ML19184A134

Transcript of the Advisory Committee on Reactor Safeguards Nuscale Subcommittee - May 14 2019 (Open)

ML19184A134
Person / Time
Issue date:	05/14/2019
From:	Michael Snodderly Advisory Committee on Reactor Safeguards
To:
Snodderly, M, ACRS
References
NRC-0328
Download: ML19184A134 (391)

Text

Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION

Title:

Advisory Committee on Reactor Safeguards NuScale Subcommittee Docket Number: (n/a)

Location: Rockville, Maryland Date: Tuesday, May 14, 2019 Work Order No.: NRC-0328 Pages 1-377 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 NuSCALE SUBCOMMITTEE 8 + + + + +

9 TUESDAY 10 MAY 14, 2019 11 + + + + +

12 ROCKVILLE, MARYLAND 13 + + + + +

14 The Subcommittee met at the Nuclear 15 Regulatory Commission, Two White Flint North, Room 16 T2D10, 11545 Rockville Pike, at 8:30 a.m., Michael L.

17 Corradini and Vesna B. Dimitrijevic, Co-Chairs, 18 presiding.

19 20 COMMITTEE MEMBERS:

21 MICHAEL L. CORRADINI, Co-Chair 22 VESNA B. DIMITRIJEVIC, Co-Chair 23 RONALD G. BALLINGER, Member 24 DENNIS BLEY, Member 25 CHARLES H. BROWN, JR. Member NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

2 1 JOSE MARCH-LEUBA, Member 2 JOY L. REMPE, Member 3 GORDON R. SKILLMAN, Member 4 MATTHEW W. SUNSERI, Member 5

6 ACRS CONSULTANT:

7 STEPHEN SCHULTZ 8

9 DESIGNATED FEDERAL OFFICIAL:

10 MIKE SNODDERLY 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

3 1 CONTENTS 2 Call to order and Opening Remarks . . . . . . . . 4 3 Opening Remarks . . . . . . . . . . . . . . . . . 7 4 Chapter 19.0, "Probabilistic Risk Assessment . . 8 5 and Severe Accident Evaluation," and 19.1, 6 "Probabilistic Risk Assessment (PRA),"

7 NuScale Design Certification Application (DCA) 8 Chapter 19.1 PRA Safety Evaluation (SE) with 9 Open Items . . . . . . . . . . . . . . . . . . 188 10 Public Comments . . . . . . . . . . . . . . . . 281 11 Subcommittee Discussion . . . . . . . . . . . . 284 12 Adjourn . . . . . . . . . . . . . . . . . . . . 377 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 8:26 a.m.

3 CO-CHAIR CORRADINI: The meeting will come 4 to order.

5 This is a meeting of the Advisory 6 Committee on Reactor Safeguards, NuScale Subcommittee.

7 My name is Mike Corradini. I am the Co-Chair with Dr.

8 Dimitrijevic for today's meeting.

9 Members in attendance are Ron Ballinger, 10 Gordon Skillman, Matt Sunseri, Joy Rempe, Jose March-11 Leuba, Dennis Bley, Dr. Dimitrijevic, and myself. And 12 I think we will have Harold Ray joining us by phone.

13 And Mike Snodderly is the Designated 14 Federal Official for this meeting.

15 The Subcommittee will review the staff's 16 evaluation of Chapter 19, "Probabilistic Risk 17 Assessment and Severe Accident Evaluation," of the 18 NuScale's Design Certification Application. Today we 19 have members of the NRC staff and NuScale to present 20 to the Subcommittee.

21 The ACRS was established by statute and is 22 governed by the Federal Advisory Committee Act, or 23 FACA. That means the Committee can only speak through 24 its published letter reports. We hold meetings to 25 gather information to support our deliberations, 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

5 1 interested parties who wish to provide comments can 2 contact our office requesting time after the meeting 3 announcement is published in The Federal Register.

4 That said, we set aside 10 minutes for 5 comments for members of the public attending or 6 listening to our meetings, and written comments are 7 also welcome.

8 I forgot to mention we have our esteemed 9 consultant, Dr. Schultz, with us. I apologize. I 10 looked past you.

11 The ACRS section of the U.S. NRC's public 12 website provides our Charter, Bylaws, letter reports, 13 and full transcripts of all full and subcommittee 14 meetings, including slides presented here.

15 The rules for participation in today's 16 meeting are announced in The Federal Register notice 17 dated on May 6th, 2019. The meeting was announced as 18 a previous open/closed meeting. We may close the 19 meeting after the open portion to discuss proprietary 20 materials and presenters can defer questions that 21 should be asked, or should be answered, in the private 22 session, in the closed session.

23 I'll just go off-script and note that, if 24 we start asking something of NuScale or the staff, you 25 guys have to keep us disciplined, and we'll hold 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

6 1 off until the closed session.

2 No written statement or request for making 3 an oral statement to the Subcommittee has been 4 received from the public concerning this meeting.

5 A transcript of the meeting is being kept 6 and will be made available as stated in The Federal 7 Register notice. Therefore, we request that 8 participants in this meeting use the microphones 9 located throughout the meeting room when addressing 10 the Subcommittee. Participants should first identify 11 themselves and speak with sufficient clarity and 12 volume so they can be readily heard.

13 We have a bridge line established for the 14 public to listen in to the meeting. To minimize 15 disturbances, the public line is being kept in listen-16 in-only mode. And to avoid disturbances, I request 17 that all members of the Committee and attendees put 18 their electronic devices on mute or noise-free mode, 19 so we don't get interrupted by noises.

20 I will note for myself that I participated 21 in a PIRT for Severe Accident Phenomena with NuScale 22 in 2009, and we've put that on the record in the past 23 when we met in March and October.

24 MEMBER REMPE: And I also need to 25 acknowledge that I participated not only in that PIRT, 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 but I performed some other activities for the 2 Applicant back in around that timeframe.

3 CO-CHAIR CORRADINI: So, I'm going to turn 4 it over to Dr. Dimitrijevic to start us off.

5 CO-CHAIR DIMITRIJEVIC: This is our third 6 meeting, right?

7 CO-CHAIR CORRADINI: Her mic is on. I 8 think you need to be louder.

9 CO-CHAIR DIMITRIJEVIC: I need to be 10 louder? That was never said to me.

11 (Laughter.)

12 All right.

13 CO-CHAIR CORRADINI: There we go.

14 CO-CHAIR DIMITRIJEVIC: So, you presented 15 the PRA which shows that you're meeting safety goals 16 with the large margins and that your risk is 17 practically zero, because we don't want to really 18 discuss the numbers, although it's 10 to the minus 9.

19 So, what I heard, that we can concentrate 20 today -- even I saw some of the slides that we saw 21 before -- and as we go through them, we can 22 concentrate on the important issues and, important, 23 staff, how sure we are that that's true, that your 24 risk is practically zero.

25 So, for example, important assumptions 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 that impact on the risk, the uncertainties, 2 sensitivities, uncertainty, multi-module issues which 3 are characteristic for your plant, that's why they are 4 interesting and important for us. So, just keep that 5 in mind when you are giving this presentation, and we 6 can just assess it. Did I say "uncertainty" three 7 times? That was my goal, to make sure that we do 8 address those. Okay? So, please.

9 Thank you.

10 MS. NORRIS: Thank you, Vesna.

11 Good morning, everyone.

12 I'm Rebecca Norris with NuScale. As has 13 been stated, we are here to present Chapter 19 for the 14 ACRS Subcommittee presentation. This is on 15 "Probabilistic Risk Assessment and Severe Accident 16 Evaluation". The packet that you have includes 17 Chapter 19 items for both today and tomorrow. As you 18 know, if you're looking up an acronym, it will be in 19 the very back of the packet. We just made one acronym 20 list.

21 Today, we will be presenting 19.0 and 22 19.1, "Probabilistic Risk Assessment and Severe 23 Accident Evaluation," and then, the general overview.

24 So, this is the presentation for both of those.

25 The presentation team for today is myself, 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 on the bottom, Rebecca Norris. I am Supervisor in 2 Licensing.

3 Go ahead. You're going to introduce 4 yourselves?

5 MS. BRISTOL: Sarah Bristol, Supervisor of 6 the PRA group at NuScale Power.

7 MR. MULLIN: Etienne Mullin. I'm an 8 Analyst in the PRA group.

9 MR. GALYEAN: I'm Bill Galyean, a 10 consultant to the PRA group.

11 CO-CHAIR CORRADINI: Just one thing. We 12 have such a terrific system here, you have to speak up 13 because our mics are of various qualities.

14 MS. NORRIS: And with that, we can go 15 ahead and get started.

16 MS. BRISTOL: Okay. An overview of 19.0.

17 This really just goes over the general regulations 18 that were looked at in describing the PRA, both the 19 PRA and severe accidents. We note that we performed 20 the PRA for a single module, and we looked at all 21 modes of operations and both internal and external 22 events. The PRA demonstrates that NuScale design 23 exceeds those safety goals with significant margin, as 24 described further in Chapters 19.1 and 2.

25 In Section 19.1 of the PRA, the objective 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 was to assess the risks associated with all modes and 2 all hazards for a single module. And that is what is 3 involved in that section. We looked at multi-module 4 risks separately using a systematic process, which we 5 will discuss further in the presentation. But, 6 overall, we looked at Level 1, core damage frequency, 7 and we looked at level 2, large release frequency.

8 And we did that for full power internal events, low 9 power and shutdown, and we did crane failure analysis.

10 We did internal fires, floods, external floods, high 11 winds. And we did a PRA-based seismic margin 12 assessment.

13 The quality process we used, being a 14 Design Certification, ISG 28 came out right in the 15 middle of the time that we were doing our PRA. We've 16 been involved with the Standard Committee work, and 17 from that work, out came this ISG 28. And so, what we 18 did was we still looked at the ASME/ANS PRA standard, 19 and we looked at all the supporting requirements. And 20 we did individual self-assessments for each element of 21 the standard, and we evaluated those independently in 22 each notebook, if you will, for the PRA we looked at 23 the standard, and when the standard directed -- or Reg 24 Guide 1200, as well as the ISG and the standard, is 25 what we used to assess the quality for the PRA.

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 We had both an independent review of those 2 independent self-assessments from outside consultants 3 and we also had an independent expert panel.

4 MEMBER BLEY: Sarah?

5 MS. BRISTOL: Yes?

6 MEMBER BLEY: You make a big point that 7 you didn't have a peer review. Those sound like peer 8 reviews to me. Can you explain the difference?

9 MS. BRISTOL: Well, from our perspective, 10 there was really no significant difference in what we 11 did and a peer review. The only difference was that 12 we didn't officially have a team with a leader come in 13 and evaluate each one of those under the NEI guidance 14 of what a peer review is. But we finished our PRA --

15 MEMBER BLEY: Did you review all the 16 things that are called out in the standard for a peer 17 review?

18 MS. BRISTOL: Yes, we did, and we 19 evaluated those, and we said -- we met and there's a 20 slide in a couple that discussed that. In general, 21 the self-assessment we did compared to the 22 consultants, we exceeded even some of the points of 23 the ASME/ANS standard.

24 And then, there were a few that we 25 couldn't meet, walkdowns data, calculations we didn'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

12 1 have at the time. And so, from our perspective, we 2 did as much as we could without an official peer 3 review team.

4 MEMBER BLEY: Okay.

5 MEMBER BALLINGER: With respect to this 6 self-assessment documented by notebook authors, I'm 7 going to be kind of a pain in the rear end about this.

8 Your claim is that the plant is extremely safe and the 9 probabilities of certain events are very, very, very 10 low. Fine.

11 Along with that claim goes a 12 responsibility, at least in my mind, that you have to 13 be very careful about and very deliberate on 14 determining the uncertainties. So, when you say 15 "self-assessment," did you folks establish, 16 independent of what I call "a murder board," within 17 the company or outside the company, where the charge 18 of that board was to find out the mistakes? In other 19 words, not verify the calculations, but take an 20 outside view and say, how can I make this fail? So, 21 did you guys do that?

22 MS. BRISTOL: And so, what we did was we 23 had an expert peer review group come in. And so, on 24 this panel, we had Dr. Apostolakis be the Chair of 25 that committee. Mark Cunningham was on there, Rick 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 Grantom, Dave Moore, Per Peterson. And they all 2 evaluated our design, not necessarily from the 3 standard model, but just what is in your PRA and just 4 from a high level.

5 Next slide, please.

6 MEMBER BALLINGER: Okay, but I guess that 7 not too many of these folks have ever actually put a 8 calculator in front of them and done the calculation 9 in a very long time.

10 (Laughter.)

11 My issue is, did you guys find some people 12 that have done calculations that are independent, that 13 actually could go and find out where the errors were, 14 or discover paths or cutsets, or whatever, that they 15 hadn't thought about?

16 MS. BRISTOL: And, yes, that was the next 17 consulting group that we had -- next slide -- was they 18 did look at all of those supporting requirements, and 19 those were individuals that had performed different 20 analyses in the industry recently for items from 21 security, operating plants, different events. So, 22 they have been applying the PRA recently and are 23 familiar with the standard. And those individuals 24 evaluated our PRA and looked at each of the individual 25 requirements, and if they thought we met, if 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

14 1 thought we didn't, if they thought there was a gap 2 that we could close between DCA and COL. And so, we 3 did have both that expert side as well as independent 4 individuals currently in the industry applying PRA 5 practices.

6 MEMBER MARCH-LEUBA: Don't go so fast.

7 MS. BRISTOL: Yes.

8 MEMBER MARCH-LEUBA: Following up on that, 9 my concern with PRA analysis -- and you know I don't 10 like it as much -- is not with the math or with the 11 reviews. It's with the input data. This reactor has 12 a lot of first-of-a-kind components, divisions of the 13 protection system, all of those valves around the 14 containment; they're one of a kind. They haven't even 15 been built yet. Okay? How do we get any confidence 16 that the input data for failure rates that you 17 assigned for this particular valve is acceptable or 18 conservative? Because not having built one and not 19 having operated these in the field, the uncertainty is 20 tremendous on the input data. Please.

21 MR. GALYEAN: So, this is Bill Galyean.

22 There's a couple of aspects of the 23 question you just asked. I mean, on one side, some of 24 the components that are in the NuScale design are 25 standard nuclear power plant components. And there, 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 we used the generic industry operating data to 2 generate failure rates. Our expectation is that the 3 performance of the components in the NuScale design 4 will be no worse than what has been experienced in the 5 industry.

6 However, there are some components in the 7 NuScale design that are relatively unique. Okay? And 8 there, we have done detailed kind of piece-part 9 analyses of these components, where we dissect it down 10 into, as I said, individual piece-part. We come up 11 with estimates for each, failure rate estimates for 12 each individual piece-part, combine them into an 13 overall failure rate.

14 Again, the expectation is not that we 15 precisely determine what the failure rate is, but that 16 we come up with a failure rate that we believe is 17 conservative. Okay? And there, we do account for 18 uncertainty. We do put uncertainty bands on the 19 failure probabilities, the failure rates, and combine 20 them.

21 An example of that is particular on the 22 initiating event data that we use. Although we do use 23 the industry experience, we deliberately expanded the 24 uncertainty bands on the failure, on the rates 25 predicted were generated in the industry. And so, by NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

16 1 expanding the uncertainty bands, we feel like we 2 encompass what the expectation is for a NuScale 3 design.

4 Getting back to the quality process and 5 the peer review, again, it was kind of a two-pronged 6 process. We did have the expert panel, the George 7 Apostolakis panel, come in to do a high-level 8 overview, to look at what we were doing globally and 9 identify, or at least point to, different aspects that 10 they thought we should be focusing more attention on.

11 And, in particular, they were very keen on looking at 12 what we did for multi-modules, for example, and a few 13 other things.

14 Sarah already alluded to the self-15 assessment that we did. Now, in the self-assessment, 16 of course, we used the ASME/ANS PRA standard which 17 goes through each of the individual supporting 18 requirements on what's expected for a PRA. And 19 NuScale staff evaluated each one of those, and then, 20 we had a separate pair of external consultants come in 21 and review the NuScale self-assessment. And they made 22 some comments about how we evaluated each supporting 23 requirement. In some cases, they thought we were 24 optimistic; in other cases, they thought we were 25 conservative in how we evaluated the supporting 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 requirement match.

2 So, NuScale has been very keen on assuring 3 that the PRA is a high-quality product, and we have 4 devoted a lot of time and effort, through the expert 5 panel and through the self-assessment, to uncover any 6 potential deficiency that may have existed.

7 CO-CHAIR CORRADINI: Maybe this is for 8 later, maybe it's Chapter 3, but I'm interested in the 9 connection with -- let's just take the IAB as an 10 example or the RVV or the RRV. Those are, from my 11 perspective, somewhat unique. And were those the ones 12 that you did more of what you call a piece-part 13 analysis?

14 MR. GALYEAN: Exactly.

15 CO-CHAIR CORRADINI: Okay.

16 MR. GALYEAN: Exactly.

17 CO-CHAIR CORRADINI: So, how does that, 18 then, feed into potential testing going into a first 19 module or any new module being constructed, such that 20 you're -- what are you looking for in terms of testing 21 -- and maybe this is not here; maybe this is in 14 or 22 in 3 -- relative to assuring that the reliability is 23 of some level in comparison to what you're estimating?

24 MR. GALYEAN: Yes, that's something that 25 we can't answer from the perspective of the PRA. 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

18 1 mean, certainly the expectation is that the vendor 2 will be doing testing of the completed valve.

3 MEMBER BLEY: Well, can I interrupt you 4 there? We've had other cases where we've had the 5 design certs with unique parts. And when they came to 6 this point, they had run extensive tests and were able 7 to report those tests, the results of them. I'm not 8 quite sure why it's not your responsibility.

9 MR. GALYEAN: No, I mean, the PRA group at 10 NuScale. I mean, NuScale will certainly, there will 11 certainly be some testing done. In fact, there 12 already has been testing of the prototype --

13 MEMBER BLEY: Have you given a prioritized 14 list of what things need to be tested and why?

15 MR. GALYEAN: We have not coordinated with 16 the design folks on that aspect.

17 CO-CHAIR CORRADINI: So, to get to my 18 question, I should wait until 14 or 3? That's what 19 you're really telling me?

20 MR. GALYEAN: Yes.

21 CO-CHAIR CORRADINI: Okay. I just want to 22 make sure that what you estimate is not -- I'll pick 23 a number -- must less than what one is testing to see 24 a level of performance; that's all. I'm just looking 25 for the connection. Because I think Jose's question 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 is a fair one, which is it's really the input 2 reliability that tends to drive the result.

3 MEMBER BLEY: One more question in this 4 area. You mentioned this group of components where 5 you had to do piece-part analysis, if you will. Are 6 you going to have a slide that shows what those were?

7 MR. GALYEAN: I believe we do have a slide 8 that alludes to that. It's just a single bullet item 9 that --

10 MEMBER BLEY: How many of those kind of 11 things were there?

12 MR. GALYEAN: Well, the ECCS valves for 13 one. I'm trying to think.

14 CO-CHAIR CORRADINI: Those are the three 15 that I thought were unique.

16 MR. GALYEAN: Three?

17 CO-CHAIR CORRADINI: Well, RRV, RVV.

18 MR. GALYEAN: Oh, yes, I was just using 19 collectively the ECCS valves.

20 MEMBER SKILLMAN: And there's a crane and 21 there's some other stuff in there that's very unique.

22 MR. GALYEAN: Yes, certainly the crane, 23 yes. We did an evaluation.

24 MEMBER SKILLMAN: I'd like to weigh into 25 this. I'd like to go back to slide 6, please. As 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

20 1 review Chapter 19, the question that is overarching in 2 my mind is, to what extent has the PRA, and those who 3 reviewed the PRA, fully comprehended that this plant 4 is aquatic? This might as well be a Jules Verne 5 plant. Everything is underwater. And this is the 6 only plant I've ever experienced in my over 50 years 7 that is 100 percent aquatic.

8 And that has some great advantages in 9 terms of source term, in terms of decay heat removal.

10 Those of us who have been around water and the aquatic 11 environment know that that environment creates 12 challenges that are unique in terms of chemistry, 13 material degradation, operability, inspectability, 14 those types of things. So, unless the people who were 15 reviewing this had that aquatic lens adjusted for 16 their review, it would seem to me that the 17 uncertainties are extraordinary.

18 MR. GALYEAN: Well, none of the components 19 that were modeled in the PRA are underwater. Okay?

20 MEMBER SKILLMAN: Oh, they are.

21 MR. GALYEAN: They're not submerged.

22 They're not -- you know, I mean, they're protected.

23 They're not -- I don't know how to say it. I mean, 24 certainly the containment vessel is underwater, but 25 things like the valves, the containment isolation 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 valves, the ECCS valves, they're not maintained 2 underwater.

3 But, certainly, the reviewers that we had, 4 both in the Apostolakis panel and in the review of the 5 self-assessment, they were all familiar with the 6 design. And, of course, I mean, they had many of the 7 questions that you all are asking, things about the 8 crane and the movement and the reliability of the 9 components.

10 So, I hope that -- I don't know what else 11 to say. I mean, the valves are not maintained 12 underwater.

13 MEMBER MARCH-LEUBA: Yes, what you are 14 saying is that containment is in a vacuum. So, most 15 of the active components are in a vacuum. My concern 16 with being underwater is that they are inaccessible, 17 that they're difficult to work with, difficult to 18 maintain.

19 MR. GALYEAN: Most of the equipment that 20 needs to be maintained, like the containment isolation 21 valves, they're at the top of the module and there's 22 a maintenance platform that allows access to them.

23 Most of that will take place during refueling, when 24 the upper module is in the dry dock.

25 MEMBER SKILLMAN: Okay. Let me ask one 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 more on this slide. The reason I asked you to put 2 this up is because you identify low power in shutdown, 3 including crane failure. When you're in the refueling 4 mode and you're moving that module -- and the module 5 is listed at 762 tons in some of your documents; it's 6 730 in a couple of others -- how does one make the 7 connection between low-power shutdown, an event 8 involving the module, and an adjacent module?

9 MR. GALYEAN: We've looked at that. We've 10 looked at the potential for crane failure resulting in 11 a module being dropped off of the crane and possibly 12 impacting an operating module. We've assessed the 13 likelihood of that. And then, we've done a 14 qualitative assessment as to what the potential impact 15 might be on the operating module.

16 The impact to an operating module is 17 insignificant compared to the hazards that are already 18 accounted for in the internal event initiating events.

19 And, of course, we have done a separate analysis on 20 the crane failure and the potential for a dropped 21 module.

22 MEMBER SKILLMAN: Okay. Bill, I 23 understand your sermonette about the equipment 24 basically being in a protected environment inside the 25 containment. I got that message and I understood 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

23 1 when I asked the question.

2 But I want to continue to project my 3 concern. Here you have 12 of these large machines 4 largely underwater for 60 years. I would suggest that 5 that is an operating environment or a physical 6 environment that is unlike any we've dealt with 7 before. And it would seem to me that there are some 8 challenges that we really haven't tumbled to because 9 of that unique situation. And that's all I'm trying 10 to communicate.

11 MR. GALYEAN: Fair enough.

12 MEMBER SKILLMAN: Thank you.

13 MEMBER BROWN: I wanted to make one other 14 along the uncertainty comment that Dennis brought up.

15 PARTICIPANT: Can anyone else on this line 16 hear what's going on at the ACRS meeting? I haven't 17 had any audio for this meeting yet, and I'm wondering 18 if they're running a little late.

19 CO-CHAIR CORRADINI: Why don't we just 20 hold a minute so you can clarify? We can hold for a 21 minute or two.

22 MEMBER MARCH-LEUBA: We need to hold. We 23 need to hold because they're not hearing it.

24 CO-CHAIR CORRADINI: Yes.

25 MEMBER MARCH-LEUBA: It's a public NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

24 1 meeting.

2 (Whereupon, the above-entitled matter went 3 off the record at 8:53 a.m. and resumed at 8:56 a.m.)

4 CO-CHAIR CORRADINI: All right, let's keep 5 on going.

6 MEMBER BLEY: Okay, Bill, I'm going to get 7 you with one more.

8 CO-CHAIR CORRADINI: Oh, wait a minute.

9 I think Charlie was in the middle of something.

10 MEMBER BLEY: Oh, I forgot Charlie was 11 talking.

12 MEMBER BROWN: That's all right.

13 Everybody forgets me. I almost forgot my question, 14 but not quite.

15 Along the line of, excuse me, 16 uncertainties, if I recall properly, the pressurizer 17 level is one of the trip or trigger functions for a 18 lot of the operations relative to the module 19 protection system and some other type of stuff, and 20 it's presently just supposed to be a radar-based 21 system if I'm not mistaken. Isn't that correct? Did 22 I get that right?

23 (Simultaneous speaking.)

24 MEMBER REMPE: That's what they told us --

25 That was --

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 MEMBER BROWN: Pardon?

2 (Off-mic comments.)

3 MEMBER REMPE: That was actually presented 4 by NuScale and acknowledged by NuScale in a public 5 meeting last month, okay?

6 MEMBER BROWN: Good.

7 (Off-mic comments.)

8 MEMBER REMPE: Right, okay. Call it an 9 unknown.

10 MEMBER BROWN: Well, this unknown system 11 that nobody knows how it operates, it hasn't ever been 12 in a system like this before in a saturated steam 600-13 pound or whatever the pressure level is, excuse me, 14 high temperature or high pressure environment. The 15 frothing level of the level that you're trying to 16 measure is a significant factor in trying to get an 17 accurate measurement.

18 And if you look up, which I did, some of 19 the products that are available, they try to institute 20 some fairly sophisticated algorithms to evaluate the 21 frothing level and get the errors down, and if you 22 don't do that, you can have errors in that signal of 23 upwards of 20 or 30 percent, and that's an uncertainty 24 relative to your level and how that operates with all 25 the rest of your 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

26 1 How -- I mean, that's another system that 2 is totally unrelated, but yet how do you factor in or 3 even determine an uncertainty because you haven't 4 developed it, you haven't applied it, and you don't 5 even know how it's going to perform?

6 MR. GALYEAN: Yeah, fair enough. You 7 know, the technology for the sensors, you know, might 8 be somewhat unique, and we don't --

9 MEMBER BLEY: Mike?

10 MR. GALYEAN: We don't --

11 MEMBER BLEY: Should we hold this for a 12 closed session or --

13 MEMBER REMPE: Yeah, NuScale can hold --

14 MEMBER BLEY: I'm a little hesitant.

15 MEMBER REMPE: -- this to a closed session 16 if you'd prefer.

17 MR. GALYEAN: Yeah, I'm not going to talk 18 about the technology, okay? All I'm going to say, all 19 I want to say is that these will be safety related.

20 They will be evaluated or proven to be safety-related 21 sensors through whatever means necessary.

22 Additionally, typically the equipment that 23 needs to be actuated automatically, the safety-related 24 equipment, for example, the ECCS and the DHRS, Decay 25 Heat Removal System and Emergency Core Cooling System, 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 will be actuated -- will rely on multiple sensors, 2 multiple physical sensors, whether they're level 3 sensors, pressure sensors, level in the reactor 4 pressure vessel, level in the containment vessel, 5 pressure. There are typically multiple means for 6 actuating these safety-related systems.

7 In the PRA, we only take credit for either 8 one or two sensors, okay, to actuate these systems.

9 We don't take credit for all of the diverse means for 10 actuating these systems, okay, and we view that as a 11 conservatism in the way we model system operation.

12 And so although we don't explicitly 13 account for uncertainty that might be produced by the 14 technology, we do incorporate conservatism in the way 15 we model the actuation of these systems.

16 MEMBER REMPE: So, Bill --

17 CO-CHAIR CORRADINI: Hold on a second. I 18 think you want to try to reestablish the public line, 19 so let's hold and see if we can get this done. Do you 20 want to test it? Can you test it?

21 PARTICIPANT: Who is on the line?

22 MS. FIELDS: This is Sarah Fields, a 23 member of the pubic.

24 MR. SNODDERLY: Okay, thank you, Sarah.

25 Okay, Sarah, yeah, we're going to leave this line 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 open. Please just be in listen mode, and that would 2 be greatly appreciated. We want to make sure you can 3 hear, okay?

4 MS. FIELDS: Okay.

5 MR. SNODDERLY: Thank you. Sarah, can you 6 still hear us?

7 MS. FIELDS: Yes.

8 MR. SNODDERLY: Okay, thank you.

9 CO-CHAIR CORRADINI: Okay.

10 MEMBER REMPE: So we're going to rely on 11 diverse sensors. It may be possible. That's what 12 industry does with boiling water reactors and 13 pressurized water reactors.

14 That requires a lot of operator guidance 15 and a lot of -- and it's not acknowledged, this 16 diverse, like you're going to use the flux detectors 17 to help you decide when the water level is to a 18 certain height.

19 That's not documented in any of the DCA 20 document sections that I remember reviewing, and a lot 21 of times when we ask about such guidance, it's been 22 cut off to the COL applicant stage.

23 Is there some place that -- I know the 24 sensor technology is in ITAAC for this water level, 25 but I've not seen anything about this guidance 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

29 1 using diverse sensors and all of that, and where is 2 that documented and how do we know that the actual COL 3 applicant will be able to instruct the operators to do 4 this?

5 MR. GALYEAN: I'm not quite such which --

6 I'm talking about automatic actuation in the module 7 protection system.

8 MEMBER REMPE: Okay, so is that actually 9 encoded in if the flux detectors have some sort of 10 change? I mean, I've not seen any of that documented 11 in the DCA.

12 MR. GALYEAN: I can't tell you off the top 13 of my head which sensors or which parameters, you 14 know, would actuate the various systems. It depends 15 on things like the initiating event, you know, whether 16 it's a transient, or a LOCA, or, you know, even a loss 17 of offsite power or whatever. Different physical 18 parameters will trigger the ECCS and DHRS system.

19 Etienne, did you want to --

20 MR. MULLIN: I mean, yeah, like one, if you 21 have a LOCA inside containment, for example, you could 22 have a low pressurizer level, low RPV pressure or 23 signal, a high CNV pressure signal. That will all go 24 off within --

25 MEMBER REMPE: But the documentation 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 usually reflects that the triggers are the pressurizer 2 water level as Charlie indicated. I have not seen 3 other sensors that would be put into the protection 4 system. Have you, Charlie?

5 MEMBER MARCH-LEUBA: You know, I'm sorry 6 to disappoint you, but we saw it two weeks ago on a 7 slide.

8 MEMBER REMPE: Yeah, but the water level 9 --

10 MEMBER MARCH-LEUBA: At least it was --

11 yeah, the water level --

12 MEMBER REMPE: Okay, and that's the DHRS.

13 MEMBER MARCH-LEUBA: No, no, the LOCA 14 pressure level triggers a secondary isolation which 15 will then eventually create the high steam pressure, 16 which will trigger DHRS.

17 MEMBER REMPE: But we're not talking about 18 DHRS in this case.

19 MEMBER MARCH-LEUBA: These four signals 20 trigger the DHRS, high pressurized pressure, high and 21 hot temperature, high stream pressure in the primary 22 or loss essay voltage.

23 MEMBER REMPE: But there are other -- the 24 DHRS is one aspect, but there are other things --

25 MEMBER MARCH-LEUBA: Many more.

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 MEMBER REMPE: Right, that the pressurizer 2 level triggers, and the PRA --

3 MEMBER MARCH-LEUBA: Pressurizer level?

4 MEMBER REMPE: I thought -- is there not 5 some systems that depend solely on the pressurizer 6 level?

7 PARTICIPANT: No.

8 CO-CHAIR CORRADINI: I think the answer is 9 no.

10 MEMBER REMPE: Okay.

11 PARTICIPANT: Right.

12 MEMBER REMPE: Then I have another 13 question if we can change the topic. The expert 14 panel, the review group that reviewed the, the folks 15 who can still punch a calculator as I've identified 16 them, we only know they can punch a calculator. What 17 level of industry folks are they and can you even give 18 names? Are they industry folks? Who are they?

19 MR. GALYEAN: Dave Blanchard and Wes 20 Brinsfield. They do a lot of consulting for the 21 industry. They've both been in the industry doing PRA 22 for a long time. In fact, Dave Blanchard, I think, 23 worked on the Big Rock Point PRA, and he's been 24 involved in the PRA business all --

25 MEMBER REMPE: Okay.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

32 1 MR. GALYEAN: -- all along.

2 MEMBER REMPE: Okay, and the, on my 3 package, it's slide nine, but I think it's a different 4 number. Oh, there is it. It's up there right now, 5 this no major concerns or objections.

6 Aren't there some assumptions also that 7 the expert panel made, like the treatment of multi-8 modules be reviewed by the greater PRA community, and 9 that's -- they had certain assumptions to come up with 10 this no major concerns or objections. It maybe makes 11 the slide a little too positive.

12 MR. GALYEAN: Yeah, well, I think that was 13 Apostolakis' comment, and he was more being a 14 proponent for expanded use of PRA in general. Doing 15 multi-module assessments is obviously new to the PRA 16 community.

17 I know there has been some work on doing 18 multi-unit PRAs, but, you know, he was just -- he was 19 very complimentary of the approach we took, but he 20 felt that it warranted a broader, how do I say, 21 engagement with the industry just to propagate the 22 methods that we developed more than anything else, so 23 he was just trying to be PRA promotional as much as 24 anything else.

25 MEMBER REMPE: Okay, so perhaps there was 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 a difference of opinion is what you're trying to 2 convey, the majority opinion of the group?

3 MR. GALYEAN: No, I think the whole group 4 was on board with that. They liked what we did and 5 they just thought that we should share it with the 6 rest of the PRA community basically.

7 MEMBER REMPE: Okay, I guess what I had 8 reviewed in the summary was a little different take.

9 That we assumed this to come up with these conclusions 10 is what I recall the quote, or perhaps I've 11 misremembered or something?

12 MR. GALYEAN: I --

13 PARTICIPANT: Yeah, I don't know, sorry.

14 MR. GALYEAN: I can't remember --

15 MEMBER REMPE: Okay.

16 MR. GALYEAN: I can't remember that, so.

17 DR. SCHULTZ: Bill, in the sequence you 18 presented today with regard to the overall QA program 19 and review of the, or quality program and review of 20 the PRA, the sequences as you've described, the expert 21 panel saw this relatively early compared to the other 22 groups?

23 MR. GALYEAN: We had the expert panel 24 engaged at multiple places --

25 DR. SCHULTZ: 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

34 1 MR. GALYEAN: -- in the development of the 2 PRA.

3 DR. SCHULTZ: Did they --

4 MR. GALYEAN: I think the --

5 DR. SCHULTZ: Did they wrap it up? Did 6 they wrap --

7 MR. GALYEAN: That's right.

8 DR. SCHULTZ: -- their review up as a 9 result of the improvements that you've described here 10 or did they finish their job before the final reviews 11 by the other teams?

12 MR. GALYEAN: I don't recall the exact 13 timeline. It seemed to me the expert panel that we 14 had was fairly close to the end of the PRA 15 development, and of course the self-assessment and the 16 external review of our self-assessment was also fairly 17 late in the process. I think the expert panel 18 actually finished up after this self-assessment.

19 DR. SCHULTZ: That's what I was getting to 20 or wondering about. All right, I appreciate that.

21 MR. GALYEAN: Yeah.

22 DR. SCHULTZ: And the level of effort of 23 the expert panel and the other teams, how would you 24 describe that?

25 MR. GALYEAN: As I said, the expert panel NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

35 1 came and visited with us a couple of times over maybe 2 a two-year period, and we had a couple of telephone 3 conference calls. We certainly sent them various 4 reports that we generated. So all in all, each 5 member, again, I'm just trying to recall, maybe spent 6 a couple of man months looking at what we've done, and 7 so.

8 DR. SCHULTZ: Thank you.

9 MEMBER REMPE: So to beat a dead horse, on 10 page 19.1.97 of the open DCA, it says the CVS -- the 11 MPS -- the safety-related MPS generates a CVCS 12 isolation signal if the high pressurizer level set 13 point is exceeded.

14 Other places, when you talk about the 15 pressurizer water level, you do say pressurizer water 16 level or some other signal, but here, you don't. Is 17 it just you omitted this in the DCA?

18 MR. GALYEAN: I can't answer that 19 question. I'd have to go back and research that. I 20 don't recall.

21 CO-CHAIR CORRADINI: Why don't we try to 22 do that offline then?

23 MR. GALYEAN: Okay.

24 MEMBER REMPE: Thank you.

25 MEMBER SKILLMAN: I'd like to ask on 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

36 1 last set of cards, the terms CDF and LRF tied to large 2 reactors' current use may be misleading a NuScale 3 design. I can interpret that two different ways.

4 I can interpret that to mean that if 5 you're kind of set on the current 10 to the minus four 6 and 10 to the minus six and you look at NuScale, you 7 find that use of the current metric doesn't quite fit, 8 but I can also interpret that to mean a 10 to the 9 minus eight to the 10 to the minus 10 on NuScale 10 should be met with great skepticism, and so I wonder 11 what was really intended by that comment?

12 MR. GALYEAN: The comment is -- there --

13 remember in the NuScale, the NuScale cores are small, 14 okay? A NuScale core is only about five percent the 15 size of a large light water reactor, okay, five 16 percent.

17 If you take all 12 NuScale cores and pile 18 them up into one big core, you are still only half the 19 size of a large light water reactor.

20 And so when you compare core damage 21 frequencies, you're comparing a core damage event in 22 this tiny NuScale five percent core to a core damage 23 event in a large light water reactor with a core 20 24 times as large, okay.

25 I mean, how is that a rational comparison, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

37 1 okay. That's what that comment is intended to convey.

2 MEMBER SKILLMAN: I can understand that 3 logic, but I come from a background where even an 80-4 megawatt reactor out of control poses a huge concern, 5 and so while I try to understand the practical 6 perspective that you've just communicated, I've lived 7 the life where a much smaller core can evacuate a 8 city.

9 So I am one member on the ACRS that is on 10 the one hand supportive of the plumbing, but cautious 11 on your conclusions on your PRA.

12 MR. GALYEAN: Well, all I can say is that 13 we're talking about, you know, 21st century technology 14 here, okay. I mean, we have a lot of information or 15 a lot of knowledge that we've gained over the years 16 of, you know, 50 years of nuclear power operation with 17 material science, with the development of safety 18 systems, digital, you know, fiber optics.

19 You know, each of these five percent 20 cores, okay, are contained in their own reactor 21 pressure vessel. Each one is contained in its own 22 high pressure containment vessel. They're all 23 submerged in the ultimate heat sink, all contained in 24 a seismic Class I reactor building.

25 You know, the design, the NuScale 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

38 1 is incredibly safe, albeit unique, but the uniqueness 2 is what makes it safe.

3 CO-CHAIR CORRADINI: So let me, I think 4 I'm going to move us on if it's all right because I 5 think you understand Member Skillman's position, but 6 let me put it in a direction, I think, that can take 7 us forward, which is what you're really saying is, "I 8 don't just look at the CDF. I essentially look at the 9 release fractions and the source term."

10 Is that going to be discussed within 19 or 11 are we going to wait for that for the topical report?

12 Because I want to make sure because there are going to 13 be questions about the source term relative to its 14 application. Are we going to talk about it today?

15 PARTICIPANT: It wasn't -- we --

16 MS. BRISTOL: No, not specifically.

17 CO-CHAIR CORRADINI: Nor tomorrow?

18 MS. BRISTOL: Correct.

19 CO-CHAIR CORRADINI: Okay, but eventually 20 we want to talk about it?

21 MS. BRISTOL: Yes.

22 CO-CHAIR CORRADINI: Okay, because I 23 think, if I understand where Member Skillman is coming 24 from, I think that's the issue.

25 MEMBER SKILLMAN: Avogadro's number 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

39 1 big number, 6.02 times 10 to the 23. I don't care how 2 big your core is. That is a big number and a little 3 bit goes a long, long way. I lived that life at TMI-4 2.

5 MEMBER BLEY: I'm stuck on two things.

6 One will be easy for you to answer, so I'll ask that 7 first. You have a section in the report on passive 8 system uncertainty that raises some very important 9 issues, I think. Do you have slides on that or are 10 you going to talk about that?

11 MS. BRISTOL: We don't have specific 12 slides in this. We talked about that quite a bit in 13 October. We can answer some questions. We have a 14 couple of bullets, but no specific slides.

15 MEMBER BLEY: You've raised the issues.

16 Have you tried to quantify them?

17 MS. BRISTOL: In the passive safety system 18 reliability?

19 MEMBER BLEY: Yes, yes, yes.

20 MS. BRISTOL: Yes, we did quantify those, 21 and that was that whole analysis was to develop those 22 values for reliability of the emergency core cooling 23 system and DHRS, Decay Heat Removal System, passive 24 reliability.

25 MEMBER BLEY: Okay, we haven't gotten 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

40 1 chapter six yet. We're going to get there, so the 2 ECCS valves are in there, but it won't talk anything 3 about test frequencies, I don't think. What did --

4 CO-CHAIR CORRADINI: I was just going to 5 help you. If you look, I think, in one of our notes 6 from the staff, I think that actually resides in 7 chapter 396, what you're looking for.

8 MEMBER BLEY: Valve test frequencies?

9 CO-CHAIR CORRADINI: I think so.

10 MEMBER BLEY: What did you -- did you 11 assume the same thing in the PRA?

12 MR. GALYEAN: Assume what?

13 MEMBER BLEY: The testing frequencies for 14 the ECCS valves. So if you're doing the PRA on the 15 equipment, you've got to consider the possibility --

16 MR. GALYEAN: Right, I believe we just 17 assumed they would be tested during refueling.

18 MEMBER BLEY: Okay, once a year?

19 MR. GALYEAN: Yeah.

20 MEMBER BLEY: How does that compare with 21 --

22 MR. GALYEAN: Once every two years.

23 MEMBER BLEY: Oh, that's right, once every 24 two years. How does that compare with the test 25 frequency for similar valves, not the same valves, 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

41 1 the current plans?

2 The reason I'm bringing this up is valves, 3 especially valves that rely on springs or other 4 passive modes of operating them, when they sit for a 5 long time, don't work as well as they did before they 6 sat for a long time, and yours are going to sit for a 7 longer time than others.

8 I had some experience a long time ago with 9 safety valves that sat for several years before they 10 got tested, and we finally got around to testing, none 11 lifted anywhere near where they were supposed to.

12 So these things sitting here for a long 13 time ought to have some kind of degradation and their 14 common cause impact ought to be affected by that.

15 And most test programs wouldn't look at 16 this, but I think this is an area where it's 17 essentially when you get a test program that it looks 18 at these kind of, I'll call them short-term aging 19 effects because these are really important valves and 20 they're in a unique place for testability, so I wanted 21 to raise that with you.

22 I wasn't here for that discussion on the 23 passive degradation. When you quantified the effects 24 for passive systems, did you make assumptions about 25 how much degradation you would have due to any 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

42 1 aging issues that could affect thermohydraulic systems 2 with a fairly delicate balance?

3 MR. GALYEAN: We treated it as an 4 uncertainty.

5 MEMBER BLEY: And how did you judge the 6 worst end of the uncertainty bound?

7 MR. GALYEAN: I can't recall off the top 8 of my head. I mean, certainly for the passive system 9 reliability, we did a series of RELAP runs to identify 10 the major dependent variables for the success of heat 11 transfer, okay.

12 And once we identified those, we developed 13 uncertainty bands on each one, and then we did Monte 14 Carlo simulations where we sampled from those 15 uncertainty bands and generated --

16 MEMBER BLEY: Where did the uncertainty 17 bands come from?

18 MR. GALYEAN: Well, again, I'd have to go 19 back and look, you know, at how those limits were 20 determined.

21 MEMBER BLEY: Certainly not from any tests 22 anywhere or were there tests --

23 MR. GALYEAN: No, I --

24 MEMBER BLEY: -- far away that you --

25 MR. GALYEAN: I would guess it was 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

43 1 engineering judgment that --

2 MEMBER BLEY: By people who probably --

3 MR. GALYEAN: -- developed a high --

4 MEMBER BLEY: -- never saw this phenomena.

5 MR. GALYEAN: -- you know, a high and a 6 low value for these physical, for these --

7 MEMBER BLEY: In the past, on the other 8 design certs we've looked at, we've recommended 9 strongly that when, should there be somebody who ever 10 builds this, when they are approaching fuel load and 11 are supposed to have a really complete PRA, that these 12 issues be thoroughly addressed, and I don't think they 13 have yet.

14 And at least from my way of thinking, 15 these valves and the possible short-term aging of them 16 and long-term aging within the systems for the passive 17 effects are things that could upset these wonderfully 18 low numbers we calculated.

19 CO-CHAIR DIMITRIJEVIC: Well, you know we 20 are still in the period of consideration and review, 21 but you can see what are the subjects we are most 22 interested in. So what Dennis just brought up is you 23 have these two events which are the ECCS and DHRS, 24 which both are failure of passive heat removal, right?

25 Their failure probability is slightly 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 smaller than valves itself. I think DHRS is 40 minus 2 six and ECCS is 20 minus seven. So, I mean, I don't 3 know how you made those guesses. Maybe it's based on 4 the thermohydraulic, but that don't factor on those 5 multi, very small. It's just two. So you obviously 6 don't think there is a huge uncertainty associated 7 with this.

8 So that's what is really -- when I look 9 here, we talk uncertainty, uncertainty, uncertainty, 10 but when we look in the results, we don't see 11 uncertainty. You know, your idle factors and total 12 numbers are also very small. Your mean value is 13 almost identical to your point estimates.

14 Somehow this uncertainty discussion didn't 15 show up when you present the total results, and input 16 data also don't show these uncertainty bands. So, I 17 mean, it's very difficult to put decisive tests to 18 determine the answers to the bands.

19 That's not easy for the PRA, but I think 20 those two events which you tried to incorporate, there 21 is some attempt to put this uncertainty, but then you 22 make them very certain.

23 MR. GALYEAN: Well, I guess I would argue 24 that just the presence of those events in the PRA 25 model represents a treatment of uncertainty, okay.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

45 1 Those events represent -- okay, given that the system 2 operates the way it's supposed to, that the valves 3 open and, you know, you've got flow going, despite 4 that fact, heat transfer to the ultimate heat sink is 5 insufficient.

6 CO-CHAIR DIMITRIJEVIC: Right.

7 MR. GALYEAN: Okay, that's what those 8 events represent.

9 CO-CHAIR DIMITRIJEVIC: They allow the 10 passive --

11 MR. GALYEAN: That's right.

12 CO-CHAIR DIMITRIJEVIC: -- and all of 13 these other things which you need for the passive heat 14 removal.

15 CO-CHAIR CORRADINI: Just you guys are 16 talking as two subject matter experts at each other.

17 Let me make sure I'm following this. What you're 18 asking, Vesna, is that there are --

19 What you're saying, Bill, is that there 20 are certain events that essentially take into account 21 partial actuation of passive systems that cover what 22 is the uncertainty that Vesna and Dennis are concerned 23 about? Am I misunderstanding?

24 MR. GALYEAN: Not partial actuation. I 25 mean, it's that these systems in the NuScale 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

46 1 ECCS and the DHRS, rely on natural circulation, right, 2 for heat transfer, and that these events that we're 3 talking about represent the failure of gravity, right, 4 I mean, that the natural circulation fails to transfer 5 heat from inside the reactor pressure vessel to the 6 ultimate heat sink.

7 CO-CHAIR CORRADINI: But if I might just 8 push the point, Dennis is -- I'm just trying to link 9 it. Dennis' point is that I might have, if I 10 understand it, I've got a valve. It's supposed to 11 open upon a signal, but it's sitting there and it's, 12 I'll use the word corroded just for want.

13 It's somehow aged and it doesn't open 100 14 percent. It opens five percent or something like 15 that. That essentially then would affect natural 16 circulation or pressure-driven flows. Am I -- I think 17 that's where Dennis was going.

18 CO-CHAIR DIMITRIJEVIC: The both, 19 actually.

20 CO-CHAIR CORRADINI: Both?

21 CO-CHAIR DIMITRIJEVIC: Both, either the 22 valve failure rate or the passive heat removal failure 23 rate --

24 CO-CHAIR CORRADINI: Right.

25 CO-CHAIR DIMITRIJEVIC: -- given the valve 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 operated right.

2 CO-CHAIR CORRADINI: And your point back 3 is that you've, I'm not going to use the word bounded, 4 but you've considered it by some sort of bounding 5 scenario?

6 MR. GALYEAN: Well, we did look at the 7 potential impact of partial opening of these valves.

8 I mean, we did a number of sensitivity studies which 9 I don't think -- you know, I don't know how we 10 documented, but we --

11 CO-CHAIR CORRADINI: Well, that's going to 12 be my next question, is where do I look for that?

13 MR. GALYEAN: Yeah, it would be in a 14 supporting PRA engineering report. It won't be in the 15 --

16 CO-CHAIR CORRADINI: Okay.

17 MR. GALYEAN: -- you know, DCA or the 18 FSAR.

19 MEMBER BLEY: And those were not made 20 available here, but during the -- we can ask the staff 21 about this too because they did audits --

22 MR. GALYEAN: Right.

23 MEMBER BLEY: -- and we looked at some of 24 those --

25 MR. GALYEAN: Okay.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

48 1 MEMBER BLEY: -- reports on the electronic 2 system, but I don't know if they ever had hard copies 3 of them, but I'm going to -- sorry.

4 (Simultaneous speaking.)

5 MEMBER BLEY: We're not talking about 6 failure of gravity. We're talking about increasing 7 friction, something like that, that has happened in 8 many places in the past.

9 And to me, graded for identifying the 10 issues, running a million thermohydraulic runs if you 11 haven't covered the range of things that could happen 12 doesn't help you. It sounds good, but it doesn't 13 help, so understanding how you set those bounds over 14 the life of the plant, trying to think this through 15 because this is a plant that has to work as well 16 later.

17 You might not have all of the valves open.

18 You might have some of them open. They could be 19 partially open and then it doesn't take as much of a 20 fouling problem to mess up the thermohydraulics. So 21 those are things that are just crucial to believe in 22 the results.

23 MR. GALYEAN: Agreed, we did. The passive 24 system reliability work was documented, again, in a 25 PRA engineering report, and that, I do believe 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

49 1 was one of the reports that was made available to the 2 staff for audit.

3 CO-CHAIR CORRADINI: Okay.

4 MEMBER MARCH-LEUBA: I have a question too 5 while we're talking about DHRS. DHRS has three, at 6 least three failure modes. One is the hardware 7 failure if the valves don't open, if they had been 8 arresting for a while, and then you have non-9 condensables, which are supposed to be prevented by 10 the level on the DHRS line.

11 But I believe that's a minor actuation and 12 there's a human error probability, and also the 13 overfill possibility, which is prevented during the 14 startup by establishing some vacuum and, you know, 15 more operation. You could consider leaving that.

16 Have you considered those other failures?

17 MR. GALYEAN: Those were part of the 18 parameters that were buried in the PSSR analysis.

19 What we call the passive safety system reliability 20 analysis looked at both DHRS and ECCS as almost two 21 separate analyses.

22 MEMBER MARCH-LEUBA: Right, but do you 23 have non-condensables --

24 MR. GALYEAN: Non-consdensables, the 25 volume.

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 MEMBER MARCH-LEUBA: It completely takes 2 the DHRS to zero.

3 MR. GALYEAN: That's right. You know --

4 MEMBER MARCH-LEUBA: Can you assume that 5 with certain probability?

6 MR. GALYEAN: We used it as a parameter 7 that we varied. We put an uncertainty on it.

8 MEMBER MARCH-LEUBA: No, I mean, if the 9 operator fail, I mean, if that pressure, the level 10 sensor on the DHRS line fails to identify that 11 nitrogen is building up there, when you turn it on, it 12 won't work. I mean, it's not that it would be 13 degraded. It won't work, zero.

14 MR. GALYEAN: Well, for the most part, you 15 know, we certainly assume that the systems operate the 16 way that they're designed, okay, and --

17 MEMBER MARCH-LEUBA: Yeah, but if it's an 18 operator action, PRA --

19 MR. GALYEAN: It's not an operator action.

20 That's not an operator action, okay. That's an 21 automatic actuation.

22 MEMBER MARCH-LEUBA: Even though the 23 operator relies on automated action by a digital 24 protection system which may or may not happen with a 25 certain probability. Was that considered on the PRA?

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 MR. GALYEAN: Well, certainly we account 2 for the fact that the automatic actuation fails to 3 operate. I mean, that has a certain failure 4 probability associated with it, and so whether that 5 fails completely or it fails in a timely fashion and 6 results in a failure as you talk about is, you know, 7 it's still a failure, so that --

8 MEMBER MARCH-LEUBA: Is one of the 9 inciting events is --

10 MR. GALYEAN: That's right.

11 MEMBER MARCH-LEUBA: -- I build up non-12 condensables and didn't notice?

13 MR. GALYEAN: It's not explicitly 14 identified in that fashion. It's identified as a 15 failure of the actuation system to operate as 16 designed.

17 MEMBER MARCH-LEUBA: And if it fails DHRS 18 to zero?

19 MR. GALYEAN: That's right. If it fails, 20 the DHRS fails, right.

21 MEMBER MARCH-LEUBA: Okay, different 22 subject, but more dear to my heart, with the failure 23 probabilities, the input data, let's assume that you 24 run all of your best estimates for these valves and 25 you come up with a failure probability of 10 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

52 1 minus four. I'll give you that.

2 And then you start to operate in the 3 reactor, and within three months, the valves start 4 leaking, meaning it failed. Does that invalidate your 5 PRA and you are supposed to go and recalculate your 6 probabilities because you obviously missed it?

7 I mean, if you're assuming that your 8 probability of failure is 10 to the minus four, and 9 now either the valve starts leaking or you run a test 10 and it didn't work properly, and you caught it.

11 The probability is not 10 to the minus 12 four because you are testing two or three times. Your 13 probability is more like 50 percent, so that 14 invalidates your PRA and you have to redo it?

15 MR. GALYEAN: Well, it's certainly 16 standard practice in the PRA industry to maintain the 17 PRAs through the operating life of the plant, and 18 there are requirements that the PRAs be updated at 19 least every two years, and certainly there is the NRC 20 reactor oversight process that monitors the 21 performance of equipment modeled in the PRA.

22 MEMBER MARCH-LEUBA: My point is every two 23 years, you would be testing these valves. If one 24 fails, did you change the input data from 10 to the 25 minus four to 0.5?

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 MR. GALYEAN: It would be --

2 MEMBER MARCH-LEUBA: Or do you keep 3 saying, oh, it really wasn't 10 to the minus four? It 4 was a fluke test?

5 MR. GALYEAN: No, certainly there are 6 techniques for updating generic data with operating 7 experience data, the Bayesian techniques for example.

8 It wouldn't change to 0.5. It would be changed to 9 something in between.

10 The expectation certainly is that that 11 failure mode would be addressed in some fashion and 12 the valve repaired. No one wants to have repeated 13 failures occurring for safety-related equipment.

14 MEMBER MARCH-LEUBA: Yeah, I'm looking at 15 the NUREG, NUREG CR 70.37. That's the industry 16 performance of relief valves and more stuff, and the 17 valve failure probability based on licensee event 18 reports was already there. It's all over the place.

19 I mean, there are some types of valves 20 that fail with a 15 percent probability and those are 21 the pilot operator valves which are more like the ones 22 you've got, and they're, all of these are 10 to the 23 minus four and 10 to the minus five.

24 So I just wanted to put on the record that 25 I don't share your confidence that you know the input 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 data for your analysis that way, and I would like to 2 see some commitment or some requirement from the staff 3 that if during testing, the testing invalidates your 4 input data, you have to do something.

5 MS. BRISTOL: Right, moving onto the 6 initiating events. We, as discussed earlier, we can 7 go over these various topics we talked about in 8 October, but the initiating event analysis was, we 9 looked at various inputs to identify the potential 10 initiating events that could affect a NuScale module.

11 It was a deliberate effort to be complete 12 and comprehensive. We developed a NuScale specific 13 master logic diagram to look at the various impacts 14 that could affect a module. We looked at FMEAs for 15 all of the systems that could impact and cause 16 potential plant upsets.

17 We looked through various applicants and 18 traditionals lists of those initiating events. And 19 continuously over the past seven years, we've 20 continued to look at the design as it's developed and 21 incorporated those insights into developing these 22 initiating event --

23 MEMBER BLEY: Can I ask you a question 24 about the magic logic there, magic?

25 (Laughter.)

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 MEMBER BLEY: This is a language issue.

2 It seems to me on a lot of these, you talk about 3 losing control and then things go either up or down.

4 On heat removal, you talk about insufficient heat 5 removal, but you have at least four events that are 6 increased heat removal, and I assume that's 7 intentional and that didn't mean just insufficient 8 heat removal. It meant upsets in heat removal.

9 Is that true or do you think that 10 inadvertent turbine control valve opening is 11 insufficient heat removal? You haven't looked at this 12 in a long time.

13 MS. BRISTOL: Yeah, sorry.

14 MEMBER BLEY: Look at it later.

15 MS. BRISTOL: Okay, I will do that.

16 MEMBER BLEY: And it's a matter, I think 17 it's just a matter of language, but there are a number 18 of those kind of things --

19 MS. BRISTOL: Okay, thank you.

20 MEMBER BLEY: -- in that logic diagram.

21 I think it's reasonably complete, but not exactly as 22 stated.

23 MS. BRISTOL: Okay, thank you. If you 24 look at the next slide, we go through, these are the 25 initiating events that were evaluated in the PRA. 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

56 1 looked at injection line breaks inside the vessel and 2 outside of the vessel.

3 We also looked at the CVCS, chemical 4 volume control system discharge line outside of the 5 containment vessel and the significance of these 6 events was injection capability through CVCS. That's 7 our main injection source to get inventory into the 8 reactor pressure vessel.

9 We looked at spurious ECCS valve 10 actuations. We looked at loss of power, both DC and 11 offsite power. We looked at two failures, secondary 12 line breaks, both in the steam and feedwater, as well 13 as the DHRS. We looked at other LOCAs inside 14 containment, whether that be a safety valve lift, a 15 pressurizer heater over pressurization event.

16 We looked at just general reactor trips 17 and we looked at loss of support system, instrument 18 air, power, so what impacts to the support systems 19 maybe CVCS or CFDS would be impacted to respond to an 20 event --

21 MEMBER BLEY: You --

22 MS. BRISTOL: -- in that loss of support 23 system.

24 MEMBER BLEY: You talk about how you were 25 able to collapse this to about 10, and I remember 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

57 1 some of the LOCA discussions, you described other 2 LOCAs that would be like the ones you've had, so they 3 were adequate models.

4 When you came up with the initiating event 5 frequency, did you get the initiating event frequency 6 for the specific event or did you accumulate it for 7 all of the things that you've thrown in that same 8 category?

9 MS. BRISTOL: We evaluated, for example, 10 the CVCS injection line was the piping in that system, 11 and so a break in that pipe. For the general reactor 12 trip, we looked at all of the various impacts that 13 could cause a trip, loss of circ water, manual trips, 14 loss of support system as we say, instrument error.

15 So we tried to incorporate the various failures of the 16 systems that would support that initiating event 17 frequency.

18 Then we, again, moved on. For the 19 accident sequence analysis, we looked at those 20 initiating events and identified the various 21 conditions that a module could experience in response 22 to those initiating events, what systems would 23 respond.

24 We identified the key safety functions, 25 that heat removal, reactivity control, 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

58 1 integrity, what responses to those initiating events 2 needed to occur in order to get us to an end state, 3 and those end states we evaluated were core damage as 4 well as large release frequency, and so it was the 5 success or failure of those sequences that we 6 evaluated.

7 MEMBER BLEY: Sarah, I had to go back and 8 find my notes.

9 MS. BRISTOL: Okay.

10 MEMBER BLEY: There's a place where you 11 talk about loss of coolant accidents, and you talk 12 about for a number and you describe some additional 13 smaller RPV penetrations. The staff finds that the 14 plant response can be expected to be similar to or 15 bounded by what you explicitly modeled in the CVCS 16 line break because they have similar mitigation.

17 What I was asking you is did you combine 18 the initiating event frequencies of all of those 19 multiple LOCAs into the CVCS line break LOCA?

20 MS. BRISTOL: Yes, we did.

21 MEMBER BLEY: Okay, thank you.

22 MS. BRISTOL: Yeah. For success criteria, 23 for the level one, we used core damage frequency, and 24 the core damage then was defined by 2,200 degree 25 Fahrenheit peak cladding temperature, and we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

59 1 demonstrated that over 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. For --

2 MEMBER MARCH-LEUBA: The criteria for one 3 node reaching 2,200 or a significant number of pins?

4 MR. MULLIN: Anywhere, anywhere in the 5 core, yeah.

6 MEMBER MARCH-LEUBA: And do you want to 7 stick with that? That's extending the -- it's going 8 to bite you later if -- it's a bad idea to do it 9 because -- I keep pushing. I want to find an event 10 that will get you 2,200 in one node and it won't be of 11 any consequence, but if you stick to that, that is bad 12 for you and it's not necessary. I mean, I think it's 13 a bad idea.

14 MR. GALYEAN: Remember, this is, we're 15 talking about beyond design basis accidents here, and 16 so no matter what kind of definition we employ for 17 core damage, you can always postulate failures that 18 would get you there, and that's all we're doing here 19 is just defining what we mean.

20 When we talk about a core damage 21 frequency, we're saying here is the frequency of 22 reaching this state in the core, okay. And again, we 23 are limited to beyond design basis accidents here, or 24 not limited, but we are talking about beyond design 25 basis accidents here.

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 MEMBER MARCH-LEUBA: I'm trying to help 2 you here.

3 MR. GALYEAN: I appreciate that. I 4 appreciate that.

5 MEMBER MARCH-LEUBA: This might come back 6 and bite you.

7 MR. GALYEAN: Yeah, and --

8 MEMBER REMPE: But he is just telling --

9 MEMBER MARCH-LEUBA: It's unnecessary.

10 MEMBER REMPE: He is just saying what they 11 did. I mean, even if he wanted to change his mind, 12 right, not today.

13 MEMBER MARCH-LEUBA: I'm just saying. The 14 same way I tell you when you do something wrong, it's 15 wrong. I think this is too conservative. In 16 operating reactors, we accept one percent failure of 17 the fuel and it's during normal operation.

18 MS. BRISTOL: Okay, and the level two 19 success criterion we used as the official definition, 20 and we'll talk about maybe a little bit different when 21 we get to the containment event tree, but we defined 22 it as the source term resulting in acute whole body 23 200 rem dose at the site boundary for an individual 24 standing there for 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br />.

25 As we go through the event trees, we did 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 various simulations with, various TH simulations with 2 RELAP, and we used our initial safety-related code and 3 model, and then we modified it to meet the PRA needs, 4 and that included adding in chemical and volume 5 control system and containment flooding and drain 6 system models to the code, as well as we developed 7 various core models as we were able to simulate beyond 8 design basis transients using our PRA codes for that, 9 including ATWS as well as other --

10 CO-CHAIR CORRADINI: So let me ask, does 11 that require a software development or just simply 12 very clever user modeling with NRELAP?

13 MR. MULLIN: No, these are just input 14 models.

15 CO-CHAIR CORRADINI: Okay, fine. I 16 guessed that. I just wanted to make sure.

17 MEMBER SKILLMAN: Sarah, why isn't the 18 containment evacuation system on that first bullet?

19 MR. MULLIN: It's not used to mitigate any 20 accident.

21 MEMBER SKILLMAN: Is the assumption that 22 you have a preexisting vacuum that ensures, or either 23 ensures or doesn't ensure heat transfer?

24 MR. MULLIN: I mean, in the model, in our 25 NRELAP5 model, there's a flow path that pulls a vacuum 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 continuously during the event until you isolate that 2 flow path, until the transient starts to actually --

3 CO-CHAIR CORRADINI: But to follow on 4 Dick's point, you evaluated both -- I seem to remember 5 in previous discussions we had in the previous 6 meetings, you evaluated both ways, one where it works 7 and one where it doesn't work. That is you actually 8 have some air pressure --

9 MR. MULLIN: Yeah, we evaluated the 10 failure to isolate that --

11 CO-CHAIR CORRADINI: Okay, that's what --

12 MR. MULLIN: -- line.

13 CO-CHAIR CORRADINI: Whatever the right 14 way of saying it is.

15 MR. MULLIN: Yeah, I think the purpose of 16 adding the CVCS and CFDS models on this slide is these 17 are used to mitigate coolant injection during events, 18 and that's not done in the design basis analyses.

19 MEMBER SKILLMAN: So without including the 20 evacuation system, are you able to depend on heat 21 transfer to the pool?

22 MR. MULLIN: Without the containment of --

23 without a vacuum?

24 MEMBER SKILLMAN: Yeah.

25 MR. MULLIN: Yeah.

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 MR. GALYEAN: Can I -- I mean, in the 2 model, we assume an initial condition of what? The 3 pressure in the containment is one PSI approximately?

4 MR. MULLIN: I will be less than that 5 actually, but --

6 MR. GALYEAN: I know, but what did we 7 model in the --

8 MR. MULLIN: That might be proprietary.

9 MR. GALYEAN: Oh, okay, yeah, fair enough.

10 So, now, we did model -- you know, how do I say this?

11 (Simultaneous speaking.)

12 CO-CHAIR CORRADINI: Can I try something?

13 MR. GALYEAN: We had an initial condition 14 in containment, okay.

15 MR. MULLIN: Yeah, you certainly don't 16 require a vacuum in containment to have successful 17 heat transfer.

18 MR. GALYEAN: Yeah, that's --

19 MR. MULLIN: It's just a normal operations 20 thing to avoid heat losses, and corrosion, and stuff 21 like that.

22 CO-CHAIR CORRADINI: Thank you.

23 MEMBER MARCH-LEUBA: If you use vacuum in 24 the containment, you transfer more heat to the UHS, 25 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

64 1 MR. MULLIN: Yeah.

2 MEMBER MARCH-LEUBA: So it is a 3 conservative assumption to assume vacuum. If you 4 assume there is steam in it or anything other than 5 vacuum, you're transferring more heat.

6 MEMBER SKILLMAN: No, if you have more 7 vacuum, the heat's in the reactor vessel. If you have 8 less vacuum, you have degradability transferred to the 9 pool.

10 CO-CHAIR CORRADINI: It's not that 11 straightforward. I think what Mr. Mullin said is 12 correct. They tried it both ways. There is public 13 data from Oregon State prior to them developing their 14 system for NuScale that's published in reports as to 15 the effect of vacuum and no vacuum. That's a way you 16 might want to answer that to help everybody look 17 something up that's in the public.

18 MR. MULLIN: Sure, I mean, there's plenty 19 -- the containment provides plenty of heat transfer 20 whether you have non-condensables in there or not.

21 MS. BRISTOL: Okay, the next slide, for 22 the human reliability analysis, just to confirm and 23 clarify, there are no human actions credited in the 24 design basis event, as design basis events, but we do 25 have them to support beyond design basis events 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

65 1 PRA.

2 And so we were using methodologies 3 previously used in the industry, both ASEP as well as 4 SPRA-H, and we used those methodologies to develop 5 blatant human errors as well as the recovery actions 6 in the PRA.

7 CO-CHAIR CORRADINI: I'm assuming that the 8 experts will ask a question about this, but a simple 9 question for me is you have the EOPs, right? This 10 goes beyond.

11 These are operator actions that they may 12 take in spite of or, and if they enter into an 13 essentially beyond design basis event, so it doesn't 14 follow strictly the emergency operating procedures.

15 Am I correct in that?

16 MS. BRISTOL: We don't have EOPs developed 17 yet.

18 MR. GALYEAN: Yeah, there's operating --

19 there are --

20 CO-CHAIR CORRADINI: If I've asked it 21 wrong, clarify for me, please.

22 MR. GALYEAN: Yeah, I don't think NuScale 23 uses the term emergency operating procedures, okay.

24 MEMBER BLEY: We've been told they don't 25 have any yet. That's what we've been told.

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 MS. BRISTOL: That's correct.

2 CO-CHAIR CORRADINI: Okay.

3 MS. BRISTOL: And so, but we have worked 4 with operations to model these operator actions in the 5 simulator.

6 CO-CHAIR CORRADINI: Oh, okay.

7 MS. BRISTOL: And so they have modeled 8 both of these actually risk significant human action 9 candidates, the operator injecting with the 10 containment flooding and drain system, as well as the 11 CVCS. They have modeled that and they have tested on 12 those, you know, risk significant operator actions.

13 CO-CHAIR CORRADINI: All right, thank you.

14 MEMBER MARCH-LEUBA: But in the case of 15 the CVCS, it's a very complex system. There are 12 of 16 them and there's probably 50 valves on it and I don't 17 know how many pumps.

18 The probability of it being misaligned 19 when you try to make it work by a human following an 20 event that trips all 12 reactors and you're trying to 21 think of what is going on is fairly high.

22 I don't think -- I'm going back to -- my 23 question is what do you use for your -- for input 24 data?

25 MS. BRISTOL: 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

67 1 MEMBER MARCH-LEUBA: And if you're having 2 a bad day, that CVCS is very complex. I mean, count 3 the number of pumps on the lineup. It's very complex.

4 MS. BRISTOL: And we looked at the valve 5 lineups for these various actions we modeled the 6 potential for spurious failures as well as, you know, 7 pumps failing to start, the operator failing in error.

8 MEMBER MARCH-LEUBA: And if you put all of 9 those conservatively, you come up with a failure 10 probability for 120 percent.

11 MS. BRISTOL: That was not the failure 12 probability we calculated, but we did model the, you 13 know, potential failures of the injection in both of 14 those lines.

15 MEMBER BLEY: I take it by latent errors, 16 you're including things like Jose asked about valves 17 being misaligned?

18 MS. BRISTOL: Correct, and so if we go to 19 the next slide, as we quantified these events, as 20 you'll see in the next two slides after this, both 21 those latent errors as well as the recovery actions.

22 We calculated them and came up with a range from about 23 4E minus three to 2E minus five based on the different 24 events.

25 And we looked at the potential 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

68 1 unisolate and flood with the CVCS and CFDS. We looked 2 at the potential to start diesels, combustion turbine 3 generator, as well as the operator failing to push the 4 buttons in the control room, and so we looked at this 5 range of different operator actions and we ended up 6 bounding the values and did some post-processing as we 7 evaluated our models.

8 So what we did was rather than assigning 9 these small probabilities to each event that would 10 come up in a cutset, we applied a human error 11 probability of one in the first.

12 The first cutset was set to 40 minus 13 three. Then if there was a second operator action in 14 that same cutset, we assigned that 1.5E minus one, and 15 then if there was a third, 0.5.

16 And so we were able to capture dependency 17 that way as well as put a conservative bound on what's 18 the most limiting conservative human action that we 19 calculate in doing our HRA, and we just assigned that 20 to the first event whether that was the most limiting 21 calculated value being unisolating or flooding or 22 whether that was an operator going out into the field 23 and an operator pushing the button to actuate ECCS.

24 MEMBER MARCH-LEUBA: So for people that 25 don't speak PRA in their sleep, you're saying that the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

69 1 human error probability of actually doing the right 2 thing is only four in 1,000 trials?

3 MS. BRISTOL: Yes.

4 MEMBER MARCH-LEUBA: And this is a system 5 that uses computer procedures on a non-safety grade 6 computer that is programmed by a couple of graduate 7 students working at Oregon State.

8 I'm just going back. I'm acting like I'm 9 a salesman. I'm repeating my topic over and over 10 again. I find the number very small, very small.

11 MR. GALYEAN: Well, in the limiting case, 12 the operators have a minimum of 30 minutes to execute 13 this action, and these actions have actually been 14 tested out in our control room simulator, okay. So 15 the operators have run through these simulations where 16 they have created an upset condition.

17 In fact, they have used our sequences from 18 the PRA and mimicked those in the control room 19 simulator, and then they had the operators respond 20 with executing these particular operator actions.

21 So this isn't something we just pulled out 22 of the air. We've actually tested these in our 23 control room simulator.

24 MEMBER BLEY: Bill, Jose mentioned the 25 large number of valves in this distributed system.

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 Are all of those valves readable in the current 2 control room design and can they all be operated from 3 --

4 MR. GALYEAN: Well, of course they don't 5 all need to be operated to --

6 MEMBER BLEY: I didn't ask if they need to 7 be --

8 MR. GALYEAN: -- function. They're --

9 MEMBER BLEY: -- operated. I asked if you 10 can see their position. That's what I meant to ask.

11 MS. BRISTOL: We'd have to look at the 12 drawing to see which ones have indication. Numerous 13 ones do have indication. Most of them fail open or 14 fail in the way that the flow path remains open.

15 Outside of that, the demineralized water 16 isolation valves are the ones that would isolate on a 17 plain trip, you know, but the system is aligned. For 18 CVCS, the system is aligned to the boron admission 19 system nominally.

20 MEMBER MARCH-LEUBA: Excuse me, and this 21 is where the multi-module comes into play for human 22 factors. Did you consider the multi-module when you 23 tested this thing? Because when it rains, it pours.

24 The day you would want to have an accident 25 is when we have some unusual event that hits operators 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 in something they did not train in the simulator. If 2 it's an event they have trained in the simulator, 3 they'll do it right, absolutely, not for 10 to the 4 minus three or 10 to the minus six probability 5 failure.

6 It's those events that they didn't train 7 for and there's two operators, 12 modules. The two 8 operators are working modules A and B, and then module 9 F in the meantime is doing something unusual and 10 they're not looking at it. Here is where multi-module 11 analysis would really help. I mean, if we're going to 12 have a failure, it would be that.

13 MR. GALYEAN: We did look at operator 14 actions in our multi-module assessment, yes, and I 15 think we'll probably get to that at some --

16 MS. BRISTOL: In a couple seconds.

17 MR. GALYEAN: In a couple seconds.

18 MEMBER MARCH-LEUBA: The concern --

19 CO-CHAIR CORRADINI: Can I just ask his 20 question a little differently? So when you did a 21 simulation with watching one module have to survive an 22 upset, did you then have all of the various modules 23 being up and being simulated such that one might be 24 then confounded by something occurring on another? I 25 think that's kind of what you're 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

72 1 MEMBER MARCH-LEUBA: It's an external event 2 that following a seismic, a hurricane is coming. It's 3 going to be something like that, that a fix on 12 and 4 number 1 valve is sticking, and you don't see it.

5 That's what we worry about.

6 And even then, the reactor's so good that 7 it would probably survive it. It's just, I'm concerned 8 that the numbers that you calculated are not 9 realistic. That you are off by several hours. Remind 10 me to, because you didn't consider all these common 11 cause failures. A failure of computer system to tell 12 you to do the right thing, and there is something we 13 brought up in a previous meeting, and I want to repeat 14 it again.

15 The computer alarm system, a non-operator 16 action, is an excellent idea. It really gives you 17 better performance than if you don't have it. But you 18 have to train your operators to assume it fails. To 19 verify with different means that what the computer is 20 telling you to do is the right thing to do.

21 So yes, because you see over there that 22 module 7 is green, don't assume it's green. Go over 23 there and check it out. It's just, we're facing death.

24 MS. BRISTOL: Thank you. As you had 25 mentioned earlier here, the failures, the latent 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 failures that may occur that we've modeled in the PRA, 2 typically CFDS, CBCS, so the valves will misalign 3 during, you know, as we've been discussing, in those 4 systems prior to it being actuated, so during test and 5 maintenance those valves were misaligned.

6 Misalignments within the generators, that 7 combustion turbine generator, the diesels, or missed 8 toleration of MPS. We model that as well, as a 9 potential to be a human error.

10 In the response to the initiating events, 11 we 12 have operator failing to unisolate and initiate, 13 injection from both the CFDS or the CVCS, and CVCS can 14 be unisolated from the control room or locally, 15 depending on the initiating events, and then we have 16 modeled operators failing to start the diesel or the 17 CTG, as well as push the button for ECCS. So you see 18 that ECCS didn't actuate, they push the button and 19 that's an action from the control room.

20 Data. So, as we've been discussing, we've 21 been using industry data from the NUREG 69.28, we've 22 looked at LERs, common cause failures are modeled 23 based on the NUREG 54.97 and as we've mentioned, this 24 is generic data for a plant with no operating 25 experience, the expectation. As well we have COL 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 items. As information becomes available we update our 2 data to be consistent with the information available 3 at the time we're doing the PRA.

4 The expectation is that when a plant is 5 operating, they will update as well. There's an 6 expectation for every other refueling outage to update 7 the data to be consistent with the current industry 8 themes.

9 We've also done the design-specific 10 analyses, the passive safety system reliability, a 11 PSSR, as we've been discussing earlier. We did ECCS 12 and DHRS. We looked at those, to include those 13 failures into the model beyond just the valve failures 14 or the condenser failures of those systems. We looked 15 at potential TH failures that could apply.

16 We also looked at unique events, tube 17 failures. We did an analysis on the tube failures and 18 how to, in a severe accident what's the potential for 19 an induced steam generator tube failure, so we did 20 design specific analyses on those components as well.

21 And then on top of that, we did 22 sensitivities for the data that we weren't sure about, 23 or maybe a little, where there was more uncertainty, 24 we did sensitivities to support those, that data that 25 we used.

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 We quantified the PRA model with using 2 SAPHIRE code. In that code we included the common 3 cause failure modes and there was correlations in 4 uncertainty analysis, and so we utilized that in 5 SAPHIRE. We ended up calculating a truncation value of 6 1E-15 using the ASME and SPRA standard guidance that 7 related to the convergence criterion.

8 So when we looked at uncertainty, we 9 looked at both quantitative using SAPHIRE as well as 10 we did sensitivity studies. On these event, PSSR 11 events, MPS, we looked at different sensitivities, 12 orders of magnitude, different calculations, to 13 address any uncertainties within those new events, 14 that new data, that we didn't have operating 15 experience to support.

16 Quantitatively, we used the SAPHIRE code 17 to propagate the parametric uncertainties when we 18 calculated the mean as well as the point estimates for 19 the CDF and LRF. Sometimes we used augmented 20 sensitivity studies, so for the initiating event 21 frequencies, where we did something a little bit 22 different than what was in the industry. We evaluated 23 what the industry data would look like with our model, 24 and we calculated and compared that against our 25 internal events core damage frequency and large 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 release frequency.

2 CO-CHAIR DIMITRIJEVIC: Did you have any 3 combined sensitivity studies? I mean, you only present 4 the 1 factor. You never try to combine how that would 5 look like if more people practice.

6 MS. BRISTOL: That's correct. We didn't 7 look at numerous ---

8 CO-CHAIR DIMITRIJEVIC: In these studies, 9 what did you find, both for the standard and this too?

10 MS. BRISTOL: So, we have a slide in a 11 couple, and we can go over those various sensitivity 12 studies.

13 CO-CHAIR DIMITRIJEVIC: Okay.

14 MS. BRISTOL: The various data parameters 15 we looked at for initiating events, they weren't, 16 again they were data from the industry, and if we 17 weren't, we looked at various error bands on those. We 18 typically use an error factor of 10. Due to the fact 19 of a new design and with those initiating events from 20 the industry effective, the new scale module, the same 21 way.

22 So we captured that uncertainty in the log 23 normal error factor, and then after the cutsets were 24 generated in SAPHIRE, we did an uncertainty analysis 25 using Latin Hypercube and came up with the 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

77 1 estimate and the means for that evaluation.

2 Coming out of the quantification, we were 3 able to look at the importance as well as of the 4 various components and systems in the PRA, and we 5 compared these against the risk criterion that we 6 presented in the topical report a few years ago, and 7 so we evaluated we had --

8 CO-CHAIR DIMITRIJEVIC: Can you take us to 9 the slide? Sorry I missed this, because I was 10 concentrating on something else. When you said that 11 SAPHIRE can build that built-in ability to perform and 12 sets the analyses which include correlating failure 13 probability.

14 MS. BRISTOL: The various components, say 15 valves or pumps, would all have the same correlation 16 class and so they were correlated when we did the 17 uncertainty evaluation using SAPHIRE.

18 CO-CHAIR DIMITRIJEVIC: So you define 19 correlation groups in SAPHIRE.

20 MS. BRISTOL: Right.

21 (Simultaneous speaking.)

22 CO-CHAIR DIMITRIJEVIC: How did you choose 23 those correlation groups based, like, on common cause, 24 or how did you choose what the correlation would 25 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

78 1 MS. BRISTOL: Yeah, the similar design. So 2 the valves, if they were HOVs or MOVs or check valves, 3 or motor-driven pumps --

4 CO-CHAIR DIMITRIJEVIC: Well, usually base 5 correlation is showing a difference within your point 6 estimate and your mean value. Given your point 7 estimate, your mean value should be usually identical.

8 That means that if you don't correlate anything in 9 your point estimate, your mean values are the same.

10 So here, even the single events don't have 11 the impact and is all common cause, so that would be 12 the result that you are mean valuing your point 13 estimates throughout almost the same, or I wasn't sure 14 that you correlated. That was my question in 15 uncertainty analyses.

16 So because there is not too many 17 independent, you know, the state of knowledge, 18 relationships with companies back when you had that 19 sampling dating of certainty, you know, your two 20 valves had the same uncertainty -- maybe billions of 21 uncertainty multiplying them with that in the state of 22 knowledge increasing your probability of failure.

23 But if you go through the common cause and 24 you just have, you know, quote unquote factors like 25 that number you don't see that. So that was one of my 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 concerns, especially here when you're defending on 2 very small number but identical elements. So I, you 3 know, that was one of my concerns, this correlation 4 type.

5 It will be interesting to see how that 6 impacts. Will that be part of them and what assets 7 have a possibility to be impacted by this.

8 MS. BRISTOL: Anybody else? These are the 9 events we provided the D-RAP panel as well as just had 10 to show that the systems that were important were in 11 fact safety-related systems coming out of the PRA, 12 containment isolation valves as part of the 13 containment system, emergency core cooling system, 14 module protection system, ultimate heat sync, these 15 are important in the PRA as well as safety related.

16 The components that had the elevated risk 17 profiles were the reactor vent valves and the reactor 18 recirculation valves, the actuation valves of the 19 decay heat removal system, the safety relief valves 20 and the CVCS and CES isolation valves, those are 21 required to open when we inject and then the 22 combustion turbine generator was also right a 23 Basically the Fussell-Vesely threshold.

24 25 Other events and initiators, we also look 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 at initiating events as well as human actions. The 2 crane was important based on our PRA, LOCA inside 3 containment, and outside containment, loss of outside 4 power, fires, floods, and the various hazards, we 5 looked at all the initiating events and provided those 6 inputs.

7 And as we've discussed, the human actions 8 being the ability to inject into the vessel were both 9 risk-important.

10 MEMBER BLEY: These are all in terms of 11 contribution to core damage treatments.

12 MS. BRISTOL: That was a clarification that 13 I wanted to make.

14 MEMBER BLEY: Did you do RAWs as well?

15 MS. BRISTOL: We did.

16 MEMBER BLEY: Do you have a slide on those?

17 MS. BRISTOL: The systems, we didn't 18 explicitly say on this slide whether it was the RAW or 19 the Fussell-Veseley. It was provided in the SR, but 20 the systems on here would be based on RAWs, not on 21 Fussell-Veseley. And just to clarify, the human 22 actions aren't, it's based on level 2 and low-powering 23 shutdown, not the internal events --

24 MEMBER BLEY: In the committee, the jargon 25 is, what's achievement worth? And that means if 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

81 1 component fails, it's guaranteed to fail, how much 2 increase in core damage frequency do you get, rather 3 than the other one that's telling you what fraction of 4 the core damage frequency is due to it.

5 So those systems up there were the ones 6 that had high RAWs, if that makes sense.

7 MS. BRISTOL: And that the human actions 8 are related to LRF and low-powering shutdowns, and in 9 the low-powering shutdown model, so not the core 10 damage frequency.

11 CO-CHAIR DIMITRIJEVIC: So this Fussell-12 Vessell is contribution to what?

13 MS. BRISTOL: To large-release frequency.

14 Level 2, the operator action associated with Level 2 15 was in the containment flooding and drain system, 16 operator action was part of Level 2, and then low-17 powering shutdown with this CVCS --

18 CO-CHAIR DIMITRIJEVIC: So they don't show 19 as important if you just look in the core damage 20 report.

21 MS. BRISTOL: That's right. Of a Level 1.

22 MEMBER BLEY: The interesting thing about 23 the RAW contribution, it's where you would expect ECCS 24 but that sum is, if we got the failure probability's 25 wrong, reliability values wrong on these valves, it's NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

82 1 a big deal It's much worse than we thought it was.

2 MR. BARBOUR: Say that again, Dennis, 3 please, I want to make sure I understand that.

4 MEMBER BLEY: If the likelihood of failure 5 of those systems up there is substantially higher than 6 we thought it was going in, there would be a big 7 change in core damage treatments.

8 MR. BARBOUR: If that's what we see.

9 MEMBER BLEY: That's what you see up there 10 in the top line.

11 CO-CHAIR DIMITRIJEVIC: It goes to 2E-5, I 12 saw that somewhere. I think it comes to the ultimate 13 heat sync and ECCS have the same values because 14 basically you cannot mitigate any LOCA events without 15 that, so it basically comes to your LOCA frequency.

16 MEMBER MARCH-LEUBA: So it's failure to 17 open when on demand?

18 CO-CHAIR DIMITRIJEVIC: If it's assumed it 19 always fails to open, then your core damage frequency 20 will be something ---

21 (Simultaneous speaking.)

22 MEMBER MARCH-LEUBA: IAB, inadvertent 23 actuation block, failure will not continue to ---

24 (Simultaneous speaking.)

25 MR. GALYEAN: Okay, the IAB is a kind of 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

83 1 special case ---

2 MR. BARBOUR: Help us with what that means.

3 (Simultaneous speaking.)

4 MR. GALYEAN: Obviously we spent a lot of 5 time looking at the performance of the IAB in the 6 context of the PRA and in the context of the thermal-7 hydraulic simulations we did to support the PRA. Of 8 course the function of the IAB is to prevent spurious 9 actuation of the ECCS system. And so it depends on 10 what kind of failure mode you're talking about, right?

11 Does the IAB fail to inhibit initial action--

12 MEMBER MARCH-LEUBA: I'm more concerned of 13 the IAB remains closed, prevents it from opening. It 14 does its job too well if it prevents it from opening, 15 even though it needs to open. Have you considered 16 that?

17 MR. GALYEAN: Yes. We do consider that in 18 the PRA.

19 MEMBER MARCH-LEUBA: With the proper 20 sequences?

21 (Simultaneous speaking.)

22 MR. GALYEAN: But I think you made your 23 point. I think what, I want to make sure I understand, 24 that the whole community understands your point by the 25 RAW and the Fussell-Veseley, is you're assuming 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

84 1 unity failure and then looking at its jump-in 2 probability.

3 MS. BRISTOL: That's right.

4 MEMBER BLEY: No. A change in probability 5 of the core damage --

6 MR. BARBOUR: I'm sorry. I'm sorry. I said 7 that wrong.

8 CO-CHAIR DIMITRIJEVIC: How do you, do you 9 have a basic event for them?

10 MR. GALYEAN: For the IAB?

11 CO-CHAIR DIMITRIJEVIC: Yes.

12 MR. GALYEAN: No. Only in the, for the ECCS 13 valve failing to operate the way it's designed.

14 CO-CHAIR DIMITRIJEVIC: So if ECCS valve 15 fails to open, you have analyzed this which concludes 16 that the block valves --

17 MR. GALYEAN: The IAB?

18 CO-CHAIR DIMITRIJEVIC: Yes.

19 MR. GALYEAN: Right. We did kind of an off-20 line piece-part analysis of the ECCS valve, which 21 included the IAB, and then we took the result of that 22 analysis and plugged it into the ECCS valve basic 23 event in the PRA model.

24 MEMBER BLEY: Doesn't that affect the 25 common cause failure of those valves?

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

85 1 MR. GALYEAN: Well, we have common cause 2 failure of the ECCS valves.

3 MEMBER BLEY: Well, I mean if you did the 4 IAB separately and just put it into the failure rate 5 of the valve, that doesn't put its failure into the 6 common cause contribution.

7 CO-CHAIR DIMITRIJEVIC: And the other 8 factors.

9 MR. GALYEAN: Again, I ---

10 MEMBER BLEY: You didn't think it affected 11 it.

12 MR. GALYEAN: Well, only in the sense that 13 if the IAB affects the failure of the ECCS valve, then 14 the common cause failure model in the PRA looks at 15 common cause failure of the ECCS valves due to the 16 same failure mechanism. I mean, the common cause 17 failure modeling, it's a parametric model, right? It's 18 not a mechanistic model. It doesn't look at all the 19 different ways the --

20 (Simultaneous speaking.)

21 MEMBER BLEY: I agree with that. But 22 there's only one IAB, right?

23 MR. GALYEAN: No. Each valve has its own.

24 (Simultaneous speaking.)

25 MEMBER BLEY: Okay. I forgot 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

86 1 MR. BARBOUR: I view it almost like they're 2 together.

3 MS. BRISTOL: It's an internal --

4 MR. GALYEAN: It's embedded in the valve.

5 MR. BARBOUR: It's an integral part of the 6 valve.

7 MEMBER BLEY: Yeah, okay.

8 CO-CHAIR DIMITRIJEVIC: What you say is 9 common cause is true. But it's also true that this is 10 completely new, so we basically don't know anything 11 about common cause. So if you looking at failure 12 mechanisms and conclude that those failure mechanisms 13 are always applicable to both valves, now you can 14 actually conclude your own common causes going to be 15 the band, there is no way the band can fail at the 16 same cause doesn't apply to another.

17 So then you, I mean that will be, because 18 you are very sensitive to the common cause. It's not 19 on your sensitivity standard, so in your other 20 assumptions that, when you did the analyzing of 21 failure modes, maybe you should include this common 22 cause there too. What is the, why would we choose some 23 common cause parameters, but not other?

24 Your -- basically fails to open for the 25 added is 2.52E-6 common cause data, which is 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

87 1 a low number, right? It's a 2E million, and this is 2 one of the most important components, right, so, it's 3 interesting, I mean I have to go back to this passive 4 system analyses, maybe you have there a break and if 5 this fails to open, based on standby failure rate, or 6 on the demand failure rate? Wherever you got that. Is 7 it based on standby?

8 MS. BRISTOL: It's based on demand. So when 9 the system demands it. And also to note --

10 CO-CHAIR DIMITRIJEVIC: It does include 11 this 2ES, that's intervallic in the fails to open 12 number. If you are using standby failure then you will 13 have to test intervally. If you are using just the 14 demand rate that you took from some data, based on 15 whatever you found, something like, I guess that BWR 16 safety valve, so something you just took the demand 17 rate, not the standby.

18 MS. BRISTOL: And also in the PRA we model 19 the potential to open at low pressures. And so, 20 indifferent of the IAB position, the valves can open 21 with a low DP. And so that's also captured in the PRA.

22 That at low pressures, these valves will open 23 indifferent of the IAB position.

24 CO-CHAIR DIMITRIJEVIC: And you assume it's 25 point 1, 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

88 1 MS. BRISTOL: Correct.

2 CO-CHAIR DIMITRIJEVIC: Probability to 3 open.

4 MS. BRISTOL: Yes. So that's also captured, 5 so if it opens or closes, the valves will open with 6 the pressure between the RPVNC is low.

7 CO-CHAIR DIMITRIJEVIC: That's in a 8 specific sequence. I did not find it in any cutsets or 9 things there.

10 MS. BRISTOL: No, it's, I shouldn't say no, 11 but it's in the fault tree model as a potential to 12 actuate PCCS valves to open is the low DP.

13 MEMBER BLEY: I just want to make a 14 comment. You may not like it. If you use the PRA along 15 the way to make decisions about the design of this 16 plant, when the PRA finds that something's kind of 17 crucial, like the ECCS valves and they're new, seems 18 to me a recommendation out of the PRA to the rest of 19 the project about what kind of testing's necessary 20 would have been very appropriate. Still would be.

21 And to me, that kind of test ought to 22 include multiple valves so that we see if we have 23 common cause problems that we don't know about yet.

24 MS. BRISTOL: Thank you, yes. And there's 25 ongoing testing plans with the design group 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

89 1 have been in conversation with and not at this point 2 in the design, but as first of a kind it goes on, we 3 have been talking to them with respect to testing 4 potential.

5 MEMBER BLEY: Okay, good. And I hope you'll 6 consider in this area, because I think your expert 7 panel is raised, as we've raised, and you must have 8 thought hard about it. The only way out is going to be 9 to get enough testing to be comfortable with the 10 results. Or let Mother Nature teach us about it later.

11 MEMBER MARCH-LEUBA: Before just that, 12 what's a HEP? H-E-P?

13 MS. BRISTOL: Human error probability.

14 Those are the human actions. And so what we did for 15 various sensitivities, here's a list of some of them, 16 we set all of the human actions to always succeed, and 17 we looked at the impact on the core damage in the 18 large release. As you can see, our base case PRA 19 values for the core damage frequency, 2.70-10, large 20 release frequency 1.7E-11, we were able to see and 21 evaluate the significance of the various sensitivities 22 to that base CDF and LRF results.

23 If we failed, if all of the operator 24 actions failed, you could see the elevated core damage 25 as a result as well as a large release frequency, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

90 1 MEMBER MARCH-LEUBA: Is that because the 2 CDF is controlled by something else, like the module 3 drop?

4 MS. BRISTOL: These are internal events --

5 (Simultaneous speaking.)

6 MS. BRISTOL: With respect to the Level 1 7 internal events. The module drop is not included.

8 CO-CHAIR DIMITRIJEVIC: Those valves, ECCS 9 valves.

10 MEMBER MARCH-LEUBA: ECCS valves failing is 11 for control, so everything you put into has nothing to 12 do with it?

13 CO-CHAIR DIMITRIJEVIC: Not too much.

14 MEMBER MARCH-LEUBA: Is that what you're 15 saying?

16 MEMBER BLEY: I don't know if you have to 17 show these to anybody else at another time, but if you 18 use success and failure instead of --

19 MS. BRISTOL: True and false.

20 (Simultaneous speaking.)

21 MEMBER BLEY: That would be helpful to most 22 people.

23 MS. BRISTOL: Thank you. We then looked at 24 common cause failures and as we discussed earlier, you 25 can see here the common cause failures is the impact 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 to the PRA.

2 CO-CHAIR DIMITRIJEVIC: Explain to us, what 3 did you set to one, common cause events or factors?

4 MS. BRISTOL: The events.

5 CO-CHAIR DIMITRIJEVIC: So, total events, 6 including a failure, right? For example, that anomaly, 7 ECCS 2-3, that was set to one.

8 MS. BRISTOL: Correct.

9 MR. GALYEAN: 2 times semi.

10 MS. BRISTOL: Sorry. 2E-3. They were set 11 to, yeah.

12 CO-CHAIR DIMITRIJEVIC: By failures you 13 mean success.

14 MR. GALYEAN: The probability.

15 MS. BRISTOL: Yes. They were set to success 16 or the failures were set to 2E-3. The EPRI guidance --

17 MEMBER MARCH-LEUBA: So you're saying the 18 probability of the two are VV or RSV, the two ECCS 19 valves failing is 2.10-3. The two of them at the same 20 time is 2.10-3. In my mind, if you designed the valve 21 wrong, it won't fail as another one is about to go, 22 wouldn't you say? I mean, I keep repeating, I'm trying 23 to sell cars, what is this 2.10-3 come from? And 24 happens if it was 2.10-2?

25 MEMBER BLEY: You can see that pretty 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 easily from the common cause failures up there. The 2 core damage frequency jumped all the way to 2D-. So E-3 6. It's essentially, as you make that number bigger, 4 as it goes up to the challenge frequency.

5 MEMBER MARCH-LEUBA: Yes, of course, then 6 if these valves are supposed to be very highly 7 reliable, if they fail, it was of course because they 8 failed. And the other one has the same course.

9 CO-CHAIR DIMITRIJEVIC: It might.

10 (Simultaneous speaking.)

11 MR. GALYEAN: Just to be clear, okay, just 12 to be clear, that's not, that 2x20-3 is not a 13 conditional probability. It is the independent 14 probability of the set of valves failing. All failing.

15 MEMBER BLEY: Oh, thank you.

16 CO-CHAIR DIMITRIJEVIC: I just asked you 17 but I am thinking about how do you do that in the 18 model? You went, because you have a, you know, 19 everything is broken on the 2, 3, 4, did you only, 20 which ones did you set to E-3?

21 MS. BRISTOL: It quantified basic events in 22 the model.

23 CO-CHAIR DIMITRIJEVIC: Every common cause, 24 basic event. That's a message to 2, 3, 4, every common 25 cause basically.

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 MR. GALYEAN: That's correct.

2 MEMBER BLEY: And so for Jose's statement, 3 it's not just the 2 valve, any set of multiple valves 4 that have a common cause failure are all being failed 5 on this --

6 MS. BRISTOL: Pumps, diesels, all of the 7 multiple component groups in the --

8 MEMBER BLEY: There is no redundancy left.

9 MEMBER MARCH-LEUBA: Yeah, but there was a 10 control of ECCS valves, right? And what's the failure 11 for, do you remember, of one ECCS valve failing?

12 MR. GALYEAN: I don't remember off the top 13 of my head.

14 MEMBER MARCH-LEUBA: Is it 2.10-3 also? Or 15 is it higher?

16 MR. GALYEAN: I think it's lower than that.

17 I think it's, yeah.

18 CO-CHAIR DIMITRIJEVIC: It's 5.88E-5.

19 MEMBER MARCH-LEUBA: E-5? Failure of one of 20 those valves?

21 MEMBER BLEY: For a fail open valve.

22 MR. GALYEAN: These are failsafe.

23 CO-CHAIR DIMITRIJEVIC: And the tree valve 24 is 3.8D-4. This is main valve, is 5.88E-5. Tree valve 25 is 3.8D-4.

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 MEMBER BLEY: In the very beginning, you 2 said our components ought to be at least as good as 3 what's out there in the industry, so when you found a 4 failure rate for fail-open, spring-operated valves, 5 was that this number or did you say that we're a 6 little better than that?

7 CO-CHAIR DIMITRIJEVIC: Like BWR SOV, which 8 you said you used, is that the number for --

9 MR. GALYEAN: We used the data as --

10 MEMBER BLEY: From the NUREG.

11 MR. GALYEAN: From the NUREG, 12 MEMBER BLEY: Okay. So that's data for all 13 of this kind of valve that's out in the current fleet.

14 MS. BRISTOL: That's correct.

15 CO-CHAIR DIMITRIJEVIC: Yes. That's why 16 they have 5.88 as the rate to -- (Simultaneous 17 speaking.) We don't know anything, how did we figure 18 out to do that --

19 MEMBER BLEY: We used to tell people, teach 20 people not to do that.

21 MR. BARBOUR: So we're into discussion, you 22 have one more slide, I think that then leads us into 23 external hazards. Do you want to take that slide now, 24 or take a break?

25 CO-CHAIR DIMITRIJEVIC: We can take break, 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 because that's the Level 2 methodology.

2 MR. BARBOUR: It leads us back to external 3 hazards, that's the thing. I think all they're doing 4 is defining how they transition in this slide.

5 MS. BRISTOL: That's right. We can do the 6 next slide. The Level 2 methodology, as we mentioned, 7 we look at containment isolation failures and that's 8 what we look at in the PRA model, SAPHIRE model, 9 leading to large release isn't necessarily the dose at 10 the site boundary but it's just the fact that 11 containment is unisolated.

12 We don't use bridge trees in all of the 13 end states that end in core damage in the Level 1 14 event trees then just transition directly to the 15 containment event tree and those all then get directly 16 linked.

17 MR. BARBOUR: Tell me to this one now, I 18 just want to raise it, I'm not sure if you'll put many 19 event trees up there later or not. You have a lot of 20 cases where you have leaks from one tree to another, 21 and some of them, like for -- Some of them are very 22 simple. I'm not sure why in the world you did that and 23 it's really hard to chase from one to the other. The 24 computer can do it easily, but I have trouble finding 25 that next link. Don't talk about it now, but if we get 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 into some of them, tell us why you did that.

2 MS. BRISTOL: Okay.

3 MR. BARBOUR: I think this is a good time, 4 because they're going to switch to a topic of external 5 hazards next. So why don't we take a break until 6 quarter-of?

7 (Whereupon, the above-entitled matter went 8 off the record at 10:27 a.m. and resumed at 10:42 9 a.m.)

10 CO-CHAIR CORRADINI: Let's reassemble so 11 we can get started. Sarah, do you want to start us 12 off again, please.

13 MS. BRISTOL: I'll go ahead and we'll talk 14 about internal events at a high level. We evaluated 15 internal fires, floods, external floods, high winds, 16 just consistent with the various parts of the 17 standard, and looked at the requirements of those and 18 evaluated those.

19 And we also did a seismic margin 20 assessment and looked at Part 5 of that standard.

21 CO-CHAIR CORRADINI: And that's what we're 22 going to hear about now?

23 MS. BRISTOL: That's the next slide, yes.

24 MR. GALYEAN: So, as Sarah just said, 25 NuScale did a PRA base seismic margin assessment, 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

97 1 SMA.

2 In the context of that, we did perform 3 some design-specific fragility calculations. We 4 actually employed a set of three different consulting 5 companies to assist us in developing these design-6 specific fragility calculations.

7 Now, we had the consultants focus on 8 those, mostly, structures that we felt would be the 9 dominant drivers of the SMA results.

10 For many of the other components in the 11 design, like valves and pumps and things, we typically 12 use generic capacities that we obtain from the 13 literature and then modified them using the in-14 structure response factors that were -- that we 15 obtained from our seismic design folks -- or the 16 design folks when they did their seismic analysis.

17 Per the Interim Staff Guidance 20, they --

18 the figure of merit -- or the metric, I guess I should 19 say, for the acceptance criteria, is that the high-20 confidence-of-low-probability-of-failure out of the 21 SMA should be at least 1.67 times the design basis 22 earthquake; or, you know, to be more precise, the 23 certified seismic design response spectra.

24 And in NuScale's case, that means our 25 acceptance criteria is that our HCLPF needs to be 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

98 1 least 0.84g or higher peak ground acceleration.

2 So, the seismic margins assessment is 3 basically layered on top of the full-power internal 4 event PRA logic.

5 The fault trees -- basically, what you --

6 what happens is all the basic events in the PRA model 7 now get an additional failure mode of seismic failure 8 attached to them.

9 And so, what you then have, basically, is 10 a model that includes the random, independent failures 11 org'd with the seismic failure, okay. So, a component 12 can fail either randomly, or it can fail by virtue of 13 the seismic event, okay.

14 As I said, the result that comes out of 15 the seismic margins assessment is the high-confidence-16 of-low-probability-of-failure.

17 And basically, it just means that your --

18 you have a 95 percent confidence that the probability 19 of failure is no greater than 5 percent.

20 An alternate way of looking at that, 21 though, is that you have your best estimate confidence 22 level that the conditional failure probability is no 23 greater than one percent, all right. So, those two 24 definitions are essentially equivalent, okay.

25 MEMBER BLEY: Can I slip in a question?

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 MR. GALYEAN: Yeah.

2 MEMBER BLEY: In the staff audit, they ask 3 you guys a question and I haven't seen the answer to 4 it yet.

5 You did what, in the past, I think, was 6 called an NRC seismic margins. There was also an EPRI 7 seismic margins approach that did a pathways to 8 success approach.

9 They ask you if both safety-related and 10 nonsafety-related equipment was included in the model 11 for the seismic margins.

12 So, I'll ask you the question, because I 13 didn't find the answer to it.

14 MR. GALYEAN: Okay. I guess I'm -- yes, 15 I mean, we included both safety-related and nonsafety-16 related --

17 MEMBER BLEY: Okay. In the margins --

18 MR. GALYEAN: Yes.

19 MEMBER BLEY: Okay.

20 MR. GALYEAN: Yes.

21 MEMBER BLEY: That's all.

22 MR. GALYEAN: So, the margins assessment, 23 then, I mean, basically what we're doing in the 24 margins assessment is determining those seismic 25 failures that would result in a conditional core 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 damage probability of greater than one percent.

2 And translated, you know, what is the 3 postulated peak ground acceleration that would produce 4 a conditional core damage probability of one percent?

5 And that's basically what the margins 6 assessment does. It determines what is that peak 7 ground acceleration.

8 Included in that assessment, we looked at 9 structural failures. And those comprise, basically, 10 the reactor building walls, the crane, the module 11 supports, that type of thing, those that would touch 12 the module directly, okay.

13 Next slide, please.

14 DR. SCHULTZ: Bill, before you leave that 15 one, when you say "we looked at the structural 16 fragilities" and you mentioned on the previous slide 17 the component fragilities, is that something that the 18 consultants have done?

19 Who did what here --

20 MR. GALYEAN: Yeah.

21 DR. SCHULTZ: -- in terms of the 22 evaluation specifically of the fragilities?

23 MR. GALYEAN: Yeah. For the most part, 24 the consultants looked at the structural fragilities, 25 okay, the fragilities of the major structures.

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 And NuScale did the component fragilities 2 using the generic fragilities modified by the in-3 structure response spectra that came out of the 4 NuScale design -- came from the NuScale design folks.

5 DR. SCHULTZ: So, the evaluation of the 6 new valves -- or the design valves, for example, that 7 fragility was determined by NuScale?

8 MR. GALYEAN: Well, again, I say for the 9 most part.

10 DR. SCHULTZ: That's what I want to find 11 out.

12 MR. GALYEAN: I'd have to check.

13 DR. SCHULTZ: Okay.

14 MR. GALYEAN: I do believe that the ECCS 15 valves and the DHR valves were evaluated by the 16 consultants.

17 CO-CHAIR DIMITRIJEVIC: And how did you 18 incorporate (speaking off mic) --

19 MR. GALYEAN: Not directly into the fault 20 trees. Those were handled at a higher level at the 21 event tree level, and we'll get to it.

22 I think in the next slide -- let's see.

23 Yeah, this is good.

24 MEMBER BLEY: I want to --

25 MR. GALYEAN: Basically, the structural 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 failures -- I'm sorry, did you --

2 MEMBER BLEY: I'm sorry, go ahead. You're 3 going ahead with this slide.

4 Back on what Steve asked you, I'm -- there 5 must be a separate engineering report on --

6 MR. GALYEAN: Right.

7 MEMBER BLEY: -- the evaluation of those 8 valves. That would be interesting. We'll ask the 9 staff if they've had a chance to look at that.

10 The reason I ask, is because it ought not 11 be the same people who were looking at your 12 structures, it ought to be people who really know the 13 guts of how the valve is put together and usually 14 that's not the same kind of people.

15 MR. GALYEAN: We -- again, we had the 16 three different consultants. And, in some cases, we 17 gave the same component or structure to more than one 18 consultant.

19 And obviously, the consultants all 20 provided their final reports to NuScale and those 21 were, you know, we do have those, as well as our own 22 fragility calculation reports.

23 MEMBER BLEY: Okay. And those aren't 24 things we can see, at least not --

25 MR. GALYEAN: If you do an audit. I 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

103 1 --

2 MEMBER BLEY: Okay.

3 MR. GALYEAN: So, the seismic response, I 4 talked about the structural failures handled at a high 5 level.

6 Basically, the assumption was made that if 7 the -- one of these major structures failed, whether 8 it's the crane or the reactor building walls or the 9 bay walls or whatever, that just automatically went to 10 core damage, okay.

11 There was no potential for mitigation or 12 recovery or anything like that. It just went right to 13 core damage.

14 If the plant survived the structures, then 15 it looked at the major -- well, LOCAs, for example.

16 And then if it survived the LOCAs, then it assumed the 17 loss of offsite power.

18 We did no pre-screening of the internal 19 event PRA results. The seismic model overlay 20 comprised the entire, you know, logic of the full-21 power initiating internal events.

22 And we evaluated the model for 14 23 different ground motions ranging from 0.5g to 3-1/2 g.

24 And the end result was that at 0.88g is where we 25 determined that there would be the one percent chance 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 of core damage.

2 And that was dominated by the structural 3 failures, you know, the crane, the exterior walls, the 4 bay walls and module supports.

5 Again, we simply assume that if failure 6 occurred -- and, in the case of the crane, it was the 7 seismic restraints on the trolley of the crane that 8 hold it to the rails that rotate.

9 If those structural seismic restraints 10 were to fail, we'd simply assume the crane falls on 11 top of the modules and it goes to core damage.

12 In the case of the exterior walls, it was 13 if there was a displacement crack occurred, we simply 14 assumed the walls collapse and it goes right to core 15 damage.

16 MEMBER BLEY: So, for all the structures, 17 you assumed it went to core damage?

18 MR. GALYEAN: That's right. That's right.

19 MEMBER BLEY: Okay.

20 CO-CHAIR DIMITRIJEVIC: And that will 21 apply to all modules?

22 MR. GALYEAN: That's right, yeah. I mean, 23 it was just -- we didn't differentiate 1 versus 12.

24 It just -- we figured, you know, we're just 25 calculating the HCLPF at this point. We're 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

105 1 calculating core damage frequency.

2 MEMBER MARCH-LEUBA: Did the seismic 3 analysis evaluate internal components?

4 I'm always worried about the control rods 5 with a very long drive on top of them.

6 MR. GALYEAN: Yeah. In our case, of 7 course, our control rods -- our fuel is only half 8 height. So, the control rods, of course, are half 9 height and --

10 MEMBER MARCH-LEUBA: No, I'm talking about 11 the driver, the spaghetti that was on top of it to 12 hold them and activate them.

13 If it misaligns by two degrees after the 14 seismic event, it won't go in and you will say that it 15 doesn't need to go that much anyway because how much 16 is of no concern, right?

17 How about additional LOCAs in the CVCS 18 lines? Because the way I see it, is you have the 19 containment, which is a very heavy equipment, on the 20 vessel, which is a very heavy equipment, two flanges 21 and a two-inch pipe in between.

22 If anything moves in there, the -- at what 23 "g" level do you break the CVCS lines?

24 MR. GALYEAN: Yeah, I can't answer that 25 question off the top of my head. As I said --

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 MEMBER MARCH-LEUBA: Doesn't take much.

2 MR. GALYEAN: Yeah.

3 MEMBER BLEY: You looked at that?

4 MR. GALYEAN: Oh, yes.

5 CO-CHAIR CORRADINI: Can we get back, 6 then, to that question if you have -- can we get some 7 response to that if you have it?

8 MR. GALYEAN: Sure.

9 MEMBER MARCH-LEUBA: Because you have two 10 heavy elements --

11 MR. GALYEAN: Right.

12 MEMBER MARCH-LEUBA: -- connected by a 13 two-inch line in which --

14 MEMBER BLEY: I don't think you answered 15 Jose on the reactor internals question. I know many 16 years ago that was looked at in great detail on 17 existing PWRs.

18 Did you do fragility analysis on the 19 vessel internals and the control rod drive?

20 MR. GALYEAN: I can't answer. I don't 21 recall exactly the details of that.

22 MEMBER BLEY: We'd like to hear if you did 23 or not --

24 MR. GALYEAN: Okay.

25 MEMBER BLEY: -- or if you took it 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

107 1 somewhere else.

2 I didn't ask it earlier, but you did -- I 3 looked through the crane stuff here pretty well.

4 You did the crane in the Level 1, right?

5 MR. GALYEAN: Right.

6 MEMBER BLEY: When you used failure of the 7 crane, the probability of failure of the crane, did 8 you also look at the accouterments, the connection 9 mechanisms, the temporary things that are brought to 10 make connections and the operator -- the hooking up of 11 those?

12 Is there an operator involved in any of 13 the hookups on the crane lifts?

14 MR. GALYEAN: No.

15 MEMBER BLEY: Not at all?

16 MR. GALYEAN: Well, I think there's an 17 auxiliary crane attached to the main bridge crane 18 that's used for incidental things, but that doesn't 19 really factor into, you know, what we did here.

20 MEMBER BLEY: Okay. But the bridge crane, 21 there's no connections that operators have to monitor 22 and manually hook up?

23 MR. GALYEAN: I -- I mean, there's a 24 design-specific -- what we call a module lifting 25 adaptor.

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 MEMBER BLEY: Uh-huh.

2 MR. GALYEAN: That connects the modules to 3 the crane.

4 MEMBER BLEY: So, they have to put that in 5 place?

6 MR. GALYEAN: Pardon me?

7 MEMBER BLEY: That has to be connected to 8 the crane before the lifts?

9 MR. GALYEAN: No, it's --

10 MEMBER BLEY: A permanent part?

11 MR. GALYEAN: Yeah, a permanent part of 12 the crane.

13 MEMBER BLEY: Okay.

14 MR. GALYEAN: Next slide.

15 This is just an illustration to show the 16 major structures that were evaluated, the reactor 17 building walls, the crane, the bioshield, the bay 18 walls and the module supports.

19 CO-CHAIR CORRADINI: Is the bioshield 20 strictly a radiation shield or does it have some 21 structural integrity?

22 MR. GALYEAN: I believe it's just a 23 bioshield. I don't believe there's any structural 24 mission to it or function to it.

25 CO-CHAIR CORRADINI: So, if shaken 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

109 1 impacted, it would just fall off. I'm asking the 2 question -- I'm trying to understand what structural 3 integrity is. That's what I'm interested in.

4 MR. GALYEAN: I mean, we did -- that was 5 part of the fragility -- I mean --

6 CO-CHAIR CORRADINI: Oh, it was?

7 MR. GALYEAN: It was part of the fragility 8 evaluation.

9 CO-CHAIR CORRADINI: So, if I enter into 10 a seismic event that's of a sufficient magnitude, it 11 would shake and fall into the --

12 MR. GALYEAN: On top of the module.

13 CO-CHAIR CORRADINI: On top of the module, 14 which then would lead to --

15 MR. GALYEAN: Core damage.

16 MEMBER BLEY: Assumed core damage.

17 MR. GALYEAN: Right.

18 CO-CHAIR CORRADINI: So, let me ask the 19 question a different way.

20 So, now I do that, all 12 would then 21 undergo -- you know what I'm asking? Because if I 22 have a sufficient seismic event, all 12 are being 23 affected by it simultaneously.

24 MR. GALYEAN: Yeah.

25 CO-CHAIR CORRADINI: Do you understand my 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 question?

2 MR. GALYEAN: That's an issue with the 3 current state of the art in seismic analysis whether 4 seismically induced failures like this are fully 5 correlated, you know, or not.

6 CO-CHAIR CORRADINI: Oh, we address that 7 in a different manner.

8 MR. GALYEAN: Yeah.

9 CO-CHAIR CORRADINI: Okay. I remember 10 this.

11 MEMBER BLEY: Well, you know, you've got 12 -- but you've kind of got six and six.

13 Aren't those the bioshields across --

14 MR. GALYEAN: But, again, we're not --

15 MEMBER BLEY: -- six of them pretty much 16 hooked together?

17 MR. GALYEAN: Yeah. Again, we're not 18 calculating core damage frequency here. We're just 19 calculating high-confidence-of-low-probability-to-20 failure.

21 CO-CHAIR CORRADINI: So, you're just 22 basically saying if I pass a threshold, it fails to go 23 to core damage.

24 MR. GALYEAN: That's right.

25 CO-CHAIR CORRADINI: So -- well, okay.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

111 1 All right. Thank you.

2 MEMBER REMPE: So, I recall seeing 3 somewhere that there's some changes in the bioshield 4 that are being discussed or are happening.

5 Could you elaborate, and did that affect 6 this analysis or not?

7 MR. GALYEAN: Again, I can't comment on 8 that, you know. All I can say is all the bioshields 9 are independent, you know. They're not connected to 10 each other.

11 MEMBER REMPE: Can you elaborate, but you 12 can't say because you want to talk about it in the 13 closed meeting?

14 What is the change to the bioshield and 15 does it affect --

16 MS. BRISTOL: It's still under -- there 17 are RAIs still out there.

18 MEMBER REMPE: So, you have not decided 19 what the change is.

20 MS. BRISTOL: Correct. The staff has not 21 evaluated the responses to the bioshield design.

22 CO-CHAIR CORRADINI: But from the 23 standpoint to help Joy, I think it's connected to 19.2 24 and hydrogen distribution.

25 MR. GALYEAN: That's 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

112 1 MS. BRISTOL: There is no --

2 MEMBER REMPE: There's a different 3 material or something or something different in the 4 design, is what I'm asking.

5 MEMBER SKILLMAN: I think polyethylene is 6 removed. Replaced with concrete.

7 MEMBER REMPE: Okay.

8 MEMBER SKILLMAN: And then I think another 9 portion of the shield curtain had polyethylene added 10 to it.

11 So, I think they were -- the material 12 changes had to do with concrete, steel and 13 polyethylene.

14 MEMBER REMPE: That wouldn't affect your 15 analysis of the --

16 MR. GALYEAN: Yeah. I guess I would 17 rather not talk about design details.

18 MEMBER REMPE: Okay. But maybe in the 19 closed session or something.

20 CO-CHAIR DIMITRIJEVIC: But in the 21 previous slide, something --

22 (Simultaneous speaking) 23 CO-CHAIR DIMITRIJEVIC: -- you know, when 24 you say crane study was least --

25 MR. GALYEAN: I said -- okay. That's 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

113 1 one of the structural failures that drives the HCLPF.

2 CO-CHAIR DIMITRIJEVIC: It's not?

3 MR. GALYEAN: It's not. It was evaluated, 4 you know. There was a fragility calculation made, but 5 it's not one of the drivers of the HCLPF.

6 CO-CHAIR DIMITRIJEVIC: Okay. Even it was 7 assumed also that it could lead to core damage.

8 MR. GALYEAN: Right.

9 CO-CHAIR DIMITRIJEVIC: All right.

10 MEMBER MARCH-LEUBA: When I look at -- are 11 you done?

12 CO-CHAIR DIMITRIJEVIC: Yes.

13 MEMBER MARCH-LEUBA: When I look at this 14 figure, I -- this is the component that worries me and 15 I direct you to it, see where it says "refueling 16 machine" and there's a yellow thing?

17 Move your eyes to the left and there is a 18 gray wall, which is adjacent to the refueling pool --

19 the spent fuel pool.

20 Can that wall fall on the spent fuel and 21 has that been evaluated?

22 I mean, when I'm looking at this, I'm 23 thinking refueling. That's the first thing that 24 points to me is that wall is going to go boom on top 25 of all your spent fuel.

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 And if it has not been evaluated, should 2 it be?

3 MEMBER BLEY: Well, you included the wall, 4 right?

5 MR. GALYEAN: The wall that separates the 6 fuel pool from the bulk of the reactor pool.

7 MEMBER BLEY: Oh.

8 MEMBER MARCH-LEUBA: No, no, no, I'm 9 talking --

10 MEMBER BLEY: He's talking about the high 11 wall.

12 MEMBER BLEY: the one labeled "reactor 13 building wall"?

14 MEMBER MARCH-LEUBA: This one right here.

15 MEMBER BLEY: Oh, that wall.

16 MEMBER MARCH-LEUBA: This one. I mean, 17 the -- if -- I'm sure that's not the seismically 18 designed wall, it's probably just masonry.

19 And if you don't worry about the sign that 20 says seismic, you can send out lot of bricks on top of 21 the spent fuel. That's what my eyes are telling me.

22 MEMBER BLEY: Do you have masonry walls 23 around the inside of this here -- unreinforced masonry 24 walls?

25 MR. GALYEAN: I don't believe 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

115 1 MEMBER BLEY: That's what he said. I 2 wouldn't have thought and -- we used to have a lot of 3 them, but we got rid of most of them.

4 MR. GALYEAN: You know, that's the dry 5 dock area and I don't know how much detail there is on 6 the dry dock.

7 MEMBER MARCH-LEUBA: But under 8 sufficiently horizontal g-force, can part of the wall 9 fall on top of the spent fuel? That's my question.

10 CO-CHAIR DIMITRIJEVIC: Is your concern 11 about spent fuel pool?

12 MEMBER MARCH-LEUBA: Yes.

13 CO-CHAIR DIMITRIJEVIC: Okay. But that's 14 not -- spent fuel pool is not part of the PRA here --

15 or is your concern about impact of the total pool?

16 MEMBER MARCH-LEUBA: If you drop that wall 17 on top of the spent fuel pool --

18 CO-CHAIR DIMITRIJEVIC: Right.

19 MEMBER MARCH-LEUBA: -- it's going to be 20 bad.

21 CO-CHAIR DIMITRIJEVIC: It's going to be 22 bad, but this is -- I mean, the -- but we are not 23 doing -- you know, the spent fuel pool is --

24 MEMBER MARCH-LEUBA: I was just asking --

25 I'm sure it's not seismic category 1, but this 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

116 1 strong enough to be seismic category 1.

2 CO-CHAIR CORRADINI: Okay. So, since 3 we've raised it, let me make sure I understand.

4 So, was -- were all of these part of the 5 structural analysis, or not, that he's mentioning or 6 do you have to check on that?

7 MR. GALYEAN: The dry dock area was not 8 part of the structural analysis, if that's the 9 question.

10 CO-CHAIR CORRADINI: Okay. That's your 11 question.

12 MEMBER BLEY: You had a full Level 1 13 model, you had fragilities, you had some kind of 14 generic hazard curve and you almost had to do a 15 seismic PRA to come up with your HCLPF following this.

16 Why didn't you?

17 MR. GALYEAN: Well, our hazard curve, of 18 course, did not have frequencies on it, right. They 19 were just postulated ground motions.

20 MEMBER BLEY: So, it was just to set some 21 kind of bound on what you expect?

22 MR. GALYEAN: That's right. It was just 23 ground motion segregated into --

24 MEMBER BLEY: Okay.

25 MR. GALYEAN: -- those 14 --

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 MEMBER BLEY: So, you left that to the COL 2 people to deal with?

3 MR. GALYEAN: Right. Yeah. So, we did 4 not have --

5 MEMBER BLEY: Okay.

6 MR. GALYEAN: -- a full hazard curve.

7 MEMBER BLEY: Okay.

8 MR. GALYEAN: So, the fragility 9 calculations, you know, for the most part, you know, 10 we used bounding conservative values.

11 These are just some details of, like I 12 said, the fragility calculations that we thought that 13 one member who's not here might be interested in.

14 CO-CHAIR CORRADINI: He's with us in 15 spirit.

16 MR. GALYEAN: Okay. So, you know, just 17 the factors that went into the structural response, 18 you know, damping and modeling and its various factors 19 that account for not only the magnitude of the motion, 20 but the frequency of the ground vibration.

21 So, next slide, please.

22 MEMBER BLEY: When you did this, did you 23 accumulate all the component and structural -- no, I 24 don't think you did -- fragilities into an overall 25 fragility curve and then apply the hazard, or did 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

118 1 have a plant fragility curve and a structures --

2 separate structures fragilities and apply the hazard 3 to both of them and look at them independently?

4 MR. GALYEAN: No, it was -- we modeled the 5 fragilities individually, if that -- in the model.

6 MEMBER BLEY: Even the components?

7 MR. GALYEAN: That's right. Even the 8 components.

9 MEMBER BLEY: Okay. So, when you come up 10 with a HCLPF, it's the highest HCLPF for the set of 11 things you were looking at?

12 MR. GALYEAN: Right. I mean --

13 MEMBER BLEY: Because you have a different 14 one for every --

15 MR. GALYEAN: Yeah. People refer to it 16 as, like, a min/max approach.

17 MEMBER BLEY: Yeah. Okay.

18 MR. GALYEAN: You know, each action 19 sequence comprises multiple cut sets. And so, the 20 maximum HCLPF of a cut set -- of each cut set is 21 combined, and then the minimum of that set is the 22 sequence.

23 MEMBER BLEY: Okay. And although Pete's 24 not here, this HCLPF is -- I remember back when people 25 came up with it, but it's a figment of a 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

119 1 You don't believe the tail is that far 2 out. In fact, essentially, that place where there's 3 a HCLPF, nothing is going to fail.

4 It would be like a truck driving past, but 5 it lets you do this kind of bounding analysis to get 6 an idea of what's going on.

7 MR. GALYEAN: Yeah. So, again, more 8 details on the fragility calculations and what went 9 into the structural response. So, I'm not going to 10 get into those.

11 MEMBER BLEY: Okay.

12 MR. GALYEAN: Low-power shutdown, you 13 know, we did the full -- we started with the full --

14 a list of initiating events from the full-power 15 internal events PRA.

16 Obviously, there is unique features in the 17 NuScale design. There's no reduced inventory 18 configuration during refueling. Everything is done 19 underwater.

20 We looked at -- we did look at all the 21 external event hazards during low-power shutdown.

22 Obviously, the dropped module came out as the most 23 significant core damage frequency contributor.

24 I think, in my opinion, that's 25 attributable to the conservatism in the crane 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

120 1 The crane design is still being refined and so a lot 2 of the details on things like the control system are 3 not final.

4 We made a lot of conservative assumptions 5 with respect to how the crane is operated.

6 MEMBER BLEY: Dick, I don't want to put 7 you on the spot, but you've raised the issue before 8 about whether there are or should be some kind of 9 physical barriers to prevent a dropped cask from 10 hitting another one.

11 Do you want to pursue that at all here?

12 Because the PRA should have thought of that.

13 MEMBER SKILLMAN: I don't know that it's 14 constructed to do so. I've made my concerns known.

15 CO-CHAIR CORRADINI: But I do think they 16 want to -- I think where Dennis is coming from, which 17 I think would be beneficial, is at least to go through 18 your concern about minimizing interaction.

19 Isn't that -- I think that's where Dennis 20 was going.

21 MEMBER SKILLMAN: Couple different things 22 --

23 MEMBER BLEY: Exactly. And I thought Dick 24 can do it better than I can.

25 MEMBER SKILLMAN: Yeah. If I look 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

121 1 NUREG-0612, asset failures are about 15 percent. Now, 2 this is a dated document. I acknowledge that.

3 My concern is a wire rope failure. It 4 might have 12 or 15 sheathes and, under that 5 condition, the load intensity is relative minor, but 6 it just takes a single snap -- asset failure to drop.

7 I was admonished by one of your 8 colleagues, you know, this is not too different than 9 removing a reactor vessel head or a steam drum, and 10 that's not exactly accurate.

11 It's rare that you move 750 tons. You do 12 in construction, but you don't after live operation, 13 and you certainly don't over a live core. And here, 14 you're moving the whole core, so my concerns are 15 really maybe twofold or threefold.

16 The reactor building liner is one-quarter 17 of an inch throughout the entire reactor building; it 18 is seismical in structure; it's QA-1, for good reason, 19 it is the envelope; but it's not much thicker than the 20 inner liner of the wheel wells in your car, if it's a 21 plastic liner.

22 And that is going to see at Module No. 11-23 12 if there's -- I don't know if it's 1, 2, 3, 4, 5, 24 6, but the 2 bays that are closest to the refueling 25 pit will see 720 module passings in 60 years of a 762-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 ton machine using the 762-ton number.

2 My belief, based on operating experience, 3 is the likelihood of bumping or scuffing the floor is 4 fairly high. I think the floor ought to be armored on 5 your key load path.

6 Independent from that, I think that the 7 trolley and bridge coming out of the module bay should 8 have some barrier to ensure that, in the transit, for 9 any module, the module is so restricted it can only 10 fall in one direction, and that direction has been 11 fully analyzed, but it cannot hit adjacent modules.

12 It either has bumpers or has a chaffing 13 design, but something assures that the module cannot 14 fall in any direction that would bump into an adjacent 15 module.

16 So, those are the concerns I have. The 17 floor, the fragility of the floor. I'm told, hey, 18 look, quarter-of-an-inch, they got them all -- all 19 plants have quarter-of-an-inch. If you drop, all 20 you're doing is maybe puncturing the membrane, but the 21 concrete's really the load-bearing surface.

22 I understand that. Concrete is 20 feet 23 thick, but I've also been in plants where just a nick 24 in the floor drives the operating crew crazy trying to 25 find a leak.

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 And I honestly believe that this design 2 would be well-served by the ability for the NSSS 3 vendor to say, we consider a module drop, but it's 4 confined.

5 It can only fall in a certain geometry and 6 we know, for that failure, whether it's banging up the 7 decay heat removal system, whether it's chaffing the 8 walls, the 11 other operating modules are safe.

9 I would say that's an argument that I 10 think really makes sense.

11 MR. GALYEAN: Fair enough.

12 I would just clarify that there are -- the 13 crane design has 2 wire ropes. Each is 100 percent 14 capacity, okay.

15 So, a failure of a single wire rope would 16 not result in a dropped module. Again, there are two 17 wire ropes, you know.

18 Your points about nicks in the liner, I 19 think, are, you know, reasonable. I view that as an 20 asset protection issue that the operator would make a 21 decision on -- the owner/operator.

22 We think we've done a pretty comprehensive 23 job at identifying the potential for crane drop, for 24 crane failure and module drop and what the potential 25 impacts 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

124 1 Again, you know, the owner/operator 2 certainly has a prerogative to make changes as they 3 see fit for asset protection, and I think that's 4 reasonable.

5 MEMBER BLEY: But the way you did it, as 6 I understand it, is you came up with a frequency of 7 drop and then assumed you'd have core damage if that 8 happened.

9 MR. GALYEAN: That's correct.

10 MEMBER BLEY: That's the extent of the 11 analysis.

12 MR. GALYEAN: For the dropped module.

13 MEMBER BLEY: Okay.

14 MR. GALYEAN: We have done other -- we 15 have taken other looks at, you know, what the 16 potential scenario might entail --

17 MEMBER BLEY: Uh-huh.

18 MR. GALYEAN: -- okay, but I think --

19 MEMBER BLEY: Somewhere I saw qualitative 20 pictures of where it might hit and it could hit other 21 modules and --

22 MEMBER SKILLMAN: Yeah. We had a 23 presentation -- it's been three years, and I've got 24 those in here somewhere -- and it showed three 25 different geometries where a module could impact one 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 or two others and either fell or slid or -- I'm kind 2 of retaining that information even though it's a 3 couple years old.

4 MEMBER BLEY: Now, just for me, I know 5 there are other loads, but when you lift the module, 6 how high off of the bottom is it?

7 MR. GALYEAN: Well, in the initial lift --

8 MEMBER BLEY: Yes.

9 MR. GALYEAN: -- okay, it's lifting up out 10 of the operating bay.

11 MEMBER BLEY: So, it's all the way up.

12 MR. GALYEAN: And so -- and then it moves 13 out of the -- you know, out of the operating bay and 14 then is lowered to where it's just a foot off the 15 floor.

16 MEMBER BLEY: Just a foot.

17 MR. GALYEAN: And so, for the majority of 18 the travel it's moving basically a foot off of the 19 floor.

20 MEMBER BLEY: Did you do any analysis 21 given the maximum velocity of travel and being a foot 22 off the floor, what would happen to it?

23 I would think it would land upright and 24 sit there, but it might not. It sounds like something 25 easy to analyze.

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 MR. GALYEAN: We did consider that and --

2 MEMBER BLEY: Uh-huh.

3 MR. GALYEAN: -- frankly, we thought it 4 was more for the context -- in the context of the FSAR 5 in the design certification, frankly, we thought it 6 was more trouble than it was worth.

7 MEMBER BLEY: Seems like a "back of the 8 envelope" calculation, but maybe not.

9 MR. GALYEAN: Yeah. It's more complicated 10 than that.

11 CO-CHAIR CORRADINI: And we may not 12 believe they're back of the envelope.

13 MEMBER BLEY: We may not believe they're 14 computer code.

15 MEMBER BALLINGER: I think it's over 20 16 feet, right, the lift? Part of the --

17 MEMBER BLEY: Oh, I wasn't talking about 18 the initial lift.

19 MEMBER BALLINGER: Oh, okay.

20 MEMBER BLEY: From the initial lift, yeah, 21 lots of stuff can happen. But once -- before it 22 starts traveling --

23 MR. GALYEAN: Yeah. I guess I don't want 24 to get into too much detail on --

25 MEMBER SKILLMAN: This is the essence 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 my concern: The thought is it's moving so slowly in 2 its hoist distances, 12 or 14 inches, so what. And 3 I'm just not in "so what" for a machine this big.

4 It's 762 tons.

5 And it might be that one wire rope parts, 6 the crane clamps as it should, the load settles, but, 7 for whatever reason, it begins to dip and I'm imaging 8 an event that I probably shouldn't be imaging.

9 But this, to me, is the real life and you 10 get into these situations and you say, golly, if we 11 had just had a bumper here or something there, that 12 would not have happened.

13 The assumption is the module falls. If 14 there is a release, it's well underwater. It's under 15 55 feet of water.

16 There is no offsite dose -- I got that --

17 but I will tell you the actions to recover it are 18 going to be extraordinary.

19 And so, it just seems to me that if we 20 have only one module in the entire plant and that 21 module fails, I'm almost at "so what."

22 If I've got 11 modules operating at 160 23 megawatts thermal and I drop one because it's gone, 24 you know, it's 24-month fuel cycle, I'm taking it to 25 refueling, I've got at least one module problem 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

128 1 terms of failure and now I have 11 that are at power.

2 And it seems to me that that is an 3 operating environment. I don't know if it's 1, 2, 3, 4 4, 5, 6 and 6A, but that is definitely an operating 5 scheme that is different than any other plant, that 6 I'm aware of. I just --

7 CO-CHAIR DIMITRIJEVIC: Yes. So, we want 8 to concentrate on this because that is something very 9 specific to your design, these module moves and things 10 like that.

11 Especially because you have a very high 12 CDF, you know, those 10 to minus 7, which leads to 13 nowhere, you know, hang in there.

14 And there's obviously some, you know, the 15 administrative decision was decided that that's all 16 right. That CDF shouldn't have been counted because 17 releases is negligible.

18 I see in your shutdown assumptions in your 19 PRA in the Table 19.1-71, says that during the ABC 20 leave (phonetic) the module is kept below the height 21 that would damage the ultimate heat sink if dropped.

22 So, does that mean they assume that this 23 assumption is hundred percent true? I mean, is there 24 -- also, you have a lot of scenarios that leads to 25 load of 10 to minus 14.

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 Is there difference between 0 and 10 to 2 minus 14? So, why we don't count the CDF?

3 Is the safety goal say the core damage 4 frequency should be below 10 to minus 4 and, you know, 5 only for the vents which could lead to some potential 6 release.

7 So, my main concern is how -- what type of 8 failure modes do you consider for ultimate heat sink 9 in general?

10 What's your failure probability to 11 ultimate heat sink?

12 Is there any, or is zero?

13 MR. GALYEAN: There are a lot of questions 14 in there.

15 CO-CHAIR DIMITRIJEVIC: Yes.

16 MR. GALYEAN: First of all, you know, a 17 lot of these questions, again, in my mind, are more 18 asset protection questions.

19 And the -- of course the design 20 certification is focused on is the public health and 21 safety -- are the public reasonably protected.

22 We do report the core damage frequency 23 associated with module drop, you know, crane failure, 24 module drop.

25 So, we keep it separate so as not to hide 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 or obscure the risk insights that might be gained from 2 the different portions of the PRA, you know. We want 3 to make sure that we're not overlooking anything when 4 we look at risk insights.

5 As far as failure of the ultimate heat 6 sink, we have done a number of internal sensitivity 7 studies looking at, well, if the ultimate heat sink is 8 damaged and potentially it gets drained away, what is 9 the safety impact on the other modules. We have done 10 these studies.

11 In one case, we simply open up the hole in 12 the bottom of the reactor pool and drain out the water 13 and do a thermal hydraulic simulation using MELCOR to 14 see what the impact would be on the operating modules.

15 I mean, we have done these analyses to 16 satisfy our own -- to educate ourselves on what the 17 impact would be.

18 These are not part of the FSAR because, 19 again, there's no credible mechanism that we could 20 identify that would damage the ultimate heat sink in 21 this fashion.

22 CO-CHAIR CORRADINI: Could I just stop you 23 there to make sure I understand?

24 Your point is that if you did the seismic 25 margin analysis, which struck me as the way in 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

131 1 I could get damage, as long as I satisfy that, it's a 2 small enough probability.

3 Am I understanding it correctly?

4 MR. GALYEAN: No. Okay. Understand that 5 the plant is going to be built on an engineered site, 6 right. And so, the only -- how do I say this?

7 There is no credible mechanism for 8 creating a drain in the ultimate heat sink of 9 sufficient size that would threaten the integrity of 10 the -- of 12 modules.

11 CO-CHAIR CORRADINI: And it's not 12 determined by a seismic -- it's determined by what can 13 drain by some sort of assumption of various 14 sensitivity calculations?

15 MR. GALYEAN: That's right. Again, we 16 just postulate a hole in the bottom of the reactor 17 pool and drain the water out, and we look at what the 18 impact would be on the operating modules.

19 CO-CHAIR CORRADINI: Okay.

20 MR. GALYEAN: There's no credible 21 mechanism that would create a hole like that.

22 CO-CHAIR DIMITRIJEVIC: But if you have a 23 hole, then you have a huge impact on the module.

24 MR. GALYEAN: Again, the hole that I'm 25 talking about is one that would not result in core 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 damage.

2 MEMBER BLEY: As long as everything is 3 operating normally.

4 CO-CHAIR DIMITRIJEVIC: Yeah. As long as 5 you don't have these LOCAs on the heated module.

6 MEMBER BLEY: To my question --

7 MR. GALYEAN: Again, we're talking about 8 an -- again, there's no credible mechanism that would 9 get us there.

10 MEMBER BLEY: And Mike's question -- we're 11 calculating numbers like 10 to the minus 9, 10 to the 12 minus 14, these crazy little numbers.

13 There's an earthquake at most sites at 14 much higher frequency than that that would be totally 15 devastating.

16 I mean, they don't hit a cutoff and you 17 never get a bigger earthquake. Something at 10 to the 18 minus 5, 10 to the minus 6, 10 to the minus 7 can be 19 massive.

20 MR. GALYEAN: Fair enough.

21 MEMBER BLEY: And that's not --

22 MR. GALYEAN: But we're not doing a 23 seismic PRA here.

24 MEMBER BLEY: And that's not incredible --

25 MR. GALYEAN: We're not doing 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

133 1 MEMBER BLEY: -- which you are implying.

2 MR. GALYEAN: -- seismic PRA here, right.

3 MEMBER BLEY: Is that incredible?

4 MR. GALYEAN: We're doing seismic margins 5 assessment.

6 MEMBER BLEY: It's a lot higher frequency 7 than what we're calculating for.

8 MR. GALYEAN: Fair enough. I'm not going 9 to argue with that. I mean, who can predict the 10 results of a peak ground acceleration of 3g, for 11 example. No one.

12 I mean, we are talking about an earthquake 13 the size that has never been experienced, to anyone's 14 knowledge, in the history of the world, right.

15 I mean, you want to postulate events that 16 have never happened. I mean, that --

17 MEMBER MARCH-LEUBA: That's the definition 18 of 10 to the minus 5 and 10 to the minus 11.

19 MEMBER BLEY: That's right.

20 MEMBER MARCH-LEUBA: And one thing you can 21 postulate is there's a hole that you didn't even know 22 suddenly the core goes like this.

23 The right side of the pool drops a couple 24 of meters from the left side of the pool. That's more 25 likely that SVG (phonetic).

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 Again, probability to 10 to the minus 6, 2 10 to the minus 7, nothing to minus 11.

3 MEMBER BLEY: Go ahead. I just wanted to 4 note there are some -- I don't know what kind of hole 5 you get or anything like that, but there are big 6 earthquakes that are very unlikely, but not as 7 unlikely as some things we're actually calculating.

8 MR. GALYEAN: I'm sure there are lots of 9 things -- supervolcanoes, for example. Asteroid 10 impacts, you know, there's all kinds of things that 11 can happen, but I don't know that --

12 MEMBER BLEY: Meteor strikes are more 13 likely.

14 MR. GALYEAN: -- I don't know that 15 analyzing these things adds value to what we're trying 16 to accomplish here.

17 CO-CHAIR DIMITRIJEVIC: Well, okay, 18 because we now are mixing apples and oranges. We are 19 not going to talk about seismic event happening during 20 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> after the, you know, module drop or something 21 like that.

22 So, if you ever do that, your seismic PRA, 23 I mean, (unintelligible) although the results with the 24 small numbers and then we are in completely different 25 ball park.

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 How I started the question, is that you 2 have here assumption the module will never reach the 3 height which can damage the pool.

4 That means that you have some height in 5 mind?

6 MR. GALYEAN: That's right.

7 CO-CHAIR DIMITRIJEVIC: So, you have some 8 height where actually pool can be damaged.

9 MR. GALYEAN: That's right.

10 CO-CHAIR DIMITRIJEVIC: My question was 11 how.

12 MR. GALYEAN: We have a design study that 13 was performed that determined what the maximum height 14 of a lift would be that would damage the pool 15 integrity.

16 CO-CHAIR DIMITRIJEVIC: So, what does that 17 mean, "damage pool"?

18 Does that mean in making ultimate heat 19 sink --

20 MR. GALYEAN: No. It means causing a 21 crack in the concrete.

22 CO-CHAIR DIMITRIJEVIC: Well, does 23 ultimate heat sink not fail to perform its functions, 24 right; is what you are saying?

25 MR. GALYEAN: That's correct -- well, 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 again, in the context of the PRA, the -- you know, the 2 size of the holes -- or the draining of the pool that 3 we have looked at doesn't compare -- you know, is far 4 more catastrophic than what would happen if the pool 5 concrete was damaged.

6 I mean, we postulate just an open hole 7 that goes out into a vacuum. No back pressure, 8 anything, just open up a hole, a drain in the bottom 9 of the pool and just drain out the water.

10 CO-CHAIR DIMITRIJEVIC: So, in your mind, 11 it's okay to have here -- just because you claim the 12 latch release probability given this is zero, it's 13 okay to exclude these sort of event but then we have 14 a 10 to minus 14 for extent of flood we are going to 15 include that CDF.

16 MR. GALYEAN: I guess I don't -- you know, 17 I take exception to "excluded." I mean, we do report 18 it. We report --

19 CO-CHAIR DIMITRIJEVIC: Yea, but it's not 20 counted in CDF.

21 MR. GALYEAN: We do report core damage 22 frequency.

23 CO-CHAIR DIMITRIJEVIC: Yes, but your 24 total core damage frequency doesn't include this core 25 damage frequency.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

137 1 MR. GALYEAN: We don't report a total core 2 damage frequency anywhere. We just report core damage 3 frequency for each individual hazard.

4 CO-CHAIR DIMITRIJEVIC: Okay. So -- okay.

5 Let's go back.

6 You mean you don't report the core damage 7 frequency for low-power shutdown or --

8 MR. GALYEAN: Yes, we do. We report each 9 individual hazard --

10 MEMBER BLEY: But you don't aggregate 11 MR. GALYEAN: We don't aggregate.

12 MEMBER BLEY: We can aggregate.

13 MR. GALYEAN: You can aggregate.

14 CO-CHAIR DIMITRIJEVIC: Wait a second.

15 You have this really significant report where you use 16 your CDF to make argument about significant 17 determinations.

18 MR. GALYEAN: And we evaluate every hazard 19 individually in the context of that risk significance 20 determination.

21 And that's why we have some that show up 22 coming out of the low-power shutdown, we have some 23 coming out of the external hazard, you know, as risk 24 significant.

25 If a component or system or operator 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 action shows up as risk significant in any of those 2 evaluations, then it's added to the list of risk 3 significance.

4 CO-CHAIR CORRADINI: So, in my mind, one 5 could think of a table that has various categories of 6 internal events, low-power shutdown, module drop, 7 whatever, and then, on here, all the systems and where 8 they appear as risk significant or not.

9 MR. GALYEAN: That's correct.

10 CO-CHAIR CORRADINI: That's what you're 11 saying?

12 MR. GALYEAN: That's correct.

13 CO-CHAIR CORRADINI: Does such a table 14 exist?

15 MS. BRISTOL: Those are in the FSAR. So, 16 there's one for Level 1, there's one for --

17 CO-CHAIR CORRADINI: Oh, they're 18 individual, but no --

19 MS. BRISTOL: -- Level 2 and then there's 20 external events.

21 CO-CHAIR CORRADINI: Okay.

22 MS. BRISTOL: And so, all of those -- and 23 then at the end of the set of tables there's also a 24 table of all the different hazards, full power, low 25 power, and then --

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

139 1 CO-CHAIR DIMITRIJEVIC: But you have it 2 here some importance. We just look in the one slide, 3 the importances, right?

4 MS. BRISTOL: Right.

5 CO-CHAIR DIMITRIJEVIC: And that --

6 MS. BRISTOL: And we took those from all 7 of the various tables.

8 CO-CHAIR DIMITRIJEVIC: And so, the ladder 9 dropping event would be there, right?

10 MS. BRISTOL: So, the crane is on there.

11 So, the --

12 CO-CHAIR DIMITRIJEVIC: Okay. Crane is --

13 MS. BRISTOL: The reactor building crane, 14 sorry, RBC, under the other events is greater than, 15 you know -- so, that one is captured as --

16 (Simultaneous speaking) 17 MS. BRISTOL: Yeah, since it's evaluated 18 a little differently than just, say, a ECCS valve.

19 CO-CHAIR DIMITRIJEVIC: Okay. All right.

20 MS. BRISTOL: So, that is captured on 21 there.

22 CO-CHAIR DIMITRIJEVIC: All right.

23 Because I saw that in some tables that you presented 24 in our previous meetings, you said, well, logically 25 they are not applicable and we should maybe say NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

140 1 "negligible," like, for others, you know, things like 2 that instead of "not applicable," because it implies 3 is not possible, which is not correct, right?

4 So, okay. All right. So, I will -- with 5 what you said in mind. I have to go look back in the 6 sum of the conclusions because this initiating event 7 of that -- of the dropping module consist of the 8 multiple -- it probably has some human actions inside 9 and things like that, right?

10 MS. BRISTOL: The dropped module looks at, 11 yeah, numerous potential failures for the crane.

12 CO-CHAIR DIMITRIJEVIC: Right. But those 13 are not -- like, those human actions are not part of 14 the human actions importances and things like that?

15 MS. BRISTOL: That's correct. We looked 16 at the crane as somewhat of a supercomponent and any 17 potential failure of the crane we just designated the 18 crane as a risk-significant component candidate.

19 CO-CHAIR DIMITRIJEVIC: Okay.

20 MEMBER BLEY: Okay. And, actually, I 21 guess your last table, 19.1-80 --

22 MS. BRISTOL: Correct. That's the one --

23 MEMBER BLEY: -- it might not sum them up, 24 but it has --

25 MS. BRISTOL: Yes.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

141 1 MEMBER BLEY: -- all the contributions on 2 the one table. So, you can see them there.

3 MR. GALYEAN: Right.

4 MEMBER BLEY: I forgot what else I wanted 5 -- what I really wanted to ask you, so go ahead. I'll 6 remember in a minute.

7 MR. GALYEAN: Okay. So, the dropped 8 module, we assumed that if the crane failed and a 9 module went horizontal, okay -- well, part of the core 10 would uncover and that it would then go to core 11 damage.

12 And then we did -- we simply assumed that 13 the containment would fail by virtue of the impact 14 with the pool floor. And we did a -- evaluated the 15 radiological release that came out of it and it was 16 small. Okay.

17 MEMBER BLEY: You showed us that analysis 18 --

19 MR. GALYEAN: Right.

20 MEMBER BLEY: -- and I don't remember just 21 when --

22 MR. GALYEAN: This is the one exception to 23 the simplification we made in the full-power internal 24 events where we simply said if there was a bypass or 25 a LOCA outside containment, that automatically 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 resulted in a large release.

2 This is the one scenario -- one sequence 3 or scenario that we explicitly evaluated the release 4 and the potential for dose at the site boundary to 5 compare to our definition of a large release.

6 And in this particular case, it was much 7 smaller than a large release.

8 MEMBER MARCH-LEUBA: So, without going 9 into too much details, you're assuming that the module 10 drops --

11 MR. GALYEAN: Right.

12 MEMBER MARCH-LEUBA: -- and then, by some 13 miracle, a gas forms on half of it or 10 percent of 14 it, and then the fuel that is uncovered would then 15 melt.

16 MR. GALYEAN: Right.

17 MEMBER MARCH-LEUBA: Why does the gas --

18 because you're moving it completely filled.

19 MR. GALYEAN: No. The module is not 20 completely filled when --

21 MEMBER MARCH-LEUBA: Why not?

22 MR. GALYEAN: -- moved.

23 MEMBER MARCH-LEUBA: If it's your primary 24 -- your primary contributor to CDF, why not require to 25 move it full and cold? Why 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

143 1 MR. GALYEAN: Well, there's a couple of 2 reasons. One, is it adds weight to the module for the 3 crane -- that the crane has to lift.

4 And for another, there is equipment at the 5 top of the module inside the containment that the 6 designers don't want to be submerged.

7 MEMBER MARCH-LEUBA: If it tips, it's 8 going to be submerged.

9 MR. GALYEAN: Well, certainly, but that's 10 not the plan.

11 MEMBER BLEY: Bill, you chastised me about 12 asking you about earthquakes bigger than have been 13 recorded in the history of the world.

14 In your set of hazards, what's the biggest 15 one you have?

16 MR. GALYEAN: In the HCLPF evaluation, we 17 went up to 3.5.

18 MEMBER BLEY: 3.5g. You did go up to 3.5.

19 MR. GALYEAN: That's right.

20 MEMBER BLEY: Okay. Thank you.

21 DR. SCHULTZ: Bill, on this slide, a 22 radiological release evaluation stops with the third 23 bullet -- in other words, the fourth bullet -- where 24 you could potentially induce LOCAs or transients. You 25 didn't go further and say the consequence is there.

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 MR. GALYEAN: Yeah. We just looked at the 2 potential for a crane failure and a module drop to 3 impact an operating module.

4 DR. SCHULTZ: Right.

5 MR. GALYEAN: So, we then looked at the 6 likelihood -- or not the likelihood, but the 7 probability of inducing an upset condition in an 8 operating module by virtue of crane failure or module 9 drop.

10 DR. SCHULTZ: Right.

11 MR. GALYEAN: That then has a, you know, 12 potential for inducing an upset condition in an 13 operating module.

14 And then we took that likelihood and 15 compared it to the likelihood of those upset 16 conditions already modeled in the PRA, in the full-17 power internal events PRA, okay.

18 And those -- and the crane failure-induced 19 contribution was orders of magnitude lower than the 20 independent likelihood of those events already modeled 21 in the full-power internal events PRA.

22 CO-CHAIR CORRADINI: So, your point is 23 they're bounded?

24 MR. GALYEAN: That's right.

25 CO-CHAIR DIMITRIJEVIC: You consider 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

145 1 to be just LOCA or LOCA with failure of containment 2 insulation?

3 MR. GALYEAN: LOCA.

4 CO-CHAIR DIMITRIJEVIC: Just LOCA?

5 MR. GALYEAN: That's right.

6 CO-CHAIR DIMITRIJEVIC: So, there is no --

7 MR. GALYEAN: A LOCA outside containment.

8 CO-CHAIR DIMITRIJEVIC: LOCA outside of 9 containment, which is the same bypass of containment.

10 MR. GALYEAN: Right.

11 CO-CHAIR DIMITRIJEVIC: And so, since the 12 frequency of that is 1E minus seven, is much less than 13 your other LOCA outside the containment --

14 MR. GALYEAN: That's correct.

15 CO-CHAIR DIMITRIJEVIC: -- which was --

16 MR. GALYEAN: That's correct.

17 CO-CHAIR DIMITRIJEVIC: I forgot what it 18 was, but I can look in there. So, basically they 19 consider that what happen is LOCA outside containment.

20 And then that component is internal LOCA outside 21 containment.

22 MR. GALYEAN: Yeah.

23 CO-CHAIR CORRADINI: And later, we'll hear 24 about that from -- you had a grouping of these that 25 you analyzed. Later, meaning tomorrow.

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 MR. GALYEAN: Okay. That's right. Next 2 slide, please.

3 Now, as the -- now, as part of our multi-4 module evaluation, our multi-module evaluation 5 comprised two parts.

6 We did a qualitative evaluation where we 7 looked at all the systems in a NuScale plant and 8 basically did a hazards assessment.

9 If that system failed, what would be the 10 impact on the operating modules? And I said that's a 11 qualitative evaluation. We looked at that and 12 documented that.

13 We also did a quantitative evaluation. We 14 started with the results of the full-power internal 15 event assessment, PRA, and we -- and for each basic 16 event in the full-power internal events PRA model, 17 for each basic event, we applied what we called a 18 multi-module adjustment factor, okay, where we said, 19 if this event happens in one module, what's the 20 conditional probability that it could occur --

21 simultaneously appear in another module, in two or 22 more modules?

23 And that's what we refer to as these 24 multi-module adjustment factors. These are 25 conditional probabilities given that the failure 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 occurs in one module, what's the conditional 2 probability that it could also occur in another 3 module.

4 MEMBER BLEY: And these are judgment 5 based.

6 MR. GALYEAN: These are engineering -- but 7 they are very -- in my opinion, I mean, they are 8 extremely conservative in the sense that every single 9 basic event in the full-power initiating event model 10 is accounted for with at least a one percent 11 probability.

12 So, think of, for example, a pipe break.

13 Okay. Even if you have a -- or a LOCA inside 14 containment, you know, just a relief valve opening or, 15 you know, something, a CVCS pipe breaks inside 16 containment.

17 We say, given it occurred in the first 18 module, there's a one percent chance it could 19 simultaneously occur in a second module, okay, just to 20 see what the impact would be.

21 And so, using a minimum value of one 22 percent and a maximum value of a hundred percent, we 23 applied these adjustment factors to every basic event 24 in the PRA.

25 CO-CHAIR CORRADINI: Can you help me 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

148 1 the second one, the 0.1 to 0.3 one? I don't think I 2 understand where you -- how you assigned those.

3 MR. GALYEAN: These -- okay. These are 4 events that -- you can think of the one percent 5 conditional probability as applying to events you 6 would, on the surface of it, think of it as completely 7 independent. Okay.

8 And then there's another class of events 9 that you could look at and say they are completely 10 dependent.

11 For example, a loss of offsite power. It 12 affects all 12 modules, right? That would be a 13 hundred percent.

14 Then there's an intermediate class of 15 events, maybe operator actions, maybe thermal 16 hydraulic conditions that maybe there's a common cause 17 failure in one module.

18 And given that a common cause failure 19 occurs in one module, there's a chance that that same 20 common cause failure could occur in a second module.

21 And that's the purpose of these 22 intermediate values of 10 percent and 30 percent, 23 okay, is to account for these things that maybe have 24 some dependency, okay, but less than complete 25 dependency, but more than completely independent, 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 okay.

2 And so, there, we went and applied these 3 values to those basic events.

4 CO-CHAIR DIMITRIJEVIC: Well, we had this 5 discussion last time when we had -- and there was some 6 issue which I had with some of those things.

7 Let's discuss shared system. Give me some 8 good example of shared system like a --

9 MS. BRISTOL: Diesel.

10 MR. GALYEAN: Yeah. Instrument air, 11 service water, you know, electric power.

12 CO-CHAIR DIMITRIJEVIC: How about the 13 important one like this --

14 MEMBER SKILLMAN: Nuke service. Nuke 15 service is the important one.

16 CO-CHAIR DIMITRIJEVIC: Which one?

17 MEMBER SKILLMAN: Nuclear service cooling 18 water.

19 MR. GALYEAN: We call it circulating 20 water.

21 MEMBER SKILLMAN: Circulating water.

22 MR. GALYEAN: Circulating water, but --

23 CO-CHAIR DIMITRIJEVIC: But from the PRA 24 --

25 MS. BRISTOL: 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

150 1 CO-CHAIR DIMITRIJEVIC: -- but something 2 which has a high importance in the PRA, but it's -- is 3 this combustion tubing common?

4 MR. GALYEAN: Yes.

5 MS. BRISTOL: Yes.

6 CO-CHAIR DIMITRIJEVIC: What does that 7 mean?

8 There's only one?

9 MS. BRISTOL: Correct.

10 MR. GALYEAN: That's correct.

11 CO-CHAIR DIMITRIJEVIC: All right. So, 12 this is what I want to tell you is my problem. You 13 have, let's say, 10 units -- 12 or whatever -- and you 14 have one turbine, right? You have loss of offsite 15 power to all the units.

16 You can only use it for one, right?

17 MR. GALYEAN: I'm sorry?

18 CO-CHAIR DIMITRIJEVIC: You can only use 19 that to backup the power to one module.

20 MR. GALYEAN: No. No.

21 CO-CHAIR DIMITRIJEVIC: You can use it to 22 all modules?

23 MR. GALYEAN: Yes. It is sized to supply 24 the entire site.

25 CO-CHAIR DIMITRIJEVIC: Okay.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

151 1 MR. GALYEAN: And same with the diesel 2 generators.

3 CO-CHAIR DIMITRIJEVIC: Okay. So, you can 4 supply all the modules with that and you have operator 5 actions to do that, right?

6 Operator action to do this -- whether it 7 supplies one unit or supply 12 unit, it has to have 8 different probability, right?

9 MS. BRISTOL: It's to start the generator 10 indifferent of how many units that that generator is 11 supporting.

12 So, the action modeled in the PRA is to 13 start the CTG, combustion turbine generator.

14 CO-CHAIR DIMITRIJEVIC: And he doesn't 15 have anything to do module-specific just as the diesel 16 generator, the load, everything goes automatically on 17 the different modules?

18 MS. BRISTOL: That's how the design is 19 currently modeled.

20 CO-CHAIR DIMITRIJEVIC: Okay. Well, then 21 my concern is mostly about these human actions because 22 I think that human actions will be different in the 23 multi-modules that affect the various uses when the 24 single modules ---

25 MR. GALYEAN: Yeah. In this 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

152 1 context here on the multi-module evaluation, which is 2 not on this slide, is the human actions were 3 multiplied by a factor of ten to increase the failure 4 probability. A factor of ten.

5 CO-CHAIR DIMITRIJEVIC: For second unit, 6 because you actually calculate --

7 MR. GALYEAN: Well, for --

8 MS. BRISTOL: Any greater than 1. 2 to 9 12.

10 CO-CHAIR DIMITRIJEVIC: Okay. Well, 11 that's why we are calculating CDF based on the one and 12 not based on two. That's my question.

13 Why is -- because you are calculating two, 14 but you are multiplying human actions only for --

15 other than one, and you are using one as your base 16 CDF.

17 Why you are not using two as your base 18 CDF? You understand which I'm asking? Because you're 19 going to have one unit where the human action will be 20 40 minus 3. And you're going to have 11 units where 21 human action is going to be 80 minus 3, but you are 22 using 40 minus 3 as your base case.

23 MR. GALYEAN: We -- in essence, we assumed 24 complete dependence on human action. We said, you 25 know, if the human action is performed, it's performed 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 for everything or nothing. And we take that base 2 human error probability and multiply it by ten.

3 So, what you saw before, the 4 times 10 to 4 the minus 3, now becomes a 4 times 10 to the minus 2 5 and --

6 CO-CHAIR DIMITRIJEVIC: And I think you 7 should use that as base case. That's my main comment.

8 Because what's happening, you are --

9 MR. GALYEAN: In this context of the 10 multi-module evaluation, that is the base.

11 CO-CHAIR DIMITRIJEVIC: No, no, because 12 you are calculating probability to fail 2 or more out 13 of 12. That's what you are calculating.

14 MR. GALYEAN: That's right.

15 CO-CHAIR DIMITRIJEVIC: But you are not 16 taking in account where you're calculating 1 for your 17 base calculation there, you are calculating 1 of 12.

18 Not just 1, 1 of 12.

19 So, therefore, you should choose that 1 of 20 12 should be representing the worst situation for 21 things like that.

22 Let's say the human action -- this is why 23 I was trying to explain this. Let's say that you have 24 a -- it's much better to understand what I mean.

25 Let's say that you have something which can only 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 supply six units and not more than six units.

2 Therefore, you know, you can put it to one 3 for the second unit, but then for the first you are 4 crediting it, and that first thing shouldn't be your 5 base case.

6 Your base case should be the one which is 7 problematic.

8 MR. GALYEAN: In this particular -- in the 9 context of the multi-module evaluation, we did what 10 you suggest.

11 We took the first module human error 12 probability, multiplied it by ten, okay, and then 13 assumed all the others were probability of one, okay.

14 They were completely correlated, okay, so --

15 CO-CHAIR DIMITRIJEVIC: But you don't 16 apply that in your base case. We are discussing all 17 the time base case, yes.

18 MR. GALYEAN: In the multi-module 19 evaluation, that's -- it is the probability for the 20 first module, okay.

21 CO-CHAIR DIMITRIJEVIC: Yeah.

22 MR. GALYEAN: That probability is 23 multiplied by ten.

24 CO-CHAIR DIMITRIJEVIC: So, why don't you 25 use that as a base case? That's my question.

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 MR. GALYEAN: We talked about the human 2 reliability assessment and we picked the value that we 3 did, you know.

4 It was the limiting value of all the 5 values we calculated. I mean, we have a basis for why 6 we picked it. I don't know what else to say.

7 CO-CHAIR DIMITRIJEVIC: Okay. I will 8 think about that and present that in a letter so you 9 better understand.

10 What I have a problem with this multi-11 modules is when you talk multi-modules, you say 1 --

12 2 or more -- 2 or more, but you don't keep in account 13 that you're calculating 1 out of 12 as a base case, 14 even you're just doing one module.

15 That module cannot be done as the -- in 16 some -- the one thing which sits there independent of 17 everything else, right? That module is part of unity.

18 So, you are calculating 1 out of 12, 19 right, or you, in your mind, you're just doing 1?

20 That's a very different question.

21 Are you calculating 1 out of 12 or you are 22 just calculating 1? You understand the differences?

23 MR. GALYEAN: I think I do. But I think 24 there's still a misunderstanding going on about what 25 we do.

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 CO-CHAIR DIMITRIJEVIC: Okay. So, what I 2 want to say, let's say the situation, like the loss 3 of, loss of power that he has to do multiple actions 4 on the multiple modules. Obviously that's a much more 5 complex action then if you had one module which losses 6 also power, and he's only concentrate on that.

7 MR. GALYEAN: Agreed.

8 CO-CHAIR DIMITRIJEVIC: So, that 4E minus 9 3 represents that. That he's dealing with multiple 10 modules doing all of those steps.

11 You cannot represent --

12 MR. GALYEAN: That's not what we did.

13 CO-CHAIR DIMITRIJEVIC: I know. But why 14 not?

15 MR. GALYEAN: Because we -- because we 16 believe that the complication imposed by having a 17 multi-module event would disrupt the response to the 18 first module.

19 So we increased the failure probability of 20 the first module by a factor of 10. And then we 21 simply just --

22 CO-CHAIR DIMITRIJEVIC: To the 4E minus 3, 23 right?

24 MR. GALYEAN: That became 4E minus 2.

25 CO-CHAIR DIMITRIJEVIC: Two, 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

157 1 MR. GALYEAN: For the first module. And 2 then we assumed that all the related actions were 3 completely dependent.

4 That if he failed the first one, he would 5 fail them all.

6 CO-CHAIR DIMITRIJEVIC: Buy why don't you 7 do this in your basic study? Because in the basic 8 study that always happens.

9 There is no loss of offsite power where he 10 had to deal with only one module. He always deals 11 with multiples.

12 So why don't you use this 40E minus 2 in 13 your basic study?

14 MR. GALYEAN: Because in basic --

15 CO-CHAIR DIMITRIJEVIC: For those 16 situations where he has to deal with the multi-module 17 problem.

18 MR. GALYEAN: In the basic study we're 19 assuming there is a plant upset, how do I say, a 20 complication with only one module. Okay?

21 I don't know, yeah.

22 CO-CHAIR DIMITRIJEVIC: But that's a --

23 MEMBER BLEY: But then loss of offsite 24 power is not an appropriate initiator then. Because 25 that can't happen, unless you're only running one 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 module.

2 CO-CHAIR DIMITRIJEVIC: Or a million 3 others cannot happen.

4 MEMBER BLEY: Yeah.

5 CO-CHAIR DIMITRIJEVIC: Extended flood, 6 fire.

7 MEMBER BLEY: Yeah. All of those.

8 CO-CHAIR DIMITRIJEVIC: Yeah.

9 MR. GALYEAN: Well, obviously a loss of 10 offsite power by itself is not a concern. Right? I 11 mean, you have to have something else going on.

12 Other failures going on for it to show up 13 in as a safety hazard.

14 MEMBER BLEY: But it still means you're, 15 where you're normally really focused on one or just 16 the overview of all the modules, now you have a 17 reactor trip on one, but on 11 others as well.

18 So it puts your three operators in a 19 different mode then you're analyzing them for all the 20 really internal events that are separate that are 21 independent events, so.

22 MR. GALYEAN: I understand.

23 CO-CHAIR DIMITRIJEVIC: That's exactly.

24 Because you did your base case as you just have -- and 25 maybe that's okay, because Joy always brings up how NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

159 1 are you going to be loading modules?

2 Maybe somebody will just decide to have 3 the one module. But your base, this is just for 4 somebody who decides to just have one module. Other 5 then this drop.

6 MR. GALYEAN: Effectively, yeah. That's 7 true.

8 CO-CHAIR DIMITRIJEVIC: Yeah. Yeah.

9 MEMBER BLEY: Human actions have come up 10 here. So I'm going to ask a question I didn't ask 11 earlier.

12 Back when we did the human engineering 13 chapter, it -- or the referenced technical reports to 14 it, talked a lot about how we considered there, errors 15 of commission and errors or omission.

16 And it says over and over again we thought 17 about errors of commission. And then when you get to 18 the list of events they have, there aren't any.

19 Did you guys look at errors of commission?

20 MR. GALYEAN: We looked for them. The 21 only place that we actually --

22 MEMBER BLEY: I don't think in the report 23 you tell us much about how you look for them, do you?

24 I might have missed it.

25 MS. BRISTOL: They were looked at. And 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

160 1 was described in the actual supporting technical 2 theory document in Journal Three.

3 MEMBER BLEY: Okay. But not a technical 4 report that we have access to?

5 MS. BRISTOL: Correct.

6 MEMBER BLEY: So there's a lot of this 7 stuff we -- you got -- can we take a minute and do 8 that?

9 Tell us how you looked for them.

10 MS. BRISTOL: We looked at --

11 MEMBER BLEY: I know lots of ways to look 12 for them.

13 MS. BRISTOL: We looked at the various 14 events that operator -- the model PRA human actions.

15 And if they were performed at a different time, would 16 they put the plant in an upset condition?

17 For instance, if they operated CFDS at 18 power, where would the impact on the module be? Or 19 any of the actions that they were trained to perform, 20 if they performed them in an inappropriate time, would 21 that contribute to a plant upset?

22 And how would the module respond?

23 MEMBER BLEY: And you did this 24 qualitatively? You built a list and said, we're not 25 going to look at these because they don't affect 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

161 1 Are there errors of commission in the PRA?

2 MR. GALYEAN: Only in the way we modeled 3 the crane, the reactor building crane.

4 MEMBER BLEY: Okay. And kind of sort of 5 in the way you did this multi-module evaluation.

6 There's some effective allowance for them, I think.

7 When you throw these factors on the 8 errors, you could make that argument. Another way to 9 look at them is for each event tree, each sequence 10 model, it's a given I have this.

11 Are there any things people could do for 12 a variety of reasons that could change the likelihood 13 of any of my top events? Did you do something like 14 that?

15 MS. BRISTOL: Qualitatively that is what 16 we did.

17 MEMBER BLEY: Okay. That is what you did.

18 MS. BRISTOL: We looked at those, and --

19 yes.

20 MEMBER BLEY: Okay. And the kind of 21 things people can do, or as you said, they're the kind 22 of things they have in their procedures, but it's not 23 in there.

24 There's also the kind of things, and you 25 don't have a history here, but there will be some kind 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 of informal knowledge about how things work that 2 builds up over time and gives the operators a 3 rationale to do things that aren't called for.

4 Because they don't omit or commit. They respond to 5 the best of their ability.

6 I don't know. I got a feeling you didn't 7 really look for that kind of thing. What could lead 8 people to do things you weren't expecting them to do?

9 Turn something off or turn something on.

10 Or cross connect something. It happens. Anyway.

11 It's a thing that we'll look for some more.

12 CO-CHAIR DIMITRIJEVIC: Well, like a 13 controlling crane speed could be one of the good 14 examples, you know.

15 MEMBER BLEY: If it's done manually, or 16 can be done manually. Yeah.

17 CO-CHAIR DIMITRIJEVIC: Yeah. Yeah.

18 MEMBER BLEY: That was all. You gave me 19 a sense.

20 MS. BRISTOL: Overall, you know, we 21 discussed in sites the -- as we've continued to 22 discuss the NuScale design exceeds the core damage 23 frequency safety goal with a significant margin.

24 Being an internal event CDF of 3 minus 10.

25 CO-CHAIR CORRADINI: So, I know you want 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 to get to the end, and I'm sorry. But, I was 2 listening to Vesna trying to express her worry. And 3 your answer to her relative to these ones where it 4 connects across it.

5 If you were to do what she's suggesting in 6 terms of re-base lining a human action for one, as a 7 community of 12, how much would that change that?

8 Or is your sensitivity where you took in 9 multiplying by a factor of 10 bounding that? Am I 10 making sense?

11 MR GALYEAN: Somewhat.

12 CO-CHAIR CORRADINI: I'm trying to get a 13 feeling for what her worry is. Because what I heard 14 from her question and your answer was, she would have 15 done it a different way.

16 And your answer is, well, you know, we 17 looked at it, and I'll misrepresent, as one 18 individually. But when they were connected for multi-19 module, you did effectively do kind of a worst case 20 sensitivity.

21 Does that worst case sensitivity bound 22 these insights?

23 MR. GALYEAN: Okay. I --

24 CO-CHAIR CORRADINI: I was kind of 25 understanding --

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 MR. GALYEAN: I just first, I want to try 2 and make clear what we're talking about. And remember 3 for the single module, full power internal events' 4 PRA, --

5 CO-CHAIR CORRADINI: Yeah.

6 MR. GALYEAN: Okay. We did do a 7 sensitivity study where we set all human actions to 8 always fail. Okay.

9 CO-CHAIR CORRADINI: that was in the 10 summary we got back a year ago.

11 MR. GALYEAN: And earlier just this 12 morning.

13 CO-CHAIR CORRADINI: Right. And early.

14 Right. Okay.

15 MR. GALYEAN: I mean, we did that 16 sensitivity study, and we showed you the results in 17 the slides.

18 MEMBER BLEY: But that's kind of an 19 extreme application.

20 CO-CHAIR CORRADINI: Of her question.

21 CO-CHAIR DIMITRIJEVIC: Yeah. 3.3 minus 22 8, I think was. Is there any shared system which 23 doesn't have a capacity to supply all modules?

24 MEMBER MARCH-LEUBA: Yes. Boron addition 25 system. And you can only boron one module while 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 supplying a little bit to the others.

2 I'm sure there are more. But this is non-3 safety grade.

4 MS. BRISTOL: The containment flooding and 5 drain system, there's two subsystems. One operates 6 six modules.

7 There's two systems. One --

8 CO-CHAIR DIMITRIJEVIC: And each operates 9 each one?

10 MS. BRISTOL: Correct. Central.

11 CO-CHAIR DIMITRIJEVIC: Yeah. Because 12 that will also, we will see in the impact of this one 13 that is 12. You know, also in this shared system.

14 Because we cannot have an initiator which 15 creates the, you know, you -- we have a new initiator 16 which will call for the, you know boration. But 17 that's not common to the unit, yeah?

18 MEMBER MARCH-LEUBA: I don't think it's 19 even is included in the PRA at all.

20 CO-CHAIR DIMITRIJEVIC: Not for the action 21 even?

22 MEMBER MARCH-LEUBA: I don't know. You 23 tell me, is BAS on the PRA?

24 MS. BRISTOL: Boron addition is not 25 modeled explicitly in the PRA at all.

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 MEMBER MARCH-LEUBA: Yeah. Because it's 2 not needed for any event.

3 MR. GALYEAN: Right.

4 CO-CHAIR DIMITRIJEVIC: Oh, okay. All 5 right.

6 MS. BRISTOL: And apparently Table 19.1, 7 Tech 76 shows all the shared system broke down by 8 common modules each time.

9 CO-CHAIR DIMITRIJEVIC: In the 76?

10 MS. BRISTOL: In the 76.

11 CO-CHAIR DIMITRIJEVIC: Okay.

12 CO-CHAIR CORRADINI: Page 278.

13 MEMBER REMPE: So I didn't hear the final 14 part of your answer, Bill, to Mike's question about 15 how important would this be. I know you said you did 16 the sensitivity for a single unit where you assumed 17 all the human actions failed.

18 But, what's the bottom line to the whole 19 question for the multi-modules?

20 MS. BRISTOL: Well, again we did not do 21 sensitivity studies on the multi-module 22 quantification. Okay.

23 Again, what we did in the multi-module 24 quantification was, we took the base case human error 25 probability and multiplied it by a factor of 10 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

167 1 account for the complication of the operators having 2 to deal with multiple modules.

3 I mean, that was -- that was the multi-4 module adjustment factor that we did. And so I don't 5 --

6 CO-CHAIR CORRADINI: Let me ask you, I 7 think I understand in terms of bounded by 8 sensitivities calculation. The difference toward a 9 different distinction typically --

10 MR. GALYEAN: Right.

11 CO-CHAIR CORRADINI: Sorry, that you did 12 it in a different manner to bound it.

13 CO-CHAIR DIMITRIJEVIC: But they have to 14 reach agreement with the reviewers. In the beginning 15 when you started these that that reviewer will do it 16 that way, right?

17 That you were just looking in one module 18 like it's independent from that.

19 MR. GALYEAN: That's right.

20 CO-CHAIR DIMITRIJEVIC: And that's 21 acceptable.

22 MEMBER BLEY: And the thing that Vesna 23 brought up, it's very legitimate. It is four these 24 multi-unit initiating events of which there is a 25 handful, the way it's modeled is not the way it's NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

168 1 carried out.

2 Because you don't have a single unit 3 operation. Yeah, you always have more than one here.

4 And how big an effect that is, that's something that 5 has to be thought about carefully.

6 But it shouldn't be any worse than the 7 sensitivity study they ran.

8 CO-CHAIR CORRADINI: That's what I was 9 saying.

10 CO-CHAIR DIMITRIJEVIC: When it comes to 11 human error.

12 MEMBER BLEY: Right.

13 MS. BRISTOL: And so the other various 14 analysis in the associated core damage frequencies is 15 we discussed the low power and shut down was dominated 16 by the conservatively performed module drop event.

17 We did a focused PRA where we only 18 credited the safety-related components in the PRA to 19 support D-RAP. And that was below the threshold for 20 those criteria.

21 And the multi-module CDF factor, as we 22 discussed, was .13. And so that was -- while we 23 didn't do, you know, a full multi-module, that was the 24 impact of applying those multi-module adjustment 25 factors to the core damage frequency.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

169 1 MEMBER BLEY: So, lest we get 2 overconfident here, we haven't done a level three PRA, 3 which is where the multi-module effects would show up.

4 And you don't know if this is a bound on --

5 MR. GALYEAN: Okay.

6 MEMBER BLEY: That or not, until it gets 7 looked at. And it will have to be looked at, at some 8 point.

9 MR. GALYEAN: What will?

10 MEMBER BLEY: Level three PRA.

11 MR. GALYEAN: Well, I mean, we have done 12 --

13 MEMBER BLEY: Before fuel load.

14 MR. GALYEAN: We have done --

15 MEMBER BLEY: Not for you guys, before 16 fuel load.

17 MR. GALYEAN: We have done dose 18 calculations for the site boundary.

19 MEMBER BLEY: Okay.

20 MR. GALYEAN: Okay. And for example the 21 dropped module, you know, that was a dose calculation.

22 MEMBER BLEY: Um-hum.

23 MR. GALYEAN: We've also done dose 24 calculations to support the environmental report.

25 Okay.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

170 1 And you know, all I can say is that even 2 if you took those dose calculations and multiplied 3 them by a factor of 20, we're still way below the 4 large release definition.

5 MS. BRISTOL: From the level 2 6 perspective, we looked at the large release frequency 7 from the various hazards. And we were well below the 8 safety goal of the large release frequency.

9 As we've discussed, module drop is blow --

10 is underwater. Didn't impact the large release 11 frequency.

12 And we did the focused PRA for LRF as 13 well. And was below the criteria for that for RTNSS 14 purposes. And evaluated the multi-module factor.

15 As requested, we tried to include these 16 insights at a high level from the FSAR. And more so 17 if there was anything of interest or to discuss.

18 But, you know, our design is a very 19 passively safe design. And there are a lot of 20 components and elements of that design that 21 contributes to these low values that we're seeing for 22 core damage frequency and large release frequency.

23 And that give us confidence that the 24 numbers, you know, while they are values, we do have 25 supporting design features that support those low 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 values.

2 For instance, you know, as we've mentioned 3 previously, failure to scram events don't directly 4 lead to core damage. You know, they progress similar 5 to trip events.

6 MEMBER MARCH-LEUBA: And do you have a 7 calculation to support that statement?

8 MS. BRISTOL: We do, yes.

9 MEMBER MARCH-LEUBA: Has anybody seen it?

10 Have you seen it? Because I've been asking for it for 11 a long time.

12 And the first time I'm seeing something 13 from the staff, which I'll drill this afternoon, which 14 I got this morning.

15 MS. BRISTOL: Okay.

16 MEMBER MARCH-LEUBA: So --

17 MS. BRISTOL: So we can add that.

18 MEMBER MARCH-LEUBA: I find it hard to 19 believe that that was a bounding calculation.

20 MS. BRISTOL: Understood.

21 MEMBER MARCH-LEUBA: And stay tuned for 22 this afternoon.

23 MS. BRISTOL: Okay. And so as we note, 24 the cycling reactor safety valve provides enough 25 inventory to provide that coolant flow path from 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

172 1 RPV to the CNV that cools the core adequately.

2 As Bill mentioned, these cores are much 3 smaller then current industry cores. And so having 4 that water available for heat transfer goes a low way 5 in our success criteria runs that we have performed.

6 Our safety systems are fail safe.

7 CO-CHAIR CORRADINI: So, just a 8 clarification. But if I cycle and freeze open, I turn 9 into essentially in containment LOCA.

10 MS. BRISTOL: Correct. And we analyze 11 that as well.

12 CO-CHAIR CORRADINI: Okay. That's what I 13 want to make sure.

14 MS. BRISTOL: Yes. The safety systems 15 passively fail safe. The ECCS functions to preserve 16 that inventory within the containment, and allows the 17 core cooling without additional inventory.

18 We then in PRA look at the beyond design 19 basis where we need that inventory. But from just the 20 containment being isolated, ECCS functioning, there's 21 no additional inventory needed.

22 We have talked about the lack of need for 23 power or operator actions because these -- the 24 containment isolations fail closed. The ECCS valves 25 fail open. The HRS actuates. All without electrical 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 power and operator action.

2 MEMBER BLEY: I have a question. And we 3 haven't gotten to six. But I'm not sure if I remember 4 seeing it in there.

5 The ECCS valves, not the reliefs, they are 6 fail open, spring operated. I assume it's spring 7 operated. Are they held shut by air? Or some other 8 fluid or mechanism?

9 Are they design?

10 MR. GALYEAN: I don't know how much detail 11 we want to get into the design of the valves.

12 MEMBER BLEY: Then let's do it later.

13 Because I really want to ask you about something 14 there.

15 CO-CHAIR CORRADINI: Okay. We'll hold 16 this for a closed session.

17 MS. BRISTOL: And then we go into the C-18 well items. Just overall completeness of the PRA as 19 we've discussed, you know, there are a lot of open 20 items and assumptions that are in the PRA.

21 And we acknowledge that at the design 22 phase. There's a lot of information that isn't 23 available. There's testing. There's operation 24 experience. There's walk downs. All of those things 25 that we can't do at this stage.

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 We've attempted to capture its 2 assumptions. And we have items so that we will go 3 back and address those items at the next phase, being 4 a COL phase.

5 And so we have a lot of reminders to go 6 back and check various critical assumptions that we 7 make in the PRA.

8 MEMBER BALLINGER: I have sort of a 9 metallurgical question, I guess. And that is, are 10 some of these items that you've identified as needing 11 to be done later on, are they -- do they have the 12 possibility of being a very big ticket item which 13 would change your results?

14 MS. BRISTOL: There's always that 15 potential if there's something. But as you've seen in 16 item eight, you know, the applicant, the COL 17 applicant, you know, will confirm the validity of the 18 key assumptions in the data use.

19 You know, and so as Bill mentioned, if 20 there's testing, or if there's additional information 21 that gets applied to the PRA, as we've discussed, we 22 believe our design, you know, the way it's 23 functioning, there aren't.

24 MEMBER BALLINGER: But somebody must have 25 had the discussion around the table at Starbucks, 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 which one of these, if they happened, would screw us 2 up big time?

3 MEMBER BLEY: Well, they had that in those 4 numbers they showed us on the Risk Achievement Worth.

5 MEMBER BALLINGER: Yeah. Yeah. But those 6 are --

7 MEMBER BLEY: But you may have others 8 you've talked about.

9 MEMBER BALLINGER: That's what I mean.

10 MS. BRISTOL: Well the ECCS valve is an 11 example. You know, we -- they're important and we'll 12 evaluate them again you know, if they're -- when 13 additional information becomes available.

14 But they'll still be important. And 15 they'll still be safety related.

16 MEMBER BALLINGER: But I would think that 17 those are not, even those are not that bad, because 18 you would then redesign it. Presumably.

19 Are there things that you can't design 20 yourself out of on this list that would cause real 21 problems? I'm assuming you can make a valve work.

22 MS. BRISTOL: Nothing that we can think 23 of, --

24 MEMBER BALLINGER: Okay.

25 MS. BRISTOL: Or that we've evaluated 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

176 1 date. So we don't.

2 (Off mic comment) 3 MEMBER BALLINGER: Huh?

4 MEMBER BLEY: I said in principle Ron, you 5 could make a plan.

6 CO-CHAIR DIMITRIJEVIC: Well, to Ron's 7 discussion, I would like to say not with the question 8 that you guys are meeting safety goals. And whatever 9 we question, and whatever uncertainty means produced, 10 you're not going to go above the 10 to minus 4 or 10 11 to minus 6, so.

12 The reason we ask so many questions is 13 because when you're coming with claiming that 14 basically you have a zero risk, and that not too many 15 things are important, like our job is to make sure 16 that not too many things are important.

17 Because if we -- if some of those 18 uncertainties and sensitivities bring something else 19 that's important, you know, there is the program, the 20 wrap, things like that, that that should be check 21 procedure.

22 You know, you will hear, we are here to 23 identify is there vulnerabilities. Is there something 24 which is essential to keep you where you say you are?

25 And it also the difference is, are you 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

177 1 10 to minus 7 or are you a 10 to minus 9 is also 2 different in concluding what is important and things 3 like that.

4 So, it's not every, you know, that's why 5 we have to clean all of these to make sure. I mean, 6 for me that means that I have to feel comfortable that 7 we did not miss something that, you know, which is 8 like, you know, the crane movement which should have 9 a procedures check or some, you know, just something 10 with testing.

11 I notice that you wrote out that you 12 assume the test is staggered. Well, but then I 13 concluded that you assume everything is tested during 14 the refueling.

15 So, I mean, I don't know what does that 16 assumption mean? I find a lot of things which makes 17 me think, you know, about things like that.

18 So, this is why I feel that you're 19 dreaming. We just want to feel secure ourselves.

20 MS. BRISTOL: With respect to the testing, 21 the -- we say non-staggered to be conservative.

22 CO-CHAIR DIMITRIJEVIC: Yeah.

23 MS. BRISTOL: And only for instance, DHRS 24 and ECCS. The other things can be tested. We don't 25 assume that all testing is done at refueling outages.

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 Only those items for PCCS.

2 CO-CHAIR DIMITRIJEVIC: For the CVCS and 3 the --

4 MS. BRISTOL: Yeah. Those can all be 5 tested. And we have testing maintenance events for 6 CVCS, demineralized water.

7 CO-CHAIR DIMITRIJEVIC: Well, CVCS is 8 operating systems. So, I mean --

9 MS. BRISTOL: Right.

10 MEMBER BALLINGER: But to put things in 11 perspective, these numbers in the words of that great 12 Greek philosopher George Apostolakis, this cost is two 13 times ten to the ninth year's old.

14 CO-CHAIR DIMITRIJEVIC: Yeah. I know. We 15 will have to have dinosaurs building a million of 16 those reactors to get there I feel.

17 Well, then one of the things is also, in 18 your review that George Apostolakis panel said that 19 this is interesting first step in multi-module. But 20 they say first step.

21 So, we have to be a little more. You 22 cannot just stay on this first step. We have to 23 identify other issues associated with multi-module 24 models to make a little dent in this, you know, multi-25 module system.

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 MEMBER REMPE: Could you put that previous 2 slide back up where it had about the key assumptions?

3 So that COL item is beyond just Section 19.1. Right?

4 Or is it just key assumptions as 5 identified in Section 19.1?

6 MS. BRISTOL: That is associated with the 7 PRA key assumptions that we make.

8 MEMBER REMPE: And just the PRA 19.1. So, 9 if there are other key assumptions that you can't just 10 identify using risk achievement work, I thought that 11 that COL item actually pertained to not just 12 frequency, but also consequence evaluations too.

13 MS. BRISTOL: That's correct. All of --

14 there's tables and tables in the FSAR. I don't --

15 MEMBER REMPE: Right.

16 MS. BRISTOL: Know that they were back.

17 Yeah, all of those key assumptions.

18 MEMBER REMPE: Okay. So then how do you 19 decide what's key and not key? Is it some -- is there 20 some sort of -- to kind of ask what Ron's asking in a 21 different way, is there some sort of a process that 22 tells you that the key assumptions are those that are 23 needed to provide reasonable assurance for adequate 24 protection of safety and health of the public?

25 And how do you complete that assump --

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 that evaluation to come up with the key assumptions?

2 MR. GALYEAN: It might be more appropriate 3 to say identified assumptions.

4 MS. BRISTOL: Yeah. Assumptions. Yeah.

5 MR. GALYEAN: You know, rather than key.

6 MEMBER REMPE: So then all assumptions.

7 MS. BRISTOL: Yes.

8 MR. GALYEAN: All assumptions. That's 9 right. There's no particular significance attached to 10 the work key other then these are the assumptions that 11 we have explicitly identified --

12 MEMBER REMPE: Okay.

13 MR. GALYEAN: In the -- for the PRA.

14 MEMBER REMPE: Yeah. Thanks.

15 MEMBER SKILLMAN: Bill, I'd like to go 16 back to a term that you used an hour ago, maybe two.

17 How do you make the distinction between what is asset 18 protection, and what is really operating experience 19 proven safety protection?

20 MR. GALYEAN: Well, obviously the 21 objective of PRA and Section 19 is to establish that 22 the public health is protected here. Right?

23 And in order to achieve that, the NRC 24 staff has identified these safety goals. And so the 25 objective of the PRA is to show that we conform 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

181 1 expectation in the context of the safety goals for 2 core damage frequency and large release frequency.

3 I don't know if that, you know, answers 4 your question. If there's something there that 5 affects the calculation of core damage frequency or 6 large release frequency, then it's more than just 7 asset protection. Okay.

8 But if it does not affect the core damage 9 frequency or the large release frequency, then that's 10 what I classify as asset protection.

11 MEMBER SKILLMAN: Yeah. I guess I would 12 just respond that brings to my mind the question of 13 vigor and thoroughness in the model.

14 And if the review teams have genuinely 15 used a thick magnifying glass and concluded this model 16 represents a credible set of assumptions. And I can 17 concur with your theorem.

18 Okay. Thank you.

19 MEMBER MARCH-LEUBA: So, less then ten 20 minutes to go of course. What tool did NuScale use to 21 calculate this obvious transient? You've seen the 22 calculation, right?

23 Was it RELAP 5?

24 MS. BRISTOL: Yes.

25 MEMBER MARCH-LEUBA: With point kinetics?

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 MR. GALYEAN: Point kinetics and nodule 2 kinetics as well.

3 MEMBER MARCH-LEUBA: With nodule kinetics 4 or with point?

5 MR. GALYEAN: Both.

6 MS. BRISTOL: Both.

7 MEMBER MARCH-LEUBA: With both. And both 8 survived?

9 MR. GALYEAN: Yeah.

10 MEMBER MARCH-LEUBA: Using minus 5 PCM for 11 a high? Which is the --

12 MR. GALYEAN: I'm not sure of that.

13 CO-CHAIR CORRADINI: Before we get into 14 this, if we can take -- we're going to take it up in 15 the closed session. Right?

16 MR. GALYEAN: Sure.

17 CO-CHAIR CORRADINI: So, my suggestion is, 18 let's take it up in closed session.

19 MR. GALYEAN: Okay.

20 CO-CHAIR CORRADINI: Because he's going to 21 start with one question. And there will be ten more 22 coming.

23 And I just have a funny feeling we're 24 going to get into things that are more precise and 25 proprietary.

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 MS. BRISTOL: Okay.

2 MR. GALYEAN: But just briefly, I mean we 3 have a long history of looking at ATWS. I mean, when 4 we first started looking at ATWS we used MELCOR using 5 our developed model, which is a point, you know, uses 6 a point kinetics model, I think, for the core.

7 MEMBER MARCH-LEUBA: But let's talk within 8 the closed session.

9 MR. GALYEAN: Okay.

10 MEMBER MARCH-LEUBA: So we can talk 11 numbers.

12 MR. GALYEAN: Okay.

13 MS. NORRIS: There were two questions 14 brought up earlier that we just got answers to. I'd 15 like to bring them up again.

16 CO-CHAIR CORRADINI: Good.

17 MS. NORRIS: So the first one was on the 18 bioshield redesign affecting the PRA hydrogen 19 analysis. So we did confirm that they will be 20 minimizing the hydrogen concentration underneath the 21 bioshield.

22 And this is discussed in our RAI response.

23 MEMBER REMPE: So again, my question was, 24 yeah, I know that there is something that it will 25 affect other phenomena. But what I want to know 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

184 1 will it affect the structural evaluation you did?

2 Or are you going to reevaluate the 3 structural integrity of it? Because you've made a 4 change in the design.

5 And again, you don't have to get into the 6 details of the design. But, just a definition.

7 MS. BRISTOL: Right. In our initial 8 analysis we provided that change in design does not 9 impact our previous analysis.

10 MEMBER REMPE: That's what I wanted to 11 hear.

12 MS. BRISTOL: Yes.

13 MEMBER REMPE: Thank you.

14 MS. BRISTOL: You're welcome.

15 MS. NORRIS: And the second question was 16 on the SMA analyzing of the different components. So, 17 we did analyze the control rods. They were shown to 18 not lead to core failure without additional random 19 other failures.

20 And also specifically the steel piping 21 with CVCS was actually not included, due to steel 22 piping ductility assumptions. So, you asked about the 23 control rods and the --

24 MEMBER MARCH-LEUBA: Yeah. So the one 25 thing is, the control rods may fail, but even if 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

185 1 are failing, it doesn't make any difference.

2 MS. BRISTOL: Correct.

3 MEMBER MARCH-LEUBA: Is that what I heard 4 you say?

5 MS. BRISTOL: Any additional failures.

6 MEMBER MARCH-LEUBA: Yeah. It turns into 7 an otherwise. And you're telling me it's okay. Which 8 we'll talk later.

9 MS. BRISTOL: Okay.

10 MEMBER MARCH-LEUBA: The CVCS pipe, you 11 say is ductile and it won't fail you. And if it -- if 12 you have high rates?

13 MS. NORRIS: Yes. Yes, and we do have 14 RSME on the phone for that if you'd like more details 15 on it, I believe.

16 MEMBER MARCH-LEUBA: No, I'm not -- the 17 guy that knows how to do this operation is not here.

18 MS. NORRIS: Right.

19 MEMBER MARCH-LEUBA: So, I (whistle).

20 MS. NORRIS: But yes, that is the 21 assumption anyhow.

22 MEMBER MARCH-LEUBA: I'm looking at 23 breaks, but.

24 CO-CHAIR CORRADINI: Any other questions 25 from the members?

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 MS. NORRIS: I believe Tom has one.

2 CO-CHAIR CORRADINI: Oh, Tom?

3 MR. BERGMAN: Tom Bergman --

4 CO-CHAIR CORRADINI: I think you're going 5 to have to get closer. Or tap it. Is it working?

6 No.

7 MR. BERGMAN: Now is it working?

8 CO-CHAIR CORRADINI: Okay. Now it is.

9 MR. BERGMAN: Tom Bergman with NuScale.

10 I did get the answer. I was out of date on our status 11 of our proprietary. The use of radar technology is 12 not proprietary.

13 So, if you had questions on that, that 14 didn't get answered, you can ask them now in the 15 public session.

16 MEMBER REMPE: Well, I had asked earlier 17 even, without worrying about what the mysterious 18 sensor is, a question about the high pressurizer level 19 setpoint that we could generate by itself according to 20 the text on 19.1-97.

21 And it does not say there's a back up 22 sensor that would give you that isolation signal for 23 the CVCS. And again, you don't have to answer it now.

24 But I just -- that was a question.

25 MS. BRISTOL: Understood. And I think 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 what was discussed earlier, is while there are 2 numerous sensors for all of -- for the various 3 setpoints in the safety analysis.

4 And it's analytical and it's in the PRA we 5 just model one. And so while we only discussed one of 6 them, there are various different setpoints and 7 sensors that would trigger, you know, a containment 8 isolation, a reactor trip, so addresses.

9 MEMBER REMPE: So, it's Section 19 in the 10 DCA. Every other place I looked at it, when it talks 11 about the DHRS, it talks about the backup sensor 12 signals.

13 Give me an idea of what is the back up for 14 these CVCS isolation signals. And you don't have to 15 do it today. But just sometime let know.

16 MS. BRISTOL: Okay.

17 MEMBER REMPE: Thank you.

18 CO-CHAIR CORRADINI: Other questions by 19 the members?

20 (No response) 21 CO-CHAIR CORRADINI: Okay. Let's take a 22 break for lunch. We'll be back at 1:30.

23 (Whereupon, the above-entitled matter went 24 off the record at 12:24 p.m. and resumed at 1:27 p.m.)

25 CO-CHAIR CORRADINI: Okay, we'll come back 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 into session and we're now going to hear from the 2 staff. Greg, are you the one that's going to lead us 3 off, or is it Alissa?

4 MR. CRANSTON: I'm just going to -- good 5 morning. I'm Greg Cranston. I'm the lead project 6 manager for the NuScale project, and I'm here on 7 behalf of Rani Franovich, who's the chapter PM for 8 this particular chapter.

9 A presentation on Chapter 19 will occur 10 over the next two days, with today's discussion 11 focusing on 19.1 PRA. And I just want to note that 12 Section 19, formulated to loss of large area due to --

13 with the plant due to explosion and fires, is part of 14 Chapter 20, which will be presented at a later date.

15 So, with that, I'd like to turn over to 16 our initial presenter, Alissa, for this presentation.

17 CO-CHAIR CORRADINI: Okay, good.

18 MS. NEUHAUSEN: Good afternoon. My name 19 is Alissa Neuhausen. I'm a technical reviewer in the 20 PRA and Severe Accident --

21 CO-CHAIR CORRADINI: And bring it close.

22 You're --

23 MS. NEUHAUSEN: This one's really far.

24 Okay, I'll just start over. My name is Alissa 25 Neuhausen. I'm a technical reviewer in the PRA 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

189 1 Severe Accidents Branch.

2 This morning we're going to talk about the 3 full scope of the PRA. We'll start with that, our 4 Internal Events Level 1. We'll touch on Level 2, the 5 power shutdown external events.

6 And then, I know this morning Alice came 7 up. We do have some slides in the open session.

8 We're going to go ahead and present those, even if we 9 need to push that discussion mostly to the closed 10 session.

11 So, I'm going to start with the staff's 12 review approach. And then, I'll present some of the 13 external events towards the end.

14 CO-CHAIR CORRADINI: And, you guys, make 15 sure your green light's on and you talk loud for 16 our -- yeah. Thank you.

17 MS. NEUHAUSEN: All right. This is 18 slide 4. Okay, so the staff's Chapter 19 SE is based 19 on Revision 2 of the DCA. We issued 31 RAIs, which 20 contains about 59 questions. And that was for 21 Sections 19.1, 19.2 and 19.3.

22 As part of the review, staff conducted two 23 regulatory audits. These provided access to the PRA 24 notebooks.

25 The first audit took place April to August 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 of 2017, and staff reviewed NuScale documents which 2 included over 50 of those notebooks supporting the 3 PRA, and we asked 31 questions to clarify information 4 in the DCA. Those were the formal questions that are 5 included as part of the audit summary.

6 The staff sampled the notebooks. Those 7 notebooks included both the self-assessment and the 8 external review of the self-assessment that NuScale 9 performed.

10 So, as a result of the audit staff was 11 able to resolve some of the questions, and then issued 12 RAIs based on some of the others.

13 So, staff determined that an extensive 14 number of calculations and auxiliary studies support 15 the description and results of the PRA that were 16 reported in the SR, and that the scope and level of 17 detail is generally consistent with the expectations 18 of the NRC. The expectations documents are in NRC Red 19 Guide 1.206 and SRP 19.0.

20 And then, the second audit occurred from 21 March to April 2018, and we evaluated and examined 22 documents to support those RAIs that haven't been 23 resolved.

24 And then, also early in the review, staff 25 acquired the Enhanced Safety Focus Review 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

191 1 that's shorthand known as ESFRA -- to support 2 integrated decision-making and increase the focus on 3 safety for effectiveness and efficiency of the review.

4 Next slide.

5 CO-CHAIR CORRADINI: When you did the 6 audits, they were a week long? Because you said one 7 in April of '17, one in March of '18.

8 MS. NEUHAUSEN: No. The first --

9 CO-CHAIR CORRADINI: Or much longer?

10 MS. NEUHAUSEN: The first audit I think 11 was two months or four months, and the second audit 12 was one month.

13 CO-CHAIR CORRADINI: And you went to the 14 local offices, or out to the Pacific Northwest?

15 MS. NEUHAUSEN: The local offices.

16 CO-CHAIR CORRADINI: Local offices. Okay.

17 So, in town.

18 MS. NEUHAUSEN: Yep.

19 CO-CHAIR CORRADINI: Okay. Thank you.

20 MS. NEUHAUSEN: So, during the review, 21 staff focused on the quality, completeness and 22 consistency of the information in the DCA to ensure 23 that the results, conclusions and insights obtained 24 from the PRA are valid.

25 The staff focused on the purpose 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

192 1 PRA, and specifically at the DCA stage. Some of those 2 include like the determining the risk insights, 3 provide information about risk contributors and 4 defense and deaths, the inputs to operational programs 5 in that some of those were incorporated into the 6 design.

7 So, at the DCA stage many aspects of the 8 PRA rely on key assumptions, and those are documented 9 in tables throughout the Chapter 19 FSAR. And then, 10 it's the responsibility of the COL applicant to 11 confirm that those results are valid.

12 For uses of the PRA beyond those 13 considered for DC purposes, the applicant or licensee 14 will need to demonstrate acceptability in accordance 15 with that intended use.

16 And so, for all the PRA topics, staff 17 focused on ensuring that the appropriate key 18 assumptions were included in those FSAR tables, and 19 the review focus was guided by the commission's goals 20 for core damage frequency, large release frequency, 21 conditional containment failure probability, and PRA 22 insights.

23 The staff used their review guidance 24 provided in SRP 19.0, which includes the acceptance 25 criteria for PRA and severe accidents, 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

193 1 guidance provided in DC/COL-ISG-28, which is endorsed 2 by reg guide 1.200, and addresses the use of the PRA 3 standard.

4 NuScale committed to using that ASME/ANS 5 PRA standard as endorsed by that reg guide and 6 modified by the ISG, which is one way to assess the 7 technical acceptability of the PRA at the DCA stage.

8 So, for the rest of the presentation, 9 staff will cover specific topics of interest to 10 NuScale that is related to NuScale-specific attribute, 11 and I'll turn it over to Ayo.

12 MR. AYEGBUSI: All right. Good 13 afternoon. My name -- sorry.

14 CO-CHAIR CORRADINI: Nice and loud.

15 MR. AYEGBUSI: That's always hard. Can 16 you hear me? All right, good afternoon. My name is 17 Ayo Ayegbusi. I'm a risk and reliability analyst in 18 the Office of New Reactors.

19 So, over the next four slides my goal here 20 is really to discuss topics or aspects of the PRA that 21 the committee has shown some interest in, or that we 22 had some further interactions beyond just what we saw 23 in the DCA or during the audit, and we thought were 24 interesting to highlight to the committee.

25 So, the first one had to do with data.

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 Right? So, the staff reviewed the applicants' data 2 analysis, and for a large portion of the failure 3 probabilities that we used, the applicant relied on 4 generic data that the agency puts out.

5 However, for some, for unique components 6 such as the ECCS valves, the applicant developed its 7 own failure probability for those values. And so, the 8 staff reviewed the reasonability of those failure 9 probabilities and the assumptions that went into them.

10 MEMBER BLEY: Did you go through -- was 11 this done during the audit? Did you go through the 12 detailed engineering reports on those valves?

13 MR. AYEGBUSI: Actually, we cannot speak 14 to that because the person who did the review and the 15 audit is not here.

16 MEMBER BLEY: Was there only person did 17 the audit?

18 MR. AYEGBUSI: On this particular area, 19 yes.

20 MEMBER BLEY: Nobody who was involved in 21 the audit can address that?

22 PARTICIPANT: Can you find out?

23 MEMBER BLEY: We'd like to hear back about 24 that some other time, then.

25 MR. AYEGBUSI: Okay.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

195 1 CO-CHAIR CORRADINI: I don't know if you 2 were in the room or not. They did an audit in '17 and 3 an audit in '18, a few months each time.

4 MEMBER BLEY: If you read the reports, it 5 was a continuation of the same audit, but there were 6 different questions.

7 MR. AYEGBUSI: So, you would like just to 8 double-check --

9 MEMBER BLEY: I want to know what was 10 looked at to decide if these things are reasonable or 11 not.

12 MR. AYEGBUSI: Okay. We do -- I mean, as 13 a result of the audit we do put out an audit report --

14 MEMBER BLEY: I got the audit report. It 15 tells me the questions you asked. It doesn't tell 16 what you found out.

17 MR. AYEGBUSI: So, typically with audit 18 reports, we raise the topics that we discuss with the 19 applicant. We don't typically make an assessment in 20 audit reports.

21 MEMBER BLEY: You didn't. That's true, 22 you didn't. I can't read the audit report and find 23 the answer to my question.

24 MR. AYEGBUSI: So -- but I think your 25 question is, you'd like to know if during the audit 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

196 1 looked at the engineering reports, engineering design 2 of the valves.

3 MEMBER BLEY: Yes. And whatever was done 4 by whoever did it to come up with the failure rates if 5 they decided to do this.

6 CO-CHAIR CORRADINI: And to put it more 7 succinctly, the applicant suggested that for things 8 that were unusual -- we'll just call them the ECCS --

9 as an example, the ECCS valves that they called the 10 piece parts analysis that developed the failure rate 11 for the valve as a whole, the question is, did 12 somebody from the staff look at that and determine 13 that it was reasonable, or had RAIs, or there was a 14 re-analysis. Is that close, Dennis?

15 MEMBER BLEY: Yeah. I want to know the 16 basis for the staff's evaluation.

17 MR. AYEGBUSI: So, I mean --

18 MEMBER BLEY: And I don't just want to 19 hear that we looked at it and it was reasonable.

20 MR. AYEGBUSI: Okay. So, I can give you 21 my -- the insights that I have from discussions with 22 the individual who did the audit. Right? So --

23 because this particular set of failure rates for the 24 ECCS valves were definitely something that we wanted 25 to take a look 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

197 1 In this case, the person who did the audit 2 looked at the applicant's assessment -- right? --

3 looked at the inputs going into the applicant's 4 assessment, and I can only speak overall to how we 5 would look at that. Right?

6 Typically, what we'd look at is, if you 7 look at the industry generic data for valves --

8 right? -- you typically would see -- they're typically 9 on the order of ten to minus three, ten to minus 4.

10 Right?

11 So, when we look at what the applicant did 12 from a reasonability standpoint, what we're looking 13 for, we look to see if the final failure rate for the 14 valve is somewhere in that ballpark, given what we 15 have --

16 MEMBER BLEY: I think the ECCS valves are 17 something like five times minus 5?

18 CO-CHAIR DIMITRIJEVIC: 5.90 minus 5.

19 MEMBER BLEY: That's not like ten to the 20 minus 3. So, why is that a reasonable result?

21 MR. AYEGBUSI: So, again, in this case, 22 there are many things we look at. So, we would look 23 at the valve --

24 MEMBER BLEY: You can't tell me what was 25 exactly --

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 MR. AYEGBUSI: That is correct.

2 MEMBER BLEY: I want to hear what was 3 looked at and why it was decided that this is 4 reasonable.

5 MR. AYEGBUSI: Okay.

6 MEMBER BLEY: So, I -- keep going --

7 MR. AYEGBUSI: Okay.

8 MEMBER BLEY: -- in generality.

9 CO-CHAIR CORRADINI: And just -- I'm 10 sorry. Just so you know where we're coming from, so 11 at the end of today we're going to come up with 12 suggestion at the June meeting what you want to 13 emphasize in the presentation. This you want to 14 emphasize.

15 MR. AYEGBUSI: Okay.

16 DR. SCHULTZ: The conclusion on the slide 17 seems to indicate that in the evaluation of the 18 component reliability or probability, could have been 19 dismissed because there was so much margin between the 20 CDF goals and that which was calculated. And we 21 prefer not to hear that. We prefer to know what was 22 determined associated with the failure rate 23 calculations and whether they were valid.

24 MR. AYEGBUSI: Okay. Our intent was not 25 to make it seem that way. Our intent was to say 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

199 1 because this is a new design -- the whole plant itself 2 is a completely new design. These unique components 3 are new designs that have no previous operating 4 experience. Right?

5 As I was trying to explain earlier, a 6 couple of ways we, as a group, we evaluate 7 reasonability of such things as to look at how it 8 compares with similar components that have operating 9 experience, but also big-picture-wise, to look at how, 10 you know, using these failure rates, how the results 11 compare with commission goals, and also looking at the 12 sensitivity studies that were performed, and looking 13 at how those compare with the commission goals.

14 And so, all we're saying here is when we 15 go through that process, is the process we in our 16 group typically do, when we go through that process we 17 still find that the results are still favorable when 18 you compare them to the commission goals. That's the 19 point of this slide.

20 DR. SCHULTZ: Thank you.

21 MR. AYEGBUSI: Yes, sir.

22 MEMBER MARCH-LEUBA: The language only on 23 the first point says that the things are reasonable 24 for the DCA stage. Was that in play that you were 25 going to revisit this at the COL stage?

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 MR. AYEGBUSI: So --

2 MEMBER MARCH-LEUBA: So, once it's 3 approved, it's going to be -- nobody's going to touch 4 it.

5 MR. AYEGBUSI: So, particularly with the 6 PRA, our expectation is if there is any additional 7 information that the next revision of the PRA would 8 consider any new information. Right?

9 So, if you needed to update some of the 10 failure rates -- right? -- our expectation is at least 11 that would be considered. Right? We can't say for a 12 fact that that would become the new failure rate when 13 you consider any new information.

14 MEMBER MARCH-LEUBA: You expect the 15 applicant to initiate that change.

16 MR. AYEGBUSI: The COL applicants.

17 MEMBER MARCH-LEUBA: The COL applicants.

18 MR. AYEGBUSI: Yes, sir. I think Ian 19 wanted to say something.

20 MR. JUNG: Yeah. So, I just want to share 21 the information that I have on audit notes. Ian Jung, 22 Senior Reliability and Risk Analyst.

23 So, those staff participate in the audit 24 left actually detailed notes. It's a proprietary 25 information here. So, I'll share some information 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 through a different channel.

2 Specifically here, it talks about the 3 staff auditing data analysis ECCS system notebooks, 4 PRA analysis of the ECCS. So, there are multiple 5 pages of detailed information what staff looked at.

6 That's one information.

7 And also, we've been working with the 8 mechanical engineering branch folks. And actually, 9 Alissa and I are currently a part of the audit team 10 participating. And Tom Scarborough, another member, 11 actually doing the audit of the ECCS actually valve 12 testing that's going to happen pretty soon.

13 So, we are following up on that to make 14 sure that there's no significant delta between the 15 submissions and the system develop --

16 CO-CHAIR CORRADINI: Can you repeat what 17 you said? You said that you're going to be part of 18 the ECCS testing that'll be what, I'm sorry?

19 MR. JUNG: There's an audit ongoing right 20 now on ECCS system valves. There's actually onsite 21 valve testing that's going -- I think that's going to 22 take place pretty soon.

23 MEMBER BLEY: The test program's defined?

24 MR. JUNG: You're asking -- I've seen the 25 audited plan defines the system testing, the plan 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

202 1 the outcome, and the detailed information behind that.

2 So, I think you're going to hear some of this 3 information through another chapter.

4 MS. NEUHAUSEN: The answer, too, is that 5 the MEB staff is reviewing that testing plan leading 6 up to the onsite testing that they're going to be 7 auditing.

8 MEMBER BLEY: We're reviewing Chapter 6 9 next month. You're going to be ready to talk about it 10 then?

11 CO-CHAIR CORRADINI: I think it's in 12 Chapter 39-something -- 396 -- but they're together 13 next month. Am I -- you guys have to correct me, but 14 I think somewhere in the notes that I can't find at 15 the very instant here -- I think it's 396 -- is where 16 valve testing or valve certification is part of.

17 MS. NEUHAUSEN: Yeah.

18 CO-CHAIR CORRADINI: I'll check.

19 MR. JUNG: Okay.

20 CO-CHAIR CORRADINI: But to answer your 21 question, we're going to ask, if we can't get it this 22 month, we're going to get it next month at the 23 subcommittee meeting. And Chapter 3 is on Tuesday, 24 June the --

25 PARTICIPANT: Eighteenth.

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 CO-CHAIR CORRADINI: -- 18th. Thank you.

2 PARTICIPANT: Okay.

3 CO-CHAIR CORRADINI: Does that help?

4 MEMBER BLEY: It does a little. And I 5 guess I -- I hope that the PRA review group that's 6 talking to the people who are going to be watching the 7 testing to make sure your concerns are being addressed 8 by the people who are going to oversee the testing.

9 You're part of the team --

10 MS. NEUHAUSEN: Yes. So, I'm part of 11 the -- I'm not overseeing the testing but I'm part of 12 the audit team.

13 MEMBER BLEY: The audit team of the 14 testing.

15 MS. NEUHAUSEN: Yes.

16 MEMBER BLEY: Okay.

17 MEMBER MARCH-LEUBA: My concern with 18 respect to the testing is that extreme reliability on 19 the order of the ten to minus 5 is claimed for this 20 complex valves, one of a kind. This distinction is 21 important. I mean, if you're in testing else once, 22 you cannot claim the ten to the minus 5 reliability.

23 Just keep that in mind.

24 I mean, unless you run another 100,000.

25 So, we need to have a problem that validates 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

204 1 this is a very high reliability for a one-of-a-kind, 2 complex system. And I'm no expert on frequency of 3 failure, but it's hard to believe.

4 MR. AYEGBUSI: Thank you. Understood.

5 CO-CHAIR DIMITRIJEVIC: I also have a 6 couple of comments.

7 CO-CHAIR CORRADINI: Microphone.

8 CO-CHAIR DIMITRIJEVIC: Sorry. This table 9 19.1-9 in PRA with the comparison to be modified 10 generic data. But they don't really give a generic 11 mean values for the half of the table.

12 I assume -- they said not applicable, but 13 I assume that's because they started something and it 14 was modified so much that they considered not 15 applicable.

16 However, I think that this table will be 17 very useful to include these generic data. For 18 example, for this hydraulically-operated ECCS, which 19 they base on the BWR, because they give a generic 20 source. They give data, too, put a star if it's 21 significantly modified, but they should know where it 22 starts.

23 So, I will complete this table it misses 24 in the six of those -- misses this generic data.

25 The second bullet there is not true.

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 Sensitivity studies were not done on the component 2 failure rates. Sensitivity studies were done on the 3 failure of the passive heat transfer. That's 4 completely different.

5 CO-CHAIR CORRADINI: Say that again, 6 please?

7 CO-CHAIR DIMITRIJEVIC: Sensitivity 8 studies were not done on component failure rates.

9 They're done on this event, which says the -- how is 10 it called -- passive heat transfer failure. That's a 11 different -- you know there is a valve.

12 And then, there is a passive heat failure, 13 and the sensitivity studies are not done on component 14 failure rates. What you said in the second bullet is 15 not true.

16 MR. AYEGBUSI: Okay. Yes, I understand 17 what you're saying on components, specifically on 18 components.

19 CO-CHAIR DIMITRIJEVIC: On component 20 failure rates.

21 MR. AYEGBUSI: Yes.

22 CO-CHAIR DIMITRIJEVIC: There's nothing 23 one in sensitivity on that.

24 MR. AYEGBUSI: Correct. Well, I guess it 25 depends on what you're referring to, because 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

206 1 sensitivity done with common cause failure of 2 components. I don't know if that --

3 CO-CHAIR DIMITRIJEVIC: Yes, but that 4 sensitivity only shows that that's important. So, we 5 cannot conclude like that.

6 CO-CHAIR CORRADINI: Vesna, can I ask the 7 applicant -- all right, I want to make sure, because 8 you're looking at slide 25 from the applicant's thing 9 where it had the sensitivity studies.

10 CO-CHAIR DIMITRIJEVIC: Yes.

11 CO-CHAIR CORRADINI: Is how Vesna 12 described it correct? Or you also did component 13 sensitivities? Is anybody from NuScale can help us 14 here?

15 CO-CHAIR DIMITRIJEVIC: Well, if they did 16 that, it's not presented anywhere.

17 CO-CHAIR CORRADINI: Well, I know. But I 18 wanted to get clear to it. Yeah, they're going to 19 turn it on. It take a while to energize. It's 20 failsafe. Just keep on hitting it.

21 PARTICIPANT: There's no switch?

22 CO-CHAIR CORRADINI: No, no. They have to 23 do it in the control room, I think. There you go.

24 MR. GALYEAN: Okay. Well, as we --

25 CO-CHAIR CORRADINI: You 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

207 1 MR. GALYEAN: I'm sorry. This is Bill 2 Galyean, NuScale PRA Group. We did do the sensitivity 3 studies on, for example, the component -- the common 4 cause failure rates. Okay? The results, of course, 5 of our PRA are dominated by common cause failure.

6 So, you talk about the five times ten to 7 minus 5 failure probability for the ECCS valves.

8 Remember that we scaled -- or the sensitivity study 9 that we did on common cause failures, we changed the 10 probability of the common cause failures to two times 11 ten to minus 3.

12 So, indirectly, that's a sensitivity study 13 on the failure rate, because the combined failure 14 of -- the common cause failure of -- the common cause 15 failure groups of the ECCS valves, both of the 16 failing, was set to two times ten to minus 3.

17 CO-CHAIR DIMITRIJEVIC: I understand. But 18 this is still not the same. You did the common cause 19 so everything -- and it was put to two minus 3, is not 20 change of the failure rate because exchange are both 21 common cause factor rates even worse change, because 22 otherwise, it's 2.5 E minus 6, you are changing to two 23 E minus 3.

24 I know you're conservative, but I don't 25 really still don't know how sensitive you are to 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

208 1 failure rate. I don't that if you make really 2 conservative assumption, you still meeting safety 3 goal, and I never doubt that. I just said that 4 doesn't show us how sensitive are you to the failure 5 rates of the valve.

6 CO-CHAIR CORRADINI: Of the valve 7 specifically.

8 CO-CHAIR DIMITRIJEVIC: Yes.

9 MR. GALYEAN: Granted. There are lots of 10 components and details --

11 CO-CHAIR DIMITRIJEVIC: Right. It's 12 all --

13 MR. GALYEAN: -- that we could do 14 sensitivity studies on.

15 CO-CHAIR DIMITRIJEVIC: -- these are -- I 16 mean, really, on common cause of change here --

17 CO-CHAIR CORRADINI: Okay.

18 CO-CHAIR DIMITRIJEVIC: -- the big change.

19 So --

20 CO-CHAIR CORRADINI: Okay. Go ahead.

21 MR. AYEGBUSI: Yeah. I mean, I agree with 22 you, we could probably clarify the slide better.

23 CO-CHAIR CORRADINI: Keep on going. If 24 there's quiet, keep on going.

25 MR. AYEGBUSI: All right. So, 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

209 1 please? All right, so this is another hot topic 2 earlier today. So, the staff reviewed the applicant's 3 passive system reliability evaluation. And, you know, 4 our assessment was documented in the safety evaluation 5 report.

6 But mainly the staff was looking at the 7 uncertainty around the passive systems that the 8 applicant relies on, the DHR system and ECCS systems.

9 And as a result of our assessment, we 10 raised some questions with applicant, one of which, a 11 good example, had to do with the non-condensable 12 gases, volume and distribution in the passive system.

13 The subsequence of that interaction with 14 the applicant, the applicant -- I would say 15 significant, but changed a significant portion of 16 their discussion, the release of that, in the FSAR.

17 And upon review of the revision of the DCA 18 in this particular area, the staff has determined that 19 the applicant's passive system, the liability 20 evaluation, was reasonable, and identified areas of 21 potential challenges to the passive systems 22 adequately. Next slide, please.

23 So, in our SER the staff documented one 24 open item, which had to do with RAI 8840, 25 question 19-2. In that RAI, what the staff identified 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 was for LOCA's inside containment the applicant did 2 not specifically identify an assumption that -- and 3 those specific events that containment isolation would 4 not be necessary for the safety systems to actuate and 5 function, and get the plant to a safe and stable 6 condition.

7 So, the staff engaged the applicant on 8 this particular issue, and through subsequent 9 supplemental responses, the applicant and audits and 10 independent evaluation -- or independent calculations 11 done on our part, we were able to find what the 12 applicant assumed reasonable.

13 And so, we're awaiting the supplemental 14 response that would capture this assumption in the 15 supplemental response and the subsequent revision to 16 the DCA.

17 CO-CHAIR CORRADINI: I don't -- I'm sorry.

18 Do you understand this?

19 MEMBER MARCH-LEUBA: I was going to ask, 20 what is the problem with not isolating --

21 CO-CHAIR CORRADINI: Thank you.

22 MR. AYEGBUSI: So, the typical prior 23 response is when you have an event -- right? -- you 24 have containment isolation. That bottles up 25 containment. 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

211 1 And when the ECCS actuation valves -- when 2 the ECCS valves actuates -- right? -- allows for that 3 passive cooling of the core through the vessel into 4 the containment, into the ultimate heat sync. Right?

5 So, when we looked at the event trees 6 specifically for LOCA's inside containment, the LOCA's 7 outside containment isolation was shown to be 8 necessary.

9 But for LOCA's inside containment, the 10 applicant did not question if containment isolation 11 was necessary to prevent core damage.

12 MEMBER MARCH-LEUBA: Did they perform a 13 calculation with and without isolation, or did you 14 perform a calculation? I cannot see how the isolation 15 of -- if you don't really work inside containment, 16 you're depressurized.

17 Whether the valves outside are open or 18 closed, it changes the pressure a little bit, but not 19 much.

20 MR. AYEGBUSI: I don't think I -- so, I 21 think --

22 MEMBER MARCH-LEUBA: Do you have a 23 calculation, a TRACE calculation, with and without 24 isolation for an inside containment LOCA that's just 25 a difference? Because I'm doing it in my head and 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

212 1 don't see much difference.

2 MR. AYEGBUSI: So, obviously for -- I 3 don't have both calculations with me, but obviously 4 for Chapter 15 -- right? -- you have to assume that 5 you have containment -- you have to go with the plant 6 design, and that's containment isolation. Right?

7 MEMBER MARCH-LEUBA: So, what you're 8 saying, in the notices of record they assume 9 isolation, and it's up to that 19 they didn't require 10 it? Is that what you're saying?

11 MR. AYEGBUSI: I guess what I'm saying is 12 the plant design is to have containment isolation.

13 Right? So, in Chapter 15 you would. And that would 14 support the first part of your question.

15 MEMBER MARCH-LEUBA: Right.

16 MR. AYEGBUSI: Right? The second part of 17 your question is, in Chapter 19, after multiple 18 interactions with the applicant, they said containment 19 isolation's not necessary -- right? -- for LOCA's 20 inside containment.

21 MEMBER MARCH-LEUBA: And the obvious path 22 for that is you issue another ISA. Show me a 23 calculation without isolation, and show me it's okay.

24 MR. AYEGBUSI: Correct. That's what we 25 want --

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 MEMBER MARCH-LEUBA: Show me the result.

2 MR. AYEGBUSI: We went through that 3 iteration, audited their calculation, and we had our 4 own staff perform our own calculations, and we got to 5 the point where we could reasonably conclude that what 6 the applicant did was adequate.

7 MEMBER MARCH-LEUBA: I'm just saying it's 8 fastest to us either to run the calculation --

9 MR. AYEGBUSI: Well, we did that. We had 10 our internal people run the calculation.

11 MEMBER MARCH-LEUBA: It's much faster to 12 run TRACE or RELAP than to talk about it.

13 MR. AYEGBUSI: Well, so, the real -- I 14 mean, the number one concern, though, was the PRA's 15 supposed to reflect the plant design. Right? So, 16 that was the number one concern there.

17 MEMBER MARCH-LEUBA: That's good.

18 MR. AYEGBUSI: Not necessarily just the 19 thermal-hydraulic response.

20 MEMBER MARCH-LEUBA: My claim is, that 21 raises to an RAI that you issue earlier in the review, 22 because there's a hole in the review.

23 MS. NEUHAUSEN: That is the case. We 24 issued this early in the review and it's just there 25 were subsequent questions.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

214 1 MEMBER MARCH-LEUBA: I'm concerned this is 2 still an open item.

3 MR. AYEGBUSI: Oh. Well, so, as I said --

4 MEMBER MARCH-LEUBA: It's easy to close.

5 Very easy to close.

6 MEMBER BLEY: Or better yet, why is it 7 still open?

8 MEMBER MARCH-LEUBA: Yeah.

9 MR. AYEGBUSI: So, I would say it's still 10 open because obviously, like you said, we perform our 11 independent calculation, and it took a while to 12 understand their inputs, and kind of -- and then 13 change around our inputs to ensure we're working along 14 the same path, basically.

15 DR. SCHULTZ: And the SER says it's still 16 open here.

17 MR. AYEGBUSI: Correct.

18 DR. SCHULTZ: So, there are differences 19 that have not yet been explained between what has been 20 submitted in response to the RAI? Or you just haven't 21 had the time to put everything together to close the 22 open item?

23 MR. AYEGBUSI: So, well, where we are as 24 of today is the supplemental response that was -- a 25 supplemental response was sent in and we need 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

215 1 review that, and then close out the open item. Right?

2 And that supplemental response was as a 3 result of a public meeting we had with applicant where 4 they explained their position and we found their 5 position reasonable, or acceptable.

6 DR. SCHULTZ: It was close but not 7 complete, in terms of the conclusion.

8 MR. AYEGBUSI: I would say we're close --

9 I would say at this point we're ready to close out 10 this open item.

11 DR. SCHULTZ: Okay, thank you.

12 MR. AYEGBUSI: Any other questions?

13 CO-CHAIR CORRADINI: Keep on going.

14 MR. AYEGBUSI: All right.

15 MS. NEUHAUSEN: Next slide.

16 MR. AYEGBUSI: Next slide. Okay, so this 17 is another hot topic that came up that we were told 18 about that the committee was interested in. So, for 19 ATWS the staff reviewed the applicant's ATWS 20 discussion. And that's actually done -- and our SER's 21 done in several places. Chapter 7 of the SER 22 documents the staff approval of the ATWS exemption 23 from 5062.

24 And then, Chapter 15 also touches on the 25 overall conclusion of compliance to that. However, 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 Chapter 19 documents the --

2 MEMBER MARCH-LEUBA: If I remember 3 correctly, Chapter 7 -- I mean, or -- of the other ten 4 to the minus 5?

5 MR. AYEGBUSI: That's correct.

6 MEMBER MARCH-LEUBA: All right. That 7 would not drop you below the goal. If ATWS would fail 8 the core and the probability to failure to scram is 9 ten to the minus 5, you are at the limit, meaning 10 you're still within the non-acceptable ban.

11 So, you have to combine the probability of 12 failure rates -- the probability of failure to scram 13 is ten to minus 5, and then you have to have 14 additional failures to have core damage.

15 MR. AYEGBUSI: Correct.

16 MEMBER MARCH-LEUBA: For which you have to 17 an on-the-record calculation that shows there's no 18 failures if you fail to scram.

19 MR. AYEGBUSI: Correct.

20 MEMBER MARCH-LEUBA: But then, in 21 Chapter 15, they have a paragraph that says, we are 22 not going to license because the probability extends 23 to the minus, I don't know, 43. An original number.

24 There is one of those logical holes in 25 there. Somebody needs to do what Pete has done, 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

217 1 is analyze it and demonstrate that you don't fail with 2 an ATWS, and you have to put it on record. You cannot 3 just say it. And I haven't seen NuScale's calculation 4 anywhere.

5 MR. AYEGBUSI: Okay.

6 CO-CHAIR CORRADINI: Don't agree with him 7 unless you really agree with him. I don't know what 8 the statement says up there. Is the statement saying 9 that because of the MPS the failure to scram, to 10 insert rods as one in ten to minus 5th and that's 11 acceptable to the staff? Is that what that statement 12 says? I'm still not sure what it says in front of me.

13 MR. AYEGBUSI: So, just to be clear, what 14 we're really doing here is setting the stage for Pete 15 to present. And all we're really saying is there's 16 some discussion of ATWS in Chapter 7, Chapter 15, and 17 Chapter 19. Right?

18 And the specific calculations that Pete --

19 well, analysis that Pete has done, he's going to 20 present that in the next slide.

21 MS. NEUHAUSEN: Yeah. And I think this --

22 Ian, correct me if I'm wrong, that his commissioned 23 CDF goal is 20 to the minus 5 per year --

24 MEMBER MARCH-LEUBA: Can you put your 25 microphone closer?

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 MS. NEUHAUSEN: Closer? There was that 2 SECY paper that we referenced in the SE that 3 documented some goal for ATWS specifically.

4 MR. JUNG: Yeah. Ian Jung. So, around 5 1983, after the Salem ATWS event, that there was a 6 whole series of conversations between the commission 7 and the staff, and this particular the SECY paper, 8 although it's not like current -- the SECY paper is 9 like 700-something pages long, presentations back and 10 forth and discussion of it.

11 But the whole basis for the ATWS rule, the 12 purpose of the ATWS rule, is to reduce the risk from 13 ATWS to be below one minus 5, based on the studies on 14 the existing plants, PRAs, and all that.

15 Most of the plants were 20 minus 5-ish.

16 So -- and some of them are higher. The commission 17 wanted that ATWS to be below 20 minus 5. That is a 18 policy acceptance goal. So, during the Chapter 7 19 review, Mark Caruso, who has retired since then, was 20 involved in looking at ATWS risk portion of the 21 exemption request to confirm that.

22 MEMBER MARCH-LEUBA: But administratively, 23 the ATWS rule is a rule that imposes a number of 24 design criteria that you must have in your reactor.

25 Like you have to have an alternate means of rod 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 insertion, and I don't remember them all right now.

2 And because -- is NuScale asking for an 3 exception to that? To the rule? Is that what you're 4 saying? They need that exception because they meet 5 the goal that the commission had when they wrote the 6 rule, but they don't really meet the rule?

7 MS. NEUHAUSEN: Yeah. So, that's what we 8 were saying, is that this is kind of the 9 administrative side to say Chapter 7 is what's 10 documenting the ATWS exemption request.

11 But then, what we reviewed was, in 12 Chapter 19, the beyond-design basis. And then, Pete's 13 going to talk about his evaluation. And then, 14 Chapter 15 makes the overall conclusion. So, this was 15 really just trying to point you to the right chapters 16 where we make different conclusions.

17 MEMBER BLEY: It just says what the 18 criteria is, and that Pete's going to tell us about 19 whether they made it or not.

20 CO-CHAIR CORRADINI: But maybe you guys 21 don't understand. I'm still confused.

22 MEMBER BLEY: I think we need to wait for 23 Pete here.

24 CO-CHAIR CORRADINI: Well -- but I want to 25 make sure what the -- what is the exemption request?

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 The request is an exemption from the ATWS rule as 2 stated?

3 MS. NEUHAUSEN: It was from a portion of 4 the ATWS rule, and I have to get back to you. I 5 didn't review --

6 CO-CHAIR CORRADINI: Okay. All right, I 7 think that's what's, at least, confusing me. Maybe 8 it's confusing him, too.

9 MEMBER MARCH-LEUBA: I know that they have 10 an ATWS rule exception from it. I think we're going 11 to get a clarification.

12 MR. GILMER: Jim Gilmer, NRO Reactor 13 Systems. The clarification is, NuScale's only asking 14 for exemption from the automatic turbine trip portion 15 of 5062, not exemption from the ATWS rule entirely.

16 In pre-application discussions, there was 17 back-and-forth on whether or not they also need 18 auxiliary feedwater automatic initiation.

19 CO-CHAIR CORRADINI: Okay.

20 MR. GILMER: And there was something 21 called the gap letter and the staff response to that, 22 which we can provide to you.

23 CO-CHAIR CORRADINI: So, they're not --

24 make sure, Jim, they're not looking for an exemption 25 to the rule, they're looking for a specific exemption 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 from some of the actions.

2 MR. GILMER: From a portion of the rule.

3 The 5062 C1, which is the automatic turbine trip. And 4 staff has agreed with NuScale's position on the need 5 for auxiliary feedwater. Basically, the decay heat 6 removal system performs the equivalent function.

7 CO-CHAIR CORRADINI: Okay. Okay.

8 MR. GILMER: So, they don't need to ask 9 for an exemption from the aux feed initiation.

10 MEMBER BLEY: My memory entirely -- when 11 we did Chapter 7, we had a fairly extensive talk about 12 the turbine trip and it seemed to me there was 13 something we were waiting for when we got to some 14 other chapter. Remember?

15 MR. GILMER: No, I'm just -- I'd have to 16 look at the notes.

17 MEMBER BROWN: I just pulled it back up 18 and yeah, it's very explicit. In 7.1.6 they talked 19 about an exemption from the diverse turbine trip --

20 CO-CHAIR CORRADINI: Right.

21 MEMBER BROWN: -- capability. And I --

22 MEMBER BLEY: I was just looking at that.

23 MEMBER BROWN: So, my memory's getting 24 jogged. I'd have to go back and -- I don't remember 25 talking about heat in the transcript. I'd have to go 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 back and look.

2 MEMBER BLEY: We did. But I thought there 3 was something reserved until later.

4 MEMBER BROWN: I'll go back and look at 5 the transcript.

6 CO-CHAIR CORRADINI: Thank you.

7 MEMBER BLEY: I think we need to do that.

8 MS. NEUHAUSEN: We can turn it over to --

9 MEMBER BROWN: But that was the only part 10 of it. It was just the discussion. And I remember 11 that discussion.

12 CO-CHAIR CORRADINI: That was -- okay, so 13 that's a lot more specific. That helps.

14 MS. NEUHAUSEN: Okay. Pete will present 15 the research ATWS.

16 DR. YARSKY: Hello. I'm Dr. Peter Yarsky 17 from the Office of Research and I'm here to discuss 18 confirmatory calculations that we performed with TRACE 19 for ATWS scenarios for NuScale.

20 In performing these calculations, the key 21 figures of merit that we considered were the peak 22 reactor vessel pressure, which is to confirm the 23 integrity of the RPV.

24 Additionally, we looked at the collapsed 25 liquid level in the riser to ensure that that liquid 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 level was above the top of active fuel, and that's to 2 confirm the core coolability. These are figures of 3 merit that we used in our analysis.

4 We go to the next slide. The base case 5 ATWS that we evaluated is initiated by a loss of AC 6 power. And this leads to an immediate turbine and 7 feedwater system trip, at which point we assumed that 8 the module protection system fails to insert the 9 control rods.

10 And for our analysis we assume that the 11 control rods remain withdrawn through the entire 12 event.

13 MEMBER BROWN: So, you mean you assumed 14 failure of all divisions of the module protection 15 system.

16 DR. YARSKY: Right.

17 MEMBER BROWN: And they do not initiate a 18 trip.

19 DR. YARSKY: Right. So, we assumed that 20 the module protection system --

21 MEMBER BROWN: -- I got it. All right.

22 DR. YARSKY: -- there's no diverse 23 protection --

24 MEMBER BROWN: I just want to make sure I 25 had your words down exactly.

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 MEMBER MARCH-LEUBA: It's either that or 2 there is a mechanical failure that presents the rods 3 from going in.

4 MEMBER BLEY: All the rods.

5 MEMBER MARCH-LEUBA: All the rods.

6 DR. YARSKY: All the rods. From the 7 standpoint of the TRACE calculation, the control rods 8 remain withdrawn from the full transcript.

9 In response to the event, the RPV pressure 10 increases due to the loss of heat sync, and the 11 reactor cooling system heats up. This higher RPV 12 pressure will initiate the actuation of the key heat 13 removal system.

14 As RPV pressure continues to increase 15 beyond that during the ATWS event, this would initiate 16 opening of the reactor safety valve. We analyzed --

17 MEMBER BLEY: You didn't -- yeah, you 18 didn't consider turbine bypass at all either.

19 Everything's bottled up.

20 DR. YARSKY: Right. So, if the way we 21 simulate the event is the turbine will trip, and then 22 we don't simulate any kind of turbine bypass.

23 Instead, we would simulate the actuation of the DHRS 24 valves to open.

25 MEMBER MARCH-LEUBA: Yeah. Plus, you've 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 assumed loss of AC power, so there is no feedwater --

2 there is no steam flow to go through a bypass.

3 DR. YARSKY: Yeah. So, the turbine bypass 4 valves were to open and DHRS was unavailable, there 5 would be no feedwater to supply any kind of liquid to 6 remove heat from the primary side.

7 MEMBER MARCH-LEUBA: There wouldn't be any 8 steam to go to the bypass.

9 DR. YARSKY: Exactly. So, seismic 10 analysis has performed a number of sensitivity 11 studies. And what we looked at were different key 12 scenarios. Beyond the base case, we looked at -- we 13 did perform calculations of both end-of-cycle and 14 beginning-of-cycle kinetics parameters.

15 We also performed a case where we assumed 16 that the RSV-1 valve was out of service. We then 17 considered a scenario where both RSV-1s --

18 MEMBER BLEY: I don't remember, what's 19 RSV-1?

20 DR. YARSKY: So, there's two reactor 21 safety valves that are at the --

22 MEMBER BLEY: Yeah, this is one of them.

23 DR. YARSKY: -- on the pressurizer. So, 24 RSV-1 is the lower pressure --

25 MEMBER BLEY: 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

226 1 DR. YARSKY: -- RSV valve. Yeah, so they 2 have lift and set pressures, and they're in different 3 bands. RSV-1 is the lower pressure one.

4 We also performed a case where we assumed 5 that RSV-1 is out of service, but in addition, DHRS is 6 out of service.

7 MEMBER MARCH-LEUBA: All the interest.

8 DR. YARSKY: All the interest.

9 MEMBER MARCH-LEUBA: On both of them?

10 DR. YARSKY: Right. So, there's no DHRS 11 at all. And lastly, we did a sensitivity calculation 12 which is like the base case, but we have artificially 13 reduced the steam generator heat transfer until we 14 achieve a very high initial RCS temperature.

15 MEMBER MARCH-LEUBA: Okay. On the PRA 16 case you still have one RSV working. Correct?

17 DR. YARSKY: Right. So, RSV-2 is in 18 service, RSV-1 is out of service. So, we go to the 19 next slide.

20 Just to summarize our findings in these 21 cases, the base case, the BOC case --

22 MEMBER MARCH-LEUBA: Sorry. Let me --

23 you're going to show us your backup slides on the 24 closed session. Right?

25 DR. YARSKY: If you would like to see 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

227 1 detailed calculation results, we can discuss them in 2 the closed session.

3 MEMBER MARCH-LEUBA: Yeah. I will 4 reserve my questions for the closed session.

5 CO-CHAIR CORRADINI: I told them you might 6 do that.

7 MEMBER MARCH-LEUBA: Clairvoyant.

8 DR. YARSKY: Yeah. So, in the closed 9 session we can discuss our results in more detail.

10 But I wanted to present a summary of our key findings.

11 The base case, the BOC case, and the 1-RSV 12 case that we've analyzed, are largely quite similar.

13 And what they all demonstrate are large margins to the 14 RPV, the pressure limit.

15 And in the long-term, we find that the 16 reactor power comes at the balance with the DHRS heat 17 removal in conjunction with a little bit of heat 18 removal through the CNV, or the containment vessel, 19 and that the long-term level remains well above the 20 top active fields. So, therefore, we find that the 21 RPV integrity is maintained, and core coolability is 22 maintained.

23 For the PRA 1-RSV case, well, this is only 24 one RSV in service and no DHRS. We find that there's 25 still large margin to the peak RPV pressure, and 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

228 1 the long-term some inventory builds up in the 2 containment vessel. And this provides a heat removal 3 pathway.

4 And so, the heat generated in the core is 5 balanced by the heat removal through the CNV, and we 6 find that the level remains well above the top of the 7 active fuel.

8 PARTICIPANT: And for the second bullet, 9 the DHRS is also out of service.

10 DR. YARSKY: Right. Correct. The DHRS is 11 out of service for what I'm calling the PRA-1 RSV.

12 CO-CHAIR CORRADINI: So, I'm only dumping 13 heat through essentially -- through the containment.

14 DR. YARSKY: Correct.

15 MEMBER MARCH-LEUBA: You're venting 16 entropy through the steam that goes into containment.

17 DR. YARSKY: Yep.

18 MEMBER MARCH-LEUBA: I guess you're losing 19 more entropy by condensation of the steam by that.

20 But done by --

21 CO-CHAIR CORRADINI: It all goes up as 22 water eventually.

23 DR. YARSKY: Early on, you'll be venting 24 from the RPV into the containment vessel. So, you're 25 burping steam through the RSVs.

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 MEMBER MARCH-LEUBA: Mm-hmm.

2 DR. YARSKY: However, because you don't 3 have the DHRS available, the RCS does reach a really 4 high temperature. When it's at that really high 5 temperature, the reactor shuts down, so it's in a 6 subcritical state. So, you're still on decay heat but 7 just at a very high temperature.

8 Once you build up a decent amount of fluid 9 inventory in the CNV, you're actually able to convect 10 about a decay heat-level worth of heat through the 11 containment. And then, RSV cycling stops.

12 So, you eventually reach a point where 13 you're at high pressure and you're hot, but you're 14 below the RSV lift pressure.

15 So, we can show more of those results, and 16 specifically like what's going on during the --

17 MEMBER MARCH-LEUBA: Let me ask on the 18 opposition and learn the details. Do you have time to 19 calculate what the moderator temperature coefficient 20 is on your simulation?

21 DR. YARSKY: We provide a table of the 22 moderator temperature coefficient as a function of the 23 reactor in a hydraulic state when that's 24 interpolating. But I don't have that value.

25 MEMBER MARCH-LEUBA: Is that document 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 somewhere? I haven't seen it.

2 DR. YARSKY: Yeah.

3 MEMBER MARCH-LEUBA: I know it's -- I'm 4 asking you where is it.

5 DR. YARSKY: The values come from the 6 design calculation --

7 MEMBER MARCH-LEUBA: Oh, no. Where can I 8 find that document to look at it?

9 DR. YARSKY: The document would be the 10 parts calculation notebook, which is the source for 11 the kinetics.

12 MEMBER MARCH-LEUBA: I would like to be 13 able to look at it before the full committee.

14 DR. YARSKY: I'll work with the staff and 15 ask to give you access to all of our calculation 16 notebooks.

17 CO-CHAIR CORRADINI: It's an open book for 18 you. Mike? Mike, Mike.

19 CO-CHAIR CORRADINI: He's writing.

20 DR. YARSKY: We'll have to figure out the 21 logistics.

22 MEMBER MARCH-LEUBA: So, you did calculate 23 an equivalent MTC for your BOC calculation.

24 DR. YARSKY: Correct.

25 MEMBER MARCH-LEUBA: Now, we're talking --

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 I, and obviously this is open session. This is what 2 I think the rules are. If you want to do a best-3 estimate cycle-specific calculation, you are perfectly 4 welcome to run a best-estimate, cycle-specific 5 calculation.

6 If you want to do a one-of-a-kind FSAR 7 calculation that demonstrates that you don't have an 8 issue, you have to use your corporate limit report 9 bound in numbers.

10 And if you're doing a genetic -- that's 11 what I think the rules are. You either do a bounding 12 calculation, or you repeat the calculation for a 13 recycle. And you're welcome to do either of the two.

14 And I think you're thinking an imaginary review cycle, 15 and call it bounding. That's what you're doing now.

16 DR. YARSKY: Well, I think that the 17 question of reload licensing is very separate from 18 what research today.

19 MEMBER MARCH-LEUBA: But -- well, not you.

20 I'm thinking ahead. Do you understand the logic? If 21 you use a cycle-specific MTC -- moderate temperature 22 coefficient -- that's completely acceptable. If the 23 cycle is specific, you have to do the calculation and 24 recycle.

25 If you want to use that calculation to say NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

232 1 I will never have a problem anymore, but in my 2 COLR -- cooperator limits report -- I'm going to allow 3 them to see to be much higher, indeed, they're allowed 4 to be plus-6, PCM per Fahrenheit, there is a 5 disconnect. So, if --

6 CO-CHAIR CORRADINI: I think he's trying 7 to -- I think Member March-Leuba is trying to say to 8 you is, you took a best-estimate and --

9 DR. YARSKY: Correct, yeah. The way I 10 would characterize the TRACE calculation is, it's 11 best-estimate based on equilibrium cycle.

12 MEMBER MARCH-LEUBA: All right.

13 DR. YARSKY: Right. As opposed to a COLR-14 limiting value or generic-limiting value.

15 MEMBER MARCH-LEUBA: Or a bounding -- I 16 mean, the FSAR in Chapter 4 I believe, there's a 17 figure of what is the maximum MTC that you can have.

18 At least negative. And it shows a curve, and it was 19 calculated using some type of procedure. That's the 20 one one should use for a bounding calculation.

21 MR. SCHMIDT: This is Jeff Schmidt from 22 Reactor Systems. So, what Pete did is a Chapter 19 23 event. For something like you're describing, which is 24 say a 15.8 event, if they have a diverse actuation 25 system, there's no reload analysis that's necessary.

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 They've shown that the probability is so 2 low with the diverse actuation system, that you don't 3 need to do a calculation, like on a reload basis.

4 MEMBER MARCH-LEUBA: I thought the failure 5 to scram priority extends to the minus 5. You get to 6 the ten to the minus 11 by adding additional failure.

7 MR. SCHMIDT: Yeah. I don't know the 8 details of the probability. But I think what was done 9 in Chapter 7 was basically show that the MPS 10 effectively meets the diverse actuation system 11 requirement.

12 And once you make that requirement, 13 there's no Chapter 15 analysis that's performed.

14 MEMBER MARCH-LEUBA: Even for the FSAR?

15 MR. SCHMIDT: Even from the FSAR.

16 DR. YARSKY: Right. So, if you look at 17 15.8, there'll be no calculation of say peak RCS 18 pressure.

19 MEMBER MARCH-LEUBA: I know there is 20 nothing on 15.8.

21 MR. SCHMIDT: Yeah. So, I'm confirming 22 that there is nothing in 15.8.

23 DR. YARSKY: Right. So, if we were to re-24 perform the TRACE analysis but were to credit the 25 diverse actuation system, then there would be control 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 rod insert. The event would look in many ways very 2 substantially similar to --

3 MEMBER MARCH-LEUBA: -- Assuming it was.

4 But they're not American reactors. You perform 5 numerous calculations with ARI that keep all the rods 6 out.

7 DR. YARSKY: Well, for the BWRs, that 8 would be the -- whether or not you credit the ARI.

9 MEMBER MARCH-LEUBA: Right. So, in 10 operating reactors, we don't credit that independent 11 actuation of the rods.

12 DR. YARSKY: So like, we'd need to swing 13 a crowbar to separate out like how the PWRs are 14 treated differently than the BWRs in this respect.

15 But in the Chapter 15 analysis for the BWRs, there 16 would be an analysis that would show failure of the 17 rods to insert following the RPS signal, but would 18 credit the insert of the rods with ARI. And then, 19 there'd be a supplemental analysis assuming also the 20 failure of the ARI.

21 CO-CHAIR CORRADINI: Remind me what the 22 ARI is.

23 DR. YARSKY: Alternate rod insert.

24 CO-CHAIR CORRADINI: Oh okay, fine. Thank 25 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

235 1 DR. YARSKY: Right. But if -- and these 2 TRACE calculations we were to credit the diverse 3 actuation system, then that credit would lead to 4 control rod insertion, and these TRACE calculations 5 assume no control rod insertion.

6 MEMBER MARCH-LEUBA: Yeah. That would be 7 an ATWS with rod insertion.

8 DR. YARSKY: Right.

9 CO-CHAIR CORRADINI: Keep on going.

10 MEMBER BROWN: Do you want the answer on 11 the ATWS thing?

12 CO-CHAIR CORRADINI: No.

13 MEMBER BROWN: Not now?

14 CO-CHAIR CORRADINI: Not now. Let's wait 15 until we go into closed session anyway. I would 16 like --

17 MEMBER BROWN: Okay. It was done in open 18 before, in our previous meeting. But that's okay. It 19 can wait.

20 MEMBER MARCH-LEUBA: If that is going to 21 be we don't need to do the calculation, why do we want 22 to?

23 MEMBER BROWN: Well, you deferred. The 24 bottom line is -- I'm going to be short -- his final 25 comment on that was, in the Chapter 7, was we'll wait 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 to see it in Chapter 15 and 19. So, we made no 2 decision and our letter did not address it.

3 CO-CHAIR CORRADINI: But I think, just so 4 we're on the same page, I think I understand what Jeff 5 is saying, based on the definition of what is allowed 6 to be credited and not credited for Chapter 15.

7 But I kind of want to get to finishing 8 Chapter 19 for the moment.

9 DR. YARSKY: And just -- my last bullet 10 that I want to discuss here is for what we're calling 11 the SGHT case. We find that there's still large 12 margin-to-peak RPV pressure limits, and that the 13 liquid level in this case drops further.

14 It reaches about the top of the riser in 15 this case, but it still maintains a significant 16 collapsed liquid level above the top of active fuel.

17 MEMBER BLEY: PRA ought to consider best-18 estimate conditions and all others, with their 19 likelihood of being what's going on at the time of an 20 event.

21 This calculation, as I understood what you 22 said, is strictly best-estimate. So, we don't -- and 23 it's a confirmatory calculation. We don't know what 24 things look like at their worst possible time, or at 25 their best possible time, which ought to be part of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

237 1 the PRA calculation.

2 They ought to look at not just best-3 estimate conditions, but all of them.

4 CO-CHAIR CORRADINI: I think he did --

5 unless I misunderstand, they did a best-estimate with 6 uncertainty?

7 DR. YARSKY: No. This is just strictly 8 best-estimate. I think that is one of the --

9 MEMBER BLEY: -- But sensitivities. Four 10 sensitivities off the best estimate.

11 DR. YARSKY: Correct, but not treatment of 12 uncertainty.

13 MEMBER BLEY: Okay.

14 DR. YARSKY: But if you were mentioning, 15 let's say point in time, we did consider different 16 points in cycle.

17 MEMBER BLEY: Oh, you did? Okay.

18 DR. YARSKY: Correct. But that would 19 primarily affect the nuclear parameters and the 20 assumptions regarding the core. But are there other 21 factors? When you said, in time, I just -- I'm not 22 sure I fully understand.

23 MEMBER BLEY: No, I meant the PRA has to 24 consider events occurring at random points in time.

25 DR. YARSKY: Okay. Yeah, here, we just 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 considered beginning a cycle and ending a cycle.

2 That concludes what I had to present on 3 the TRACE confirmatory calculations.

4 CO-CHAIR CORRADINI: Until we go closed.

5 DR. YARSKY: Until we go closed.

6 PARTICIPANT: Conclusion?

7 MS. POHIDA: Good afternoon. I'm Marie 8 Pohida. I'm the senior PRA analyst and NRO, and I'm 9 going to be talking about Level 2 and module drop.

10 CO-CHAIR CORRADINI: Great.

11 MEMBER BLEY: May I ask you a question 12 right at the outset? I was reading both your 13 inspection -- I'm sorry, your audit reports. And the 14 second audit, they don't tell us quite how, but they 15 told us most all of the events, RAIs, that were looked 16 at were either closed or no real outstanding issues, 17 except for two.

18 And one of those is about corium retention 19 in the reactor pressure vessel, and the other is the 20 potential for high-pressure melt injection, both of 21 which talk about due to phenomenological 22 uncertainties.

23 This staff is continuing to evaluate 24 these. Are you going to talk about those, too?

25 Because I didn't find anything more about it 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

239 1 MS. POHIDA: No. That will be discussed 2 tomorrow in our 19.2 discussion.

3 MEMBER BLEY: 19.2. Okay.

4 MS. POHIDA: Yeah. But thank you for 5 providing a good lead-in to my slide.

6 MEMBER BLEY: You're welcome.

7 MS. POHIDA: I appreciate that. The 8 containment event tree is very simple. There's 9 basically two end states. There's leakage from an 10 isolated containment, and release from an unisolated 11 containment.

12 Severe accident phenomenon, other than a 13 severe accident induced entire tube rupture, was 14 screened from the containment event tree. So, if the 15 RPV fails or the CNV fails due to corium, a large 16 release does not occur due to pool scrubbing. But the 17 details of our assessment are going to be discussed 18 tomorrow in the 19.2 discussion. And also, hydrogen 19 detonation is also addressed in Chapter 6 of the SER.

20 CO-CHAIR CORRADINI: So, just a question.

21 MS. POHIDA: Sure.

22 CO-CHAIR CORRADINI: My impression is that 23 we're also then going to get --

24 MS. POHIDA: Thank you.

25 CO-CHAIR CORRADINI: -- some audit -- 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

240 1 going to call it audit calculations -- independent 2 calculations by the staff tomorrow.

3 MS. POHIDA: Yes.

4 CO-CHAIR CORRADINI: Okay. Thank you.

5 MS. POHIDA: Next slide, please. Okay.

6 Now, I'm going to be talking single module drop, as to 7 differentiate between multi-module drop.

8 Okay, I audited the reactor-building crane 9 PRA notebook. There was a calculated drop probability 10 and it's dominated by operator errors of commission.

11 It's over-speed, you know, over-rays, over-travel, and 12 a failure of instrumentation, the inner locks or the 13 switches, to provide a safety stop.

14 Then, I went to review the NUREGs and load 15 drops. I went and reviewed the EPRI PRA and cask 16 drops, and recent events, to further evaluate the drop 17 probability.

18 So, when I went and reviewed NUREG 1774 19 and looked at operating data from 1980 to 2002, and 20 when you're looking at load drops were greater than 21 30 tons, there was an estimated drop probability of 5b 22 minus 5 quick demand.

23 And the events that went into that drop 24 probability, they were rigging failures. They weren't 25 crane failures, they were rigging failures.

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 The EPRI PRA for both the cask drops, they 2 reported a drop probability in the order of E to 3 minus 6 per lift. And then, I went to look at the 4 details of the STATOR drop, the 525-ton STATOR drop at 5 ANO, and that resulted from a temporary hoist 6 assembly. The calculation was not reviewed or load 7 tested.

8 MEMBER BLEY: We don't have, or at least 9 I haven't seen detailed information on, the design of 10 the crane. I think they told us that's coming at some 11 point --

12 MS. POHIDA: That information was given to 13 us during the audit.

14 MEMBER BLEY: Oh, it was? So, there is --

15 you've seen the design of the crane.

16 MS. POHIDA: I saw the PRA. I audited the 17 PRA of the crane.

18 MEMBER BLEY: What I heard this morning, 19 and I might not be getting this right, is that there 20 are no rig devices that are our usual source of 21 failure, but there's some kind of coupling mechanism 22 that doesn't depend on people going out and hitching 23 up the crane. And I don't have a clue of how that 24 works.

25 MS. POHIDA: The nuclear power module 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

242 1 transported from the operating bed --

2 MEMBER BLEY: How it gets hooked.

3 MS. POHIDA: Yes.

4 MEMBER BLEY: To the crane.

5 MS. POHIDA: The module lift adapter.

6 MEMBER BLEY: Yeah.

7 MS. POHIDA: And that is a permanent 8 feature. You know, a permanently designed feature for 9 using -- for moving the module.

10 CO-CHAIR CORRADINI: I think what he's 11 asking is, how does that permanent feature get 12 attached to a particular module?

13 MEMBER BLEY: Yes. Is there a --

14 CO-CHAIR CORRADINI: Did somebody walk up 15 there and kind of put A to B?

16 MEMBER BLEY: Does it screw itself on or 17 how does it hook up?

18 MS. POHIDA: I'm going to defer to NuScale 19 for those type of details.

20 CO-CHAIR CORRADINI: Okay.

21 MEMBER BLEY: The last time we talked 22 about it they said there was no design.

23 CO-CHAIR CORRADINI: If it's in closed 24 session tell us.

25 MR. GALYEAN: Yes, we can talk about 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

243 1 in the closed session.

2 MEMBER BLEY: Thanks, Bill.

3 MR. GALYEAN: We have, we'll get some 4 pictures that you can take a look at.

5 CO-CHAIR CORRADINI: Thank you.

6 MS. POHIDA: Thank you.

7 MEMBER BLEY: Thanks a lot.

8 MS. POHIDA: Okay, my last bullet on the 9 slide is, NuScale committed to the guidance that's 10 used by operating plants, and that's NUREG-0554, 11 supplemented by ASME NOG-1, for single failure proof 12 of the crane. And that's consistent with operating 13 plants.

14 And I just wanted to note that, in DCA 15 Section, Chapter 9.1.5, there is a table that 16 documents the max speeds and lift heights of the 17 reactor building crane.

18 Can I go to the next slide please?

19 MEMBER SKILLMAN: Are those --

20 MEMBER BLEY: Leads me to the question, 21 oh, go ahead.

22 MEMBER SKILLMAN: Are those the 23 parameters, Marie, that you used? Those heights and 24 lift speeds.

25 You said you did the audit?

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 MS. POHIDA: Yes.

2 MEMBER SKILLMAN: Are the heights and 3 speeds from the table in Chapter 9, the ones that you 4 used for that audit?

5 MS. POHIDA: Let me see, I don't know how 6 to describe this. What I will say, is the top cutsets 7 are dominated by an operator of commission, followed 8 by a failure of a limit switch or an interlock that is 9 assumed to cause a module lock. And that's about what 10 I can say.

11 CO-CHAIR DIMITRIJEVIC: The cutsets for 12 the crane failure, for the module drop?

13 MS. POHIDA: I'm treating the crane and 14 the module lift as a lift fixture. As an integrated 15 lift mechanism.

16 CO-CHAIR CORRADINI: So, some combination 17 of an operator of commission plus a limit, some sort 18 of safety latch that doesn't latch?

19 MS. POHIDA: Yes.

20 MEMBER SKILLMAN: I was going after 21 momentum.

22 MS. POHIDA: I'm sorry?

23 MEMBER SKILLMAN: I was going after the 24 issue of momentum. Speed and height. The amount of 25 energy tacked to the floor or to a lateral bumper 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

245 1 some sort.

2 So, when you said you did the audit, and 3 you referred to the table that has the bridge trolley 4 and the height, I was asking whether or not those are 5 the numbers that you used in the audit. And I hear 6 you said, no, not quite, it would be something else.

7 MS. POHIDA: From what I saw in the 8 reactor building here, there wasn't direct correlation 9 between those cutsets and the speed limits, if you 10 will, that are documented in Chapter 9. Does that 11 help?

12 MEMBER SKILLMAN: If you used those speed 13 limits, that helps a lot.

14 MS. POHIDA: Yes.

15 MEMBER SKILLMAN: You did?

16 MS. POHIDA: I beg your pardon?

17 MEMBER SKILLMAN: You did use those speed 18 limits?

19 MS. POHIDA: It helped to substantiate the 20 drop probability those speed limits were documented in 21 Chapter 9 of the DCA. Does that answer your question?

22 MEMBER SKILLMAN: No. And I'm confused.

23 MS. POHIDA: Okay.

24 MEMBER SKILLMAN: We don't have to debate 25 it here.

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 MS. POHIDA: Okay.

2 MEMBER SKILLMAN: I don't understand.

3 MR. GALYEAN: Can I chime in and ask that 4 we defer some of these details to the closed session?

5 MEMBER SKILLMAN: Sure. Yes. I'm not 6 trying to be a bulldog here, I'm trying to understand.

7 You've got a 762 ton load, how fast is it going, how 8 high was it and what did you evaluate to determine 9 everything is okay? That's a big load.

10 MS. POHIDA: I understand.

11 MEMBER BLEY: Marie, this probably, I'm 12 not being aimed at you, but you and everybody else, 13 when I read the audit reports it mostly said, during 14 the course of the audits we reviewed documents and had 15 discussions with them.

16 In past design certs, we, on the 17 Committee, at least some of us have had the 18 opportunity to review the event trees as, I'm sorry, 19 the fault tress as well as the event trees, and to ask 20 the Applicant to manipulate the PRA models so we can 21 see importance of things.

22 During the audit, did you or some of the 23 other people, especially in level, in the internal 24 events activities, actually get a chance to do that?

25 Watch the PRA model being manipulated 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

247 1 changes and making sure by looking at some of the 2 fault trees that you were convinced things were 3 modeled well? And I don't know who to direct that to.

4 MS. POHIDA: Oh, for the reactor building 5 crane notebook, I just did an inspection of the 6 dominate cutsets.

7 MEMBER BLEY: Okay. But no looking at the 8 actual computer model of the PRA?

9 MS. NEUHAUSEN: Yes, we didn't look at the 10 actual.

11 MS. POHIDA: No.

12 MEMBER BLEY: Nobody did?

13 MS. NEUHAUSEN: No.

14 MEMBER BLEY: Every? Okay.

15 MS. POHIDA: All right. The risk 16 significance of the reactor building crane did result 17 in additional ITAACs. So there was, added was a rated 18 load test of the nuclear power module lifting fixture 19 and the module lift adapter.

20 And an inspection of the as-built nuclear 21 power module lifting fixture and the model lift 22 adapter.

23 There were also changes to the key 24 assumptions table to state that the interlocks and 25 limited switches will be functional during module 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 movement.

2 This drop probability needs to be 3 reevaluated for risk informed decision making, but the 4 analysis did meet what our needs for SRP Chapter 19 5 and ISG-028, that we recognize that this is risk 6 significant and that additional ITAACs were added and 7 additional detail was added to the PRA assumptions.

8 MEMBER BLEY: You said something, you said 9 it needs to reevaluated. you mean just these, you 10 need to look at the ITAACs and see that they're meet 11 or that you need to do something more on the analysis 12 of the drop frequency?

13 MS. POHIDA: In the future, if this PRA is 14 used for an application that's risk informed, then 15 this module drop probability will need to be 16 reevaluated.

17 MEMBER BLEY: I'm curious --

18 MS. POHIDA: We got --

19 MEMBER BLEY: -- as to what needs to be 20 reevaluated and why. You've seen the design now, so 21 what needs to be reinforced or added to in the future?

22 MS. POHIDA: As I learned from the audit, 23 the design is not finalized.

24 MEMBER BLEY: Okay. But it's more 25 finalized then I guess we had heard earlier, 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

249 1 MS. POHIDA: I beg your pardon?

2 MEMBER BLEY: At least there was a design 3 that you got to look at?

4 MS. POHIDA: There was a design that I 5 reviewed.

6 MEMBER BLEY: So it could be new and 7 that's what needs --

8 MS. POHIDA: As I understand it, it was 9 not finalized and it's being --

10 MEMBER BLEY: Okay.

11 MS. POHIDA: -- it's evolving.

12 MEMBER BLEY: Fair enough.

13 MS. POHIDA: May I continue? Okay.

14 CO-CHAIR CORRADINI: Please do.

15 MEMBER SKILLMAN: Do you want to?

16 (Laughter.)

17 MS. POHIDA: Yes, I do. I do.

18 MEMBER SKILLMAN: Let's get it done.

19 MS. POHIDA: I want to get to multi-module 20 drop. Okay.

21 For external events, once again, I'm 22 talking about single module drop. It's given a loss 23 of AC power, and that could be either from external 24 flooding or a high wind event.

25 There are redundant reactor building 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 breaks which will set and stop the motion. Each break 2 is rated to hold the maximum allowable crane load.

3 And it is assumed that this load can remain suspended 4 until AC power is restored.

5 So, on that basis, we believe that this 6 analysis is consistent with our guidance in ISG-028.

7 CO-CHAIR DIMITRIJEVIC: I'd like to ask 8 you something. The calculate, they have a flow stream 9 which calculates probability of the drop, which then 10 was changing initial event sequence by calculating 11 number of the movements through the air, right?

12 MS. POHIDA: Yes.

13 CO-CHAIR DIMITRIJEVIC: And what was that 14 time? What was the mission time for that probability?

15 MS. POHIDA: You mean the transit time?

16 CO-CHAIR DIMITRIJEVIC: Well, you have a 17 fault tree, which you analyze. What was the mission 18 time in that fault tree?

19 MS. POHIDA: I think I need to defer this 20 discussion because I can, to the closed session.

21 CO-CHAIR DIMITRIJEVIC: All right.

22 MS. POHIDA: If you're looking at duration 23 of time, the module is in transit.

24 CO-CHAIR DIMITRIJEVIC: Yes.

25 MS. POHIDA: And for the three module drop 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 scenarios, single module drop scenarios that were 2 evaluated in the DCA. So if I may defer to the closed 3 session I'd appreciate that.

4 CO-CHAIR DIMITRIJEVIC: Okay. All right.

5 MS. POHIDA: Okay. Thank you. All right, 6 next slide.

7 Now I'm going to multi-module risk. And 8 with our SRP, the applicant doesn't need to quantify 9 a CDF and a LRF. But the Staff needs to look at the 10 module drop assessment and ensure that there is no 11 vulnerabilities from a multi-module event that is 12 greater than an accident happening in a multi-unit 13 site.

14 We also need to look at to ensure that 15 there's no significant operator errors that could lead 16 to a multi-module core damage event. Okay.

17 All right, so we believe that the 18 applicant used a systematic process to evaluate multi-19 module risk. And I'm looking at internal events now.

20 And we believe, the assumptions on the 21 multi-module factors, they are based on engineering 22 judgment, but the design relies on an independent 23 module specific safety related system that prevents 24 and mitigate core damage.

25 In external events, they are evaluated 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 qualitatively.

2 Next slide. Okay, now I'd like to 3 continue with multi-module drop. Okay.

4 CO-CHAIR DIMITRIJEVIC: Do you mean, it's 5 not going to be multi-module drop because --

6 MEMBER SKILLMAN: Microphone.

7 CO-CHAIR DIMITRIJEVIC: -- more than one.

8 CO-CHAIR CORRADINI: No, she --

9 MS. POHIDA: Yes. As the module is being 10 moved, I'm trying to be careful that I don't trip into 11 proprietary space here, as the module is being moved 12 from its bay, in the operating bay to the refueling, 13 if it's dropped it can impact up to two operating 14 modules.

15 CO-CHAIR DIMITRIJEVIC: Right. So you 16 mean impact module drop. You said multi-module drop.

17 MS. POHIDA: You drop a module that's 18 being removed, that's being moved, excuse me, for 19 refueling, and it strikes an operating module.

20 CO-CHAIR DIMITRIJEVIC: Right. So it's 21 impact of multi-modules.

22 MS. POHIDA: Okay.

23 MEMBER BROWN: Can you back a slide?

24 MS. POHIDA: Oh, I'm sorry.

25 MEMBER BROWN: Could you go back a 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

253 1 MS. POHIDA: Yes.

2 MEMBER BROWN: Multi-module risk, there's 3 a whole another briefing on Chapter 21 --

4 MS. POHIDA: Yes.

5 MEMBER BROWN: -- tomorrow. So, it just 6 seems redundant relative to, you said go away, I mean, 7 there's not a problem.

8 MEMBER BLEY: It's kind of content free 9 compared to what we're hearing today --

10 MEMBER BROWN: Yes.

11 CO-CHAIR CORRADINI: So let's --

12 MEMBER BROWN: Well, I don't know. I 13 guess it's got more slides as opposed to five bullets, 14 or four bullets.

15 MEMBER BLEY: We talked about multi-module 16 risk earlier today.

17 MEMBER BROWN: I know, I didn't bring it 18 up then and --

19 MS. POHIDA: Oh, I'm sorry. Yes, the 20 results of the Applicant's analysis was brought up 21 this morning.

22 MEMBER BROWN: Yes, I understand that, but 23 that was a summary. But then there's still another 24 briefing on it tomorrow.

25 CO-CHAIR CORRADINI: But that one goes 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 beyond just this, that goes for --

2 MEMBER BROWN: Multi-module design 3 considerations.

4 CO-CHAIR CORRADINI: Right. Which 5 involves construction --

6 MEMBER BROWN: Not just drops.

7 CO-CHAIR CORRADINI: Not just drops.

8 MEMBER BROWN: This is just drops?

9 CO-CHAIR CORRADINI: Correct.

10 MEMBER BROWN: Okay, then I'll shut up 11 until tomorrow. They're just going to get ready, 12 that's all.

13 MS. POHIDA: Oh, I believe, we're talking 14 about multi-module risk, the overview. What I just 15 covered was the module adjustment factors that the 16 Applicant used to come up with a multi-module CDF and 17 LRF for internal events. And external events were 18 evaluated qualitatively.

19 MEMBER BROWN: Good enough. So tomorrow 20 was going to be addressing shared systems and all that 21 other type of stuff, correct?

22 MS. POHIDA: Yes. Yes, that's the plan.

23 MEMBER BROWN: Okay, I'll wait.

24 MS. POHIDA: Thank you. Okay. All right.

25 In DCA, in Revision 1, it's stated in Chapter 19.1.7, 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 that if a module that's being moved for refueling 2 drops on an operating module near the top, it could 3 damage DHRS piping or the heat exchangers.

4 In Revision 2, there was an addition, and 5 it states is, that additional pipe breaks may occur 6 that could lead to a CVCS line break outside of 7 containment.

8 So, we asked and RAI, and the RAI is, once 9 again, it's to make sure our risk insights are 10 complete, is what pipes are assumed to fail, which is 11 CVCS, DHRS and the cavity flood and drain system.

12 And I guess more importantly is that if 13 you have a strike to an operating module that's 14 sufficient to cause pipe breaks, is the capability of 15 the containment isolation valves to close compromised.

16 And so, we are evaluating this event. This RAI is 17 still under Staff evaluation.

18 CO-CHAIR CORRADINI: So, can I ask the 19 question a little bit differently?

20 MS. POHIDA: Thank you.

21 CO-CHAIR CORRADINI: What we heard from 22 the Applicant prior to that was there were other 23 accident scenarios that in some sense bounded the 24 damage from other approaches. You want to make sure 25 that so that they analyze essentially what would occur 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 with the module drop? That's what I hear you saying.

2 MS. POHIDA: We want to make sure that 3 those conclusions, we can confirm those conclusions.

4 CO-CHAIR CORRADINI: Okay. Got it.

5 MS. POHIDA: Okay. All right, if I may 6 continue, I'm going to talk about other external 7 hazards.

8 For external flooding, the DCA states that 9 there are no flooding penetrations, such as flood 10 doors, that are credited in the analysis, so therefore 11 no flooding penetrations were found to be risk 12 significant.

13 For the high winds' assessment, the Staff 14 verified that all important accident mitigation 15 features are housed within a seismic Cat 1 structure, 16 the reactor building structure. And thus, are 17 protected from the effects of high winds.

18 MEMBER BLEY: I'm a little confused by the 19 first statement, which is theirs and I didn't ask them 20 about it.

21 MS. POHIDA: Oh, okay.

22 MEMBER BLEY: Since there are no flood 23 doors, we don't have to worry about flooding. Well, 24 flood doors are designed to keep the water out.

25 There are other doors, and are they all 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 sufficiently above grade that we don't have to worry 2 about ingress of water from those or from ventilation 3 systems?

4 I don't know where they're going to plant 5 one of these things but.

6 MS. POHIDA: I'm going to have to defer to 7 the Applicant on those specifics.

8 MEMBER BLEY: Okay. I didn't ask them 9 because we didn't get to this level of detail.

10 MS. POHIDA: I asked an RAI on this, on 11 the status of anything, if the operators are required 12 to do anything for an external. You know, any doors, 13 penetrations, anything needed to change state.

14 And as a result, there was an addition to 15 the DCA that no flooding penetrations, external 16 flooding penetrations were found to be risk 17 significant. But I defer to the Applicant if there is 18 more detail that's needed.

19 MEMBER BLEY: I apologize for not asking 20 that this morning, but if you guys can respond to it.

21 It just snapped for me there are no flood doors.

22 Well, flood doors are designed to keep the water out.

23 A lot of plants have sand bags they put up 24 to keep the water out. You have doors somewhere to 25 get into this dang on thing and how do you know 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

258 1 are not under any risk of flooding or have ventilation 2 ducts, that sort of thing?

3 MR. GALYEAN: We treat flooding as 4 basically a loss of all AC power, okay. So we made 5 certain assumptions --

6 MEMBER BLEY: So as the water comes in, it 7 may short things out and lead you to something 8 approximating a loss of offsite power?

9 MR. GALYEAN: Exactly. Exactly.

10 CO-CHAIR DIMITRIJEVIC: You haven't --

11 MR. GALYEAN: We simply assume that if a 12 flood occurs, a beyond design basis flood, that it 13 simply results in a loss of all electrical --

14 MEMBER BLEY: And you did some --

15 CO-CHAIR DIMITRIJEVIC: With point one, 16 you have a factor there, which I asked last time, the 17 point one, in ten percent cases you assume results in 18 loss of loss of power. External flood.

19 MR. GALYEAN: Right.

20 MEMBER BLEY: So ten percent of external 21 --

22 MR. GALYEAN: But it's --

23 MEMBER BLEY: -- floods or the floods of 24 both --

25 MR. GALYEAN: The flood 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

259 1 MEMBER BLEY: -- design flood?

2 MR. GALYEAN: I think the design basis 3 flooding. Ten percent probability that given you have 4 a flood that exceeds the design basis that results in 5 loss of all AC power.

6 MEMBER BLEY: Are the doors, the main 7 access ways and equipment access areas, are those well 8 above grade or where do they sit?

9 MR. GALYEAN: Obviously, we don't have a 10 site.

11 MEMBER BLEY: If you, if they were up a 12 little high, then ten percent of the floods above 13 design basis floods is kind of reasonable. If they're 14 right at grade level and design flood is anywhere near 15 grade level, there's not such a good assumption, I 16 think. But that's where we are.

17 MR. GALYEAN: Right.

18 MEMBER BLEY: It's an assumption --

19 MR. GALYEAN: That's right, it's an 20 assumption.

21 MEMBER BLEY: -- and the COL is going to 22 have to take a look at that.

23 MR. GALYEAN: That's right.

24 MEMBER BLEY: And see if it's reasonable 25 for them.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

260 1 MR. GALYEAN: Once there's a site that can 2 be evaluated for --

3 MEMBER BLEY: Okay. So this ten percent 4 thing is --

5 MR. GALYEAN: -- floods.

6 MEMBER BLEY: -- just an artifice for now?

7 MR. GALYEAN: That's right.

8 MEMBER BLEY: Okay. Okay, for now.

9 MS. POHIDA: Well, that concludes my 10 presentation, are there any more questions?

11 MEMBER BLEY: Could you finish the high 12 wind?

13 MEMBER SKILLMAN: I thought she did.

14 MS. POHIDA: Oh, yes, I am missing a 15 conclusion here. So, for both external flooding and 16 the high winds analysis, we found the analysis to be 17 consistent with our Staff guidance, with our guidance 18 in ISG-028 and SRP Chapter 19. Thank you.

19 MEMBER BLEY: So I may as well have fun 20 with this. I don't know of anybody whose included it 21 in their PRA, but I don't know why not.

22 Some of the existing plants, and you 23 usually think of things like a large dry containment 24 as being almost a Faraday cage with all the steel in 25 it, but some of them have had lightning be brought 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 inside through some penetration can really do some 2 bizarre things inside the containment.

3 Did you, any chance you looked at, is that 4 a possibility here?

5 MS. POHIDA: That was not part of my 6 review, no. The effect of lightning.

7 MEMBER BLEY: Yes.

8 MS. POHIDA: Yes.

9 MEMBER BLEY: You know, for just an aside, 10 for some weapons, bunkers, you see those things that 11 the Military has, they thought they had Faraday cages, 12 but they had a metal support or something else that 13 extended inside and outside of containment, and if the 14 lightning hit that, it can kind of bypass the Faraday 15 cage and get inside. And once it gets inside it just 16 jumps all around and burns stuff. And that's 17 interesting.

18 At least one of our large dry containments 19 has had that same kind of event with some pretty 20 spectacular damage inside. You can design around it.

21 But personally, I don't know anybody whose 22 looked hard at that, but it's an interesting failure 23 mode.

24 MS. POHIDA: Would you like us to follow-25 up on 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

262 1 CO-CHAIR CORRADINI: He's trying to have 2 fun with you.

3 MS. POHIDA: Okay.

4 (Laughter.)

5 MEMBER BLEY: Well, I'm a little bit 6 trying to have fun with you but it's a fairly serious 7 thing and it's been absent from most of our PRAs.

8 MS. POHIDA: Okay.

9 MEMBER BLEY: Maybe all of them. And I 10 don't why it's absent. Except maybe most of us think, 11 yes, it's a Faraday cage, nothing can get inside.

12 There's a lot been learned about lightning 13 in the last 20, 30 years that wasn't known before. So 14 you can give it a little thought and come back to us 15 the next time around --

16 MS. POHIDA: Okay.

17 MEMBER BLEY: -- that would be 18 interesting.

19 MS. NEUHAUSEN: All right. There's a 20 still a couple more slides, so, next slide.

21 Okay. This is at power and internal fire 22 and internal flood. So for the internal fire PRA, we 23 focus on the assumptions used in the FPRA and the 24 consistency with methods in NUREG-CR 6850.

25 And so we found that because this is the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

263 1 DCA, many tests were omitted or simplified analyses 2 were used. So, for example, specifics of cable 3 routing, ignition sources, target locations weren't 4 known.

5 And so, we still now reviewed what the 6 assumptions were and made sure that the assumptions 7 were included in these tables. And found FPRA 8 sufficiently consistent with the SRP and ISG.

9 Similarly, for the internal flood, a lot 10 of design details are unknown. Staff considered that 11 the design is less dependent on active systems and 12 that the mitigating functions of the active systems 13 aren't credited for flood in the reactor building.

14 Next slide.

15 MEMBER REMPE: Could you move the 16 microphone a little closer to you?

17 MS. NEUHAUSEN: I can.

18 MEMBER REMPE: Some of us are old and hard 19 hearing.

20 MS. NEUHAUSEN: Sorry. I have two in 21 front of me, so.

22 And then the PRA based seismic margins.

23 For the PRA based SMA, we focused on the review of the 24 scope of SSCs included in the fragility evaluation and 25 the analysis methods that were used to determine 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 seismic fragility.

2 All of the SSCs were included in the 3 fragility evaluation. There were two methods, 4 conservative deterministic failure margins and 5 separation of variables, which are endorsed by the SRP 6 were used for the seismic fragility for the PRA 7 critical SSCs.

8 And then the component boundary includes 9 all failure mechanisms affecting component functions.

10 So, Staff found that the plant-level HCLPF capacity 11 demonstrated adequate margin in accordance with the 12 SECY 9387 in the SRP.

13 Next slide. So, due to the open items 14 mentioned earlier, Staff hasn't made a finding on the 15 acceptability of the PRA for NuScale's Chapter 19 yet.

16 And we can take any more questions.

17 MEMBER BLEY: Marie just told us about one 18 of the open items. Is it a short list? Can you give 19 us the other ones?

20 MS. NEUHAUSEN: Yes, we already talked 21 about both of them. So, Marie talked about one and 22 then OI spoke to the, it was the containment.

23 MEMBER BLEY: And that's all there are?

24 MS. NEUHAUSEN: Yes. For 19.1 there's 25 just those two.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

265 1 MEMBER BLEY: Okay.

2 CO-CHAIR DIMITRIJEVIC: I have a couple 3 questions on something that you stated in SSC. That 4 which I have a problem with the sum of the statement.

5 But boundaries related to Level 2, are you 6 going to discuss the, are we done with the, all right, 7 let me just ask you this --

8 MS. NEUHAUSEN: That's our whole 9 presentation.

10 CO-CHAIR DIMITRIJEVIC: What?

11 MS. NEUHAUSEN: That's our whole 12 presentation.

13 CO-CHAIR DIMITRIJEVIC: That's your whole 14 presentation? You're done with the --

15 MS. NEUHAUSEN: There's the closed session 16 on the ATWS, but --

17 CO-CHAIR DIMITRIJEVIC: Yes, okay. So let 18 me just ask you, on the Page 19.16, in the end of the 19 third paragraph says, the LOCA inset containment 20 initiating event end loop are also embedded by very 21 significant initiating event because they meet the 22 least achievement mode to failure.

23 The initiating events are not ran by these 24 achievements, they're only ran by Fussell-Vesely, so 25 I don't know how that found the place there 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

266 1 least achievement doesn't make sense for an initiating 2 event.

3 MS. NEUHAUSEN: We'll take it back and 4 make sure that it's correct.

5 CO-CHAIR DIMITRIJEVIC: All right.

6 Achievement means what you will put in frequency to 7 one, it just doesn't make sense.

8 The other thing, which I have on the Page 9 19.22, I didn't understand these statements. It says 10 in the first paragraph, 19.1457 said, Applicant 11 defines CCFP as the ratio LRF to CDF to the solvent 12 certainties regarding potential failure of the RPV and 13 CNV bottom half. What does that mean?

14 I mean, isn't the CCFP definition, I mean, 15 what does this statement mean?

16 MR. AYEGBUSI: I'm sorry, what page is 17 that?

18 CO-CHAIR DIMITRIJEVIC: The Page is 22 in 19 the first section, 19.1457.

20 MR. AYEGBUSI: 19.1457.

21 CO-CHAIR DIMITRIJEVIC: Yes.

22 MS. POHIDA: May I answer that?

23 MR. AYEGBUSI: Go ahead.

24 MS. POHIDA: CCFP is Conditional 25 Containment Failure Probability. And what I, I wrote NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

267 1 that, what I'm referring to in there is when, 2 remember, containment event tree is really simple.

3 CO-CHAIR DIMITRIJEVIC: Okay.

4 MS. POHIDA: Its, you know, all severe 5 accident phenomena, other than severe accident 6 induced, steam generator tube rupture was screened.

7 Okay.

8 And so, that tree was, if you're looking 9 at the screened phenomena and you're looking at its 10 likelihood of causing containment failure, that will 11 be discussed tomorrow. The uncertainty regarding 12 severe accident phenomena --

13 CO-CHAIR DIMITRIJEVIC: Okay, but I'm 14 going --

15 MS. POHIDA: -- as it impacts containment 16 failure and RPV failure.

17 CO-CHAIR DIMITRIJEVIC: Okay, but it's not 18 the CCFP failure or review of the LRF to the CDF, how 19 does this resolve essentially?

20 I just don't understand what, I mean, 21 you're not trying to bring in the definition of CCFP 22 because that's also sort of controversial issue, 23 right?

24 MS. POHIDA: Oh, okay. I'm sorry, I 25 believe I misspoke.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

268 1 Yes. We have the commission goals for new 2 reactors, and we have a subsidiary goal of having a 3 conditional containment failure probability of .1.

4 CO-CHAIR DIMITRIJEVIC: Right.

5 MS. POHIDA: But in this, but for this 6 application, instead of looking at because the Staff 7 is going to explain tomorrow in the 19.2 discussion, 8 there's uncertainty with a severe accident 9 phenomenology.

10 What we're talking about here is using, 11 that definition is not conditional containment failure 12 probability, it's meant to be the ratio of the LRF to 13 the CDF.

14 CO-CHAIR DIMITRIJEVIC: Well, what do you 15 consider to be condition of containment failure 16 probability?

17 MS. POHIDA: I'm sorry?

18 CO-CHAIR DIMITRIJEVIC: How do you define 19 conditional containment --

20 CO-CHAIR CORRADINI: They don't.

21 CO-CHAIR DIMITRIJEVIC: -- failure 22 probability?

23 MS. POHIDA: For this application it's --

24 CO-CHAIR DIMITRIJEVIC: Just in general, 25 how do you define 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

269 1 MS. POHIDA: Oh. I'm going to take an 2 example for shut down, that's typically what I do.

3 It's the likelihood of an operator failing to isolate 4 containment.

5 CO-CHAIR DIMITRIJEVIC: Go ahead. I know 6 every isolation failure is considered LRF --

7 MS. POHIDA: Yes.

8 CO-CHAIR DIMITRIJEVIC: -- but I just 9 discussed, okay, even if you consider CCFP always to 10 be LRF of a CDF, or there is some other definition and 11 they're using this as a, to cover for something.

12 I doubt, I mean, we can consider CC, 13 containment failure probability to be annually, right?

14 It doesn't have to be large release.

15 So let's say that we arguable hear the 16 containment failure probability is just large release, 17 right?

18 MS. POHIDA: Okay.

19 CO-CHAIR DIMITRIJEVIC: I assume that you 20 can reach this agreement before --

21 MS. POHIDA: I think I'm going to defer to 22 Jason on this.

23 CO-CHAIR CORRADINI: Yes, I was going to 24 say, I think the Staff is going to help her. James.

25 MR. SCHAPEROW: So, different, oh, Jason NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

270 1 Schaperow of NRO Staff. Different applicants, over 2 the years, have actually used different definitions 3 for what they consider conditional containment failure 4 probability.

5 For example, ABWR actually used two 6 definitions. They use one which is actually given a 7 core damage accident, the actual chance that the 8 containment fails. Like there's an actual hole in it 9 or a tear or something.

10 But they also used a dose base definition 11 that if you have a severe accident that dose at the 12 site boundary is more than 25 rem.

13 So NuScale's application uses a definition 14 of containment failure, given a core damage, 15 definition of containment failure that we have a large 16 release. And they gave their large release definition 17 this morning.

18 Tomorrow morning we'll go into more detail 19 why that works for the NuScale design review. Because 20 of some uncertainties with regard to in-vessel 21 retention of core debris sitting in the lower plenum 22 of the reactor.

23 CO-CHAIR DIMITRIJEVIC: So, to make a, to 24 be clear, that's not what I mean. I don't want to 25 bring this definition of the CCFP.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

271 1 Let's assume they're using large release 2 to CDF, right, that's fine with me. This is the least 3 conservative definition of containment failure 4 probability.

5 So I don't understand this sentence that 6 says, this ratio resolve uncertainties. Why would the 7 least conservative definition of CCFP resolve any 8 uncertainties? That's what I don't understand.

9 So this is resolve uncertainties is which 10 I have a problem.

11 MS. HAYES: This is Michelle Hayes. I'm 12 with the staff as well. We're going to look at -- I 13 understand what you're saying. We're going to look at 14 that. I think it's a grammatical thing.

15 MEMBER BLEY: I have two kind of quick 16 things. One is for 30 years LERF was L-E-R-F and now 17 it's become LRF which leaves me confused half the 18 time.

19 The other is I'm going to have to look at 20 this some more. The staff in their audits reviewed 21 many of NuScale's engineering reports. One we talked 22 about earlier on the ECCS valves and I think you're 23 going to talk more about that in closed session.

24 Another that we heard about this morning 25 I'm not sure. I couldn't find in the audit report if NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

272 1 you had reviewed this or not. There's an engineering 2 report on passive system uncertainties. And I don't 3 know if I got that right. PSSR, PSSR, passive safety 4 system reliability. That's the other one. But I know 5 I would like to hear more about it. If the staff has 6 anyone who can talk about that second one, I'd be very 7 interested.

8 There are probably others that we'll ask 9 about later. I don't think they rise to the 10 importance of these two, but you can expect at least 11 one member is going to ask again, maybe at the full 12 committee meeting, then sometime between then and when 13 we come around in the next phase we can hear something 14 about those. That's that. PSSR.

15 CO-CHAIR CORRADINI: Anything else?

16 MEMBER MARCH-LEUBA: Yes, I have some.

17 MR. AYEGBUSI: First of all, so I guess, 18 when you say engineering reports, that was confusing 19 earlier. I think a lot of those refer to NuScale's 20 PRA notebooks, right?

21 MEMBER BLEY: Some might have been 22 notebooks and others were like ERPO 60 dash something 23 else rev something and I understood that those were 24 some kind of engineering reports. There were 25 engineering reports on selection of initiating 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

273 1 and a bunch of other things that were discussed this 2 morning. And I think I saw that a number of those who 3 are in things people listed as being reviewed on the 4 electronic system during your --

5 MR. AYEGBUSI: So we discussed the passive 6 safety system reliability evaluation earlier today and 7 we talked about our review and what we found and the 8 iterations with the applicant.

9 Did you have specific questions that you 10 would like us to address?

11 MEMBER BLEY: No, I want to know what was 12 in your internal report on your audit of that report.

13 MR. AYEGBUSI: Internal report?

14 MEMBER BLEY: Well, we heard earlier that 15 on at least the ECCS valves somebody back here was 16 looking through the audit report by the guy who did 17 the audit and had a lot of detail on what they looked 18 at and --

19 MS. NEUHAUSEN: Yes, those are just 20 personal notes, not any sort of internal report.

21 MEMBER BLEY: Well, they're still a sort 22 of internal report. It had information on your audit 23 for heaven's sake. So that's what I'm getting at and 24 many of those items that you looked at on the audit 25 were based on, at least when I read the audit report NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

274 1 were based on your reviewing and discussing with the 2 applicant their detailed engineering calculations for 3 various technical topics.

4 So those two areas I would like to hear 5 some more details on and they're probably going to be 6 a couple of others where under other conditions I'd 7 just get to look at those reports and read them myself 8 and see if I had any problems.

9 MS. POHIDA: May I ask, did you want more 10 details of the inputs to the study?

11 MEMBER BLEY: I want to understand why the 12 staff felt that the approach NuScale has taken to 13 developing failure rates for the ECCS valves and the 14 process they use for considering the uncertainties in 15 passive systems are adequate. And I don't want to 16 hear that we read them and thought they were adequate.

17 I'd like to hear exactly technically why. And the 18 real reason is because I haven't had a chance to read 19 the reports. I don't know what's in them. I can't 20 make my own judgment about them, so I'd like to more 21 thoroughly understand what issues the staff pursued 22 and why you have been happy with those right now, 23 those two issues.

24 CO-CHAIR CORRADINI: Jose.

25 MEMBER MARCH-LEUBA: Yes, I wanted to save NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

275 1 you some time and I'm going to say for the record that 2 I don't need a closed session for ATWS. I've seen the 3 data that Peter is going to show and the rest are not 4 not interested in it.

5 But I would like for the staff to confirm 6 that I understand their position so -- the microphone 7 is there.

8 MEMBER BLEY: Well, I really think your 9 assumption is that nobody else is interested is wrong.

10 (Laughter.)

11 And I think that if he would get into some 12 of that would be very useful.

13 CO-CHAIR CORRADINI: Yes, why don't we 14 defer Jeff to when we see some numbers. I want to see 15 some graphs.

16 MEMBER MARCH-LEUBA: It's important but 17 the numbers make no difference. The staff position is 18 that 10 CFR 5062, the ATWS rule, is a capability-based 19 rule. It's not a performance-based rule. Thou shalt 20 have ARI, turbine trip, and a number of things. And 21 once you have those capabilities implementing your 22 plan, you satisfy the rule. You don't need to do any 23 analysis. Is that your position?

24 MR. SCHMIDT: This is Jeff Schmidt from 25 Reactor Systems again.

NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

276 1 So there are multiple parts of that rule 2 like we're talking about where they did take an 3 exemption to the turbine trip.

4 MEMBER MARCH-LEUBA: Sure.

5 MR. SCHMIDT: So that's --

6 MEMBER MARCH-LEUBA: But your evaluation 7 --

8 MR. SCHMIDT: But the 15.8 evaluation 9 you're referring to?

10 MEMBER MARCH-LEUBA: Right. So I was 11 going to argue with Pete that he's done a cycle 12 specific ATWS calculation when he should have done a 13 bounding ATWS calculation for the FSAR, Chapter 15.8 14 on 19. But it's not required because they satisfy the 15 rule with a coolant design and they don't have to do 16 an analysis.

17 MR. SCHMIDT: That's right. So they 18 satisfy the rule as written except for the exemption 19 which --

20 MEMBER MARCH-LEUBA: Correct.

21 MR. SCHMIDT: Yes, that's correct.

22 MEMBER MARCH-LEUBA: And the fact that we 23 have -- they have run some calculations with -- I 24 haven't seen and Pete has run some calculations which 25 I have seen that show that for cycle-specific numbers NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

277 1 you're safe. It's icing on the cake. But it's not 2 required to add additional conservatism and bounding 3 assumptions.

4 MR. SCHMIDT: I think it's still 5 informative and I haven't read Pete's report in a 6 while at this point because he did it quite a while 7 ago, but I think it's still very representative of 8 what the module would do even on a reload basis.

9 So you're right. MTC does change with the 10 reload. Right and MTC has a reasonably big effect on 11 the response.

12 MEMBER MARCH-LEUBA: Okay.

13 MR. SCHMIDT: But you know, what I heard 14 Pete say and maybe Pete can speak to it, too, I think 15 is that the relief capability of the safety relief 16 valve is pretty substantial and that protects your 17 over pressure from the RCS being over pressure. So it 18 allows the system to basically come to equilibrium 19 with a heat decay removal system.

20 MEMBER MARCH-LEUBA: Okay, go back, sit 21 down. We will have a closed session and we will talk 22 about it.

23 MR. SCHMIDT: Okay.

24 CO-CHAIR DIMITRIJEVIC: I have one comment 25 here also, Mr. Chair, which is very important and 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

278 1 don't really know how I want to treat this from here.

2 On the Chapter 9 in the SER on page 19.15, 3 you said the following: The uncertainty base on the 4 CDF reported by applicant accounts for only parameter 5 uncertainty not model uncertainties. Therefore, the 6 staff finds that a design state uncertainty could be 7 very large. Okay? You are right. It should be very 8 large. That's not what applicant reported. Applicant 9 reported very actually small uncertainty and if you 10 presented the numerical data with 95 and 5 percent, 11 you will see because there is only base unremittant 12 parameter uncertainty.

13 So my question is even with the latch 14 potential uncertainty, velocity estimated a flight 15 barely initiating the same effort to reduce or 16 eliminate the contribution to CDF found in the 17 previous PRA. So my question is, first, do you offer 18 even the table you presented in the beginning you can 19 only use mean values. That doesn't give uncertainty 20 range. And if you are expecting large uncertainty and 21 you didn't see large uncertainty, why didn't you 22 question that?

23 MS. NEUHAUSEN: I think that what the 24 staff was trying to convey here is just that even with 25 large uncertainties and we're measuring this against NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433

279 1 the Commission's safety goal. And so NuScale reported 2 their uncertainties. We did our review, but there's 3 really -- we can't ask any questions. And so even 4 though there may be -- even if there is more 5 uncertainty than is shown in those -- what they've 6 included, it's still not going to reach the safety 7 goal as we're measuring that against.

8 CO-CHAIR DIMITRIJEVIC: That's why I don't 9 really know. I just want to say you say here large 10 uncertainty, but it's not true. You could say 11 surprisingly we saw very low uncertainties with all of 12 these uncertainties. Something is -- I'm not sure 13 really where this leads because you would assume as 14 they go through the collar, this uncertainty will 15 reduce, but you don't see it. I mean you cannot 16 really go back to do the more complete uncertainty or 17 you don't want to question others factors of the sum 18 of those elements which you know are very 19 questionable.

20 CO-CHAIR CORRADINI: There's somebody --

21 staff wants to say --

22 MS. HAYES: This is Michelle Hayes. And 23 I think that paragraph that you're referring to is 24 we're trying to say there's more uncertainties than 25 they included in their parametric uncertainties 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

280 1 that even given model uncertainties that haven't been 2 identified yet, there's still a large margin.

3 Your point about not going further into 4 the 90.55 percent, we can take that back and come back 5 to you on that. I think it has to do with the large 6 margin and that our finding is that it meets the 7 Commission goals. We're not certifying every number 8 that they provide. We're making sure they follow the 9 process and they meet the Commission goals and we did 10 that with -- while recognizing there's a large margin 11 even given this extreme CCF sensitivity study.

12 CO-CHAIR CORRADINI: I think she's 13 agreeing with you.

14 CO-CHAIR DIMITRIJEVIC: I think you don't 15 want to say we have a zero risk and we know this very 16 well. Yes.

17 MS. HAYES: We are not saying that, but we 18 are saying it's the same as the safety goals and we 19 know that very well.

20 CO-CHAIR DIMITRIJEVIC: All right. Okay.

21 We meet the safety goals and we know that very well, 22 that's true.

23 MS. HAYES: That's what we're trying to 24 say.

25 CO-CHAIR CORRADINI: Other comments by 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

281 1 committee? I want to get public comments if we could 2 before we go into closed session.

3 Okay, so as we open up the phones lines 4 and people can speak from the phone, is anybody in the 5 room that wants to make a comment? If not, I have my 6 wanted helper here, Mr. Snodderly.

7 MR. SNODDERLY: Is there anyone on the 8 public line? Good afternoon, Marvin. Would you like 9 to make a comment?

10 MR. LEWIS: Not at this time. I'm torn.

11 Yes, I would like to make a comment.

12 MR. SNODDERLY: Please do.

13 MR. LEWIS: Respectfully. Respectfully.

14 I'm trying to be most respectful, but I'm having a 15 very difficult time of it.

16 The reason is this. You're talking 17 probabilistic. Yes, there is such history as 18 probability. However, I must remind you Three Mile 19 Island, 1979; Fukushima, a few years ago. That day 20 when those accidents happened, probabilistically it 21 was a hundred percent because it did happen and you 22 refuse to see that there is a glitch in your 23 probabilistic assessment and analysis where it 24 actually happens, the accident actually happens and 25 has happened. 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

282 1 MR. SNODDERLY: So Marvin, I appreciate 2 that comment and I encourage you on June 18th and 3 19th, we'll be reviewing Chapter 6 and 15 that deal 4 more with the deterministic in a given design-basis 5 accident and it complements the probabilistic. So 6 hopefully, you'll be able to join us for that and 7 again, we appreciate your comment.

8 Are there any other -- is there anyone 9 else on the public line who would like to make a 10 comment?

11 MS. FIELDS: Yes. This is Sarah Fields.

12 One of my comments is it was very hard to hear some of 13 the speakers. There's that difficulty. And then in 14 one of the discussions appeared to be disconnected 15 from the slide presentations that were sent out. So it 16 was -- you kind of get lost because you think that the 17 NRC staff is reading from something, but whatever 18 they're reading from is not part of the slide 19 presentation and for someone who is not a technical 20 person, it's very easy to get lost and I appreciate 21 that you will be putting a transcript up as soon as 22 possible. It would be very useful to get a handle on 23 what the full discussion was about.

24 MR. SNODDERLY: So Sarah, this is Mike 25 Snodderly. I appreciate your comment and you 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

283 1 right. We could -- we'll make an announcement to make 2 sure you are aware of what slide set we're on. So I 3 understand because I sent you I think three sets of 4 slides, so it was probably hard for you to figure out 5 what slide set we were on.

6 MS. FIELDS: Not only the sets, but 7 particularly where you think you're on the right page 8 and you are, and then it goes to discussion, but the 9 discussion isn't exactly what's on the next page.

10 There's just so much more coming from the staff that 11 is not really reflected on the slides. I mean I know 12 they're hard to put together and they have a lot to 13 say and the discussion will go this way and that. So 14 the transcript will be important. Thank you.

15 MR. SNODDERLY: You're welcome. Thank you 16 for your time.

17 Are there any other comments from people 18 on the line?

19 CO-CHAIR CORRADINI: Okay, thank you very 20 much. Let's close the public line and we're going to 21 take a break and go into closed session. You want to 22 do that? I'm sorry.

23 MEMBER BLEY: There's nobody there.

24 CO-CHAIR CORRADINI: Why don't we take a 25 break and we'll come back. Fifteen minute break at 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

284 1 of.

2 MR. SNODDERLY: And we'll be in 3 proprietary closed session so we close this line, and 4 clear --- make sure. So anyone that returns 5 afterwards, you'll have to be a member of NuScale or 6 you have to be a member of the staff or a need to know 7 and there will be no members of the public.

8 (Whereupon, the above-entitled matter went 9 off the record at 3:17 p.m.)

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

LO-0519- 65373 May 09, 2019 Docket No.52-048 U.S. Nuclear Regulatory Commission ATTN: Document Control Desk One White Flint North 11555 Rockville Pike Rockville, MD 20852-2738

SUBJECT:

NuScale Power, LLC Submittal of Presentation Materials Entitled ACRS Subcommittee Presentation: NuScale FSAR Chapter 19, Probabilistic Risk Assessment and Severe Accident Evaluation, PM-0519-65372, Revision 0 The purpose of this submittal is to provide presentation materials for use during the upcoming Advisory Committee on Reactor Safeguards (ACRS) NuScale Subcommittee meeting on May 14 and 15, 2019.

The materials support NuScales presentation of Chapter 19, Probabilistic Risk Assessment and Severe Accident Evaluation, of the NuScale Design Certification Application. is the nonproprietary presentation entitled ACRS Subcommittee Presentation: NuScale FSAR Chapter 19, Probabilistic Risk Assessment and Severe Accident Evaluation, PM-0519-65372, Revision 0.

This letter makes no regulatory commitments and no revisions to any existing regulatory commitments.

If you have any questions, please contact Rebecca Norris at 541-452-7539 or at rnorris@nuscalepower.com.

Sincerely, Zackary W. Rad Director, Regulatory Affairs NuScale Power, LLC Distribution: Robert Taylor, NRC, OWFN-7H4 Michael Snodderly, NRC, TWFN-2E26 Gregory Cranston, NRC, OWFN-8H12 Samuel Lee, NRC, OWFN-8H12 Rani Franovich, NRC, OWFN-8H12 : ACRS Subcommittee Presentation: NuScale FSAR Chapter 19, Probabilistic Risk Assessment and Severe Accident Evaluation, PM-0519-65372, Revision 0 NuScale Power, LLC 1100 NE Circle Blvd., Suite 200 Corvallis, Oregon 97330 Office 541.360-0500 Fax 541.207.3928 www.nuscalepower.com

LO-0519- 65373 :

ACRS Subcommittee Presentation: NuScale FSAR Chapter 19, Probabilistic Risk Assessment and Severe Accident Evaluation, PM-0519-65372, Revision 0 NuScale Power, LLC 1100 NE Circle Blvd., Suite 200 Corvallis, Oregon 97330 Office 541.360-0500 Fax 541.207.3928 www.nuscalepower.com

NuScale Nonproprietary ACRS Subcommittee Presentation:

NuScale FSAR Chapter 19 Probabilistic Risk Assessment and Severe Accident Evaluation May 14 and 15, 2019 1

PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Chapter 19 Section Title Comment Probabilistic Risk Assessment and Severe 19.0 Overview Accident Evaluation 19.1 Probabilistic Risk Assessment Level 1, 2 Thermal hydraulic &

19.2 Severe Accident Evaluation phenomenological analyses 19.3 Regulatory Treatment of Nonsafety Systems No RTNSS SSCs Strategies and Guidance to Address Loss of Addressed in 19.4 Large Areas of the Plant due to Explosions and Chapter 20 Fires Adequacy of Design Features and Functional 19.5 Capabilities Identified and Described for Overview Withstanding Aircraft Impacts 2

PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary ACRS Subcommittee Presentation:

NuScale FSAR Chapter 19.0 and19.1 Probabilistic Risk Assessment May 14, 2019 3

PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Presentation Team Sarah Bristol Supervisor, Probabilistic Risk Assessment Etienne Mullin Probabilistic Risk Analyst Bill Galyean Probabilistic Risk Assessment Consultant Rebecca Norris Supervisor, Licensing 4

PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Section 19.0: Probabilistic Risk Assessment and Severe Accident Evaluation

• Developed in accordance with applicable regulations, regulatory guidance, and industry standards

• Performed for a single module

• Considered all modes of operation for both internal and external initiating events

• Provides risk insights including those related to risk-significant systems, components, human actions, relevant programs (e.g., RTNSS, SAMDA), and multiple module risk

• PRA demonstrates that the NuScale design exceeds NRC safety goals with significant margin 5

PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Section 19.1: Probabilistic Risk Assessment

• Objective: to assess risks associated with all modes and all hazards for a single NuScale Power Module (NPM)

• Level-1 (CDF) and Level-2 (LRF)

- Full power, internal events (FP-IE)

- Low power and shutdown (LPSD)

• Include crane failure

- Internal fire

- Internal flood

- External flood

- High winds

- Seismic margins assessment (PRA based) 6 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary PRA Quality Process

• NuScale PRA quality procedure

- Follows guidance provided in NRC Regulatory Guide 1.174

• NuScale PRA follows guidance provided by

- ASME/ANS PRA standard

- NRC Regulatory Guide 1.200 and Interim Staff Guidance 028

• Each PRA notebook reviewed for conformance with PRA standard

- Self-assessment documented by notebook authors

- Self-assessment independently reviewed/verified by outside consultants

• PRA reviewed by outside, independent expert panel 7

PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary PRA Expert Peer Review Group

• Separate and independent from PRA standard self-assessment reviewers

• Expert review group members:

- George Apostolakis (chairman)

- Mark Cunningham

- Rick Grantom

- Dave Moore

- Per Peterson 8

PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Expert Panel Findings

• Review group authored a final report

- No major concerns or objections

• Minor points that were raised include

>> NuScale multi-module risk approach represents an important first step in advancing the state-of-the-art

>> There are more detailed and sophisticated HRA methods available compared to what was done in the NuScale PRA

>> The terms CDF and LRF are tied to current large reactors and use of these terms in the NuScale design may be misleading 9

PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Independent Self Assessment

• External review of the NuScale PRA self-assessment against the high level and supporting requirements of the ASME PRA Standard

• In general, there was agreement, and in fact, in some cases, a higher capability category than identified was believed to be met. However, there were also some instances of a lack of concurrence, and possible enhancements were provided

• NuScale was able to incorporate those recommendations into the design certification PRA 10 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Initiating Event Analysis

• Multiple sources of input used to identify potential initiating events (IEs)

- NuScale design-specific master logic diagram

- NuScale design-specific simplified system-level failure modes and effects analysis (FMEA)

- Traditional lists of PRA initiating events

- Continuous focus (over the years of NuScale design and PRA development) on identifying potential initiating events and hazards 11 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Full Power Internal Initiating Events

• CVCS LOCA (injection line) inside containment vessel (CNV)

• CVCS LOCA (injection line) outside CNV

• CVCS LOCA (discharge line) outside CNV

• Spurious opening of ECCS valve

• Loss of DC power

• Loss of offsite power

• Steam generator tube failure

• LOCA (other) inside CNV

• Secondary-side line break (i.e., feedwater or main steam)

• General reactor trip

• Loss of support system (e.g., instrument air, AC power bus) 12 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Accident Sequence Analysis

• Initiating events and subsequent plant responses evaluated

• Key safety functions identified

- Fuel assembly heat removal, reactivity control, containment integrity

• End states of the accident sequences defined

- Level-1: core damage frequency (CDF)

- Level-2: large release frequency (LRF)

• Event trees constructed for each of the initiating events associated with system successes or failures to accomplish the applicable safety functions 13 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Success Criteria

• The Level 1 PRA overall success criterion is the prevention of core damage, defined by maintaining a peak cladding temperature less than 2,200 degrees Fahrenheit

- This is demonstrated for a 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> mission time

• System success criteria is determined by the minimum system availability required to prevent core damage

• The Level 1 success criteria evaluation is built upon a comprehensive simulation suite of more than 40 unique accident sequences

• The Level 2 success criterion is large release defined as a source term resulting in acute whole body 200 rem dose to the maximally exposed individual stationary at the reactor site boundary for 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br /> 14 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Success Criteria

• PRA success criteria simulations use NuScales safety-related NRELAP5 code with an input model that starts with NuScales safety-related input model

- The PRA simulations augment the safety-related input model with additional nonsafety-related models for beyond-design-basis phenomena

• Chemical and volume control system (CVCS) and containment flooding and drain system (CFDS) models

• Multi-dimensional core thermal hydraulic and neutronic models are used to simulate complex beyond design basis transients such as ATWS 15 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Human Reliability Analysis

• Human actions are not credited in the evaluation of design basis events

- Human actions only relevant to beyond design basis analyses

• Human error probabilities for beyond design basis events based on methodologies provided in NUREG/CR-4772 and NUREG/CR-6883

- Latent human errors and recovery actions

• As a modeling convenience, when quantifying the PRA model, the bounding human error probability of the complete set of post-initiator human failure events, is used for all independently modeled post-initiator human failure events

• Risk significant human action candidates input to D-RAP

- Operator fails to initiate CFDS injection

- Operator fails to initiate CVCS injection 16 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Post-initiator HEPs in PRA Quantification

• Post-initiator final human error probability (HEP) values range from 4E-3 to 2E-5

- Time available (based on bounding scenarios) for human actions range from 30 minutes to 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />

• To simplify the quantification of the PRA model, bounding value of the set of HEPs used to quantify all post-initiator HEPs Event Description Value EF HEP01 Human error probability for first HFE in cutset 4.0E-03 10 HEP02 Human error probability for second HFE in cutset 1.5E-01 3 HEP03 Human error probability for third HFE in cutset 0.5 -

17 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary NuScale PRA Human Errors Modeled (Pre-Initiator)

Name Description Operator misaligns MDP 0001A CFDS train A manual CFDS--HFE-0001A-UTM-N valves during test and maintenance Operator misaligns MDP 0001B CFDS train B manual CFDS--HFE-0002A-UTM-N valves during test and maintenance Operator misaligns MDP 0002A CVCS train A manual CVCS--HFE-0001A-UTM-N valves during test and maintenance Operator misaligns MDP 0002B CVCS train B manual CVCS--HFE-0002A-UTM-N valves during test and maintenance Operator misaligns CTG 0003X EHVS combustion EHVS--HFE-0001A-UTM-N turbine generator during test and maintenance Operator misaligns DGN 0001X ELVS standby diesel ELVS--HFE-0001A-UTM-N generator during test and maintenance Operator misaligns DGN 0002X ELVS standby diesel ELVS--HFE-0002A-UTM-N generator during test and maintenance Operator miscalibrates safety function modules during MPS---HFE-0001A-UTM-S test and maintenance 18 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary NuScale PRA Human Errors Modeled (Post-Initiator)

Name Description Context Operator fails to unisolate and Used for LOCA-OC (2 IEs), SGTFs, CFDS--HFE-0001C-FOP-N initiate CFDS injection and transients (1 IE)

Used for LOCA-IC (3 IEs), LOCA-OC (letdown) (1 IE), transients (1 IE)

Operator fails to unisolate and CVCS--HFE-0001C-FOP-N and secondary steam line break (1 initiate CVCS injection IE) upon failure of ECCS, and SGTFs Operator fails to locally Local unisolation due to lack of CVCS--HFE-0002C-FOP-N unisolate and initiate CVCS control from a partial loss of DC injection power Operator fails to open ECCS Backup action to MPS autofunction ECCS--HFE-0001C-FTO-N valves failure Backup local action to control room Operator fails to start/load EHVS--HFE-0001C-FTS-N initiation failure during loss of offsite combustion turbine generator power Backup local action to control room Operator fails to start/load ELVS--HFE-0001C-FTS-N initiation failure during loss of offsite backup diesel generator power 19 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Data Sources

• Industry information (e.g., NUREG/CR-6928, LERs) where applicable

- Common cause failure (CCF) modeling based NUREG/CR-5497

• Design-specific analyses

- Passive safety system reliability (i.e., ECCS, DHRS)

- Unique events (e.g., steam generator tube failure) 20 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Quantification

• Quantification of the PRA model was performed with the SAPHIRE code

- Including CCF models, failure data correlations and uncertainty analyses

• Using the ASME/ANS PRA Standard convergence criterion, a truncation value of 1E-15 per module year was used for the CDF 21 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Uncertainty Analyses

• Addressed using both quantitative uncertainty analyses and sensitivity studies

- SAPHIRE PRA code has capability for propagating parametric uncertainties

- Sometimes augmented using sensitivity studies (e.g., SGTF)

- Thermal hydraulic analyses typically use bounding inputs

• Uncertainty addressed in all modes and all hazards of single module PRA

- Multi-module risk quantification uses conservative, bounding estimates 22 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Parametric Uncertainty

• The data parameters include initiating event frequencies, component failure probabilities, CCF events and their alpha factors, and human error probabilities

- Initiating event frequencies that rely on generic industry data were assigned an expanded uncertainty distribution (i.e., lognormal error factor = 10)

• SAPHIRE has the built-in ability to perform an uncertainty analysis

- Includes correlating failure probabilities

• After cutsets were generated in SAPHIRE, an uncertainty analyses was performed using the Latin Hypercube uncertainty sampling methodology.

23 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Importance

• Systems

- CNTS (containment isolation valves), ECCS, MPS, and UHS

• Components

- ECCS RVVs and RRVs

- DHRS actuation valves

- RSVs

- CVCS and CES containment isolation valves

- Combustion turbine generator

• Other events and initiators (FV>20%)

- RBC, LOCA inside CNV, LOCA outside CNV, LOOP, internal fires, internal flood

• Human actions (FV>20%)

- CVCS actuation and CFDS actuation (Level 2 and LPSD) 24 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Sensitivity Studies Parameter Parameter Change CDF Result LRF Result Base Case 2.7E-10 1.7E-11 HEP All HEPs set to FALSE 2.0E-10 1.0E-11 HEP All HEPs set to TRUE 3.2E-8 2.8E-9 CCF All CCFs set to FALSE 5.4E-12 1.2E-12 CCF All CCFs set to max value of 0.002 4.2E-6 3.7E-8 LOOP-IE LOOP frequency set to 1 per year 2.2E-9 1.7E-11 (base = 3.1E-2 per year)

LOCA-IC-IE LOCA inside CNV frequency increased 1 3.4E-10 1.7E-11 order of magnitude SGTF-IE SGTF frequency increased to generic value 2.8E-10 2.2E-11 ECCS & ECCS and DHRS passive heat transfer 3.2E-10 1.7E-11 DHRS PSSR failure increased 1 order of magnitude I&C sensors Failure probability of sensors was increased 2.8E-10 1.7E-11 an order of magnitude 25 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 2 Methodology

• Analysis indicates that the only applicable containment vessel (CNV) failure mechanisms are containment bypass events and failure of containment isolation

• No bridge trees or Level 1 plant damage state binning

- Level 2 event tree is directly linked to the Level 1 event trees Core Damage Sequences Core Damage Cutset Mapped to Containment Isolation - CIVs # End State Comments Release Size Close (Phase - PH1) (Phase - PH1)

CD CD-T01 CNTS-T01 1 CD Core Damage LEVEL2-ET 2 NR RC1:CD with Isolation 3 LR RC2:CD with Release 26 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary External Hazards

• External events are evaluated using Level 1 PRA model and the following methodologies

- Internal fire: NUREG/CR-6850

- Internal flood: Part 3 of ASME/ANS RA-Sa-2009

- External flood: Part 8 of ASME/ANS RA-Sa-2009

- High winds: Part 7 of ASME/ANS RA-Sa-2009

- Seismic margin assessment: Part 5 of ASME/ANS RA-Sa-2009 27 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Seismic Risk Evaluation

• NuScale performed a PRA-based seismic margin assessment (SMA)

• Design-specific fragility calculations were performed for SSCs that contribute to the seismic margin

- Consulted with Simpson, Gumpertz & Heger, Rizzo Associates, and Stevenson and Associates

• Generic capacities with design-specific response factors were used for other SSCs

• DC/COL-ISG-020 seismic margin goal: high confidence of low probability of failure (HCLPF) value of 1.67 times the certified seismic design response spectra (CSDRS)

- Corresponds to 0.84g peak ground acceleration (PGA) 28 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary SMA Methodology

• PRA-based SMA uses internal event logic, seismically-induced initiators, and maps seismic failures to random failures

• HCLPF: high confidence (95%) of low probability (5%) of failure

- HCLPF can also be interpreted as a 1% probability of failure at the mean (or best-estimate) confidence level (i.e., at the HCLPF PGA there is a 1% probability of core damage)

- Evaluated at the sequence level using min-max criteria

• Seismic margin determined by those seismic failures that would result in a conditional core damage probability of greater than 1%

• Structural fragilities evaluated for those SSCs that contact the module, are located above the module, or where collapse might damage the module (which is assumed to result in core damage) 29 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Seismic Risk Evaluation Seismic plant response SMA results

• Induced initiator event trees

• Plant level HCLPF: 0.88g

- Structural failures

• Structural failures dominate

- LOCAs - Crane

- Loss of offsite power - Exterior walls

• Seismic failure mapping - Bay walls

- No pre-screening (all PRA cutsets - Module supports included in the SMA)

• At lower PGAs, LOOP combined

- Evaluated at 14 ground motion with random failures dominate levels ranging from 0.05g to 3.5g results

• Negligible seismic risk from low power and shutdown states 30 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Select NuScale SMA Structures 31 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Fragility Calculation Parameters

• Design calculations for demand/capacity (D/C) ratio inputs

- Uses bounding, conservative values

- For fragility purposes, design calculates are adjusted to median-centered values, uncertainties quantified

• Structural response factor variables

- Ground motion response

- Damping

- Modeling

- Mode combination

- Time history simulation

- Foundation-structure interaction

- Earthquake component combination 32 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Fragility Calculation Parameters

• Capacity variables

- Strength

- Ductility

• Earthquake scale factor (ESF)

- Used in wall calculations, where capacity changes with demand

- Ratio by which the seismic demand must increase for overall demand to equal capacity

• Static demand + ESF

• seismic demand = static capacity +/- ESF

• dynamic capacity (sign is dependent on load in compression / tension)

- Used to calculated median capacity Am 33 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Low Power and Shutdown

• Potential initiating events are those considered for full power and those unique to LPSD

- Reduced inventory (drain down) events not applicable

• No reduced inventory operations in the NuScale design

• Evaluated external events shown to be not important

• Dropped module event most significant CDF contributor

- Relatively high level of conservatism embedded in analysis 34 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Dropped Module Evaluation

• Drop probability developed based on conceptual reactor building crane design

• Core damage conservatively assumed for dropped module

- For a horizontal module the core partially uncovers

- Containment assumed to fail in a manner that prevents pool water incursion but allows radionuclide release

• Maximum radiological release much less than large release due to pool scrubbing effect

• Up to two operating modules theoretically could be struck by free-falling module, potentially inducing LOCA or transient in struck module 35 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Postulated Dropped Module Impacts

• Potential damage to the decay heat removal system (DHRS) because the heat exchangers are located external to the containment and face central pool channel

- Likelihood is an insignificant contributor to the modeled frequency of secondary side line break initiating event

• Potential damage to the chemical and volume control system (CVCS) piping where the piping penetrates the bay wall as a result of movement of the struck module

- Likelihood is an insignificant contributor to the modeled frequency of the CVCS pipe break outside containment initiating event

• Considering the probability of a load drop, the contribution of a potential module drop to the initiating event frequencies of an operating module is judged to be negligible both in absolute terms and in comparison to the frequency of a randomly occurring initiating events 36 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Multiple Module Evaluation

• Each NPM comprises a separate, independent RPV and CNV, and is serviced by separate, independent safety systems

• Systematic evaluation performed per SRP 19.0

• Single module PRA with bounding multi-module adjustment factors (MMAF) applied to each and every basic and initiating event

- MMAF value of 1.0 for SSCs shared amongst multiple modules and plant wide initiating events (e.g., LOOP)

- MMAF values from 0.1 to 0.3 for SSCs with potential coupling mechanisms between modules (e.g., potential for common cause failures)

- Smallest applied MMAF of 0.01 to events that would nominally be considered independent (e.g., pipe failures) 37 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 1 Insights

• NuScale design exceeds NRC core damage frequency safety goal with significant margin

- Full power internal event CDF 3.0E-10/mcyr

• External initiator CDFs: 1.0E-09 to 6.1E-11/mcyr

- LPSD CDF dominated by module drop event: 8.8E-08/mcyr

- Focused PRA CDF (no credit for nonsafety-related systems): 3.1E-06/mcyr

• Approximately equivalent to a long-term station blackout with no recovery of ac power

- Multiple module CDF factor: 0.13 38 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 2 Insights

• NuScale design exceeds NRC large release frequency safety goal with significant margin

- Full power internal event LRF 2.3E-11/mcyr

• External initiator LRFs: 4.3E-11 to <1E-15/mcyr

- Module drop event does not result in large release

- Focused PRA LRF (no credit for nonsafety-related systems): 1.6E-07/mcyr

• Approximately equivalent to a long-term station blackout with no recovery of ac power

- Multiple module LRF factor: 0.01 39 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 1 Key Insights (1 of 2)

Design Feature/Insight Comment Failure to scram events (ATWS) do Core characteristics result in ATWS power levels that are not lead directly to core damage. comparable to decay heat levels. Heat transfer from the containment vessel (CNV) to reactor pool is adequate to prevent core damage and most ATWS sequences require approximately the same system success criteria as non-ATWS events.

Passive heat removal capability is RSV cycling transfers adequate RCS water to the CNV to allow sufficient to prevent core damage if heat transfer through the RPV to the CNV and ultimately to a reactor safety valve (RSV) cycles. reactor pool to remove decay heat.

Post-accident heat removal through The steam generators and DHRS provide effective heat removal steam generators or decay heat paths to prevent core damage, but are unnecessary if RSV removal system (DHRS) is cycling allows heat transfer to reactor pool. Passive, fail-safe unnecessary if RSVs cycle. DHRS provides a natural circulation closed loop system that does not require pumps, power, or additional water.

Passive, fail-safe emergency core The ECCS consists of 5 valves that fail-safe on a loss of power cooling system (ECCS) functions to and provides a natural circulation path through the core and preserve RCS inventory, which is CNV, thus providing heat transfer to the reactor pool. The sufficient to allow core cooling closed-loop system does not need additional inventory.

without RCS makeup from external source.

40 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 1 Key Insights (2 of 2)

Design Feature/Insight Comment Containment isolation preserves Containment isolation eliminates the potential for breaks outside RCS inventory for core cooling of containment to result in loss of RCS inventory. For breaks without external makeup. inside of containment, containment isolation is not necessary to support passive core cooling and heat removal.

Passive, fail-safe safety systems Safety-related mitigating systems are fail-safe on loss of power (ECCS, DHRS, RSVs) include and do not require supporting systems such as lube oil, air or redundancy and do not need support HVAC to function. No single failure results in a loss of system systems, including electric power or function.

operator actions.

There are no risk significant, post- No operator actions, including backup and recovery actions, are initiator human actions associated risk significant to the CDF because of passive system reliability with the full-power PRA. and fail-safe system design.

Risk significant structures, systems The module response to external events is comparable to the and components (SSCs) for external response to internal event due to the passive features of the events are largely the same as those design and independence from support systems such as power.

found risk significant for internal Additional systems and components have been identified as risk events. significant for external events due to a conservative evaluation.

Active systems providing makeup Inventory addition is possible by the active systems chemical inventory to the RPV are not risk and volume control system (CVCS) and containment flooding significant. and drain system (CFDS). Due to the reliability of the passive safety systems, the active systems providing this backup function were found not to be risk significant.

41 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Section 19.1 COL Items Item Number Description COL Item 19.1-1 A COL applicant that references the NuScale Power Plant design certification will identify and describe the use of the probabilistic risk assessment in support of licensee programs being implemented during the COL application phase.

COL Item 19.1-2 A COL applicant that references the NuScale Power Plant design certification will identify and describe specific risk-informed applications being implemented during the COL application phase.

COL Item 19.1-3 A COL applicant that references the NuScale Power Plant design certification will specify and describe the use of the probabilistic risk assessment in support of licensee programs during the construction phase (from issuance of the COL up to initial fuel loading).

COL Item 19.1-4 A COL applicant that references the NuScale Power Plant design certification will specify and describe risk-informed applications during the construction phase (from issuance of the COL up to initial fuel loading).

42 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Section 19.1 COL Items Item Number Description COL Item 19.1-5 A COL applicant that references the NuScale Power Plant design certification will specify and describe the use of the probabilistic risk assessment in support of licensee programs during the operational phase (from initial fuel loading through commercial operation).

COL Item 19.1-6 A COL applicant that references the NuScale Power Plant design certification will specify and describe risk-informed applications during the operational phase (from initial fuel loading through commercial operation).

COL Item 19.1-7 A COL applicant that references the NuScale Power Plant design certification will evaluate site-specific external event hazards (e.g.,

liquefaction, slope failure), screen those for risk-significance, and evaluate the risk associated with external hazards that are not bounded by the design certification.

COL Item 19.1-8 A COL applicant that references the NuScale Power Plant design certification will confirm the validity of the key assumptions and data used in the design certification application probabilistic risk assessment (PRA) and modify, as necessary, for applicability to the as-built, as-operated PRA.

43 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary ACRS Subcommittee Presentation:

NuScale FSAR Chapter 19.2 Severe Accident Evaluation May 15, 2019 44 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Presentation Team Sarah Bristol Supervisor, Probabilistic Risk Assessment Etienne Mullin Probabilistic Risk Analyst Bill Galyean Probabilistic Risk Assessment Consultant Rebecca Norris Supervisor, Licensing 45 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Section 19.2: Severe Accident Evaluation

• Accident sequences resulting in core damage are evaluated in the Level 2 PRA for potential to challenge containment vessel (CNV) integrity and result in a large radionuclide release

- Large release defined as source term resulting in acute whole body 200 rem dose to the maximally exposed individual stationary at the reactor site boundary for 96 hours0.00111 days <br />0.0267 hours <br />1.587302e-4 weeks <br />3.6528e-5 months <br />

- MACCS off-site consequence calculations demonstrate that sequences with intact CNV are not large release

- CNV bypass accidents counted as large release (simplification for convenience)

• Potential challenges to CNV integrity identified from SRP, PRA standard, and NUREGs

• There are no unique phenomenological challenges that are introduced by the NuScale design 46 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Use of MELCOR

• Provides a best estimate evaluation of severe accident challenges to CNV

• Informs conservative evaluations of severe accident challenges

- Provides a physical basis for parameters

• Timing of core damage, core relocation

• Quantity of relocated material, composition of relocated material

• System pressures, temperatures, quantity of hydrogen produced

- Evaluations use limiting values from database of simulations that each involve bounding/conservative simplifications

• End of cycle decay heat load

• DHRS not credited to slow down accident progression

- Evaluations also consider parameters that bound all results observed from database of simulations

• 100% of fuel UO2 relocates at first observed relocation time from database

• Assume debris is molten, pure UO2 composed of no filler materials (e.g., steel, zirconium)

• No credit for water in lower plenum at time of relocation 47 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary MELCOR Model Development

• Thermal-hydraulics modeling developed from NRELAP5 model

- Matching elevations, volumes, flow areas, frictional losses, heat structure material, surface area, thickness, heated diameters, etc

• Benchmarking of steady-state operation and transients demonstrate reasonable to excellent agreement with NRELAP5

- Goal is to approximately match NRELAP5 accident simulation to the point of core damage and then extend simulation into severe accident space

• Severe accident modeling based on appropriate and accurate modeling of NPM design characteristics

- Decay power curve, core component masses and locations, radionuclide inventory, core flow geometry

• Incorporates modeling best practices from

- MELCOR code development staff and industry leading subject matter experts

- State-of-the-Art Reactor Consequence Analyses (SOARCA) reports

- MELCOR guides, manuals, assessments 48 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary In-Vessel Retention (IVR)

• Conservative analysis demonstrates that RPV lower head integrity is maintained if core debris relocates to lower plenum

• Maximum heat flux remains below critical heat flux (CHF) on exterior surface

- Heat generation rate based on conservative assumptions/inputs (e.g., 100% core UO2 - no upward radiation heat losses)

- Assumed CHF threshold conservatively does not credit high absolute pressure and large subcooling in CNV

• With effective external vessel cooling, the lower head remains intact and the severe accident progression is stabilized in RPV 49 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Severe Accident Phenomena CNV integrity not challenged by severe accident phenomena

• Hydrogen combustion not challenging due to limited oxygen concentration

• In-vessel fuel-coolant interactions (FCI) (i.e., steam explosion) are not sufficiently energetic to induce alpha mode failure due to factors including:

- Small core size, low debris temperatures, small drop height, shallow pool, relatively high system pressure

• Containment overpressure does not occur

- High pressure steel CNV designed for most limiting LOCA blowdown which exceeds maximum severe accident pressures

- Submergence of CNV in UHS provides highly effective pressure suppression

- No concrete interactions to generate non-condensable gases 50 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Consideration of Uncertainty

• If IVR in RPV fails

- High pressure melt ejection (leading to direct containment heating) does not occur because there is no driving pressure differential

- Energetic ex-vessel FCI not likely for similar reasons as in-vessel FCI

- Debris relocated to CNV would be retained by CNV lower head

• Effective external cooling of CNV by reactor pool

• If lower CNV fails

- Pool scrubbing minimizes release

• If upper CNV fails

- Instantaneous release of entire airborne radionuclide inventory in module at time of postulated CNV failure would not constitute a large release 51 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 2 Insights

• Core damage events are stabilized within the RPV

• Severe accident phenomena do not challenge CNV integrity

• Large release does not occur even if RPV and CNV are postulated to fail

• The large release frequency is dominated by containment bypass events 52 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 2 Key Insights (1 of 5)

Design Feature/Insight Comment Containment Isolation The primary purpose of CNTS is to If coolant remains primarily within the RPV, then the core retain primary coolant inventory is covered. If the core is not covered in the RPV then within the CNV. With primary sufficient primary coolant is in the CNV to submerge the coolant inventory maintained in the outside of the lower RPV and establish conductive heat RPV or CNV, cooling of core debris removal from the core debris to the coolant in the CNV is ensured. through the RPV wall.

CNTS terminates releases through Containment penetrations through which releases are penetrations leading outside assumed to occur that dominate risk include those that containment. bypass containment such as CVCS (injection and discharge) and paths through the steam generator tubes (main steam and feedwater piping). Isolation of normally open valves in these penetrations prevents releases from bypassing containment.

53 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 2 Key Insights (2 of 5)

Design Feature/Insight Comment Passive Heat Removal The RPV has no insulating material Retaining primary coolant in the containment results in collection and passive heat removal capability of sufficient RCS water in the CNV to allow heat transfer through from the RPV to the CNV is RPV to CNV and ultimately UHS to remove heat generated in sufficient to prevent core debris from the fuel regardless of its location.

penetrating the reactor vessel.

The CNV is uninsulated and passive heat removal capability from the CNV to the UHS is sufficient to prevent the containment from pressurizing and or core debris from penetrating the containment 54 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 2 Key Insights (3 of 5)

Design Feature/Insight Comment Severe Accident Containment Challenges (1 of 2)

Primary coolant system Addition of water to the containment from external overpressure failure cannot lead to sources (CFDS) results in submergence of the reactor vessel overpressurization of containment and establishes passive heat removal through the containment (i.e., loss of decay heat removal wall to the reactor pool. Even if containment flooding is not through the steam generators plus successful, the RPV failure mode is such that containment failure of the RSVs to open). ultimate capacity would not be exceeded.

Hydrogen combustion is not likely as There is very little oxygen available (oxygen generated from the containment is normally radiolysis is only a long-term issue) and containment is steam evacuated. inerted under severe accident conditions. In addition, conservative AICC analyses predict containment pressures that do not exceed the design pressure.

In-vessel steam explosions are not Core support failure is expected before the fuel has a chance to likely due to core support design and become molten. With the core uncovered there is little water in volume of lower vessel head. the bottom of the RPV with which core debris can interact.

HPME cannot occur Submergence of the lower RPV With passive heat removal establishes passive heat from the reactor to removal and prevents core containment established, the debris from exiting the RPV. No reactor is depressurized ex-vessel challenges occur if the even if core debris is core remains within the vessel. postulated to exit the vessel.

55 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 2 Key Insights (4 of 5)

Design Feature/Insight Comment Severe Accident Containment Challenges (2 of 2)

Ex-vessel steam explosion does not Submergence of the lower RPV establishes passive heat occur with a submerged RPV. removal and prevents core debris from exiting the RPV. No ex-vessel challenges occur if the core remains within the vessel.

Overpressure of containment due to There is no concrete in the containment with which the core non-condensable gas generation is debris could interact and generate non-condensable gases.

not applicable to the NuScale design.

Basemat penetration is not There is no basemat making up the containment boundary. This applicable to the NuScale design. issue is addressed as a part of considering protection against contact of core debris with the containment wall.

56 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Level 2 Key Insights (5 of 5)

Design Feature/Insight Comment Support Systems, Human Action, External Events Support systems are not needed for Safety-related mitigating systems are fail-safe on loss of power safety-related system functions (i.e., and do not require supporting systems such as lube oil, containment isolation) important to instrument air, or HVAC to function.

the Level 2 PRA.

With one exception, there are no risk Operator actions, including backup and recovery actions, are significant, post-accident human not significant to the Level 2 analysis because of passive actions associated with the full- system reliability and fail-safe system design. The operator power internal events Level 2 PRA. action to align CFDS during a CVCS break outside containment The exception is alignment of CFDS meets the risk significance thresholds because of a during accident sequences in which mathematical limitation of the calculation of the Fussell-Vesely isolation of a broken CVCS line measure of importance outside containment fails, ECCS is successful but coolant inventory in containment needs replenishment in order to maintain natural circulation between CNV and the RPV.

Risk significant SSC for external The module response to external events is comparable to the events are largely the same as those response to internal event due to the passive features of the found risk significant for internal design which are not affected by the external events and plant events systems that are protected against external event challenges.

57 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Section 19.2 COL Items Item Number Description COL Item 19.2-1 A COL applicant that references the NuScale Power Plant design certification will develop severe accident management guidelines and other administrative controls to define the response to beyond-design-basis events.

COL Item 19.2-2 A COL applicant that references the NuScale Power Plant design certification will use the site-specific probabilistic risk assessment to evaluate and identify improvements in the reliability of core and containment heat removal systems as specified by 10 CFR 50.34(f)(1)(i).

COL Item 19.2-3 A COL applicant that references the NuScale Power Plant design certification will evaluate severe accident mitigation design alternatives screened as not required for design certification application.

58 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary ACRS Subcommittee Presentation:

NuScale FSAR Chapter 19.3 Regulatory Treatment of Nonsafety Systems May 15, 2019 59 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Presentation Team Sarah Bristol Supervisor, Probabilistic Risk Assessment Etienne Mullin Probabilistic Risk Analyst Bill Galyean Probabilistic Risk Assessment Consultant Rebecca Norris Supervisor, Licensing 60 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Section 19.3

• There are no RTNSS SSCs in the NuScale design

- None of the five RTNSS criteria were met by any NuScale SSC

• RTNSS is also discussed in FSAR 17.4.3.3 61 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Section 19.3 COL Item Item Number Description COL Item 19.3-1 A COL applicant that references the NuScale Power Plant design certification will identify site-specific regulatory treatment of nonsafety systems (RTNSS) structures, systems, and components and applicable RTNSS process controls.

62 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary ACRS Subcommittee Presentation:

NuScale FSAR Chapter 19.5 Adequacy of Design Features and Functional Capabilities Identified and Described for Withstanding Aircraft Impacts May 15, 2019 63 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Presentation Team Amber Berger Civil/Structural Engineer Rebecca Norris Supervisor, Licensing Marty Bryan Licensing Project Manager 64 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Introduction and Background

• Plant design for potential effects of beyond design basis large commercial aircraft impact [10 CFR 50.150(a)]

- The reactor core remains cooled, or the containment remains intact

- Spent fuel cooling or spent fuel pool integrity is maintained

• Design-specific impact assessment per RG 1.217, which endorses NEI 07-13

• NEI 07-13 methods followed with no exceptions

• Aircraft impact informed the plant design 65 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Assessment Scope

• Reactor Building assessed for effects in three areas for postulated aircraft impact

- Physical damage

- Shock damage from shock-induced vibration on structures, systems, and components

- Fire damage from aviation fuel-fed fire 66 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Assessment Methodology

• NEI 07-13

• Reactor Building is structure of concern

- NuScale Power Modules

- Ultimate heat sink

- Spent fuel pool

• Impact locations

- Screening by NEI 07-13

- Radioactive Waste Building (RWB) is intervening structure to mitigate physical damage to RXB, conservatively do not credit RWB in shock assessment

- No credit taken for CRB or TGB 67 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary NuScale Site Plan North NuScale DCA Tier 2 Figure 1.2-1 Conceptual Site Layout 68 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Assessment Results

• Physical damage

- Local assessment per NEI formulas for perforation and scabbing

- Global response performed using detailed finite element models and NRC specified force-time history

- RXB external walls prevent physical damage from entering RXB

- No internal missiles for secondary impact

- No impact on containment boundary

- Spent fuel pool protected inside RXB below grade

- Reactor Building crane trolley cannot be dislodged 69 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Assessment Results (contd)

• Shock damage

- Aircraft impact causes short duration, high acceleration, high frequency vibration

- Core cooling

• At-power and shutdown scenarios considered

• No active equipment required for success

• Adequate heat removal is shown for all strikes

- Spent fuel

• SFP integrity maintained for all strikes 70 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Assessment Results (contd)

• Fire damage

- Design and location of 3-hr fire barriers and 3-hr, 5-psid fire barriers prevent propagation of fire into RXB

- Design and location of 5-psid, fast-acting blast dampers at RXB HVAC key design feature

- Concrete shrouds protect exterior wall pipe and HVAC penetrations from physical damage and prevent fire propagation into the RXB

- Fire that enters through external personnel doors at grade level does not propagate beyond stairwells

• All required operator actions occur prior to impact 71 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Assessment Conclusions

• Design and functional capabilities provide adequate protection of public health and safety

• NuScale plant meets 10 CFR 50.150 regulation

- Maintain containment integrity AND core cooling capability (only required to meet one)

- Maintain SFP integrity

• For most postulated aircraft impact strikes, spent fuel pool cooling maintained, meeting all four CFR requirements 72 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Acronyms (1 of 3)

• ATWS anticipated transient without scram

• CRB Control Building

• BDB beyond design basis

• CVCS chemical and volume control system

• CCF common cause failure

• CSDRS certified seismic design response

• CD core damage spectra

• CDF core damage frequency

• CTG combustion turbine generator

• CES containment evacuation system

• D/C demand/capacity

• CFDS containment flooding and drain

• DGN diesel generator system

• DHRS decay heat removal system

• CFR Code of Federal Regulations

• ECCS emergency core cooling system

• CHF critical heat flux

• EHVS 13.8 kV and switchyard system

• CIV containment isolation valve

• ELVS low voltage AC electrical

• CNV containment vessel distribution system

• CNTS containment system

• ESF earthquake scale factor

• COL combined license 73 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Acronyms (2 of 3)

• FCI fuel-coolant interaction

• LPSD low power and shutdown

• FMEA failure modes and effects analysis

• LR large release

• FP-IE full power, internal event

• LRF large release frequency

• FSAR Final Safety Analysis Report

• mcyr module critical year

• HCLPF high confidence of low probability

• MDP motor driven pump of failure

• MMAF multi-module adjustment factor

• HEP human error probability

• MPS module protection system

• HPME high pressure melt ejection

• NEI Nuclear Energy Institute

• HVAC heating ventilation and air

• NPM NuScale Power Module conditioning

• NR no release

• IE initiating event

• NRC Nuclear Regulatory Commission

• IVR in-vessel retention

• PGA peak ground acceleration

• LOCA loss of coolant accident

• PRA probabilistic risk assessment

• LOOP loss of offsite power 74 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Acronyms (3 of 3)

• RBC reactor building crane

• SAPHIRE Systems Analysis Programs for Hands-on Integrated Reliability

• RCS reactor coolant system Evaluations

• RG Regulatory Guide

• SGTF steam generator tube failure

• RPV reactor pressure vessel

• SMA seismic margin assessment

• RSV reactor safety valve

• SOARCA State-of-the-Art Reactor

• RTNSS regulatory treatment of nonsafety Consequence Analysis systems

• SSC structures, systems, and

• RVV reactor vent valve components

• RWB Radioactive Waste Building

• SFP spent fuel pool

• RXB Reactor Building

• SRP Standard Review Plan

• SAMDA severe accident mitigation design

• TGB Turbine Generator Building alternative

• UHS ultimate heat sink 75 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

NuScale Nonproprietary Portland Office Richland Office 6650 SW Redwood Lane, 1933 Jadwin Ave., Suite 130 Suite 210 Richland, WA 99354 Portland, OR 97224 541.360.0500 971.371.1592 Arlington Office Corvallis Office 2300 Clarendon Blvd., Suite 1110 1100 NE Circle Blvd., Suite 200 Arlington, VA 22201 Corvallis, OR 97330 541.360.0500 London Office 1st Floor Portland House Rockville Office Bressenden Place 11333 Woodglen Ave., Suite 205 London SW1E 5BH Rockville, MD 20852 United Kingdom 301.770.0472 +44 (0) 2079 321700 Charlotte Office 2815 Coliseum Centre Drive, Suite 230 Charlotte, NC 28217 980.349.4804 http://www.nuscalepower.com Twitter: @NuScale_Power 76 PM-0519-65372 Revision: 0 Copyright 2019 by NuScale Power, LLC.

Template #: 0000-21727-F01 R5

Safety Evaluation with Open Items:

Chapter 19 Probabilistic Risk Assessment and Severe Accident Evaluation for New Reactors NuScale Design Certification Application ACRS Subcommittee Meeting May 14-15, 2019

Agenda

• Presentation Topic for May 14, 2019:

- Section 19.1, Probabilistic Risk Assessment

• Presentation Topics for May 15, 2019:

- Section 19.2, Severe Accident Evaluation

- Section 19.3, Regulatory Treatment of Nonsafety Systems for Passive Advanced Light Water Reactors

- Section 19.5, Adequacy of Design Features and Functional Capabilities Identified and Described for Withstanding Aircraft Impacts

• Presentation Topic Included in Chapter 20 (for future discussion):

- Section 19.4, Strategies and Guidance To Address Loss of Large Areas of the Plant Because of Explosions and Fires May 14, 2019 Non-Proprietary 2

Section 19.1 Probabilistic Risk Assessment

Staffs Review - Overview

• SER is based on DCA Revision 2

• Staff conducted two regulatory audits

• Applied the enhanced safety focused review approach to:

- Support integrated decision making

- Increase focus on safety for effectiveness and efficiency of the review NuScale Chapter 19 - PRA and Severe May 14, 2019 4 Accident Evaluation

Staffs Review - Overview (Continued)

• Reviewed the quality, completeness, and consistency of the information in the DCA

• Paid an increased attention to key assumptions

• Review guided by the Commission goals for CDF, LRF, conditional containment failure probability, and PRA insights NuScale Chapter 19 - PRA and Severe May 14, 2019 5 Accident Evaluation

Overview of Review Guidance

• SRP 19.0 - Staff guidance, including acceptance criteria, for PRA and severe accident evaluation

• DC/COL-ISG Guidance on the use of ASME/ANS PRA Standard (as endorsed by RG 1.200) for a DCA NuScale Chapter 19 - PRA and Severe May 14, 2019 6 Accident Evaluation

At-Power Internal Events Level 1 PRA Data Sources

• Basic event data

- PWR generic failure probabilities are reasonable for the DCA stage

• NuScale unique components

- Failure rates and probabilities are key assumptions

- Sensitivity studies using conservative assumptions for component failure rates show results that compare favorably with the Commissions CDF and LRF goals NuScale Chapter 19 - PRA and Severe May 14, 2019 7 Accident Evaluation

At-Power Internal Events Level 1 PRA Passive System Reliability Evaluation

• Staff audited the inputs for NRELAP5 thermal-hydraulic simulations

• Applicant adequately considered effect of passive system reliability for the DHRS and ECCS consistent with the level of detail at the DCA stage NuScale Chapter 19 - PRA and Severe May 14, 2019 8 Accident Evaluation

At-Power Internal Events Level 1 PRA Event Tree - LOCA inside containment

• NuScale assumes containment isolation is not necessary for LOCAs inside containment (Open Item: RAI 8840, Question 19-2)

• Staff audited NuScales analysis

• Staff is evaluating the NuScale response NuScale Chapter 19 - PRA and Severe May 15, 2019 9 Accident Evaluation

At-Power Internal Events Level 1 PRA ATWS Sequences

• SER Chapter 7 documents staff approval of ATWS exemption request for 10 CFR 50.62(c)(1), acceptability of the module protection system design, and the reduction of ATWS risk below the Commission CDF goal of 1E-5 per year NuScale Chapter 19 - PRA and Severe May 14, 2019 10 Accident Evaluation

Non-Proprietary Office of Nuclear Regulatory Research TRACE Confirmatory Analysis of Anticipated Transients without SCRAM for NuScale Power Module NuScale Design Certification Application ACRS Subcommittee Meeting May 15, 2019

TRACE Confirmatory Calculations

• The staff performed confirmatory calculations for ATWS using TRACE.

• Key figures of merit:

- Reactor pressure vessel (RPV) pressure to confirm RPV integrity, and

- Riser collapsed liquid water level above the top of active fuel (TAF) to confirm core coolability.

NuScale Chapter 19 - PRA and Severe May 15, 2019 2 Accident Evaluation

Base Case Scenario Evaluated Event Progression for Margins to RPV Pressure and Level Criteria

• Initiated by LOAC which leads to immediate turbine and feedwater trips.

• Control rods fail to insert and remain withdrawn for the full transient.

• RPV pressure increases as reactor coolant system heats up.

• High RPV pressure triggers decay heat removal system (DHRS) actuation.

• Higher RPV pressure initiates reactor safety valve (RSV) cycling.

NuScale Chapter 19 - PRA and Severe May 15, 2019 3 Accident Evaluation

Key Alternate Scenarios Analyzed Staff Analyzed Several Cases in Addition to the Base Case

• BOC: Like the base case but with BOC kinetics parameters.

• 1RSV: Like the base case but with RSV-1 out-of-service (OOS).

• PRA-1RSV: Like BOC case but with RSV-1 OOS and DHRS OOS.

• SGHT: Like base case but RCS initial temperature is much higher.

NuScale Chapter 19 - PRA and Severe May 15, 2019 4 Accident Evaluation

TRACE Results Demonstrated Large Margins in All Cases

• Base/BOC/1RSV Cases: Large RPV peak pressure margin. In the long term, reactor power is balanced by DHRS at low levels and riser level remains well above the top of active fuel (TAF).

• PRA-1RSV Case: Large RPV peak pressure margin.

In the long term containment (CNV) inventory increases and core power is balanced by CNV heat removal. Level remains well above TAF.

• SGHT Case: Large RPV peak pressure margin. In the long term, RPV level drops to top of the riser but well above TAF. Core power balanced by DHRS and CNV heat removal.

NuScale Chapter 19 - PRA and Severe May 15, 2019 5 Accident Evaluation

At-Power Internal Events Level 2 PRA

• Two Containment Event Tree end states

• Other severe accident phenomena are addressed in the presentation on Section 19.2 NuScale Chapter 19 - PRA and Severe May 14, 2019 11 Accident Evaluation

LPSD Internal Events Level 1 PRA Single Module Drop

• Staff audited Reactor Building Crane PRA notebook

• Staff reviewed NUREGs on load drops, EPRI PRA on cask drops, and recent events to evaluate drop probability

• NuScale committed to: NUREG-0554 as supplemented by ASME NOG single failure proof

- consistent with operating plants NuScale Chapter 19 - PRA and Severe May 14, 2019 12 Accident Evaluation

LPSD Internal Events Level 1 PRA Single Module Drop (continued)

• Risk significance of Reactor Building Crane resulted in additional ITAACs:

- Rated load test and inspection of NuScale power module lifting fixture and module lifting adapter

• Analysis consistent with SRP Section 19.0 and DC/COL-ISG-028 NuScale Chapter 19 - PRA and Severe May 14, 2019 13 Accident Evaluation

LPSD External Events Single Module Drop

• Reactor Building Crane design adequately considers loss of AC power due to external flooding and high winds

• Module drop analysis consistent with SRP Section 19.0 and DC/COL-ISG-028 NuScale Chapter 19 - PRA and Severe May 14, 2019 14 Accident Evaluation

Multi-module Risk Overview

• Systematic process is used to evaluate multi-module risk

• Approach relies on assumptions made based on engineering judgement

• Design relies on independent, module-specific safety-related systems to prevent and mitigate core damage

- DHRS, ECCS, CIVs

• Impact of external events is addressed qualitatively NuScale Chapter 19 - PRA and Severe May 14, 2019 15 Accident Evaluation

Multi-module Risk Module Drop Event

• Module dropped during refueling can impact up to two operating modules (Open Item: RAI 9659, Question 19-39)

• Staff is evaluating NuScales analysis of this event NuScale Chapter 19 - PRA and Severe May 14, 2019 16 Accident Evaluation

External Flooding and High Winds Analyses

• For the external flooding PRA, staff finds the applicants approach reasonable

• For the high winds analysis, staff verified that all important accident mitigation features are housed within seismic Category I reactor building structure and are protected from the effects of high winds.

• Staff finds these external hazard analyses sufficiently consistent with DC/COL-ISG-028 and SRP Section 19.0.

NuScale Chapter 19 - PRA and Severe May 14, 2019 17 Accident Evaluation

At-Power Internal Fire and Internal Flood PRAs

• Staff focused its review on the appropriateness of assumptions used to address incomplete aspects of the design and operating procedures

• Staff finds:

- the estimated risk is reasonable for the DCA stage

- the internal fire and internal flood PRAs for at-power and LPSD operations are sufficiently consistent with DC/COL-ISG-028 and SRP Section 19.0 NuScale Chapter 19 - PRA and Severe May 14, 2019 18 Accident Evaluation

PRA-based SMA

• Staff focused its review on the scope of SSCs included in the fragility evaluation and the analysis methods used to determine seismic fragility

• Staff finds the plant-level HCLPF capacity demonstrates adequate margin in accordance with SRP Section 19.0 NuScale Chapter 19 - PRA and Severe May 14, 2019 19 Accident Evaluation

Summary PRA

• Due to the open items, the staff cannot make a finding on the PRA description in DCA Part 2, Tier 2, Sections 19.0 and 19.1

- RAIs 8840 and 9659 NuScale Chapter 19 - PRA and Severe May 14, 2019 20 Accident Evaluation

Abbreviations and Acronyms

• AC - alternating current

• ISG - Interim Staff Guidance

• ANS - American Nuclear Society

• ITAAC - inspections, tests, analyses,

• ASME - American Society of and acceptance criteria Mechanical Engineers

• LOCA - loss-of-coolant accident

• ATWS - anticipated transient without

• LPSD - low power and shutdown scram

• LRF - large release frequency

• CDF - core damage frequency

• PRA - probabilistic risk assessment

• CIV - containment isolation valve

• PWR - Pressurized water reactor

• COL - combined license

• RAI - request for additional

• DC - design certification information

• DCA - design certification application

• RG - Regulatory Guide

• DHRS - decay heat removal system

• SER - safety evaluation report

• ECCS - emergency core cooling

• SMA - seismic margin assessment system

• SRP - standard review plan

• HCLPF -high confidence low

• SSCs - structures, systems, and probability of failure components NuScale Chapter 19 - PRA and Severe May 14, 2019 21 Accident Evaluation

Questions/comments from members of the public before the closed session starts?

22 Non-Proprietary