 education process. We gave them a training program on 11 the LMP process and they came back with questions and 12 insights.
13 It culminated in not a public but a 14 meeting that was attended by the NRC staff and some of 15 the other advanced reactor developers just a few weeks 16 ago.
17 As I understand it a public domain report 18 will eventually be available on this exercise.
19 So, these are the steps that they 20 performed. They processed event sequence families 21 from their PRA into AOOs, DBEs and BDBEs. They did 22 sensitivity studies to derive what the required safety 23 functions were. And I'll let Gary speak to the 24 specifics there.
25 They were able to come up with 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
204 1 classification of not only safety-related but non-2 safety-related with special treatment SSCs based on a 3 defense-in-depth adequacy evaluation of their plant 4 capabilities.
5 They were able then to formulate what 6 their DBAs would look like following this process.
7 The NSRST SSCs benefitted from the plant capability 8 defense-in-depth evaluation.
9 Gary, would you like to elaborate a little 10 more?
11 MR. MILLER: As you said you described the 12 process, the steps that we went through. This was 13 limited in scope. It was internal events at power.
14 And particularly as we went down the line in looking 15 at defense-in-depth we looked solely at the heat 16 removal function. So we did a good deep dive but it 17 wasn't broad because scope, resources and things like 18 that.
19 And as has been mentioned here before we 20 found a lot of areas where the methodology made sense.
21 We learned a lot. We added in describing how but not 22 necessarily changing the methodology. I think it was 23 pretty sound.
24 MR. FLEMING: This was the safety-related 25 SSCs on the top part of this slide that they 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
205 1 with for performing the required safety functions 2 which basically was controlled core heat removal, 3 controlled heat generation. By doing those functions 4 for this reactor they assure the retention of 5 radionuclides in the fuel.
6 And also a scope limitation was that they 7 only looked at the source of radioactivity in the 8 fuel. They didn't look at all the sources of 9 radioactivity. They looked at some other sources but 10 not all of them. Do you want to elaborate further?
11 MR. MILLER: You said it well, Karl.
12 MR. FLEMING: That's all right. I meant 13 to tee up something for you that I forgot to. They 14 found that none of their non-safety-related SSCs were 15 risk-significant. So zero risk-significant SSCs. But 16 applying the defense-in-depth adequacy criteria and 17 particularly focusing on table 5-2 in the guidance 18 document that talks about adequacy of plant capability 19 defense-in-depth they came up with four additional 20 items on here that would be examples of what could be 21 added to NSRST SSCs.
22 MR. MILLER: We used the frequency 23 consequence plot quite extensively to look at not only 24 as you mentioned before borderline cases where we can 25 do sensitivity studies and evaluate taking a component NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
206 1 out of service one at a time, and then looking at the 2 resulting plot to see if it was in the DBE region or 3 in the cross hatched region.
4 And as Karl said when we took it one by 5 one on each component none of them made it into the 6 cross hatched region based on that criterion.
7 However, because this is a very integrated process we 8 went through the defense-in-depth process and asked 9 those questions.
10 We did in fact come up with what you see 11 there, the four areas of functions that did have some 12 safety significance. So those were NSRST.
13 ACTING CHAIRMAN CORRADINI: So, help me --
14 so this to me is important. So you used the technique 15 which I am still fuzzy about to use defense-in-depth.
16 What did you exactly do? Can you go back to that 17 slide? The four systems, the steam generator shell 18 and tube design. I still don't -- can you help me 19 what those four systems are and how they work?
20 MR. MILLER: Okay. If you follow the 21 process and the detailed steps there's one step, I 22 think it's layer two, but whatever layer it is it 23 talks about what equipment do you need to maintain 24 this within the DBE region. Not make it worse.
25 So you kind of look at one at a time what NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
207 1 would happen.
2 If we took certain equipment out of 3 service would it make it worse. In that case what we 4 found was that in general our heat removal is 5 adequately covered by the reactor vessel auxiliary 6 cooling system, that RVACS.
7 When you look at defense-in-depth in the 8 scope that the methodology mandates you determine that 9 not only that but the backup function would be having 10 to do with forced air cooling and along with that the 11 natural circulation from your intermediate heat 12 transfer system, and from there your steam generator 13 tubes and shell.
14 ACTING CHAIRMAN CORRADINI: Okay. So let 15 me say it back to you so I make sure I understand. So 16 your point is RVACS is safety significant and 17 therefore at a one at a time application the 18 intermediate heat transfer system was not important 19 and the shell and tube steam generator wasn't 20 important. But if you took them as a combination they 21 provided a defense-in-depth to the RVACS or vice 22 versa.
23 MR. MILLER: Yes.
24 ACTING CHAIRMAN CORRADINI: Have I got it 25 approximately 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
208 1 MR. MILLER: Yes.
2 ACTING CHAIRMAN CORRADINI: Okay. So then 3 because of that they would appear as -- they would be 4 treated as non-safety treatment of --
5 MR. MILLER: Non-safety-related with 6 special treatment.
7 ACTING CHAIRMAN CORRADINI: I want to say 8 RTNSS, but I'm not allowed to say RTNSS.
9 MR. FLEMING: Yes, non-safety-related with 10 special treatment. NSRST.
11 ACTING CHAIRMAN CORRADINI: Okay, fine.
12 Thank you. And then what is SWRPS?
13 MR. MILLER: That is sodium water reaction 14 protection system.
15 ACTING CHAIRMAN CORRADINI: So you're 16 looking for sodium leakage?
17 MR. MILLER: Yes. In the steam generator.
18 MR. FLEMING: From the intermediate.
19 ACTING CHAIRMAN CORRADINI: Are these 20 double walled steam -- I'm sorry to get to details but 21 it matters. Are these double walled steam generator 22 tubes where you have the helium gap that you're 23 monitoring the helium gap?
24 MR. MILLER: No, that was not -- the 25 design was not a double wall.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
209 1 ACTING CHAIRMAN CORRADINI: So what is 2 that? Is it a pressure measurement? How do I detect 3 it if I'm not tracking some sort of intermediate layer 4 in the steam generator tube?
5 MR. MILLER: Go ahead.
6 MR. GRABASKAS: Typically it monitors for 7 hydrogen production up in the top of the steam 8 generator.
9 ACTING CHAIRMAN CORRADINI: Okay, fine.
10 So it's a hydrogen sampling system. Okay, thank you.
11 MR. FLEMING: Would you back up a slide?
12 I wanted to also mention that on the third bullet down 13 here questions that often come up, how do you deal 14 with passive component reliability. Another question 15 that will come up is you don't have any experience, 16 how are you going to develop a database. And then 17 what about mechanistic source terms.
18 So I'm going to have David say a few words 19 about that. We actually have some public domain 20 papers out there on this. What it's trying to do here 21 is meet the requirements in our non-light water 22 reactor standard for these activities.
23 MR. GRABASKAS: It's interesting. I 24 mentioned that we kind of foresee these as issues but 25 it really goes back to the nineteen eighties and 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
210 1 PSID of PRISM and the NRC review.
2 If you look in NUREG-1368 kind of three 3 big issues the NRC called out with the PRISM PRA were 4 a simplified optimistic look at passive system 5 reliability, lack of a detailed treatment of source 6 term and then also questions about the component 7 reliability database.
8 So that was part of the reason why we had 9 been focusing so much on that at Argonne, kind of 10 developing methodologies. But then also with the new 11 non-light water reactor PRA standard developing 12 methodologies that also meet the requirements of the 13 standard.
14 Because the standard can be really strict 15 in some of these areas, for example with passive 16 system reliability. It's a requirement in the 17 standard that you need to mechanistically model the 18 response of the passive systems but also using models 19 that have been empirically validated through 20 experimentation too.
21 So it's really quite a strong step in this 22 PRISM PRA update and the LMP really gave us a good 23 chance to demonstrate or actually run through the 24 methodologies we had developed and we've refined them 25 because of the lessons learned because of 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
211 1 Same with mechanistic source term too. We 2 had come up with the methodologies but this was a good 3 chance to actually apply them and go through the 4 actual research and do the analyses.
5 MR. FLEMING: Thanks a lot.
6 CHAIRMAN BLEY: This is Dennis Bley. Two 7 things. The first is beneath the slide of -- papers 8 Karl referred to and you referred to and the other 9 issue is on the passive component reliability at least 10 to my thinking it's not so much component reliability 11 on passive systems as it is potential degradation over 12 time of the passive functions because some are fairly 13 delicate balances. Did that paper go into a 14 discussion of that area as well?
15 MR. GRABASKAS: Yes. You're absolutely 16 right and I can provide a list of references. We have 17 a whole bunch of public Argonne reports on mechanistic 18 source terms but also the passive system reliability 19 approach too.
20 And you're right, that's the real big 21 tradeoff with passive systems is yes, you're running 22 on inherent phenomena but then your driving force 23 instead of being megawatt powered pumps is now just 24 buoyance differences and things like that. So 25 properly characterizing those differences has a big NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
212 1 effect.
2 But yes, I'll provide a whole list of the 3 open source references.
4 CHAIRMAN BLEY: Okay, and do that through 5 our staff at ACRS. Thank you.
6 MEMBER REMPE: And the tools you used for 7 the mechanistic source term were validated based on 8 EBR -- it's a metal fuel reactor, right. So it's EBR2 9 data?
10 MR. GRABASKAS: A couple of different 11 tools we used. Depending on what the tool did we 12 validated different ways. But on EBR2 data 13 experimentation. Unfortunately in the SFR world we 14 also have some past accidents that we were able to 15 pull data from too.
16 But there are other tools that we actually 17 developed ourselves by demonstrating that the 18 importance was low for the outside consequences we 19 didn't have to validate to an extent that we might 20 have to validate other codes too.
21 PARTICIPANT: It would be helpful if 22 people that are asking questions would use the 23 microphones.
24 MR. FLEMING: And to wrap up the GE PRISM 25 part of this show this was some of Gary's thoughts NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
213 1 about their feedback on the process.
2 MR. MILLER: Okay, to wrap it up we did 3 find that it was very systematic and repeatable 4 although it may have seemed like it was advertised up 5 front. We did actually find that out.
6 It's pretty clear when a process step is 7 complete as we went through the methodology. I say 8 that sensitivity studies are easy to perform but to 9 get there it was a lot of work. Setting up the logic, 10 the file structure, quantifications and all that, it 11 was quite a lot of work to get there. But once you do 12 then you'd have a very easy way of doing a lot of 13 sensitivity studies.
14 And this comes in handy in a lot of these 15 steps later on as well as in tradeoff studies that you 16 might have later on down the road.
17 And then the results are traceable to key 18 risk and performance drivers. If you're familiar with 19 event sequences and cut sets I think you know it's 20 easy to go back and look at what are the drivers, what 21 are the dominant failures and come back to the risk 22 and performance drivers.
23 Another thing as a developer we appreciate 24 -- it's more visual. It's more meaningful than 25 talking about very low frequency numbers because 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
214 1 you know you lose interest right away. It's not 2 relatable.
3 Where we can show an FC plot with a point 4 or a group of points and then we can vary those based 5 on sensitivity studies it's much more meaningful and 6 it's much more relative. You do a sensitivity study, 7 you look at how much it moved, it's very clear to the 8 people.
9 And then iterative. Of course again as a 10 developer I think in the early design phase with a 11 conceptual design and a conceptual PRA there are a lot 12 of uncertainties and assumptions and we document 13 those. And we get to the point where something may be 14 on the line or something may have a very high 15 uncertainty distribution and that gives us a lot of 16 options. We can look at design changes or 17 programmatic changes there as well.
18 So we iterate that into the design and 19 then we update the model of course.
20 And overall it just clarifies a path to 21 regulatory engagement.
22 MR. FLEMING: Thanks a lot, Gary.
23 MEMBER MARCH-LEUBA: Just a question for 24 clarification. If a component is non-safety grade 25 according to this do you need to do seismic analysis NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
215 1 of it? I'm asking specifically about the steam 2 generator in PRISM. If it's non-safety grade you 3 don't have to do the seismic for it because it can 4 fail.
5 MR. FLEMING: As part of our process after 6 we defined the required safety functions and our 7 safety-related SSCs once you've selected your external 8 hazard levels for your external events then there's a 9 requirement, it's an implied requirement that you have 10 to protect all of your safety-related SSCs so that 11 they would be able to perform the required safety 12 function in the event of an external event.
13 MEMBER MARCH-LEUBA: According to your 14 methodology --
15 MR. FLEMING: And other non-safety-related 16 components would have to be protected like the seismic 17 two over one and those types of issues come into play.
18 So there are ways for seismic requirements to creep 19 into the non-safety-related area through that pathway.
20 MEMBER MARCH-LEUBA: Okay. I'm just 21 surprised that when you apply the methodology it came 22 out that your steam generator is not safety grade.
23 It's not safety component. Steam generators are 24 things that fail everywhere and that's the thing that 25 separated your sodium from your water. I just cannot NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
216 1 believe it came out no, we don't need it. I can't 2 believe it.
3 MR. MILLER: It's an advanced reactor 4 passive. There's a lot of thermal capacity in the 5 sodium. In reactivity --
6 MEMBER MARCH-LEUBA: If it breaks you have 7 a fire.
8 MR. FLEMING: Let's see. I think we 9 should go on to the next and final part of our 10 demonstration activity having to do with the molten 11 salt reactor experiment. And Steve Krahn is with us 12 to amplify on this.
13 There's a couple of different activities 14 that have been done. There's a report indicated on 15 the right, an Oak Ridge National Laboratory report and 16 a chart in the center here which identifies some of 17 the processes that they're going through.
18 The report on the right is an example of 19 taking the technology they've been collecting and 20 analyzing for the molten salt reactor experiment and 21 building a PRA model using the guidance that's in the 22 PRA white paper and then summarized more briefly in 23 the guidance document.
24 The diagram in the middle identifies a 25 process for performing process hazards analyses given NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
217 1 the fact that these molten salt reactors resemble more 2 of a process plant than a standard or a typical type 3 of power generation reactor facility. And they're 4 using a HAZOPs technology to build the knowledge base 5 to build a PRA model and a deterministic safety 6 analysis model for the MSRE.
7 Steve, would you like to amplify on that 8 a bit?
9 MR. KRAHN: I'll also loop back and 10 discuss the source term question which was asked 11 earlier because obviously that's a primary concern.
12 And also if you look at the dates on these reports 13 we're looking at early work in process. So I would 14 also state that up front.
15 The source term in the molten salt reactor 16 experiment was similar to most molten salt reactors is 17 split up into three large sections. The vast majority 18 of the radioactive material is in the salt itself.
19 There's also radioactive material in the offgas system 20 because the offgas system is continuously hooked up to 21 most of the -- is continuously hooked up to all of the 22 primary plants that I've seen.
23 And then finally there is some means in 24 place to either polish or chemically clean the salt.
25 That is the third major source of radioactive NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
218 1 material.
2 The characterization of those three 3 radioactive material inventories is in the joint 4 Vanderbilt-Oak Ridge technical report on the right 5 which kind of started this effort about two years ago, 6 a joint effort with Oak Ridge and Vanderbilt.
7 And one of the things that that showed was 8 the hazard analysis for the molten salt reactor 9 experiment was a very limited scope and very focused 10 on what was going on just in the salt system. So one 11 of the conclusions of the report was the need to do a 12 broader hazard assessment that took into account all 13 of the other potential radioactive material sources as 14 well.
15 So that has been worked on in parallel 16 with an Electric Power Research Institute project that 17 is working to document the process to move from early 18 stage safety analyses such as HAZOP analysis, such as 19 failure modes and effects analysis through to 20 probabilistic risk assessment. So that's where those 21 two projects are being funded from.
22 If we go to the next slide I can walk 23 through what we've learned to date and I'll expand a 24 bit on this summary.
25 The MSR lack any significant PRA legacy.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
219 1 So we're basically starting with a clean sheet of 2 paper to look at what a PRA for a molten salt reactor 3 would look like.
4 That's why we -- after completing the case 5 study on the molten salt reactor experiment that was 6 documented in the Oak Ridge technical report we have 7 dropped back to do a comprehensive hazard assessment 8 of the molten salt reactor experiment.
9 That effort is winding down now. We've 10 completed HAZOP studies on four major systems. One of 11 those has gone through peer review. The other three 12 are going through peer review right now.
13 And the next stage -- one of the things 14 that that showed us was that the HAZOP is amenable to 15 providing the quote "comprehensive" hazard analysis 16 that's desired by standards like the non-LWR PRA 17 standard.
18 It also though supports early stage safety 19 analysis providing insights back to the design team 20 and allows preliminary modeling to be done for 21 probabilistic risk assessment.
22 It also supports ready identification of 23 potential risk important initiating events. And 24 that's the parameter or the outgrowth of HAZOP that 25 we're using to move forward to the next stage which 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
220 1 going to be quantifying event trees, at least one 2 major event tree for each of these radioactive 3 material inventories in the molten salt reactor 4 experiment.
5 A couple of more lessons learned on this 6 early stage safety analysis for the molten salt 7 reactor experiment is it is valuable for providing 8 near term design and operability information. One of 9 the things we identified in the Oak Ridge technical 10 report was where based on their simplistic -- simple, 11 I don't want to say simplistic. Simple hazard 12 analysis in the mid-nineteen sixties they had 13 identified five operating modes for the reactor.
14 In our detailed review of their operations 15 report it turns out that there were really closer to 16 seven or eight operating modes that they used on a 17 regular basis. That would have allowed a much more 18 nuanced understanding of what their probabilistic risk 19 assessment would look like.
20 And then we also identified -- one of the 21 other things that the early HAZOP analysis does is, 22 and I think some of the members have been pointing out 23 this important factor, is it points out the need for 24 additional analyses.
25 Early on it shows things that we 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
221 1 need and lets us evaluate whether or not they need an 2 experimental program to be addressed or whether they 3 can be addressed by deterministic analyses.
4 For example, one of the ones that we are 5 in the middle of on the program with EPRI is looking 6 at freeze valves. Every molten salt reactor design 7 you look at uses freeze valves and they show up on 8 schematic diagrams looking just like a standard gate 9 or globe valve but they are in fact a pretty complex 10 combination of an air system, an I&C system to 11 maintain the temperature of that freeze valve and 12 continue to maintain its isolation function or when 13 demanded melt and allow the molten salt to leave the 14 reactor.
15 That identification was done by going 16 through the HAZOP study for the molten salt reactor 17 experiment and with some support from Southern Company 18 we're now in the process completing a failure modes 19 and effects analysis for the important component that 20 freeze valves are.
21 The next steps on this front are we're 22 working with Karl and Amir to look at how we would 23 characterize and move forward to do licensing basis 24 event identification and safety-significant component 25 identification and potentially if we don't run out 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
222 1 time before now and the middle of March trying to do 2 some DID assessment as well. So that's where we are 3 on molten salt reactor work. I'm happy to answer any 4 questions.
5 ACTING CHAIRMAN CORRADINI: So you chose 6 what's called the MSRE?
7 MR. KRAHN: Correct.
8 ACTING CHAIRMAN CORRADINI: Because there 9 was enough information. What about some of the 10 current conceptual designs?
11 MR. KRAHN: So it wasn't the only criteria 12 we used to select the MSRE. The other major criteria 13 is that not only was there enough design information, 14 it was all publicly available and not covered by 15 intellectual property. So it was a quick way to get 16 things into the public domain.
17 MEMBER REMPE: But we were told I believe 18 earlier you're going to have a Kairos evaluation 19 coming soon.
20 MR. FLEMING: Yes. We're just in the 21 beginning stages of putting together a Kairos 22 demonstration and also a micro reactor eVinci that 23 Westinghouse is developing. So those are on the books 24 and we're launching off to get those completed by the 25 spring of 2019.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
223 1 MEMBER REMPE: The public information 2 question. Are all of these demos publicly available 3 documents, or at least available to ACRS? Earlier 4 Mike had asked for one and you said yes, we'll get you 5 that document.
6 MR. FLEMING: Well, the two that have been 7 completed, the GE PRISM and the XE-100 will be --
8 well, XE-100 is available now and GE PRISM will be 9 available in the near future. They're preparing it 10 now.
11 MEMBER REMPE: Thank you. Go ahead.
12 CHAIRMAN BLEY: This is Dennis. I was a 13 little surprised on the discussion of the MSRE that it 14 focused on kind of starting from nuclear power plant 15 PRAs and that this was so different. There have been 16 very, very many chemical processing plant PRAs that it 17 kind of follows the way you described it, so it would 18 have the (telephonic interference) probabilistic 19 hazards and how it -- and eventually to the PRA.
20 Did you look at what's been done on the 21 chemical process industry in any depth?
22 MR. KRAHN: Yes, the HAZOP process which 23 does the initial qualitative hazard identification 24 work we took directly out of the chemical processing 25 industry. It is the standard for doing the initial NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
224 1 stages of hazard assessment and event sequence 2 identification in chemical processing plants.
3 We will then move on to doing PRA using 4 the LMP structure. But we definitely took all 5 advantage that we could from chemical processing 6 industry experience.
7 CHAIRMAN BLEY: Okay, that makes sense and 8 it isn't a great departure when you think of it from 9 that point of view. Thank you.
10 MR. FLEMING: Yes, to amplify on Steve's 11 answer to Dennis's question being a consultant to 12 their project we helped them put together a body of 13 knowledge of prior work that would be relevant to 14 supporting the project.
15 Among the many things that we looked at 16 there was in fact a PRA done on the low activity waste 17 facility at Hanford. It's part of their vitrification 18 facility that was developed not only to look at 19 radiological event sequences but also toxicological 20 event sequences.
21 And that provided some inputs in the 22 knowledge base report.
23 CHAIRMAN BLEY: Okay, thanks.
24 MR. FLEMING: Thank you very much, Steve.
25 So coming -- this sort of concluding our technical NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
225 1 presentation today we come back to these questions 2 that are the LMP process was designed to address what 3 are the initiating events, event sequences and so 4 forth. How does the design and the SSC respond to the 5 event sequences. What kind of margins do we have in 6 the response. And how the defense-in-depth philosophy 7 is implemented.
8 We give you a lot of examples of different 9 applications at different levels of development so 10 far. And if we have any more questions we'd be glad 11 to answer them.
12 ACTING CHAIRMAN CORRADINI: Committee 13 questions? Okay. At this point let's take a break 14 because there was none shown in the agenda but we need 15 a break. So we'll come back at 10 after 3.
16 (Whereupon, the above-entitled matter went 17 off the record at 2:54 p.m. and resumed at 3:09 p.m.)
18 ACTING CHAIRMAN CORRADINI: Okay, why 19 don't we get started. Everybody settle down so we can 20 have Herr Reckley lead us through this portion.
21 MR. RECKLEY: Okay, so to close out we 22 wanted to go through the draft Commission paper and 23 the draft regulatory guide because as I mentioned this 24 morning in the end this is what the staff is producing 25 and it's what we would be asking the ACRS to comment NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
226 1 on, realizing it's inseparable from NEI 18-04 because 2 that's what we're proposing to endorse.
3 Before I get started though as personal 4 soapbox I guess, they gave me the microphone these 5 processes that have been described and as you're going 6 to see the staff is comfortable with there's a couple 7 of points to point out here I think.
8 One, just because you can define some flow 9 charts and processes for what needs to be considered 10 doesn't mean that we think that this is simple. The 11 development of a mechanistic source term with the 12 modeling of specific radionuclide groups across 13 barriers, which ones -- if you're talking about molten 14 salt which ones will stay in the salt, which ones will 15 escape the salt. Then for the ones that escape how 16 will they either be retained or escape from a 17 particular barrier. That's a complex physical 18 question.
19 We can model this out and say yes, the 20 developers need to do A, B and C and we're comfortable 21 saying that. At the same time we're not implying one, 22 that it's been done in all cases, and two, that it's 23 particularly easy in any case. So I just wanted to 24 lay out that as we lay out these processes we didn't 25 want to confuse the ability to define a process with NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
227 1 the fact that the science still needs to be done, 2 still needs to be proven.
3 Or at least the uncertainties in the 4 science need to be accounted for. And that's what 5 Karl was talking about in much of the assessments in 6 some cases, how do you address the uncertainties that 7 might exist for some of these designs.
8 So with that I'll get right into the two 9 documents that the staff provided along with the 10 working draft of NEI 18-04.
11 The first one was a Commission paper. The 12 staff's view is that although much of this has been 13 brought before the Commission before that time period 14 is measured in decades. And you can see that we made 15 great strides. One paper was followed by another 16 paper albeit that paper was 10 years after the first 17 one.
18 And so this will be really the first time 19 that the process has been consolidated and applied or 20 available in a relatively integrated approach that we 21 want to bring before the Commission and say although 22 we think almost everything in here you've accepted in 23 previous papers from the nineties or the early two 24 thousands this is the result of actually applying 25 those decisions and what it looks like in a 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
228 1 And we thought the Commission would want to see that 2 and have a shot at either saying yes, that's working 3 the way that was envisioned or not.
4 So the paper as it's defined here is to 5 seek the Commission approval. And it's divided into 6 a standard format. In enclosure 1 it gives the 7 background. I won't talk a lot about that one.
8 And then enclosure 2 which summarizes this 9 approach really from NEI 18-04 and puts it in the 10 context of where we think there are references to 11 previous Commission decisions and where there might be 12 in one or two cases a remaining unanswered question 13 that this would provide the vehicle for the Commission 14 to answer.
15 So going in to the background this is very 16 similar to a slide I had earlier this morning. It 17 does start with the Advanced Reactor Policy Statement.
18 Whereas we don't assume any particular 19 design at this point can make it through the process.
20 We're not pre-judging the ability of any design and 21 how it would turn out we are assuming that the 22 Advanced Reactor Policy Statement defines attributes 23 of advanced reactors and we're assuming that there is 24 an ability to design a reactor that has those 25 attributes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
229 1 And what that assumption gives us is the 2 ability to go forward without a particular design.
3 So that's supported by some of our 4 previous interactions like the pre-application 5 evaluations that were mentioned on PRISM on MHTGR.
6 The SECY paper 93-092, the Commission made a few steps 7 in the direction that we're currently in but there are 8 also some differences in what was proposed in 1993 and 9 what's being proposed now.
10 SECY-03-0047 was the closest and that 11 probably makes sense. That was at the time when some 12 other gas reactors were being proposed and we were 13 interfacing with both developers, the Department of 14 Energy and others.
15 And so in SECY-03-0047 they proposed some 16 policy issues to the Commission or some resolution of 17 policy issues that are directly applicable to today.
18 And again it's not surprising because as 19 Karl mentioned this methodology has been evolving 20 since the eighties starting with the MHTGR.
21 At the same time as I mentioned this 22 morning the related initiatives on risk-informed 23 performance-based regulation and those were largely 24 incorporated into the proposals and the policy issues 25 that were communicated to the Commission both during NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
230 1 the development of the licensing strategy for NGNP and 2 actually in SECY-03-0047.
3 So the three big bullets from SECY-03-0047 4 that I want to mention was that the staff asked 5 specifically and the Commission stated in its staff 6 requirements memorandum their approval of these three 7 things which is that a greater emphasis can be placed 8 on the use of risk information and the use of 9 probabilistic risk assessments to identify events --
10 and here's the balancing of that -- provided there's 11 sufficient understanding of plant and fuel performance 12 and that deterministic engineering judgment is used to 13 bound uncertainties. So that's a general consensus of 14 a risk-informed performance-based approach using a mix 15 of risk-informed insights and deterministic 16 assessments including engineering judgment where 17 necessary.
18 The second is that a probabilistic 19 approach for safety classification SSCs is allowed.
20 And the last bullet there, that the single 21 failure criterion can be replaced with a probabilistic 22 reliability criterion.
23 So now the paper is organized into the 24 three primary elements of the methodology, event 25 selection and analysis, SSC classification 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
231 1 performance criteria, and defense-in-depth 2 assessments.
3 And the key points in the paper are that 4 we think this process is consistent with that 5 recommendation and Commission approval from SECY 6 0047 to use a probabilistic approach to identify 7 events and to back that up with deterministic and 8 engineering judgment.
9 One thing that wasn't specifically 10 addressed in previous papers and that is that as 11 you'll notice on the frequency consequence target 12 figure there is a lower frequency range and that is 13 often interpreted -- we try to caution not to 14 interpret this way as a hard PRA type cutoff.
15 But it is on the curve. The 5 times 10-7 16 value. And what we say in the paper is we think that 17 those kind of values and considerations of when is a 18 frequency low enough that it need not be considered is 19 inherent in a risk-informed approach, but as we also 20 state in the guide and in NEI 18-04 also states that's 21 not a hard cutoff. You do need to look at 22 uncertainties. You need to look at potential cliff 23 edge effects as was mentioned. So you do need to look 24 at the lower frequency events and make a conscious 25 decision if you're going to say something 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
232 1 residual risk that doesn't need to be addressed within 2 the licensing basis events.
3 MEMBER SKILLMAN: Yes, Bill. For that 4 fourth bullet. Is there a backstop? I could see a 5 clever analyst making the case for no containment 6 based on that fourth bullet.
7 MR. RECKLEY: The single failure criterion 8 bullet?
9 MEMBER SKILLMAN: I could see analyses 10 that indicate probability so low that one would then 11 say what had been a single failure criterion really no 12 longer applies because the -- I'm down to E-7, E-8.
13 A question is is there a backstop. Is 14 there something that one would simply say 15 deterministically I really don't care how low that 16 number is, by golly we're going to have a strong box.
17 MR. RECKLEY: I would say the closest 18 within the methodology to that is the fact that you 19 don't rely on a single system or a single feature 20 within the process.
21 And this was mentioned a little bit during 22 the other parts of the assessment, that really you're 23 looking at multiple failures and you're looking at it 24 at frequency ranges that go below the traditional 25 approach that was used for light water 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
233 1 So, I see Ed standing there. Did you want 2 to?
3 MR. WALLACE: I wasn't going to let you 4 dangle out there. The consideration here is looking 5 also at layers of defense available in the design and 6 having a single monolithic reactor with no moving 7 parts that could take care of itself and start up and 8 shut down and do all the things it had to do would be 9 one layer and that's all you'd ever get to potentially 10 which is crazy. It's not sensible.
11 So part of the strategy that's described 12 in defense-in-depth looks beyond just the numbers that 13 are showing up on the frequency consequence curve and 14 saying what other layers of defense do I have starting 15 with normal operations to keep the plant in good shape 16 there, working through strategies of startup, 17 shutdown, AOOs and so forth to really understand the 18 robustness of the design.
19 And when you get to your design basis 20 event category and you establish what your DBAs are 21 you're still looking beyond them for other things that 22 could (a) go wrong, part of the defense-in-depth 23 strategy at the end is go back to the risk triplet and 24 say what can go wrong, what's not in the PRA, all 25 those other kinds of things and say am I satisfied NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
234 1 that these questions that arise because of the 2 uncertainties at that stage of the development have 3 been adequately taken care of.
4 So what's below the 5 times 10-7 number 5 they're still in the PRA but is there anything in 6 there that really is showing a significant issue until 7 you're looking at catastrophic --
8 MEMBER SKILLMAN: Thank you. That helps.
9 Thanks.
10 MR. WALLACE: I'm sorry, Bill.
11 MEMBER MARCH-LEUBA: We went through this 12 discussion during the functional containment. We all 13 agreed that a big strong box is the best containment 14 you could have. But I guess as long as the 15 containment functions it doesn't need to be a big 16 strong box. We had that discussion before.
17 MR. WALLACE: If I could add one thought 18 to that comment. We're trying to design a process 19 that would accommodate from test reactor size 20 commercial reactors to full fleet big reactors with a 21 common logic that you could follow as a designer and 22 developer and licensing reviewer.
23 So the flexibility is in there to look at 24 all these things and to use the risk insights you can 25 garner from all of this information to say is 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
235 1 really a threat to the public or not and then take 2 appropriate actions.
3 And it would be different at the small end 4 of the spectrum. Your answer might be in the large 5 end of the spectrum.
6 So the notion of functional containments 7 versus physical single barrier containments and things 8 like that, somewhere in the middle probably come into 9 play when your hazard gets large enough and then you 10 have to look at the other phenomena such as chemical 11 retention and the fuel or other things that will 12 affect the outcome.
13 MR. RECKLEY: As we look -- for any of 14 these designs as we look at the mechanistic source 15 terms across the barriers going back to that First 16 Principles kind of approach and using the assessment 17 of the release fraction or the attenuation factor 18 against each barrier for each radionuclide group, for 19 each event family is the way in the end will determine 20 what is needed at the end of that process perhaps for 21 a final structural barrier to the release. And then 22 also whether that needs to be a safety-related 23 structure or if it is only being there to protect 24 against the beyond design basis events whether it 25 would be a structure with special treatment but 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
236 1 necessarily safety-related.
2 The process would enable you to answer 3 those questions we believe.
4 MEMBER SKILLMAN: Thank you, Bill.
5 MR. RECKLEY: Going to safety 6 classification again within the paper and as the 7 primary element of the process. This was specifically 8 addressed in the previous SECY from the 2003 time 9 frame and we think that it's consistent with that SRM, 10 staff requirements memorandum from the Commission that 11 allowed a probabilistic approach for the 12 classification of SSCs. So it really was not too much 13 of an issue there we didn't think from the Commission 14 policy standpoint.
15 In assessing defense-in-depth again as 16 we've talked about numerous times today the paper 17 provides a framework, it looks at both probabilistic 18 and deterministic approaches, has a role for the 19 integrated decision-making process, it does include 20 the verification that I think came up during the June 21 meeting that we agree with and I don't think was ever 22 really a technical issue but I think the guidance more 23 clearly states now that you'll never rely solely on a 24 particular plant design or operational feature.
25 The reason I bolded -- it's kind of hard NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
237 1 to tell but the last bullet is bolded because this is 2 something we want to bring up to the Commission 3 specifically.
4 In the following Fukushima and also there 5 was another initiative, the risk management regulatory 6 framework there were papers provided to the Commission 7 recommending that we define and come up with criteria 8 for defense-in-depth.
9 The Commission's SRM came back and 10 specifically said don't do that. And that was largely 11 in the context of the operating fleet and the 12 determination of whether doing that could be 13 introduced basically as a change to how we were going 14 to look at the operating fleet.
15 So we want to point out to the Commission 16 that this process does have an assessment of defense-17 in-depth and is making a determination on the adequacy 18 of defense-in-depth. And we point out we're not 19 proposing that this be universal. We're not proposing 20 that it be forced on anyone.
21 However, for those people using this 22 process it does include a check on the adequacy of 23 defense-in-depth and the Commission should be aware of 24 it.
25 We don't think that's necessarily an NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
238 1 issue. In most of the discussions during the risk 2 management regulatory framework and even during the 3 recommendation 1 out of the Fukushima work there was 4 usually a distinction of what we would force on the 5 operating fleet and what would be a good idea going 6 forward for example for advanced reactors.
7 It was generally acknowledged that a 8 voluntary approach like this for advanced reactors was 9 actually probably a good idea. It was just the 10 Commission wasn't going to mandate it.
11 But in any case the reason again we wanted 12 to point this out to the Commission. You said don't 13 define adequate defense-in-depth. This process for 14 these reactors using this methodology does include 15 that step.
16 MEMBER REMPE: Before you leave this slide 17 didn't you have an IOU that you promised me from this 18 morning about the integrated decision panel and any 19 sort of other interactions you'd had with such a panel 20 in the past.
21 MR. RECKLEY: I did, but I didn't fulfill 22 it.
23 MS. CUBBAGE: If Hanh is still here we did 24 have a little bit of a side discussion about the 25 integrated panel. He may be able 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
239 1 insights.
2 MR. PHAN: Hanh Phan. I am the lead PRA 3 analyst in NRO. Regarding the expert decision panels 4 the staff expected the applicant will follow the 5 guidance in NUMARC 93-01 that's the guidance for the 6 Maintenance Rule 50.65, and for new reactors we expect 7 the application would follow the guidance of the 8 process they use for the reliability assurance 9 programs in chapter 17.4.
10 MEMBER REMPE: Okay, so when I get my IOU 11 -- or you're saying we have no experience. But what 12 I'm wanting to know is how well did it work. Not what 13 they should do --
14 MR. RECKLEY: And what I didn't do during 15 lunch was to actually track down some people from --
16 that were involved either in that 50 -- unless Marty 17 or Hanh if you've been involved in like a 50.69 review 18 or some other review that included a similar panel.
19 MR. PHAN: Yes, but at this point from the 20 NRO's or from the new reactor's perspective up to this 21 point the staff had the opportunity to look at the 22 meeting minutes from the expert panel conducted for 23 other applications. We not directly participate in 24 any of those meetings but we review the minutes and we 25 have insights and information from those.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
240 1 MS. CUBBAGE: And I think that's 2 appropriate as our role as the regulator. We 3 shouldn't be participating in those panels. Yes, it's 4 on I hope. So I think it should be an auditable 5 process. It should follow the guidance that's 6 established. We wouldn't be participants.
7 MEMBER REMPE: I'm not asking you to 8 participate. I want to know was it effective.
9 MR. RECKLEY: And I've got to get to the 10 right people who were involved in that kind of a 11 review that used a similar panel.
12 MEMBER REMPE: And again the reason I'm 13 asking this is I think there may be some devils in the 14 detail that haven't been fully fleshed out.
15 MR. RECKLEY: Yes. On informing the 16 content of applications the draft guide does go into 17 a little bit more detail than NEI 18-04 on how we 18 think that these insights can inform both the scope 19 and the level of detail. So there was some discussion 20 of that during today.
21 We generally agree with the discussion.
22 If I can say I got a sense of the meeting if you will 23 that you should be able to use this process and if 24 things are less important than the description can be 25 boiled down to maybe some interface requirements or 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
241 1 least less detail on those systems.
2 An example that we have used throughout 3 the development of this has been on the power 4 production side. And for light water reactors the 5 final safety analysis reports include a fair amount of 6 discussion on the power conversion systems.
7 And that makes sense because the power 8 conversion systems can involve failures that feed back 9 to the primary system relatively quickly require the 10 actuation of safety equipment.
11 If a reactor design, an advanced reactor 12 design includes the particular attribute within the 13 Advanced Reactor Policy Statement that the thermal 14 response of a reactor should be much slower perhaps 15 the sensitivity to the power production systems is 16 much less and therefore the FSAR would not need to 17 provide as much information on the power production 18 systems, but just on the interface and whatever 19 analysis is done to show that an upset doesn't feed 20 back to the primary in quite as challenging a way as 21 it does for light water reactors.
22 MEMBER MARCH-LEUBA: But this is something 23 the staff proposes to do on their own in your letter.
24 NEI 18-04 does not have it.
25 MR. RECKLEY: NEI 18-04 hints at it but NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
242 1 it's not as clear. One of the things that we're 2 talking about now is what will follow this particular 3 guide.
4 And to the degree -- and this is a little 5 bit of what we're hearing, but I can't commit to it.
6 But one of the things that we're hearing is that the 7 developers would like a little more detail and a 8 little more certainty that we would be comfortable 9 with that kind of an approach.
10 And so this might be an area where we pick 11 to either do it from the staff or what we would prefer 12 is to work with an industry group to develop guidance 13 that we could endorse similar to this process.
14 MEMBER MARCH-LEUBA: The easiest -- the 15 least resistant path would be to hint in your letter 16 that it would be acceptable and then bring me a ROC 17 (phonetic). The next item that comes in bring me a 18 ROC. Now, the guy that has to bring the ROC will be 19 risking a lot.
20 MR. RECKLEY: And that's one of the 21 reasons we're hearing that they would prefer to have 22 a little more guidance in this area.
23 And it's one of the reasons that we tried 24 to expand on it in the draft guide to at least say 25 that we're amenable to it. But in the time that 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
243 1 had we weren't able to provide much more detail than 2 actually what I'm giving here.
3 ACTING CHAIRMAN CORRADINI: So let me take 4 this a bit further. So I asked the industry group 5 about this idea of pilots and classes and completing 6 the pilots so this provides a basis. What's your 7 feeling about how that helps? Because in some sense 8 that puts it back in industry's court but essentially 9 they would develop enough of a pilot such that they 10 would help out the other parts of the industry in 11 terms of what's expected of them to actually go 12 through this effort. Go through this exercise. It 13 can be your personal opinion.
14 MR. RECKLEY: Well, it's going to be my 15 personal opinion. The difficulty to some degree is 16 that even within a technology group the designs can 17 vary significantly.
18 And what my personal thought is that what 19 would be useful to everyone is to keep it technology-20 inclusive as this guidance is in which case you come 21 up with a methodology.
22 And what I just talked about, the designer 23 would say -- the process would say what do you put in 24 chapter 10, that's typically power conversion, you do 25 an assessment. What's the feedback from the secondary NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
244 1 to the primary and if you meet this then you don't 2 need to describe.
3 However, if you do have feedback and some 4 of the discussions were on chemicals so it may not be 5 thermal feedback, it might be chemical feedback. If 6 you have these kind of concerns then you do have to 7 describe in more detail what's in that adjacent system 8 because it has the potential to affect the primary 9 side.
10 And it would lay out that kind of a 11 process or methodology versus trying to define 12 specifically what needs to be in for example chapter 13 10 for any design because all of those things become 14 dependent on the technologies, on the power levels, on 15 more factors than typically just one.
16 MR. TSCHLITZ: So I would just add that 17 the industry recognizes that we need to do more as far 18 as risk-informing the content of applications beyond 19 what DG-1353 does and beyond what NEI 18-04 does.
20 There needs to be more guidance on this.
21 It's one of the things that we're looking at working 22 on in the near future to develop that extra guidance 23 on what goes into the content of the application that 24 the NRC could review and endorse as an acceptable 25 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
245 1 ACTING CHAIRMAN CORRADINI: So, can I say 2 that differently. So instead of them leaving it 3 general you might put some examples out there as to 4 what would be in and get their reaction as a group.
5 MR. TSCHLITZ: Yes. I would say even more 6 that would be more of a guidance document to provide 7 how to go about doing this rather than just simply 8 examples.
9 ACTING CHAIRMAN CORRADINI: But I guess 10 I'm still back with the examples strike me as 11 important because within a class of systems there's 12 going to be some commonality and certain things, 13 chemical reactions you have to consider, the fact that 14 I don't have solid fuel and I have moving fuel, these 15 sorts of things are going to be similar enough that I 16 would expect some sort of pilot would be beneficial 17 for them to do and you to at least see to try to get 18 a reaction to it.
19 MR. RECKLEY: I generally agree. It's 20 just a caution that the designs can vary and that can 21 lead you -- there was a question earlier on about the 22 steam generator. Well, if your design uses double 23 walled steam generators and the water is only a little 24 bit away from the primary sodium loop that's one level 25 of concern.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
246 1 If you're a design that uses an 2 intermediate loop and the water is one whole loop away 3 from the primary side it's a different concern. Those 4 are both fast reactors, sodium coolant but the designs 5 are significantly different.
6 So I'm glad to hear Mike say that. We've 7 heard it but now it's public.
8 This is another area, it's a little hard 9 here again to take that that's highlighted on the 10 slide. But this is another area that we don't think 11 the Commission has -- we don't think there's an issue, 12 but it's also not an issue that was brought up to our 13 knowledge in the previous Commission papers and that 14 we want them to acknowledge that we're going to use 15 this approach not to scale the NRC review but to scale 16 what's in the application.
17 The discussion this morning on the 18 enhanced safety-focused review for example, that was 19 things the staff does different. Once we get an 20 application in, but the guidance on what goes in an 21 application was basically the same. So NuScale gave 22 us a full application and then we said how can we 23 scale that back if you will. I'm shorthanding. How 24 can we adjust the review given the risk insights.
25 I think as John Monninger pointed out or NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
247 1 Ian that gets complex because now you are giving a 2 staff a chapter and saying we don't think you need to 3 look at this in quite as much detail. That's an 4 engineering practice that's hard to come across to 5 give something to somebody and say but we don't really 6 need you to look at it in quite as much detail as you 7 typically have done in the past.
8 And so we think actually a better idea is 9 to scale back what's in the application and what's 10 given to the staff to review versus giving them the 11 whole book and then telling them but you don't need to 12 look at this in quite as much -- it's not human nature 13 to actually do that.
14 But that's an area we're going to ask the 15 Commission.
16 So again the recommendation is for the 17 Commission to approve the use of this methodology 18 that's described in 18-04 and as reflected in the 19 draft guide.
20 You have a working draft of the guide so 21 I'm just going to kind of quickly go through what's in 22 there and the staff findings.
23 The staff has taken no exceptions to 24 what's in NEI 18-04. We offer a number of things that 25 we want to emphasize or perhaps clarify but at 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
248 1 time we're not proposing any exceptions. So this is 2 again just the scope of the draft guide and it is 3 applied to those rules that are associated with the 4 content applications and they're listed there, 50.34, 5 52, 47 and so forth.
6 In regards to the licensing basis events 7 again the staff position as it's stated in the working 8 draft of the guide is that it's an acceptable method 9 as described in 18-04.
10 We caution or emphasize that the FC target 11 does not depict acceptance criteria for the actual 12 regulatory limits. I think as Karl pointed out the 13 anchors that are used are surrogates. They don't 14 correlate to NRC regulations per se. So it's a useful 15 tool but you have to look at it for what it is and not 16 confuse it with actual acceptance limits.
17 The other point I already pointed out, the 18 figure includes a cutoff of 5 times 10-7 for inclusion 19 as a licensing basis event. The staff again just 20 cautioning that's not a hard and fast cutoff. You 21 need to look below it. You need to address some 22 certainties. You need to look for cliff edge effects.
23 You need to be very deliberative in what you're not 24 including in the licensing basis events.
25 We touched on this or Karl touched on 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
249 1 The methodology does address external events. It has 2 a definition of a design basis external hazard level.
3 That is basically the same as the design basis 4 earthquake, design basis flood, other external hazards 5 for which safety-related equipment needs to be 6 protected. It sets that kind of definitive limit. It 7 needs to be protected at least up to this point.
8 ACTING CHAIRMAN CORRADINI: Can you help 9 me here? If I'm in your -- I guess you've got a name 10 for the diagram. If I'm in the Reckley-Cubbage 11 diagram.
12 MR. RECKLEY: Segala.
13 ACTING CHAIRMAN CORRADINI: I'm sorry.
14 Segala-Reckley-Cubbage diagram. Is it just safety-15 related equipment or is it risk-significant? I'm 16 trying to understand what's covered under this.
17 MR. RECKLEY: Karl, be prepared. Because 18 I will give you the way I think it works and then Karl 19 can correct me if I'm wrong.
20 So for -- this is the alignment with the 21 current arrangement. For safety-related equipment 22 they'll need to be protected against the design basis 23 external hazard level which is analogous to and 24 determined using our existing methodology for defining 25 those kind of external hazards.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
250 1 In addition to that within the PRA it's 2 looking at a fuller range of external events including 3 down into the beyond design basis arena and to the 4 degree that beyond design basis external hazard can 5 influence the frequency of an event or a malfunction 6 it's going to be also addressed in that category of 7 events. So is that right, Karl?
8 MR. FLEMING: Yes, that's basically 9 correct. We start with -- when we talk about the 10 design basis external hazard levels we have a 11 requirement, a deterministic requirement that says 12 that you have to protect your safety-related SSCs in 13 the performance of your required safety functions to 14 achieve safe shutdown against any -- assuming the 15 occurrence of any design basis external hazard level.
16 And that's just for safety-related SSCs.
17 However, at some point in time and there's 18 flexibility on when this might occur, at some point in 19 time there will be external hazards included into the 20 PRA and then that would then talk to the potential for 21 creating maybe additional risk-significant SSCs or 22 perhaps additional SSCs that because of the external 23 hazard may have a defense-in-depth adequacy 24 consideration.
25 So for those -- and therefore getting 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
251 1 the NSRST categories. And for all NSRST categories 2 whatever hazard they may have come from we set 3 reliability and capability requirements to basically 4 start the process of the special treatments. And then 5 the integrated decision process would consider is 6 there anything beyond setting reliability and 7 capability requirements which it may have to do with 8 protecting against an external hazard or may not 9 depending on the nature of the LBE that produced the 10 risk significance or the defense-in-depth concern.
11 And the integrated decision panel would 12 then decide what kind of special treatments beyond 13 capability reliability requirements and a monitoring 14 program to make sure that these are enforced through 15 the life operation of the plant.
16 MR. RECKLEY: So, the other findings or 17 clarifications. As we've already discussed the single 18 failure criterion as it's applied traditionally to 19 safety-related equipment within chapter 15 of light 20 water reactors we think is not needed and it's 21 consistent with the Commission's decision in SECY 22 0047.
23 We do offer again that the methodology in 24 NEI 18-04 does in our view use PRA a little beyond 25 what is currently done. We require PRAs to be done.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
252 1 We require the results to be shown within chapter 19.
2 It's used to support things like 3 determinations of regulatory treatment of non-safety 4 systems. But in this particular case it's a little 5 more integrated into the process.
6 And so we just offer the maybe obvious 7 observation that to the degree that the ASME ANS 8 standard is completed and to the degree that that 9 standard is endorsed by the NRC that would make the 10 process much easier.
11 And the staff does currently plan -- the 12 NRC is engaged in that standard. Our understanding is 13 that that standard will be provided to the NRC for 14 endorsement when it's completed, and the NRC will 15 review it for potential endorsement when it's 16 completed.
17 So all of these things are planned to be 18 looked at. We're just saying if it all works out as 19 planned it would help tremendously in the process.
20 ACTING CHAIRMAN CORRADINI: Let me -- can 21 I ask a little bit different question. Is this PRA 22 standard for advanced reactors or advanced -- implying 23 a certain level of completeness of the design?
24 CHAIRMAN BLEY: We can't hear you.
25 ACTING CHAIRMAN CORRADINI: Is the level NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
253 1 of completeness of the design implied in this PRA 2 standard?
3 MR. RECKLEY: Since Karl's on the 4 committee let me.
5 MR. FLEMING: I'd be happy to handle that.
6 The standard does not enforce a given application. So 7 the standard is available to support a variety of user 8 applications.
9 So the user decides and perhaps with 10 negotiation with the regulator what parts of the 11 standard need to be applied to that application, what 12 level of detail has to be supplied and so forth.
13 And then the standard has requirements to 14 clarify whether certain requirements haven't been 15 addressed or whether there's been assumptions made in 16 lieu of actual inputs that would create the necessary 17 model fidelity.
18 So the standard documents the basis for --
19 requires you to document the basis for the PRA and 20 then whether or not that's sufficient is really a 21 matter for the application process, i.e., negotiation 22 with the regulator.
23 MR. RECKLEY: Because keep in mind from 24 the staff's point of view we have the luxury of being 25 at the tail end of the design process. For the actual NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
254 1 application.
2 Interactions can occur throughout the 3 design process but by the time they give us the 4 application the assumption is the design is completed, 5 the requirements for things like PRAs are completed.
6 I would suggest though if you're looking 7 at how during the design process even before an 8 application is submitted that the designers can be 9 thinking in the context of the PRA what was mentioned 10 earlier, the EPRI body of knowledge on going from 11 process hazard assessments to PRA and how you kind of 12 -- it's especially applicable to molten salts, but 13 it's not only limited to molten salts. It talks about 14 how you might start off doing PIRTs and again on 15 particular systems failure modes and effects or 16 HAZOPs. You'd use those tools that might be more 17 readily available for a design that's still being 18 developed and you mature into doing the PRA through 19 iterations and in both the analysis and in the design 20 as you go along.
21 But I found that EPRI body of knowledge 22 document that was shown on the slides to be pretty 23 insightful of how a designer might do it.
24 MR. FLEMING: If I might just add a couple 25 of more comments on that topic. When the Board 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
255 1 Nuclear Codes and Standards decided we needed some 2 more standards for different kinds of reactors they 3 set in place two working groups, one for advanced 4 light water reactors and one for advanced non-light 5 water reactors.
6 And those projects were going on in 7 parallel. And we were guided by the JCNRM, the Joint 8 Committee on Nuclear Risk Management to take a 9 consistent approach to dealing with the same issues.
10 So this whole process of how do you write 11 a standard for a PRA that's done in the maybe 12 different stages of design was also faced with the 13 advanced light water reactor working group. And it 14 just turned out that our non-light water reactor 15 standard got issued for trial use before the ALWR 16 standard got out.
17 But there's an ALWR trial use standard 18 that will be out pretty soon and it follows the same 19 logic as far as how do you deal with PRA requirements 20 for a design stage PRA.
21 A final comment is that we also have a PRA 22 white paper that was drafted several years ago, or a 23 couple of years ago I guess and one of our tasks in 24 the LMP framework is to bring our white papers up to 25 date and get them in alignment with what's currently NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
256 1 in the guidance document, taking into account lessons 2 from these pilot applications.
3 And that includes some of the standards Jas 4 (phonetic) has talked about there.
5 MR. RECKLEY: Then moving on to the second 6 element, the safety classification. Again, the staff 7 position is that what's described in NEI 18-04 8 provides an acceptable method.
9 And the only clarification or point of 10 emphasis here again is these things need to be looked 11 at with all three elements as an integrated process.
12 Just again offering a caution that we didn't want a 13 designer to pick out an element like safety 14 classification and think that that was a standalone 15 process they could use.
16 Then lastly, defense-in-depth. Again the 17 staff position, we're not taking any exceptions and 18 saying that it's an acceptable method.
19 The only clarification here that we're 20 offering and I'll be honest. These things were 21 developed in parallel so I have to go back and make 22 sure that NEI 18-04 as we've given it to you includes 23 the same statement.
24 But the revision right before that had 25 included a statement that talked about considering NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
257 1 plant capability and programmatic defense-in-depth 2 measures and change control processes that would go 3 into the operating phase of a plant.
4 And we think that's a good idea, but this 5 guidance document didn't really lay out much in terms 6 of how that would carry into the operating phase. And 7 we think that that is a good candidate for another 8 guidance development in terms of how do you maintain 9 this.
10 There was some discussion for example on 11 all the programmatic measures that we would consider 12 during licensing if you will to make sure that the 13 SSCs were actually delivering as advertised. But how 14 we roll that into the operating phase and how we 15 include it in requirements like technical 16 specifications or plant procedures or regulations or 17 whatever form it takes we weren't ready to address at 18 this point. So we're just leaving that open that this 19 only addresses up to the licensing stage, not into 20 operations.
21 Two more slides here. We mention in the 22 draft guide the same thing I mentioned this morning.
23 There are interfaces between this process and NEI 18-24 04 and other arenas.
25 One is emergency planning. And 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
258 1 trying to make sure that the Draft Guide 1350 on 2 emergency planning and the Draft Guide 1353 on 3 licensing basis events marry up because as I mentioned 4 that's where the events will be identified that you 5 then compare to the protective action guidelines in an 6 application that includes a proposed reduction in 7 emergency planning zones.
8 We've talked numerous times about 9 mechanistic source term. Mechanistic source term is 10 key to this. It didn't get a whole lot of discussion 11 in NEI 18-04. It's an inherent assumption that you 12 have the ability to assess the consequences or as 13 previously stated the release fractions across all the 14 barriers.
15 So we're just pointing out that link and 16 that importance.
17 This is another area that we envision it's 18 very possible that we'll have an additional guidance 19 document on the development of mechanistic source 20 term. And if for no other reason than the ACRS kind 21 of suggested that that might be a good idea in the 22 context of the emergency planning proposed rule.
23 So we don't really disagree with that and 24 we're talking about it. And that is another good 25 candidate for another guidance document that would 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
259 1 developed.
2 ACTING CHAIRMAN CORRADINI: It was pointed 3 out in that session. Dennis was the chair of that 4 session also and he can remind me if I have it wrong.
5 In Reg Guide -- now I'll get the reg guide 6 wrong, 1.18 -- 1.83, 1.183. I can't remember the reg 7 guide for essentially alternative source term. There 8 was a set of seven or eight attributes that if the 9 applicant wanted not to use what is in the reg guide 10 but wanted to use something of their own making it 11 ought to meet a series of attributes. And I thought 12 at least that's a good starting point.
13 MR. RECKLEY: That is a good starting 14 point. Under NGNP there was a white paper on 15 mechanistic source term. For other designs there's 16 also for fast reactors Argonne has produced a report 17 on mechanistic source term.
18 So there is -- we actually are working 19 with -- under our contract arrangements we're working 20 with some national labs in a similar context to say 21 can we develop a fairly generic way to describe the 22 development of a mechanistic source term.
23 So it was a good observation and I think 24 it's likely that we'll be here sometime down the road 25 to talk about a draft guide on mechanistic source NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
260 1 term.
2 I've already talked numerous times about 3 informing the content applications. There is a short 4 section in the draft guide that starts to talk about 5 it as we've talked about before. Maybe it doesn't go 6 far enough but it was at least a starting point to 7 include in the draft guide that you can scale the 8 format and the content and the level of detail in an 9 application based on the insights you get from this 10 methodology.
11 So going right to the bottom line here.
12 Checking off that we were here today, October 30.
13 Full committee the first week of December. I'm not 14 sure it's the 6th, but whatever date gets set for that 15 first week of December we'll come back to the full 16 committee.
17 And again what we're asking for is 18 feedback on the draft Commission paper and at your 19 leisure or at your discretion feedback on the draft 20 guide.
21 We then plan after the full committee to 22 issue the draft guide by the end of the year is our 23 current plan. Issue the SECY to the Commission in 24 early 2019.
25 In mid-2019 depending on the feedback 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
261 1 we get from the solicitation of public comments on the 2 draft guide and whatever feedback we get from the 3 Commission on the SECY paper we would be in a position 4 to finalize the guide and then start to engage the 5 ACRS on the review of the final guide and issue the 6 final guide we hope by the end of 2019.
7 ACTING CHAIRMAN CORRADINI: Thank you, 8 Bill. Questions by the committee before we go to 9 public comments? Okay. I think the line is open in 10 our new high-tech room. So first let's go with 11 there's comments from the members of the public that 12 are in the room. Any additional comments by members 13 of the public in the room? Okay.
14 So let's turn to the phone line, bridge 15 line. Are there any comments from members of the 16 public? Okay, hearing none. Oh, I'm sorry. Mr.
17 Redd. Oh, you have a homework assignment. Let's make 18 sure we have no public -- so there's no public 19 comments from the bridge line.
20 Okay. Come up with your homework 21 assignment.
22 MR. REDD: Jason Redd, Southern Nuclear.
23 We talked several times today about the public report 24 that has been issued on the X-Energy demonstration.
25 I'd like to read that ADAMS session number into 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
262 1 record so it will be available in the future.
2 That is ADAMS number ML18228A779 dated 3 August 1, 2018. Thank you.
4 ACTING CHAIRMAN CORRADINI: Thank you very 5 much. I am pulling it up as we speak just to see if 6 it really is there. I think what I got with that ML 7 number is presentation September 13, 2018 public 8 meeting on regulatory improvements. But not that the 9 ADAMS system is disorganized.
10 MR. REDD: All right. That may be the 11 overall package number.
12 ACTING CHAIRMAN CORRADINI: Oh, it's the 13 whole package. Okay, excuse me.
14 MR. REDD: I will re-verify this again.
15 ACTING CHAIRMAN CORRADINI: I think that's 16 the best thing to do.
17 MR. RECKLEY: We'll get it and the other 18 Argonne reports and the things that were mentioned.
19 We'll get to ACRS staff.
20 ACTING CHAIRMAN CORRADINI: Okay. Thank 21 you very much. Dennis, I want to kind of turn to you 22 since you're the actual chair. I'm just the in room 23 chair. Do you have any final comments you want to 24 make, Dennis?
25 CHAIRMAN BLEY: I was on mute. Thanks, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
263 1 Mike, and thanks for chairing the meeting in my 2 absence. I appreciate it.
3 I think we need to talk a little bit about 4 the full committee meeting. Today's meeting had 5 almost the whole committee, I think we're missing 6 three people.
7 So right now we're scheduled for an hour 8 and three quarters. And I think that's going to be 9 okay.
10 Bill, I think pretty much a summary of 11 what you presented today and I don't know if Karl 12 Fleming can be there but there may be some detailed 13 questions on the methodology and I think that would be 14 really good if you had somebody to take that.
15 So we'll -- our staff and the NRC staff 16 will work together to get an agenda set up for this 17 meeting.
18 I think we're probably going to draft a 19 letter on both the Commission paper and the new 20 guidance document. I don't see why we wouldn't 21 include them both.
22 And I'd like to thank everybody for a 23 great meeting. A lot of good information. So I think 24 that's where we're headed. If any members have any 25 thoughts about the full committee meeting or 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
264 1 letter I'd love to hear them.
2 ACTING CHAIRMAN CORRADINI: Okay. We'll 3 come back then to the staff and try to prepare for the 4 full committee. Okay. With industry input of course.
5 Other than that I think we're done and 6 we're adjourned. Thank you.
7 (Whereupon, the above-entitled matter went 8 off the record at 4:09 p.m.)
9 10 11 12 13 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
ACRS Future Plant Designs Subcommittee Draft Regulatory Guide (DG) 1353 and Related Commission Paper Technology-Inclusive, Risk-Informed, Performance-Based Approach to Inform the Content of Applications for Licenses, Certifications, and Approvals for Non-Light Water Reactors, October 30, 2018 (AM) 1
Outline
- Background
- Enhanced Safety Focused Review Approach (ESFRA) for Light-Water Small Modular Reactors
- Non-Light Water Reactor Program
- Context and overview for technology-inclusive methodology
- NEI 18-04 (Licensing Modernization Project)
- Draft SECY paper
- Draft Regulatory Guide 1353 2
Enhanced Safety Focused Review Approach (ESFRA)
- Staff approach used for NuScale application review to focus on safety
- Tools and strategies for defining the scope and depth of reviews
- Companion to NUREG-0800 (Standard Review Plan), Introduction - Part 2 as well as Design-Specific Review Standards
- Intended to be used during both pre-application and review stages 3
ESFRA Background
- Objective
- Increased effectiveness and efficiency for staff reviews
- Directed by the Commission
- SRM to COMGBJ-10-0004/COMGEA-10-0001
- SRM to SECY-11-0024
- Review focus and resourcesto risk-significant structures, systems, and components (SSCs) and other aspects of the design that contribute most to safety
- ACRS presentations in 2011, 2016, and 2017 4
ESFRA Review Tools
- Considerations
- Safety Significance (e.g., A1/A2/B1/B2)
- Regulatory Compliance
- Novel Design
- Shared SSCs/Nonsafety-Safety Interactions
- Unique Licensing Approach
- Safety Margin/Defense-in-depth
- Operational Programs
- Additional Risk Insights 5
ESFRA Status and Future
- Applied in multiple areas with varying degrees of success
- Developing lessons learned
- Can be used for future reviews including advanced reactors
- Coordination with LMP
- The underlying concept is consistent with the agencys risk-informed, performance-based approach 6
Advanced Reactor Program 7
Implementation Action Plans Strategy 1 Strategy 2 Strategy 3 Strategy 4 Strategy 5 Knowledge, Skills Strategy 6 Computer Codes Flexible Review Consensus Codes Policy and Key and Capability Communication
& Review Tools Processes and Standards Technical Issues ONRL Molten Salt Reactor Training Identification &
Assessment of Regulatory Roadmap ASME BPVC Section III Siting near densely populated NRC DOE Workshops Available Codes Division 5 areas Knowledge Management Prototype Guidance ANS Standards 20.1, 20.2 Insurance and Liability Periodic Stakeholder 30.2, 54.1 Meetings Competency Modeling Non-LWR Design Criteria Non-LWR PRA Standard Consequence Based Security NRC DOE GAIN MOU (SECY-18-0076)
Updated HTGR Environmental EP for SMRs International and Fast Reactor Reviews and ONTs Coordination Training (SECY-18-0103)
Licensing Functional Modernization Containment Project (SECY-18-0096)
Potential First Micro-Reactors Movers 8
Integrated Design/Review Siting near densely populated Functional areas EP for SMRs Containment and ONTs Insurance and Liability Environmental Reviews Consequence Based Security 9
Revisit First Principles 10
Integrated Design/Review Siting near densely populated Functional areas EP for SMRs Containment and ONTs Licensing Modernization Project Insurance and Liability Environmental Reviews Consequence Based Security 11
Other Requirements
- Associated requirements include:
- Quality Assurance
- Maintenance Rule
- Interfaces with requirements for:
- Siting
- Environmental Reviews
- Additional requirements for design/operation include:
- Routine Effluents
- Worker Protections
- Security
- Aircraft Impact Assessments 12
NEI 18-04, General Approach
- Licensing Basis Events
- Probabilistic Risk Assessment
- Deterministic
- SSC Classification
- Function and Risk Considerations
- Safety Related
- Non-Safety Related with Special Treatment
- Defense-in-Depth Assessment
- Structures, Systems and Components
- Programmatic
- Integrated Decision-making Process 13
Key Considerations
- Evolution of Approach
- Advanced Reactor Policy Statement
- SECY-93-092, Issues Pertaining to the Advanced Reactor (PRISM, MHTGR, and PIUS) and CANDU 3 Designs and Their Relationship to Current Regulatory Requirements
- Risk-Informed, Performance-Based Regulation
- SECY-03-0047, Policy Issues Related to Licensing Non-Light-Water Reactor Designs
- NUREG-1860, Feasibility Study for a Risk-Informed and Performance-Based Regulatory Structure for Future Plant Licensing
- Next Generation Nuclear Plant (NGNP)
- Similarities to traditional LWR structure, but also differences
- including terminology challenges with different definitions for some phrases 14
Key Considerations (continued)
- Integrated methodology consisting of three primary elements
- Licensing Basis Event Selection and Analyses
- SSC safety classification and performance requirements
- Assessing defense-in-depth adequacy
- Uses existing regulatory criteria, including guidelines for offsite dose and NRC safety goals
- Assessments performed using risk-informed and deterministic approaches, including Integrated Decision-making Process
- Includes methodology for assessing defense in depth provided by plant capabilities and programmatic controls 15
Event Selection & Analysis The F-C Target values shown in the figure should not be considered as a demarcation of acceptable and unacceptable results. The F-C Target provides a general reference to assess events, SSCs, and programmatic controls in terms of sensitivities and available margins.
AOOs Note that DBAs DBEs (Chapter 15) derived from DBEs BDBEs
- F-C Target considered along with cumulative risk metrics, safety classification, and assessment of defense in depth 16
Safety Classification and Performance Criteria
- Safety-Related (SR):
o SSCs selected by the designer from the SSCs that are available to perform the required safety functions to mitigate the consequences of DBEs to within the LBE F-C Target, and to mitigate DBAs that only rely on the SR SSCs to meet the dose limits of 10 CFR 50.34 using conservative assumptions o SSCs selected by the designer and relied on to perform required safety functions to prevent the frequency of BDBE with consequences greater than the 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):
o Non-safety-related SSCs relied on to perform risk significant functions. Risk significant SSCs are those that perform functions that prevent or mitigate any LBE 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.
o Non-safety-related SSCs relied on to perform functions requiring special treatment for DID adequacy
- Non-Safety-Related with No Special Treatment (NST):
o All other SSCs (with no special treatment required) 17
Assessing Defense in Depth 18
Informing the Content of Applications
- NEI 18-04 provides useful guidance for applicants to identify and provide the appropriate level of information needed to satisfy parts of the regulatory requirements in 10 CFR 50.34, 10 CFR 52.47, 10 CFR 52.79, 10 CFR 52.137, and 10 CFR 52.157.
- Combination of deterministic evaluations and probabilistic risk assessments
- Information needed on fuel, primary, and other barriers to define limitations, performance characteristics, and as input to mechanistic source term
- Information needed on SSCs and programmatic controls associated with key safety functions
- Scope and depth for other information (e.g., ancillary plant systems) to be determined based safety/risk significance (i.e., roles in preventing or mitigating licensing basis events)
- Level of detail can also reflect potential performance-based approaches (see Introduction, Part 2, to NUREG 0800) 19
Next Presentations
- NEI 18-04, Risk-Informed Performance-Based Guidance for Non-Light Water Reactor Licensing Basis Development, (Draft Report Revision N) and Related Tabletop Exercises
- Requested ACRS Feedback
- Draft SECY, Technology-Inclusive, Risk-Informed, and Performance-Based Approach to Inform the Content of Applications for Licenses, Certifications, and Approvals for Non-Light Water Reactors
- Draft DG-1353, Guidance for a Technology-Inclusive, Risk-Informed, and Performance-Based Approach to Inform the Content of Applications for Licenses, Certifications, and Approvals for Non-Light Water Reactors 20
NEI 18-04 AND THE LICENSING MODERNIZATION PROJECT Mike Tschiltz, Jason Redd, and Karl Fleming October 30, 2018
© 2018 NEI. All rights reserved.
VISION FOR THE FUTURE -
CREATING A STREAMLINED AND PREDICTABLE LICENSING PATHWAY TO DEVELOPMENT Mike Tschiltz Senior Director New Plant, SMRs and Advanced Reactors Nuclear Energy Institute 2
© 2018 NEI. All rights reserved.
VISION FOR THE FUTURE -
A STREAMLINED AND PREDICTABLE LICENSING PATHWAY TO DEPLOYMENT To ensure that advanced reactors are licensed and built in the U.S., near-term regulatory reforms are necessary.
These reforms should focus on achieving the following near-term objectives:
- Reversing the trend of increasing regulatory costs and excessively long reviews;
- Aligning the regulatory framework for advanced reactors with their inherent enhanced safety;
- Defining licensing options clearly, including options for staged applications and approval; and
- Providing additional flexibility for changes during construction.
3
VISION FOR THE FUTURE -
A STREAMLINED AND PREDICTABLE LICENSING PATHWAY TO DEVELOPMENT 4
© 2018 NEI. All rights reserved.
VISION FOR THE FUTURE -
A STREAMLINED AND PREDICTABLE LICENSING PATHWAY TO DEVELOPMENT Reverse The Safety Focused Staged Flexibility in Trend Reviews Approvals Construction RI PB Licensing Revise Security Standardize Regulatory TI Lessons and EP ITAAC and Engagement Learned Regulations Streamline Planning Implementation Guidance for Risk Inform Illustrating Reduce Pathways for Simplify Advanced Application Approvals Changes Reactor Unnecessary during Licensing Basis Level of Detail Construction 5
© 2018 NEI. All rights reserved.
NEI 18-04 Jason Redd NEI 18-04 Guidance Document Lead Southern Nuclear Development 6
© 2018 NEI. All rights reserved.
OVERARCHING OBJECTIVES OF THE LMP METHODOLOGY NEI 18-04 guides prospective applicants in answering the following questions:
- What are the plant initiating events, event sequences, and accidents that are associated with the design?
- How does the proposed design and its structures, systems, and components (SSCs) respond to initiating events and event sequences?
- What are the margins provided by the facilitys response, as it relates to prevention and mitigation of radiological releases within prescribed limits for the protection of public health and safety?
- Is the philosophy of Defense-in-Depth (DID) adequately reflected in the design and operation of the facility?
7
© 2018 NEI. All rights reserved.
Karl Fleming NEI 18-04 Senior Technical Lead 8
© 2018 NEI. All rights reserved.
PRINCIPAL FOCUS OF LMP METHODOLOGY
- Systematic, reproducible, robust ,and integrated processes for:
- Identification of safety significant licensing basis events (LBEs) appropriate for each non-LWR design through an integrated decision process informed by a design specific PRA.
- Demonstrating enhanced safety margins consistent with Advanced Reactor Policy;
- Identification of key sources of uncertainty;
- Evaluation of the adequacy of plant capabilities and programs for defense-in-depth.
- Appropriate balance of deterministic and probabilistic inputs to risk-informed decisions involved in design, operations, programs and licensing.
- Performance-based approach to setting plant and SSC performance requirements and monitoring performance against requirements.
- SSC performance requirements linked to balancing prevention and mitigation functions identified in LBEs.
9
SELECTION AND EVALUATION OF LBES BY DESIGN TEAM IS SYSTEMATIC AND REPRODUCEABLE
- Anticipated Operation Occurrences (AOOs), Design Basis Events (DBEs), and Beyond Design Basis Events (BDBEs) defined in terms of event sequence families with input from a reactor design-specific PRA that is integrated into the design process.
- To ensure consistency with the reactors safety design approach;
- Individually for risk significance against a Frequency-Consequence (F-C)
Target;
- Collectively by comparing the total integrated risk against cumulative risk targets.
- DBEs and high consequence BDBEs are evaluated to define Required Safety Functions (RSFs) necessary to meet F-C Target.
- DBAs are derived from DBEs by crediting only SR SSCs and evaluated conservatively for meeting Chapter 15 Design Basis Accident (DBA) requirements.
10
SSC SAFETY CLASSIFICATION AND RISK SIGNIFICANCE APPROACH
- SSC Safety Classes:
- Safety Related (SR) - selected to perform Required Safety Functions;
- Non-Safety Related with Special Treatment (NSRST) - non SR SSCs that are risk significant or perform functions necessary for DID adequacy;
- Non-Safety Related with no Special Treatment (NST).
- Risk Significant SSCs based on absolute metrics
- Perform functions necessary to keep LBEs inside F-C Target;
- Contribute at least 1% to cumulative risk targets selected to meet Quantitative Health Objectives (QHOs) and 10 CFR 20 annual dose limits.
- Risk Significant LBEs
- Doses exceed 2.5 mrem, and,
- Frequency of the LBE dose within 1% of the F-C Target.
11
SSC CATEGORY RELATIONSHIPS 12
LMP APPROACH TO SAFETY MARGINS
- Plant Level Safety Margins
- Reflected in the margins between LBE frequencies and consequences and the F-C target;
- One way to demonstrate enhanced margins consistent with NRC Advanced Reactor Policy.
- SSC Level Safety Margins
- Margins in design codes selected to provide a robust capability to support the mitigation function of safety significant SSCs;
- Margins in the performance requirements selected to ensure that SSC will perform their prevention functions with adequate reliability.
13
LBE RISK-SIGNIFICANCE CRITERIA 14
DEFENSE-IN-DEPTH (DID) ADEQUACY EVALUATION Evaluation of DID involves
- Attributes of DID
- Evaluation of attributes
- Guidelines for adequacy of Plant Capability and Programmatic DID
- Special considerations
- Integrated Decision Process
- Compensatory action determination
- DID Baseline documentation 15
DEMONSTRATING LMP APPLICABILITY TO NON-LWRS
- High Temperature Gas-Cooled Reactors
- MHTGR-1980s PSID, PRA, NUREG-1338;
- ANS 53.1 Design Standard for MHRs (PBMR, NGNP applications);
- Xe100 LMP Demonstration (completed).
- Liquid Metal Cooled Fast Reactors
- GEH PRISM -1980s, PSID, PRA, NUREG-1368;
- GEH LMP Demonstration (completed).
- Molten Salt Reactors
- EPRI PHA-to-PRA Project using MSRE Case Study;
- Vanderbilt MSRE LMP Demonstration (planned for 2019).
- Other Advanced non-LWRs
- Kairos FHR LMP Demonstration (planned for 2019);
- Westinghouse eVinci Micro reactor LMP Demonstration (planned for 2019).
16
CURRENT EXPERIENCE IN APPLYING LMP PROCESS TASKS (AS OF 10/30/2018) 17
LESSONS LEARNED FROM LMP APPLICATIONS
- LMP methodology demonstrated for the three major families of advanced non-LWRs: gas-cooled, liquid metal-cooled, and molten salt reactors.
- Developers involved in demonstrations found the methodology to be useful and to provide reasonable results consistent with safety design approach.
- Performance-based aspects enhanced by use of absolute, versus relative, metrics for LBE and SSC risk significance.
- Relationships and distinctions among safety-related, risk-significant, and safety-significant SSCs clarified.
- Importance of integrating the tasks of selecting and evaluating LBEs, safety classification and performance requirements of SSCs, and evaluation of DID adequacy into Risk-Informed, Performance-Based (RIPB) decisions demonstrated.
- Implementation feedback to be incorporated into LMP white papers. 18
XE-100 LMP DEMONSTRATION HIGHLIGHTS
- Example of LMP application at early state of design.
- Limited scope high level PRA developed during preconceptual design to guide conceptual design.
19
GEH PRISM LMP DEMONSTRATION
- Example of LMP application after previously completed conceptual design and NRC pre-application review.
- Included passive component reliability, component reliability database development, and mechanistic source term assessments.
- Inspired NEI 18-04 approach to LBE and SSC risk significance criteria based on absolute risk metrics.
- NEI 18-04 application included:
- Identification / confirmation and evaluation of RSFs
- Classification of SR and NSRST SSCs
- Preliminary selection / confirmation of DBAs
- Evaluation of plant capabilities for defense-in-depth 20
PRISM SAFETY-SIGNIFICANT SSCS
- The selected SR SSCs can be grouped into the following high level categories:
- Control rods and drives and associated operator actions;
- EM pump supply breakers and associated operator actions;
- 120 VAC equipment;
- 125 VDC equipment;
- Reactor vessel & internals;
- RVACS;
- Supporting structures.
- SG shell and tubes;
- IHTS features supporting heat transport;
- Forced air cooling mode of ACS and supporting 480 VAC electrical equipment;
- SWRPS detection and mitigation SSCs.
21
GEH PRISM DEMONSTRATION OBSERVATIONS
- Systematic and Repeatable
- It is clear when a process step is complete;
- Sensitivity studies are easy to perform;
- Results are traceable to key risk and performance drivers.
- Visual
- Provides an point of reference for conveying PRA insights to Designers and Reviewers;
- F-C plot illustrates results relative to risk targets;
- More meaningful than displaying very low frequency numbers.
- Iterative
- Complements the Design Phases;
- Identifies vulnerabilities and trends early in the design;
- Facilitates design optimization sensitivity studies;
- Clarifies path to regulatory engagement.
22
PRELIMINARY MSRE PRA DEVELOPMENT 23
MSRE
SUMMARY
- Comprehensive PHA (HAZOP) evaluations being performed to create body of knowledge for safety case and PRA development
- Project benefits from EPRI PHA-PRA Project
- MSRE PRA is at early state of development
- Event trees (with fault trees) were constructed for a total of three interesting initiating events;
- 2 of 8 total event sequences had greater than minimal consequences;
- Systematic review of auxiliary systems revealed reliance on single feature.
- Next Steps
- Definition of intermediate risk metrics;
- LMP Demonstration.
24
LMP METHODOLOGY
SUMMARY
This presentation has provided a demonstration of the LMP approach to answering the following questions:
- What are the plant initiating events, event sequences, and accidents that are associated with the design?
- How does the proposed design and its SSCs respond to initiating events and event sequences?
- What are the margins provided by the facilitys response, as it relates to prevention and mitigation of radiological releases within prescribed limits for the protection of public health and safety?
- Is the philosophy of DID adequately reflected in the design and operation of the facility?
25
© 2018 NEI. All rights reserved.
Questions?
26
Backup Slides 27
NRC DESIGN CERTIFICATION REVIEW COSTS REPORTED TO CONGRESS IN 2015 28 Costs have been normalized to 2017 dollars
SYSTEMATIC
- From the start of a reactor design project, the NEI 18-04 methodology systematically provides a clear plan to identify / confirm LBEs, classify SSC, and evaluate the adequacy of defense in depth.
- Knowledge gaps are recognized early and addressed in a deliberate, logical manner.
- Process is reproducible such that different design teams should reach similar conclusions for the same inputs.
29
© 2018 NEI. All rights reserved.
COHERENT
- LMP methodology holistically considers the identification of LBEs, the classification of SSCs and associated special treatment, and the adequacy of defense-in-depth all together, rather than as independent, sequential actions.
- Incorporates data and insights from a wide variety of diverse sources to guide decision making.
30
© 2018 NEI. All rights reserved.
CONSISTENT
- The LMP methodology can be consistently applied across different technologies.
- All technologies are evaluated against the same risk-informed, performance-based targets for safety.
- Technology-neutral, the process does not favor or penalize any particular method for satisfying regulatory outcome objectives and meeting the Commissions safety goals.
- Innovative methods to satisfy safety performance objectives are encouraged.
31
© 2018 NEI. All rights reserved.
TOP DOWN APPROACH IS NEEDED FOR OVERALL COHERENCY AND CONSISTENCY Examples: Generic Methods for being
- NEI 18-04 able to Demonstrate Option 1- Top Down Approach Option 2- Bottom Up Approach
- ARDCs Compliance Requirements Examples:
- Etc.
Design Design Specific Info Req Design and Compliance Basis (e.g., PDCs) 32
EXECUTABLE
- Demonstrations of the NEI 18-04 methodology have been performed successfully on different reactor technologies.
- Methodology accommodates designs at any stage of the design process by accommodating early design and risk information and incorporating feedback loops (which may be entered anytime) throughout as the design matures. Some designers may choose to use the methodology to confirm decisions made previously in the design process.
- NEI 18-04 is logical and within the typical technical capabilities of the designer at each stage of the design process.
33
© 2018 NEI. All rights reserved.
PRACTICAL
- NEI 18-04 elicits diverse sources to guide RIPB decision making, ensuring that viewpoints from throughout an organization are incorporated systematically.
- The process identifies and addresses gaps in knowledge and uncertainties that may otherwise go unnoticed.
- The systematic nature of the NEI 18-04 process is widely understandable, readily integrated with any engineering process, produces a more robust record of safety decision-making, and remains a useful framework throughout the life of the plant.
34
© 2018 NEI. All rights reserved.
DID Backup Slides 35
NRC DEFENSE IN DEPTH PHILOSOPHY
...an approach to designing and operating nuclear facilities that prevents and mitigates accidents that release radiation or hazardous materials. The key is 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. Defense in depth includes the use of access controls, physical barriers, redundant and diverse key safety functions, and emergency response measures.
36
NRC DEFENSE-IN-DEPTH CONCEPT 37
LMP 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.
38
PLANT CAPABILITY DEFENSE-IN-DEPTH ATTRIBUTES Attribute Evaluation Focus PRA Documentation of Initiating Event Selection and Event Sequence Modeling Initiating Event and Event Sequence Insights from reactor operating experience, 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 safety Functional Reliability 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 39
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 unknowns Compensation for Uncertainties (performance variability)
Compensation for unknowns (knowledge uncertainty)
Off-Site Response Emergency response capability 40
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 41
INTEGRATED PROCESS FOR INCORPORATION AND EVALUATION OF DID
- Tasks are not necessarily sequential.
- Tasks can begin early in the conceptual design process and mature with the design evolution.
- All of the attributes included in the DID adequacy evaluation are completed when the design baseline for the license application is submitted.
- Programmatic confirmation of performance and sustained DID continues for life of the plant.
42
EVALUATION OF DID ATTRIBUTE SATISFACTION TASKS
- Plant Capability DID
- Programmatic DID
- IDP Evaluation of Quality and Reliability outcome objectives
- IDP Evaluation of Sources of Uncertainty
- IDP Evaluation of Residual Risk Management strategies 43
GUIDELINES FOR ESTABLISHING ADEQUACY OF PLANT CAPABILITY DEFENSE-IN-DEPTH (TABLE 5-2)
Layer Guideline Overall Guidelines Layer[a]
Quantitative Qualitative Quantitative Qualitative Maintain frequency of plant transients within designed
- 1) Prevent off-normal operation and AOOs cycles; meet owner requirements for plant reliability and availability[b]
- 2) Control abnormal Minimize frequency of No single Maintain frequency of all design or operation, detect failures, challenges to safety-and prevent DBEs DBEs < 10-2/ plant-year operational related SSCs Meet F-C No single design or feature,[c]
Target for all
- 3) Control DBEs within the operational feature[c] no matter analyzed design basis Maintain frequency of all LBEs and relied upon to meet how robust, conditions and prevent BDBEs < 10-4/ plant-year cumulative quantitative objective for is BDBEs risk metric all DBEs exclusively targets with relied upon sufficient[d]
- 4) Control severe plant to satisfy conditions, mitigate No single barrier[c] or plant margins the five consequences of BDBEs Maintain individual risks feature relied upon to limit layers of from all LBEs < QHOs with releases in achieving
- 5) Deploy adequate offsite defense sufficient[d] margins quantitative objectives for protective actions and all BDBEs prevent 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 44 meet this objective.
SSC LAYERS OF DEFENSE 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 fdp0p1p2 LBE-3 No 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 45
GUIDELINES FOR EVALUATION OF PROGRAMMATIC DID (1/2)
Evaluation Attribute Implementation Strategies Evaluation Considerations Focus
- 1. Is there appropriate bias to prevention of AOOs progressing to postulated accidents?
- 2. Has appropriate conservatism been applied in bounding deterministic safety analysis of more risk significant LBEs?
- 3. Is there reasonable agreement between the deterministic safety analysis of DBAs and the upper bound consequences of risk-informed DBA Design Conservatism with Bias to included in the LBE set?
Testing Prevention 4. Have the most limiting design conditions for SSCs in plant safety and risk Quality /
Manufacturing Equipment Codes and Standards analysis been used for selection of safety-related SSC design criteria?
Reliability Construction Equipment Qualification 5. Is the reliability of functions within systems relied on for safety overly O&M Performance Testing dependent on a single inherent or passive feature for risk significant LBEs?
- 6. Is the reliability of active functions relied upon in risk significant LBEs achieved with appropriate redundancy or diversity within a layer of defense?
- 7. Have the identified safety-related SSCs been properly classified for special treatment consistent with their risk significance?
- 1. Have the insights from the Human Factors Engineering program been included in the PRA appropriately?
- 2. Have plant system control designs minimized the reliance on human Operational Command and performance as part of risk-significant LBE scenarios?
Control Practices
- 3. Have plant protection functions been automated with highly reliable Compensation Training and Qualification systems for all DBAs?
for Human Plant Simulators
- 4. Are there adequate indications of plant state and transient performance Errors Independent Oversight and Compensatio for operators to effectively monitor all risk-significant LBEs?
Inspection Programs n for 5. Are the risk-significant LBEs all properly modeled on the plant reference Reactor Oversight Program Uncertainti simulator and adequately confirmed by deterministic safety analysis?
es 6. Are all LBEs for all modes and states capable of being demonstrated on the plant reference simulator for training purposes?
- 1. Are all risk-significant LBE limiting condition for operation reflected in Operational Technical Compensation plant Operating Technical Specifications?
for Specifications
- 2. Are Allowable Outage Times in Technical Specifications consistent with 46 Allowable Outage Times Mechanical assumed functional reliability levels for risk-significant LBEs?
Part 21 Reporting Errors 3. Are all risk-significant SSCs properly included in the Maintenance
GUIDELINES FOR EVALUATION OF PROGRAMMATIC DID (2/2) 47
SPECIAL CONSIDERATIONS OVERVIEW
- Metrics
- LBE Metrics;
- SSC Metrics.
- Margins
- Plant performance margins (LBEs);
- SSC design performance conservatism.
- Uncertainties
- Completeness;
- Analyzed Uncertainties;
- Residual Risks.
- Compensatory Action Decisions
- Choices;
- Impact on Risk;
- Timing;
- Practicality.
48
MARGINS Plant Performance Margins
- Best Estimate
- Doses below low dose threshold
- Event sequence families below QHOs
- With Uncertainty Bands
- Compared to 10CFR 50.34
- Compared to 10 CFR 100
- SSC design performance conservatism
- Use of Consensus Standards
- Deterministic Margins around BE performance 49
LBE RISK-SIGNIFICANCE CRITERIA 50
EVALUATING MARGINS AGAINST F-C TARGET 51
EXAMPLE RISK MARGINS FOR MHTGR Limiting LBE[a] F-C Target LBE Freq. at LBE Mean Dose at LBE Mean Freq. Mean Dose Dose Category Name Dose/plant- Frequency Freq. (Rem)
/plant-yr. (Rem) Margin[e]
yr. [b] Margin[c] [d]
AOO AOO-5 4.00E-02 2.50E-04 4.00E+02 1.00E+04 1.00E+00 4.00E+03 DBE DBE-10 1.00E-02 2.00E-03 6.00E+01 6.00E+03 1.00E+00 5.00E+02 BDBE BDBE-2 3.00E-06 4.00E-03 2.50E+01 8.30E+06 2.50E+02 6.00E+04 Notes:
[a] The Limiting LBE is the LBE with the highest risk significance in the LBE category
[b] Frequency value measured at the LBE mean Dose level from the F-C target, See [2] in Error! Reference source not found.
[c] Ratio of the frequency in note [b] to the LBE mean frequency, mean frequency margin
[d] Dose value measured at the LBE mean frequency from the F-C target, See [4] in Error! Reference source not found.
[e] Ratio of the Dose in Note [d] to the LBE mean dose, Mean Dose Margin Limiting LBE[a] F-C Target 95th 95th 95th 95th LBE Freq. at LBE LBE Percentile Percentile Percentile Dose at LBE Percentile Category Dose/plant-Name Freq./plant- Dose Frequency Freq.(Rem)[d] Dose yr.[b]
yr. (Rem) Margin[c] Margin[e]
AOO AOO-5 8.00E-02 1.10E-03 9.00E+01 1.13E+03 1.00E+00 9.09E+02 DBE DBE-10 2.00E-02 6.00E-03 2.00E+01 1.00E+03 1.00E+00 1.67E+02 BDBE BDBE-2 1.00E-05 1.50E-02 8.00E+00 8.00E+05 1.00E+02 6.67E+03 Notes:
[a] Limiting LBE is LBE with highest risk significance in LBE Category
[b] Frequency value measured at the LBE 95th percentile Dose level from the F-C target, See [6] in Error!
Reference source not found.
[c] Ratio of the frequency in note [2] to the LBE 95th percentile frequency, 95th percentile Frequency Margin
[d] Dose value measured at the LBE 95th percentile frequency from the F-C target, See [8] in Error! Reference source not found.
[e] Ratio of the Dose in note [d] to the LBE 95th percentile dose, 95th percentile Dose Margin 52
UNCERTAINTIES
- Completeness
- PRA completeness for identified hazards
- Sources of risk-significant uncertainties
- Treatment of radiological and other hazards not included in PRA
- Analyzed
- Data Availability
- Model Maturity
- Performance History
- Residual Risks
- EPZ basis
- EP response effectiveness
- Tech Spec Completeness
- AOT basis
- Monitoring of Plant Long Term Performance
- Etc.
53
COMPENSATORY ACTION DECISIONS
- Choices
- Plant Capability
- Programmatic
- Mix
- Impact on Risk
- Improve Plant Capability
- LBE Outcome Changes
- Layers of Defense increase or independence improvements
- Improve Plant Performance Assurance
- Programmatic actions
- Reduction of Risk Significant Sources of Uncertainty
- Reduce Residual Uncertainties
- Siting and Emergency Planning performance
- External Independent Oversight
- Timing - Life Cycle Considerations
- Practicality
- When is enough, enough?
54
SSC CLASSIFICATION
SUMMARY
- LMP retains the NGNP SSC safety categories of SR, NSRST, and NST.
- All safety significant SSCs classified as SR or NSRST.
- SR SSCs are not necessarily 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 event sequences.
- Special treatment defined / confirmed via integrated decision process using forward fit adaptation of 10 CFR 50.69 process.
55
ACRS Future Plant Designs Subcommittee Draft Regulatory Guide (DG) 1353 and Related Commission Paper Technology-Inclusive, Risk-Informed, Performance-Based Approach to Inform the Content of Applications for Licenses, Certifications, and Approvals for Non-Light Water Reactors, October 30, 2018 (PM) 1
Draft SECY Paper
- Paper
- The purpose of this paper is to seek Commission approval of the U.S.
Nuclear Regulatory Commission (NRC) staffs recommendation to adopt a technology-inclusive, risk-informed, and performance-based methodology for informing the licensing basis and content of applications for licenses, certifications, and approvals for non-light-water-reactors (non-LWRs).
- Enclosure 1, Background
- Enclosure 2, Technology-Inclusive, Risk-Informed, Performance-Based Approach 2
Policy Background
- Advanced Reactor Policy Statement
- Pre-application evaluations (e.g., PRISM, MHTGR)
- SECY-93-092, Issues Pertaining to the Advanced Reactor (PRISM, MHTGR, and PIUS) and CANDU 3 Designs and Their Relationship to Current Regulatory Requirements
- SECY-03-0047, Policy Issues Related to Licensing Non-Light Water Reactor Designs
- Related initiatives to develop and implement risk-informed, performance-based regulation 3
Policy Background SECY-03-0047, Policy Issues Related to Licensing Non-Light Water Reactor Designs, and the related staff SRM dated June 26, 2003.
- Greater emphasis can be placed on the use of risk information by allowing the use of a probabilistic approach in the identification of events to be considered in the design, provided there is sufficient understanding of plant and fuel performance and deterministic engineering judgment is used to bound uncertainties;
- A probabilistic approach for the safety classification of structures, systems, and components is allowed; and
- The single-failure criterion can be replaced with a probabilistic (reliability) criterion.
4
Event Selection
- Consistent with SRM approving the use of a probabilistic approach to identify events provided there is sufficient understanding of plant and fuel performance and engineering judgment is used to address uncertainties
- Including a lower frequency range for licensing basis events, when combined with other considerations and engineering judgement, is an inherent part of a risk-informed approach and is consistent with the Commissions SRM
- The F-C targets support defining needed SSC capabilities and reliabilities to support the design process and to inform the content of applications, considering uncertainties and multi-module issues
- Consistent with the Commissions SRM approving replacement of the single-failure criterion with a probabilistic (reliability) criterion 5
Safety Classification &
Performance Criteria
- The safety classification of SSCs and determination of performance criteria are directly related to and performed in an iterative process along with the identification and assessment of LBEs and the assessment of defense in depth
- Systematic approach to assessing and determining appropriate relationships between the needed capabilities and reliabilities for SSCs and the role of those SSCs in mitigating and preventing LBEs 6
Assessing Defense in Depth
- Framework that includes probabilistic and deterministic assessment techniques to establish defense in depth using a combination of plant capabilities and programmatic controls
- Assessments performed using several approaches to assess a reactor design and determine if additional measures are appropriate to address an over-reliance on specific features or to address uncertainties
- Includes verification that two or more independent plant design or operational features are provided to meet the guidelines for each licensing basis event
- Methodology includes use of an Integrated Decision-Making Process
- Staff is not proposing to more universally define DID criteria and seeks Commission acceptance of the NEI 18-04 approach for this specific case.
7
Informing Content of Applications
- NEI 18-04 provides useful guidance for reactor designers and the NRC staff for selecting and evaluating licensing basis events, identifying safety functions and classifying SSCs, selecting special treatment requirements, identifying appropriate programmatic controls, and assessing defense in depth
- Taken together, these activities support documenting the safety case and determining the appropriate scope and level of detail in applications for licenses, certifications, or approvals for non-LWRs 8
Recommendation The staff recommends that the Commission approve the use of the technology-inclusive, risk-informed, and performance-based approach described in NEI 18-04 and DG-1353 for identifying LBEs, classifying SSCs, and assessing the adequacy of defense in depth.
These key aspects of the proposed approach will also be used to inform the appropriate scope and level of detail for information to be included in applications to the NRC for licenses, certifications, and approvals for non-LWRs.
9
Working draft DG 1353 Scope
- Methodology supports identifying the appropriate scope and depth of information provided in applications for licenses, certifications, and approvals
- 10 CFR 50.34, Contents of applications; technical information, describes the minimum information required for (a) preliminary safety analysis reports supporting applications for a construction permit, and (b) final safety analysis reports supporting applications for operating licenses.
- 10 CFR 52.47, Contents of applications; technical information, describes the information to be included at an appropriate level in final safety analysis reports supporting applications for standard design certifications (DCs).
- 10 CFR 52.79, Contents of applications; technical information in final safety analysis report, describes the information to be included at an appropriate level in final safety analysis reports supporting combined licenses (COLs).
- 10 CFR 52.137, Contents of applications; technical information, describes the information to be included at an appropriate level in final safety analysis reports supporting standard design approvals (SDAs).
- 10 CFR 52.157, Contents of applications; technical information in final safety analysis report, describes the information to be included at an appropriate level in final safety analysis reports supporting manufacturing licenses (MLs).
10
Working draft DG 1353 Findings Licensing Basis Events
- Staff Position: NEI 18-04 provides an acceptable method for identifying and categorizing events with the following clarifications:
a) The staff emphasizes the cautions in NEI 18-04 that the F-C target figure does not depict acceptance criteria or actual regulatory limits. The anchor points used for the figure are surrogates for other measures that may be expressed in different units, time scales, or distances. The F-C target provides a reasonable approach to be used within a broader, integrated approach to determine risk significance and support SSC classification and confirm the adequacy of DID [defense in depth].
b) The F-C target and related discussions in NEI 18-04 include a frequency of 5x10-7 per plant-year to define the lower range of beyond design basis events.
This demarcation of lowest event frequencies on the F-C target and category definitions should not be considered a hard and fast cutoff but should instead be considered in the context of other parts of the methodology described in NEI 18-04.
These other considerations include the role of the integrated decision-making panel, DID assessments, accounting for uncertainties, and assessing for potential cliff-edge effects.
11
Working draft DG 1353 Findings Licensing Basis Events c) NEI 18-04 describes a set of DBEHLs that will determine the design basis seismic events and other external events that the safety related SSCs will be required to withstand. When the DBEHLs are determined using NRC-approved methodologies, this approach is generally consistent with current practices and provides acceptable protection of safety-related SSCs. When supported by available methods, the PRA model is expected to address the full spectrum of internal events and external hazards that pose challenges to the capabilities of the plant, including external hazard levels exceeding the DBEHLs. The inclusion of external events within the BDBE category supports the overall risk-informed approach in NEI 18-04 and the DID assessments described in subsequent sections. NEI 18-04 states: When supported by available methods, data, design and site information, and supporting guides and standards, these DBEHLs will be informed by a probabilistic external hazards analysis and included in the PRA after the design features that are included to withstand these hazards are defined. To the degree that applicants propose methods to identify DBEHLs that have not been previously reviewed and approved by the NRC, the staff would review the proposed methodologies on a case-by-case basis.
12
Working draft DG 1353 Findings Licensing Basis Events d) NEI 18-04 describes how the application of a single failure criterion is not deemed to be necessary for non-LWRs using the methodology because they will employ a diverse combination of inherent, passive, and active design features to perform the required safety functions across layers of defense and will be subjected to an evaluation of DID adequacy. The process described in NEI 18-04 includes assessing event sequences (including reliability and availability of SSCs and combinations of SSCs) over a wide range of frequencies and establishing risk and safety function reliability measures.
The approach described in NEI 18-04 is consistent with the Commissions SRM approving the recommendation in SECY-03-0047 to replace the single-failure criterion with a probabilistic (reliability) criterion.
e) The methodology in NEI 18-04 includes a potentially expanded role for PRA beyond that currently required by 10 CFR Part 52. The staffs review of the PRA prepared by a designer could be facilitated by the NRC endorsement of consensus codes and standards (e.g., ASME/ANS RA-S-1.4, Probabilistic Risk Assessment Standard for Advanced Non-LWR Nuclear Power Plants) and the use of that approved standard by the designer.
13
Working draft DG 1353 Findings Safety Classification & Performance Criteria
- Staff Position: NEI 18-04 provides an acceptable method for assessing and classifying SSCs as safety related, non-safety related with special treatment, or non-safety related with no special treatment. The staff offers the following clarification:
a) The SSC classifications and logic outlined in NEI 18-04 are part of an integrated methodology, which includes a defined relationship between licensing basis events, equipment classification, and assessments of DID. The classifications and related outcomes may not be applicable for alternative approaches that do not follow the other parts of the methodology described in NEI 18-04.
14
Working draft DG 1353 Findings Evaluation of Defense-in-Depth Adequacy
- Staff Position: NEI 18-04 provides an acceptable method for assessing the adequacy of DID to be provided by plant capabilities and programmatic controls, with the following clarifications:
a) Section 5.9.6, Considerations in Documenting Evaluation of Plant Capability and Programmatic DID, discusses change control processes following the issuance of a license, certification, or approval. The staff plans to address such change control processes in future guidance documents and therefore makes no findings on this section of NEI 18-04.
15
Working draft DG 1353 Findings Other Considerations
- For non-LWRs, the spectrum of events is expected to be the LBEs as described in NEI 18-04, adjusted as necessary to reflect the specific criteria in the emergency planning decisionmaking process (e.g., dose calculations over 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> from the release of radioactive materials in DG-1350 versus 30 days in NEI 18-04 for plotting on the F-C target).
- Mechanistic Source Term
- While not addressed in detail within NEI 18-04, the development of mechanistic source terms for designs and specific event families is another element of an integrated, risk-informed, performance-based approach to designing and licensing non-LWRs. Applicants are expected to provide in their applications or related reports a description of their mechanistic source terms, including the retention of radionuclides by barriers and the transport of radionuclides for all barriers and pathways to the environs. Where applicable, a facility may have multiple mechanistic source terms and specific event sequences to address various systems that contain significant inventories of radioactive material.
16
Working draft DG 1353 Findings Informing Content of Applications
- NEI 18-04 provides useful guidance for applicants to identify and provide the appropriate level of information needed to satisfy parts of the regulatory requirements in 10 CFR 50.34, 10 CFR 52.47, 10 CFR 52.79, 10 CFR 52.137, and 10 CFR 52.157.
- Combination of deterministic evaluations and probabilistic risk assessments
- Information needed on fuel, primary, and other barriers to define limitations, performance characteristics, and as input to mechanistic source term
- Information needed on SSCs and programmatic controls associated with key safety functions
- Scope and depth for other information (e.g., ancillary plant systems) to be determined based safety/risk significance (i.e., roles in preventing or mitigating licensing basis events)
- Level of detail can also reflect potential performance-based approaches (see Introduction, Part 2, to NUREG 0800) 17
DG 1353 & Related SECY
- Target Schedule
- ACRS Subcommittee June 19, 2018
- Draft NEI 18-04, DG-1353, SECY Sept 28, 2018 (public, to ACRS)
- ACRS Subcommittee Oct 30, 2018
- ACRS Full Committee Dec 6, 2018
- Issue DG-1353 Dec 21, 2018
- Issue SECY early 2019
- ACRS Interactions mid 2019