ML18254A164
| ML18254A164 | |
| Person / Time | |
|---|---|
| Issue date: | 08/21/2018 |
| From: | Weidong Wang Advisory Committee on Reactor Safeguards |
| To: | |
| Wang W | |
| References | |
| NRC-3858 | |
| Download: ML18254A164 (508) | |
Text
Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION
Title:
ACRS Thermal-Hydraulic Subcommittee Docket Number: N/A Location: Rockville, Maryland Date: August 21, 2018 Work Order No.: NRC-3858 Pages 1-508 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 THERMAL-HYDRAULIC SUBCOMMITTEE 8 + + + + +
9 TUESDAY 10 AUGUST 21, 2018 11 + + + + +
12 ROCKVILLE, MARYLAND 13 The Subcommittee met at the Nuclear 14 Regulatory Commission, Two White Flint North, Room 15 T2B1, 11545 Rockville Pike, at 8:30 a.m., Michael L.
16 Corradini, Chairman, presiding.
17 COMMITTEE MEMBERS:
18 MICHAEL L. CORRADINI, Chairman 19 DENNIS C. BLEY, Member 20 WALTER L. KIRCHNER, Member 21 JOSE A. MARCH-LEUBA, Member 22 JOY L. REMPE, Member 23 GORDON R. SKILLMAN, Member 24 JOHN W. STETKAR, Member 25 MATTHEW W. SUNSERI, 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 DESIGNATED FEDERAL OFFICIAL:
2 WEIDONG WANG 3 ALSO PRESENT:
4 STEVE BAJOREK, RES 5 JON CARMACK, INL 6 DAN FUNK, DOE NE 7 JESS GEHIN, INL 8 JASON HALES, INL 9 STEVE HAYES, INL 10 SHANE JOHNSON, DOE 11 RICHARD LEE, RES 12 ELIA MERZARI, ANL 13 DAVE POINTER, ORNL 14 EVERETT REDMOND, NEI 15 TANJU SOFU, ANL 16 CHRIS STANEK, LANL 17 JEFF WHITT, Framatome*
18 RICH WILLIAMSON, INL 19 *Present via telephone 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 Introduction . . . . . . . . . . . . . . . . . . 4 3 Opening Remarks . . . . . . . . . . . . . . . . . 6 4 DOE Codes for Accident Tolerant Fuels (ATF) in LWR 5 Safety Analysis 6 Introduction . . . . . . . . . . . . . . . 22 7 Fuel Performance Modeling for ATF . . . . . 34 8 Neutronics and Thermal Hydraulics 9 for ATF . . . . . . . . . . . . . . 128 10 DOE Codes for Advanced (non-LWR) Reactor Safety 11 Analysis 12 Introduction . . . . . . . . . . . . . . 186 13 Neutronics Analysis Capabilities 14 for Non-LWRs . . . . . . . . . . . 187 15 Fuels Modeling Capabilities 16 for Non-LWRs . . . . . . . . . . . 224 17 Thermal-Hydraulic/System Analysis 18 Capabilities for Non-LWRs . . . . . 255 19 Source Term Assessment Codes 20 and Methods . . . . . . . . . . . . 292 21 Concluding Remarks (ATF and Advanced Reactors) 328 22 Public Comment Period . . . . . . . . . . . . . 333 23 Adjourn . . . . . . . . . . . . . . . . . . . . 347 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
4 1 P R O C E E D I N G S 2 (8:30 a.m.)
3 CHAIRMAN CORRADINI: Okay, the meeting 4 will now come to order. This is a meeting of the 5 Thermal-Hydraulics Phenomena Subcommittee, a standing 6 committee of the Advisory Committee on Reactor 7 Safeguards.
8 My name's Mike Corradini; I'm chairman of 9 the subcommittee. ACRS members in attendance are Dr.
10 Ron Ballinger, Dennis Bley, Matt Sunseri, Joy Rempe, 11 Jose March-Leuba, and Walt Kirchner. And Weidong Wang 12 is the designated federal official for this meeting.
13 In this meeting, the subcommittee will 14 conduct a information briefing regarding the potential 15 use of Department of Energy computer codes in risk-16 informed safety analysis for accident-tolerant fuels 17 and light water reactors as well as in non-light water 18 reactors.
19 The subcommittee will hear presentations 20 by and hold discussions with the Department of 21 Energy's personnel and other interested persons 22 regarding this matter.
23 We've received no written comments or 24 requests for time to make oral statements from 25 members of the public regarding today's meeting. 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
5 1 entire meeting will be open to public attendance.
2 The subcommittee will gather information, 3 analyze relevant issues and facts, and formulate 4 proposed positions and actions as appropriate for 5 deliberation by the full committee.
6 The rules for participation in today's 7 meeting have been announced as part of the notice of 8 this meeting previously published in the Federal 9 Register.
10 The transcript of the meeting is being 11 kept and will be available as stated in the Federal 12 Register's notice. Therefore, we request the 13 participants in this meeting use the microphones 14 located throughout the meeting room when addressing 15 the subcommittee.
16 Participants should first identify 17 themselves, speak with sufficient clarity and volume, 18 so they may well be heard. Let me add some 19 extemporaneous comments.
20 First of all, make sure everybody silences 21 their devices. We have a lot of folks in here with a 22 lot of various appliances. Make sure things don't 23 bong, bing, ring, whatever.
24 Secondly, since this is an information 25 meeting and it's the first of what I expect will be 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
6 1 couple, at least. There's no intention to write a 2 letter report with the full committee.
3 So this subcommittee will report back to 4 the full committee, but it's my intent not to have a 5 letter at this point. It's too early in the game 6 since this is kind of more of a gathering.
7 And because of that, let me make it known 8 to the DOE folks that you're going to hear at least 9 seven different opinions from this committee, all of 10 which are not the ACRS position.
11 They are individual comments by the 12 individual members which you can or not take at will.
13 So that's the final point. Other than that, I think 14 the person to turn to is Shane. Are you going to lead 15 us off today?
16 MEMBER MARCH-LEUBA: You will need to go 17 through the microphone procedure?
18 CHAIRMAN CORRADINI: Yes, there's a little 19 green light that goes on. There you go.
20 MEMBER BLEY: If you push the button.
21 CHAIRMAN CORRADINI: If you push the 22 button.
23 MR. JOHNSON: All right, good morning.
24 Mr. Chairman, members of the subcommittee it's a 25 pleasure to be here today to discuss the Department 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
7 1 Energy's work over the last nine years to develop, 2 demonstrate, and deploy a suite of world-class 3 modeling and simulation capabilities in support of 4 commercial nuclear energy in the United States.
5 That includes both the existing fleet of 6 light water reactors and the emerging non-water based 7 advanced reactors. Since 2010, the Department has 8 invested nearly $400 million in advanced nuclear 9 energy modeling and simulation.
10 Executing two similar but distinctly 11 different federal programs, the energy innovation hub 12 for modeling and simulation and the nuclear energy 13 advanced modeling and simulation programs, the 14 Department is committed to developing and deploying 15 state of the art computational platforms for light 16 water-based reactor systems and advanced non-water 17 based reactor systems.
18 Today's briefings which will be provided 19 by the Department's leading subject matter authorities 20 are focused on modeling accident-tolerant fuels for 21 possible use in the existing fleet of U.S. light water 22 reactors and advanced reactor technologies, molten 23 salt, high-temperature gas, and liquid metal which are 24 currently under development by the private sector.
25 I believe today's briefings will provide NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
8 1 the committee a good overview of the current 2 capabilities and limitations of the codes, and the 3 analytical enhancements afforded by these codes as 4 compared to the legacy codes currently used by both 5 the regulator and licensees.
6 A key DOE goal of our advanced modeling 7 and simulation program is to enable industry to 8 accelerate reactor and fuel development and 9 commercialization.
10 For example, we believe these tools can 11 help accelerate licensing as they can help the reactor 12 and fuel vendor community design, execute, and analyze 13 more effective high-value experiments to support the 14 licensing of their technologies.
15 While these advanced tools will not 16 replace the need for experimentation, they can help 17 identify the most critical experiments and help focus 18 data acquisition from the experiments.
19 The technical teams working on both the 20 hub products and the NEAMS products represent the best 21 nuclear energy modeling and simulation talent within 22 the DOE laboratory complex.
23 A very important strength of these teams 24 is the multi-disciplined, multi-laboratory composition 25 of the team membership. In the early days of our NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
9 1 program, we established an overarching success metric 2 centered on creating modeling and simulation platforms 3 that would be used by industry for both in-house 4 computational needs and for the development of 5 licensing products for submission to the Nuclear 6 Regulatory Commission.
7 I'm happy to report that our advanced 8 codes are being used by industry in support of the 9 existing fleet and in support of advanced reactors.
10 Five test stands have been deployed and 11 are being used by the private sector to analyze and 12 evaluate the operations of the light water fleet.
13 Four development companies within the 14 advanced reactor community are using our tools for 15 non-water technologies, and several additional 16 companies are evaluating their use.
17 We also know one private sector firm will 18 be submitting licensing documents later this year 19 which are based on analyses derived from our advanced 20 modeling and simulation codes.
21 It is my hope that today's discussions 22 will serve to accelerate the technical cooperation 23 between the two agencies, especially in the area of 24 advanced modeling and simulation.
25 With limited federal budgets at both NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
10 1 agencies, it is important that we cooperate to the 2 fullest extent possible on the development and 3 qualification of these tools which can be used 4 independently in the execution of each agency's 5 missions.
6 I would be remiss if I didn't acknowledge 7 the important technical contributions of the NRC's Dr.
8 Steve Bajorek and Dr. Kim Webber to our modeling and 9 simulation activities.
10 While Dr. Webber's contributions have been 11 more recent as part of her professional development 12 assignment to the department, Dr. Bajorek has been 13 intimately involved in our efforts from almost the 14 very beginning of the program.
15 With regards to the two agencies working 16 together on a common set of computational tools, I do 17 not underestimate the challenge this represents. For 18 far too long, vested and personal interest within the 19 DOE national labs and within the federal staff have 20 opposed our efforts to work together in the execution 21 of our respective responsibilities.
22 The mere fact that the ACRS requested 23 today's briefing demonstrates that the Department's 24 nuclear energy computational efforts are being 25 recognized by the U.S. technical community, and our NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
11 1 modeling and simulation program is on a success path.
2 A decade ago, the Office of Nuclear Energy 3 had little to no modeling and simulation capabilities.
4 Today you will hear about state of the art 5 computational capabilities that do not exist within 6 the classical legacy codes used to date by licensed 7 applicants for the regulator.
8 When we began our efforts at developing a 9 suite of world-class tools for analyzing nuclear fuel 10 with reactor technologies, we had many more 11 distractors than supporters.
12 But here we are today, briefing the 13 Nuclear Regulatory Commission's independent advisory 14 body. We still have plenty of distractors, but as the 15 technical community familiarizes itself with our 16 advanced modeling and simulation capabilities, those 17 distractors are finding it harder to argue against the 18 work we are doing.
19 Mr. Chairman, this concludes my opening 20 remarks. I appreciate the interest that the committee 21 has in the Department's work, and I look forward to a 22 very spirited discussion of the modeling and 23 simulation work that my colleagues here are going to 24 present to you this morning and this afternoon.
25 MEMBER MARCH-LEUBA: Shane, Mike has gone NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
12 1 over on the procedures for the microphone and the fact 2 that we speak for ourselves and not the committee. He 3 forgot to tell you that we like to interrupt, and so 4 get used to it. I'm going to be the first one to do 5 it.
6 MR. JOHNSON: Well, you interruption is 7 not an interruption because I'm done. Thank you.
8 MEMBER MARCH-LEUBA: What is the vision?
9 What vision do you have for this cause? I mean is 10 the DOE going to become a vendor, a nuclear vendor 11 that we license the code and then perform calculations 12 for individual companies?
13 Are you going to just provide the FORTRAN 14 code and let them figure out what to do? Do you have 15 a vision for the application? How is this going to be 16 implemented?
17 MR. JOHNSON: Well, for starters, our 18 vision for the codes that we've been producing both 19 for the light water application and the non-water 20 application is simply put together a world-class 21 capability in the codes.
22 Make them available to industry, let 23 industry decide whether or not they see merit in the 24 code, whether they want to use the code moving forward 25 whether it's for their in-house purposes only, whether 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 it's for, you know, informing where the companies are 2 going with their development plans or hopefully, to 3 ultimately be used in licensing space in interactions 4 with the NRC.
5 That's the vision, that's the goal. It's 6 to see a common set of tools that could be used by 7 both licensees and regulator so that the intent isn't 8 in setting up an industry that continues to create 9 codes that are being shopped and sold as mine's better 10 than yours and it's better or whatever.
11 But to have a common set of codes that 12 everyone understands their limitations, their 13 applicability, but that they can be used independently 14 the results of which can be analyzed and gain decision 15 space independent.
16 MEMBER MARCH-LEUBA: My point is 17 historically the DOE, and the staff of NCR can use 18 codes that as long as they themselves feel confident 19 that they're going to have a benchmark and they 20 provide sufficient confidence to provide complementary 21 calculations.
22 But a vendor has to use a license code, I 23 mean it has to have been reviewed and approved by the 24 NRC. Certainly, not for everything and they will have 25 to negotiation but anything that goes into tech specs 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 for the final plan has to have an approved version of 2 the NCR with it.
3 So your vision is not to license the code, 4 you would provide it to the vendors, and the vendors 5 will have to license their own applications.
6 MR. JOHNSON: Correct.
7 MEMBER MARCH-LEUBA: That's a significant 8 amount of work.
9 MEMBER REMPE: Has any vendor said, "Hey, 10 I'd like to take this licensed environment, I'd like 11 to take it for ATF and get it licensed and go 12 forward." Has GE, Westinghouse, or AREVA or 13 Lightbridge said, "We're going to do this"?
14 MR. JOHNSON: No, that has not been said.
15 MEMBER REMPE: Why not?
16 CHAIRMAN CORRADINI: Well, hang on.
17 MR. JOHNSON: Because of the reasons we've 18 said.
19 CHAIRMAN CORRADINI: Why don't we, before 20 we get to the end game which I was figuring we're 21 going to get here. But we're here before -- but I do 22 think, I do think I want to hear from the DOE folks 23 but I do thing Joy, and Jose are raising a question 24 that somewhere today we've got to address which from 25 the industry side who has decided to adopt whatever 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 for their use in a particular application?
2 MEMBER REMPE: Well, where I'm going 3 through with this question is that it seems like DOE 4 has put a lot of money in here and the end game is 5 getting it licensed and approved by NCR.
6 If Westinghouse or AREVA said, "I'd like 7 to do this, I'm willing to be the guinea pig." I 8 mean, they've gotten money for other activities from 9 DOE. Why I'm surprised they're not saying to DOE, I'm 10 willing to do this with the NRC but none of them, they 11 all want to keep using their own codes for ATF instead 12 of doing that.
13 MR. JOHNSON: That's my understanding.
14 MEMBER REMPE: Okay --
15 MR. JOHNSON: But they can use our codes 16 to do the foundational work that then they can migrate 17 over to their approved licensed codes for actual 18 application.
19 MEMBER REMPE: The long-term vision, 20 you've put a lot of money into this in the past. I 21 know a lot of laboratories, I think the laboratory I 22 used to work for put in about a million a year.
23 Of course, they had a three year limit on 24 LDRD, so it was for a different task on the code, they 25 didn't break any laws. But there's been a lot other NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
16 1 money beyond what DOE has put into the code, would it 2 continue?
3 If the NRC were to make this decision, 4 it's a lot; their resources are a lot smaller than 5 DOE's. So knowing how the DOE continuity or their 6 focus on research can change, there's a strong 7 commitment you think in the future that DOE will 8 continue to put in that much money per year?
9 MR. JOHNSON: That's our hope. We won't 10 pretend to know what future --
11 MEMBER MARCH-LEUBA: That will be up to 12 leadership.
13 MEMBER REMPE: Yes, and also, that's the 14 other question.
15 MR. STANEK: I'll just make one quick 16 comment as far this goes, and we will talk about these 17 issues during the technical presentations today. The 18 comments around accident-tolerant fuel are specific 19 to, so in the morning today we'll talk about accident-20 tolerant fuel, and in the afternoon we'll talk about 21 advanced reactors.
22 The interest level in users for the codes 23 that we're developing vary in terms of the maturity 24 level of the concept. And so that's something that we 25 will discuss, but if existing tools are good enough 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 for near-term accident-tolerant fuel concepts, then 2 there's less of an interest in using the code itself.
3 But there are things that we're doing that 4 are, at a research level, that are influencing vendor 5 codes. And so that's for ATF it's a bit more 6 complicated than it is advanced reactors where we'll 7 see probably more full-adoption of tools.
8 I will note that for the advanced reactors 9 we have an industry council that consists of the 10 technical working groups. They've asked us to not 11 mention their own interests specifically that they 12 will do that to you. DOE doesn't license reactors.
13 And so we will today talk about the codes 14 that we're developing and the capabilities. I think 15 we're comfortable in saying that there is interest 16 like Shane said in his remarks. But we will not 17 mention vendors by name as per their requests.
18 CHAIRMAN CORRADINI: So that's interesting 19 because that's essentially what I was going to ask 20 somewhere today which was you had an industry council, 21 what does the industry say?
22 So they've explicitly instructed you not 23 to talk about who's interested in what?
24 MR. STANEK: They want to talk to the NRC 25 themselves.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
18 1 CHAIRMAN CORRADINI: Okay, then let me as 2 at least a different question because you've gotten 3 ahead of me which is good then I don't have to ask my 4 question later about that.
5 Can we get the list of who's on your 6 industry council?
7 MR. STANEK: Of course.
8 CHAIRMAN CORRADINI: Okay, second part of 9 it is. I think a couple of us were involved in the 10 CASL review a few years ago. So we maybe have some 11 insights there.
12 One question that's going to come up and 13 I'm assuming it's going to be somewhere in your 14 presentations, is at what point does verification and 15 validation end in your view and then is passed off to 16 the adopter?
17 Because at least a few years ago that left 18 me personally a bit fuzzy. All right? Because I do 19 think there's going to be a significant amount of 20 validation and verification necessary that might go 21 beyond what you have chosen to do in terms of your 22 programs.
23 So I think somewhere, whether you do it 24 one by one or at the end, I think that's important for 25 us to understand. Particularly, if you're not going 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 to give us the bottom line answer which is A is 2 adopting X. Okay?
3 MEMBER KIRCHNER: May I add, Mike, I think 4 the other thing is how NQA-1-ready, if that's a valid 5 phrase, are these codes at the point where DOE decides 6 that they put enough into them and now it's time to 7 hand it over to a vendor, assuming that there's an 8 interested party out there?
9 And what's the list in terms of cost to 10 get from what's more of a developmental code, some of 11 them I understand are much more mature, to the place 12 they need to be for use in the licensing area?
13 I'll just throw that one thought on it; I 14 won't use my source term, you know, of course, the 15 department is much involved in this as supporting 16 advanced reactor technologies.
17 Something that the commission is looking 18 and this committee will review is things like 19 emergency planning. And going to a, for lack of the 20 right terminology, dose-based approach to determining 21 those kinds of things.
22 The requires a mechanistic source term and 23 how ready those codes are that you're developing for 24 analyses of mechanistic source terms are to be used in 25 the licensing area is going to be very important for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
20 1 any early movers of those advanced technologies.
2 Thank you, Mike.
3 CHAIRMAN CORRADINI: Okay.
4 MEMBER MARCH-LEUBA: Can I ask? I know 5 we're wasting time, Mike, but --
6 CHAIRMAN CORRADINI: You get the last 7 question.
8 MEMBER MARCH-LEUBA: -- okay, good, 9 excellent. It is going to be a long question, though.
10 What's the status of these codes, in the sense, I mean 11 is it going to be open source or is it going to be --
12 my concern is in the '70s the DOE developed some 13 codes, well written.
14 And then vendors on TRAC and the vendors 15 took it and made a couple of modifications, and it 16 became their proprietary code, and that's the one 17 that's licensed.
18 So the work of validating TRAC has been 19 repeated dozens of times, every time somebody uses it 20 with their proprietary modifications. So are going to 21 deliver this code as open source so they can do any 22 modifications and then it becomes their proprietary 23 code or are you going to close the code so that you 24 can do some verification for it? Do you have any 25 plans for 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
21 1 MR. JOHNSON: Great question, and it's a 2 question we've been wrestling with a bit. I think the 3 original intent of many were open source. But for 4 various reasons, we're finding that that's not a good 5 proposition for the future. So most likely these 6 codes will not available open source.
7 But if there are modifications that are 8 necessary or desired in the codes, we'll figure out a 9 mechanism by which we can incorporate those changes.
10 MEMBER MARCH-LEUBA: DOE can do the 11 changes instead of letting them.
12 MR. JOHNSON: Right, the bottom line is we 13 are, we don't want to see the source codes for many 14 aspects of these tools out in the marketplace.
15 MEMBER MARCH-LEUBA: I mean, this year, in 16 October now has slipped to December we've reviewed the 17 obligation of RELAP for a vendor. And I mean RELAP 18 has been for 50 years.
19 But it keeps coming, our proprietary 20 version code X that I have modified for LOCA and it 21 would be nice if you guys don't fall into that trap.
22 Export control, are these codes going to 23 be export control? Any of these slides are export 24 control? We would have an issue with that.
25 CHAIRMAN CORRADINI: No, they're open.
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 This is an open meeting. But since, are you done?
2 MEMBER MARCH-LEUBA: Yes.
3 CHAIRMAN CORRADINI: Okay, so if we start 4 getting into areas that require either proprietary or 5 export control issues you're going to have to tell us 6 because we're going to keep on asking.
7 Okay, I think we're onto Chris. Am I 8 correct?
9 MR. JOHNSON: Yes.
10 CHAIRMAN CORRADINI: Okay.
11 MR. STANEK: Good morning, I should have 12 introduced myself when I was making comments. My name 13 is Chris Stanek. I'm a staff scientist at Los Alamos 14 National Laboratory.
15 On behalf of all us involved in the 16 technical presentations or the briefing today, let me 17 thank the ACRS for the opportunity to present the 18 status of the DOE advanced modeling and simulation 19 codes.
20 Before we begin those technical 21 presentations ,I thought I'd say a few brief words 22 that will hopefully provide some useful context and 23 some of those contextual points have already been 24 discussed this morning. Let me make a few of those 25 comments now.
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 First, as for the guidance from the 2 Chairman and the HERS, today well will limit our 3 presentations to only the advanced code being 4 developed by the DOE.
5 As you all have noticed from the agenda, 6 and as I mentioned previously, we will focus this 7 morning on accident-tolerant fuel, and in the 8 afternoon we will talk about non-light water reactors.
9 Since we have limited our presentations to 10 only DOE codes, we will not cover any other codes that 11 may address similar phenomena to the codes we'll be 12 talking about.
13 However, even though those other codes 14 will not be discussed, we are well aware of these 15 codes and in some cases, very well aware of these 16 codes. And you'll even several instances during the 17 day of how the DOE codes may be being interfaced with 18 these.
19 But we're simply limited today with the 20 amount of information we want to convey. We've 21 established a boundary condition; we'll focus our 22 comments today only codes that we're developing. We 23 have a lot more to say about that other topic, but 24 that's for another time.
25 Like I said, the amount of material 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
24 1 intend to cover today is significant. We've taken 2 care to produce presentations that hopefully provide 3 a clear and concise picture of the codes for you.
4 However, in the case, that additional 5 information is required, or of interest, we're more 6 than happy to provide at a later date.
7 The technical presentations themselves 8 will predominantly focus on the descriptions of 9 capability, status, maturing level of the codes, but 10 not discuss deployment of them and this has already 11 been discussed.
12 Let me say a few words about that. The 13 strategy for deployment is simple. It is that the DOE 14 makes its codes available to the NCR and U.S.
15 companies from which they may generate proprietary 16 versions using their own validation data. There are 17 cases where these commercial licenses are currently 18 being executed.
19 MEMBER MARCH-LEUBA: I like the word may.
20 They may make their own proprietary codes for a large 21 light water reactor, a vendor may have 50 reactors in 22 operation, and for this aggregate fueling, they may 23 want to go this route.
24 MR. STANEK: Right.
25 MEMBER MARCH-LEUBA: But for a small 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 company that is moving a specialized new mobile 2 reactor, they will not have the resources to do that.
3 So it will be nice if DOE has a package that those 4 little companies could use.
5 MR. STANEK: In that division, the may 6 there is also intended to provide companies that will 7 be investing a lot of their own resources and 8 experimental data that they can protect data in a 9 proprietary version of the code. Because there are 10 companies that have expressed that to us and we wanted 11 to make that available.
12 Related to deployment are obviously 13 software quality assurance and the validation basis of 14 the codes that we will discuss. Although in each of 15 the presentation there will be brief mention of these 16 topics, let me say a few words now.
17 Software quality is taken very seriously 18 in the DOE programs, and all of our code development 19 efforts adhere to very strict software quality 20 principles.
21 Regarding validation and hopefully we 22 clear up your fuzziness here, Mike, but our approach 23 is to perform sufficient validation that a user then 24 adopts then code.
25 The user will then very likely, if not be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
26 1 required to perform additional validation for a 2 specific application and that will vary from 3 application to application. We'll hear about a lot of 4 that in the following presentations.
5 CHAIRMAN CORRADINI: So, maybe this is a 6 good time to ask the question. You don't have to 7 answer this question, maybe just think about it. So 8 I have code X, I don't know what it is, and you guys 9 have some sort of experimental benchmarking list that 10 you now check it against.
11 And Company A says I'm very interested in 12 code X, code X is either closed or open ,and I don't 13 appreciate that part. But I'm interested in code X, 14 your idea is to basically pass over the tool as well 15 as the base benchmarking that you've done and leave it 16 to them to decide that's enough for them to take the 17 risk to go further, use it, and then apply for a well 18 I was going to say a license. Essentially, apply to 19 the NRC to get an SER for approval to use in a safety 20 analysis. That's your logic.
21 MR. STANEK: Correct.
22 CHAIRMAN CORRADINI: Okay, so now I have 23 code X, and I have the NRC, is it really appropriate 24 or is there a conflict of interest issue when I take 25 code X, and both vendor and the regulator uses 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
27 1 same tool?
2 Historically, the only time I'm aware of 3 that was done was by congressional fiat for EPAct 2005 4 where they were told that had to do that.
5 Prior to that, it's been historic that, at 6 least in my memory, is that test data might go in 7 every DOE. NRC may say, "This is important test data 8 that we need to understand a phenomena." The data's 9 there, and then the vendor or the regulator will say, 10 "Here's our evaluation model to look at that data and 11 interpret it and then use it."
12 Unless I misunderstand, we're 13 philosophically changing; the suggestion is to 14 philosophically change that sort of paradigm. Am I 15 missing something?
16 MR. STANEK: No, it's a good question and 17 one that we are openly discussing. The DOE role there 18 though I would say is to provide the codes. What the 19 NRC and vendors do is sort of out of our purview.
20 But I think that is a worthy discussion to 21 have; it's one that we've been involved in with the 22 Office of Research in particular. And I'd direct you 23 to them for the better answer than we can give.
24 MEMBER MARCH-LEUBA: Let me put in the 25 record some individual member's opinion about that.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
28 1 In the 1960s, the codes were very simplified by 2 necessity because they had less computer power than 3 your watch.
4 So you had to have a code which use one 5 kinetics and you will want to use a different code for 6 configuratory what it would be then you use for one be 7 and see that they both agree. So they're 8 approximations, you were trying to benchmark the codes 9 with different representations to see if you got the 10 same answer.
11 Codes now are so good, especially on the 12 neutronics side that --
13 CHAIRMAN CORRADINI: Especially when?
14 MEMBER MARCH-LEUBA: On the neutronics 15 side, and I totally into neutronics too. That there 16 is no need to benchmark one A versus one B.
17 Everybody's had three D, the question is do a six-inch 18 node or a two-meter node.
19 So it's more on the applications, there is 20 not so much need of independent codes for 21 complementary analysis. There is more need for 22 independent persons running it, independent 23 assumptions, independent input deck.
24 The code itself becomes so good there is 25 no need to use a different FORTRAN line but that's 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
29 1 opinion.
2 MEMBER REMPE: So I have --
3 MR. STANEK: For what it's worth, we agree 4 with that.
5 MEMBER REMPE: -- so that's, your comment 6 about that you're going to make the codes available to 7 NCR and industry and that you've done sufficient 8 validation is why I asked Shane earlier, what's the 9 vision the DOE has about funding this in the future.
10 It sounds like you guys are about done and 11 you're just going to be a software maintenance 12 organization. And is that the plan, I mean are you 13 going to expand and go to severe accidents too?
14 It sounds like, you know, you've gotten 15 there and you're about done.
16 MR. STANEK: No, I think what you'll hear 17 today, especially in the area of advanced reactors is 18 that there's considerable work to do. That we've 19 developed capabilities that are applicable to the 20 different reactor designs, but there's some key 21 development that needs to be done going forward.
22 MEMBER REMPE: Okay, so you'll help us 23 understand where the gaps are, I mean, some things I 24 can guess, but I just was kind of wondering about the 25 long-term vision.
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 But eventually, I guess there's always 2 something else to analyze, and you'll always continue 3 having a program, that's your hope.
4 MR. STANEK: In the specific technical 5 presentations, there will be mention of the gaps for 6 each of the codes going forward where effort is 7 required.
8 MEMBER REMPE: Okay.
9 MEMBER KIRCHNER: Can I go back to your 10 oh, next to the last bullet or sub-bullet? Which 11 strict SQA principles are they adhering to, NQA-1?
12 MR. STANEK: Yes, and I think I'd leave 13 that for the technical presentations because there is 14 some differences between the codes in terms of --
15 MEMBER KIRCHNER: Is the strategy to pick 16 benchmarking such that the final product, although 17 usually the codes, having coming out of the world, 18 they're never final. They're always being, evolving.
19 But it is going to be, the first order 20 here of strategy or objective to provide for both the 21 NRC, well, or the industry, an NQA-1 ready code. In 22 other words, do sufficient benchmarking against 23 experimental data of relevance such that the code 24 would be to post-order ready to be licensed.
25 Or to be reviewed by the NRC and SER 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 issues, is that the objective or is that going to be 2 left to industry?
3 MR. STANEK: Yes, I would say to first 4 order, that is the goal. We do have NQA-1 advisors 5 working with the programs to make sure that that goal 6 is met.
7 I would say that that's a work in 8 progress. I'm making general statements because there 9 are a number of different codes that I'm referring to.
10 And there will be some discussion of that in the 11 technical presentations regarding the codes and we can 12 follow-up if those we don't get good enough answers to 13 you.
14 MEMBER KIRCHNER: Thank you.
15 CHAIRMAN CORRADINI: Why don't you keep on 16 going, Chris.
17 MR. STANEK: Okay, so we've generated a 18 table that I hope will be a useful key or a legend for 19 the presentations that will follow on accident-20 tolerant fuel today.
21 The reason we've done this is that you'll 22 hear a number of different codenames referred to 23 during the two presentations this morning. And the 24 idea is that this table could be referred to as a 25 guide to associate a code name with a specific 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 capability and its intended use.
2 In the interest of time, I won't introduce 3 the codes. You'll hear plenty about them this 4 afternoon. This table also serves an additional 5 purpose which is essentially a more detailed version 6 of the agenda that we've provided.
7 And so first this morning we will hear 8 from Steve Hayes from Idaho National Laboratory. He 9 will present an overview of fuel modeling for 10 accident-tolerant fuel.
11 Given the importance of fuel modeling for 12 that particular topic, this will take up most of the 13 morning. And so Steve's presentation will be split by 14 the coffee break.
15 After the fuel modeling presentation, Jess 16 Gehin also from Idaho National Laboratory will present 17 both thermal-hydraulics and neutronics efforts for 18 accident-tolerant fuel.
19 CHAIRMAN CORRADINI: Recent, he went west.
20 MR. STANEK: Recently, in Idaho National 21 Laboratory.
22 MEMBER REMPE: So I have question that 23 probably Steve could answer later but whenever I see 24 something like this you have MARMOT and BISON, but 25 later like when you're closing slides you only talk 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 about validation of BISON.
2 And I think when I see in your 3 presentation too Steve, mainly the validation has been 4 done with BISON, not much on MARMOT.
5 MR. HAYES: That's an issue I'm going to 6 address here.
7 MEMBER REMPE: Hard to get data for 8 MARMOT. Right, so I'd like to understand that --
9 MR. HAYES: I will talk about that.
10 MEMBER REMPE: -- and how you can have a 11 validated BISON without a validated MARMOT.
12 MR. HAYES: I'll explain that.
13 CHAIRMAN CORRADINI: So since you 14 volunteered that one, let me see if I can get you to 15 volunteer something else. So looking at this, I see 16 this for normal operation for AOOs, maybe for certain 17 sets of DBAs but I see nothing here that can do beyond 18 design-based as accident source term. Is that a fair 19 statement?
20 MR. HAYES: Yes.
21 CHAIRMAN CORRADINI: Okay, fine.
22 MR. STANEK: If there are no other 23 questions, please allow me to introduce Steve Hayes 24 who will present the status of our fuel modeling 25 effort for accident-tolerant 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
34 1 CHAIRMAN CORRADINI: You can drive for him 2 or are you going to make him drive by himself?
3 MR. HAYES: We're going to change seats.
4 MEMBER MARCH-LEUBA: You can get the 5 mouse.
6 CHAIRMAN CORRADINI: You can just look at 7 the screen, the mouse will let you talk.
8 MEMBER MARCH-LEUBA: You don't have to 9 move, and you can look at his screen. There's nothing 10 on it; you have to close.
11 MR. HAYES: Okay, let me also say thank 12 you. I'm Steve Hayes from Idaho National Lab. We do 13 appreciate this opportunity to address the committee 14 and answer questions.
15 So I'll be talking about ATF fuel 16 modeling. Here's the outline of the presentation I'm 17 going to use, and I'm going to start not specifically 18 on modeling.
19 I'm going to talk about development and 20 testing of a ATF fuels that are going on because 21 that's an important element on validation. How do we 22 foresee getting the data needed to validate these 23 fuels?
24 One slide on the multiscale, mechanistic 25 modeling approach that DOE's taking in advanced fuels 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 modeling because this is the approach we've been 2 working on for nine years and this is the approach we 3 intend to use for ATF.
4 So we'll set the stage there and then just 5 literally three or four slides to overview BISON. I 6 think most people are probably familiar, but if not, 7 we'll spend a few minutes on that and certainly answer 8 any questions.
9 An overview and then specifically we'll 10 update you on the verification validation status and 11 approach. And then the majority of the presentation 12 will be on what we call model enhancements for 13 accident-tolerant fuels.
14 We're modifying BISON to address all the 15 things that we see potentially coming, doped UO2, 16 chrome-coated zirconium, FeCrAl cladding, silicon 17 carbide cladding, silicide fuel, and non-cylindrical 18 metallic fuels.
19 And then we'll end with some remarks about 20 validation, although we'll say some things along the 21 way, and summary and conclusions.
22 Okay, starting with development and 23 testing of accident-tolerant fuels. Just to set the 24 stage here, we all know that this got started back 25 shortly after Fukushima.
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 The Fukushima accident was in 2011 and the 2 very next year, 2012, Congress directed the Department 3 of Energy to begin developing fuels that would offer 4 some enhanced accident-tolerance in accidents like 5 Fukushima and some other situations.
6 And the direction from Congress was pretty 7 specific. They challenged the DOE to insert lead test 8 rods or lead test assemblies of an ATF concept into an 9 operating commercial light-water reactor within ten 10 years.
11 This came out in 2012 and DOE took it as 12 our direction to have LTAs or LTRs of some ATF concept 13 in an operating reactor by 2022, three years from 14 today.
15 We regarded that as very challenging at 16 the time. So DOE obviously not taking a lead role in 17 developing and licensing LWR fuels historically, 18 immediately reached out to the industry through what 19 they call, a funding opportunity announcement.
20 Solicited industry teams to propose 21 concepts that DOE could sponsor. And the three 22 concepts or teams that were selected are shown on this 23 slide. So these are not my most recent slides, sorry 24 about that.
25 These are the three teams: Framatome, 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 General Electric, and Westinghouse. And from left to 2 right they go in alphabetical order, but there's 3 another logic to the order too.
4 And that is from left to right the 5 development and qualification licensing process would 6 appear to become more challenging as you proceed from 7 left to right.
8 So Framatome, primarily in the near-term 9 is working on chrome-coated Zircaloy cladding and 10 doped UO2 concepts for improved thermal-connectivity 11 and fuel performance.
12 Now they also are working on a longer-term 13 track on silicon carbide cladding. I don't list it 14 here because that's a longer-term option for them.
15 General Electric is working on iron-based 16 cladding. This is the iron-chrome-aluminum variance, 17 FeCrAl, they call it. So that's their main line, 18 although I will say they have in more recent years 19 also started work on coated cladding concepts as well.
20 And then Westinghouse has got a suite of 21 things they're working on. They are working on 22 chrome-coated Zirlo cladding. They're working on 23 silicon carbide cladding on perhaps a more aggressive 24 track than Framatome which is the reason it makes this 25 slide.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
38 1 And then completely new fuels that offer 2 improved thermal-connectivity and hopefully higher 3 density. So that's the Westinghouse concepts.
4 Now all three of these teams have multiple 5 partners with them. They have utility participants, 6 and some of them have lab participants.
7 MEMBER MARCH-LEUBA: That's what the 8 icons, the lab icons show?
9 MR. HAYES: Exactly. So this may be a bit 10 dated. For instance, Idaho and Los Alamos were a 11 formal part of the Westinghouse team when it was 12 awarded, when the FOA was awarded.
13 Oak Ridge was a formal part of GE. Those 14 are still true, but now you'll find the labs being 15 directed to support all of the teams in certain areas 16 as they request.
17 And that's the other thing to be said 18 here. So under this FOA relationship that DOE has 19 with these three industry teams, it really is the 20 industry teams that are leading the development, 21 qualification, ultimately licensing of their own ATF 22 concepts.
23 The DOE and the labs working under the 24 direction of DOE are supporting the vendors where they 25 need and ask for support and providing testing of 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
39 1 the concepts.
2 MEMBER BALLINGER: I have a question about 3 the little item at the bottom. Have you traced the 4 ownership of Lightbridge to its origin?
5 MR. HAYES: Yes.
6 MEMBER BALLINGER: What is it?
7 MR. HAYES: So this is related to my 8 comment, we had thought we provided a revision of 9 these slides that don't have Lightbridge on it. DOE 10 doesn't have a formal relationship with Lightbridge, 11 so we're not really prepared to talk about 12 Lightbridge, I think.
13 MEMBER BALLINGER: So you pass.
14 MR. HAYES: I'll pass.
15 MEMBER REMPE: But, again, I'm not sure I 16 saw it in this slide later on. You have put in BISON, 17 some Lightbridge-type work. Why did you do that if 18 you have no formal relationship? Yes.
19 MR. HAYES: Yes, so this is what is meant 20 by non-cylindrical metallic fuel. We see some 21 entities out there proposing concepts, like a 22 Lightbridge, that may come along.
23 And so we're informing the committing on 24 DOE's tools capability, the capability of DOE's tools 25 to address concepts like 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
40 1 MEMBER REMPE: Okay.
2 MR. HAYES: Okay, so I mentioned the 3 industry teams have the lead in developing their 4 concepts. DOE and especially the lab are providing 5 testing of all the concepts, and very early in the ATF 6 development cycle DOE stood up a conceptual radiation 7 testing program to support the development of ATFs.
8 And you see here, each of these columns 9 represent what we call a test series. So ATF-1 was 10 the first one stood up. It's a test series in the 11 advanced test reactor at Idaho.
12 It makes use of drop-in, double 13 encapsulated testing configurations. So it's not a 14 prototypic condition where the cladding is exposed to 15 coolant. But it is appropriate for collecting some 16 very early data on new fuels, fuel cladding 17 interactions.
18 And this test series began in 2015; it's 19 still ongoing. We've gotten some of the early low 20 burn-up test articles out already but many are going 21 on to higher burn-up, and we expect that new test 22 articles will be added to this. So this isn't just 23 one experiment, it's a test series that continues.
24 ATF-2 is a prototypic PWR condition. So 25 this is in a pressurized water loop in the ATR. 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
41 1 designed to test integral fuel and cladding concepts 2 under PWR conditions.
3 It just started this year in May, at the 4 end of May, early June the loop went operational with 5 test articles in it, and this is going to continue for 6 some time.
7 And it's designed so that new articles, 8 new rodlets can be introduced as time goes one and 9 rodlets will be removed for examination.
10 MEMBER REMPE: So Steve, hold on for a 11 second, let's ask some questions about ATF-2. Which 12 vendor fuels from the prior slide are in ATF-2 now as 13 rodlets? And which will be added in the near-term?
14 I also have a lot of other questions about 15 what data are you going to get out of ATF-2 ever?
16 MR. HAYES: So ATF-2 when it started at 17 the end of May, not every concept was ready to take 18 off at the same time. Westinghouse has coated 19 cladding concepts in, Framatome has coated cladding 20 and doped UO2 in. So that's what's in it today. The 21 first few rodlets.
22 MEMBER REMPE: And those are the 23 prototypic ones that you showed on the prior slide 24 from Framatome and AREVA, Westinghouse?
25 MR. HAYES: So Framatome has all of their NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
42 1 stuff in ATF-2 currently. Westinghouse has coated 2 Zirlo in there now. GE is expected to deliver FeCrAl 3 cladding test articles in the upcoming year.
4 Hopefully, by about January.
5 Westinghouse is expected to deliver some 6 fuels and possibly even some silicon carbide cladding 7 that specimens next year.
8 MEMBER REMPE: Okay, and then what data, 9 I think I heard you should have a thermocouple in ATF-10 2? That's it?
11 MR. HAYES: So ATF-2 is capable of doing 12 instrumented fuel tests, in fact, before we started up 13 ATF-2 we did what was called SQT, sensor qualification 14 test, which was the ATF-2 test train without fuel but 15 with instrumentation.
16 And thermocouples in fuels are possible, 17 LBDTs for measuring gas pressure and gas release are 18 possible, but none of those have been implemented yet 19 in the test articles that are in ATF-2.
20 CHAIRMAN CORRADINI: So I'm, Joy is much 21 more familiar with this than many of us, so I'm not 22 exactly clear if the instrumentation is just sitting 23 there what exactly is AFT-2 looking at?
24 MR. HAYES: It's testing miniature fuel 25 rodlets.
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 CHAIRMAN CORRADINI: But what is being 2 measured? More integral quantities?
3 MR. HAYES: So there's currently no in 4 situ measurement.
5 CHAIRMAN CORRADINI: Oh, it's post-test.
6 MR. HAYES: These are going to be taken --
7 CHAIRMAN CORRADINI: Okay.
8 MR. HAYES: -- to target burn-ups and then 9 taken to the hot cell for destructive examines.
10 CHAIRMAN CORRADINI: Okay.
11 MEMBER REMPE: So again, because I'm 12 getting ready to talk about the Halden issues. Where 13 are you going to get the data for PWR conditions since 14 Halden is going to shut down?
15 MR. HAYES: So you see by --
16 MEMBER REMPE: Yes, but that only says BWR 17 conditions.
18 MR. HAYES: I know.
19 MEMBER REMPE: And it used to be Jon 20 Carmack's flight had ATF-HX last time we heard him had 21 P and B, and you've changed it to just B. So where 22 are you going to get the P stuff?
23 MR. HAYES: Exactly, so all the vendors 24 anticipated doing some testing in Halden, maybe quite 25 a bit. Obviously, with the shutdown of Halden, Halden 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 needs to be, these tests need to be redirected.
2 And actually, the DOE's had an activity 3 over the last couple of months looking at that. Some 4 of these needed tests that were expected to be in 5 Halden have to be moved. Some of them are going to be 6 moved to ATR. Some of them might be moved to other 7 places.
8 Cladding only corrosion testing can 9 probably be done at MIT. The BR-2 reactor may do some 10 of these tests. That's something that we're working 11 through right now. We don't have every answer, but we 12 see ways to get there.
13 MEMBER REMPE: So, okay, you said, "Well, 14 we put LBDT into ATF-2." I didn't hear you say we can 15 put a diameter gauge into ATF-2. I mean the 16 standardized test rigs I didn't see come out of the 17 draft report you guys issued this summer about what 18 you're going to do about Halden.
19 MR. HAYES: Exactly.
20 MEMBER REMPE: And that needs to be done 21 where you can get all of the data real-time, not just 22 cook-and-look.
23 MR. HAYES: You know, I would challenge 24 that a little bit. The in situ instrumentation is the 25 preferred way to go. We don't question that. But as NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
45 1 we look at where to move things and get the same sort 2 of data that's not Halden, it's not going to be able; 3 it's not going to be possible to reproduce everything 4 that Halden does somewhere else. We may have to skin 5 the cat in a little bit different ways in some cases.
6 So the ATF-2 loop is not set up to make 7 online cladding strength measurements. So we may have 8 to get that at interim steps in the canal or even the 9 hot cell.
10 MEMBER BALLINGER: With respect to Halden 11 as NQA-1, the other two that you mentioned are not.
12 MR. HAYES: I won't speak for BR-2, but I 13 would say in ATR we are fully capable --
14 MEMBER BALLINGER: Okay, I was not 15 thinking about MIT and BR-2.
16 MR. HAYES: Okay.
17 MEMBER BALLINGER: I think ATR, I'm 18 assuming it's in NQA-1.
19 MR. HAYES: Yes, ATR is certainly capable 20 of doing NQA-1.
21 MEMBER BALLINGER: So if you have to go to 22 these other places other than ATR, how do you deal 23 with the quality control issue?
24 MR. HAYES: So we are talking with the NRC 25 along those lines ,and we dealt with it a little bit 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 in ATF-1 and ATF-2 fabrication of test articles which 2 in some cases vendors did and some cases labs did, and 3 in some cases, even universities contributed.
4 And so the way we've made it work is the 5 vendor sponsoring the work has to come in and cover 6 all of its subcontractors with its QA program and do 7 whatever it takes to elevate their product.
8 MEMBER BALLINGER: So the DOE would have 9 to cover the cost of upgrading or whatever needs to be 10 done to NQA-1 for wherever you go, BR-2, MIT, 11 whatever.
12 MR. HAYES: Well, the upgrade would be 13 necessary. I'll stop short of saying who ought to 14 cover the cost, perhaps DOE, perhaps the vendors.
15 MEMBER REMPE: So I didn't, again, that 16 will be something that you and the regulator can argue 17 about cook-and-look when you take the fuel out and try 18 to decide about swelling and all of that without a 19 diameter gauge real-time.
20 But what about thermo-connectivity 21 degradation, how are you going to get that?
22 MR. HAYES: So as I've said, it is 23 possible to have thermocouples in ATF-2 so we can 24 approach it that way. We also, and I'll speak to this 25 at the end, we also plan to do thermo-connectivity 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 measurements of irradiated fuel specimens in the hot 2 cell.
3 MEMBER REMPE: Okay.
4 MR. HAYES: Okay, so that's ATF-1 and 2, 5 ATF-3 is also a test theory starting up probably next 6 month. Just make '19 by a hair's breadth. And this 7 is a test series to be conducted in TREAT, both the 8 static and the loops. And so being conducted in 9 treat, this is looking specifically at off-normal 10 conditions.
11 MEMBER BLEY: How off-normal are you 12 pushing it?
13 MR. HAYES: So TREAT will look at both 14 LOCAS and RIAs, as severe as someone wants them to be.
15 MEMBER BLEY: As someone wants them to be?
16 MR. HAYES: Obviously, that's not the 17 near-term focus. The near-term focus would be on less 18 aggressive testing.
19 MEMBER BLEY: Okay.
20 MR. HAYES: And of course what's going to 21 start up later this year is probably just some 22 shakeout testing on UO2 and Zirc. But next year, in 23 '19, the first ATF concepts will be tested.
24 These will be unirradiated ATF concepts 25 but the idea is as time progresses rodlets that 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
48 1 condition in ATF-1 and ATF-2 will be tested in TREAT.
2 We've already talked about the Halden test 3 series that has to find a new home. There's a test 4 series out here called commercial reactors, and that's 5 LTRs and LTAs. And arguably that started this year, 6 but it will certainly start next year with no dispute.
7 And then --
8 CHAIRMAN CORRADINI: So if I might just 9 interrupt, for the commercial, maybe I'm 10 misremembering but what's gone in right now are lead 11 test rods unfueled. Correct?
12 MR. HAYES: Correct.
13 CHAIRMAN CORRADINI: Okay, you'll come 14 back to that.
15 MR. HAYES: I'll address that seriously in 16 about two slides. And then this test series out here 17 envisions rods or rod segments that come from the 18 commercial irradiations being refabricated, 19 instrumented, and going through TREAT testing in the 20 future.
21 MEMBER BALLINGER: Excuse ignorance, but 22 are there any parallel test programs that are actually 23 being wholly fund by the vendors?
24 MR. HAYES: I do not believe so.
25 MEMBER BALLINGER: Not even the chrome 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 coating?
2 MR. HAYES: The vendors working through 3 the Halden joint program, I think, got a few things in 4 the Halden over the last couple of years. That's 5 probably the only thing that would fall into that 6 category.
7 MEMBER BALLINGER: None in Europe?
8 MR. HAYES: Well, so Framatome is doing, 9 has done a lot of doped UO2 testing in Europe. Okay, 10 that is true. And Westinghouse may have done some 11 chrome coated cladding in Europe.
12 MEMBER REMPE: Later on I was going to ask 13 this question but when they just do --
14 MR. HAYES: But that's somewhat 15 speculation for me, you should address those questions 16 to the vendors.
17 MEMBER REMPE: Later in your talk, I 18 thought I saw some comparisons of data from the doped 19 UO2 but it was not with the cladding for ATF is what 20 I was kind of wondering.
21 MR. HAYES: That's correct.
22 MEMBER REMPE: And so how they did because 23 of high burn-up fuel interest probably. Right?
24 MR. HAYES: Right. So this is an overview 25 of the irradiation testing that DOE is performing 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
50 1 plans to perform on ATF concepts. There's also LOCA 2 testing that can be performed outside of the hot cell, 3 outside of the reactor in a hot cell.
4 So in recent years DOE is has stood up a 5 LOCA testing station in the hot cell at Oak Ridge that 6 is modeled after a similar test stand that has been 7 used for years at Argon.
8 So that's going to provide testing of ATF 9 concepts as well. And then this slide talks about or 10 highlights the near-term plans for leap test rods or 11 assemblies in commercial reactors.
12 So here we are in '18, and as Dr.
13 Corradini mentioned GE already this year, I think in 14 February, put in some FeCrAl cladding tubs, they did 15 not have fuel in them, into the Hatch reactor. But 16 they have plans to come back next year with LPAs that 17 include ironclad, which is their name for their FeCrAl 18 cladding now.
19 As well as ARMOR which is their name for 20 coated Zirc cladding that will have fuel in it in the 21 Clinton reactor.
22 CHAIRMAN CORRADINI: But maybe I've lost 23 it, what is the coating?
25 CHAIRMAN CORRADINI: 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
51 1 MR. HAYES: I actually don't know what it 2 is.
3 CHAIRMAN CORRADINI: It hasn't been 4 revealed?
5 MR. HAYES: It has not been revealed. You 6 would need to direct that question to them.
7 CHAIRMAN CORRADINI: Okay, that's fine.
8 It was new to me, so I thought maybe you knew.
9 MR. HAYES: I do not know.
10 CHAIRMAN CORRADINI: It had a nice name.
11 MR. HAYES: Yes, it has a great name.
12 Westinghouse this year established a fabrication line 13 at Idaho for making silicide fuel. And next year 14 their plans currently have beginning testing of 15 chrome-coated ZIRLO and silicide fuel rod segments in 16 Byron.
17 MEMBER REMPE: I was curious about this, 18 they are just starting to get some fuel made out there 19 in at MFC and now they're going to directly take it to 20 Byron without any lab testing or ATR testing first?
21 MR. HAYES: No, no, silicide fuel has 22 been, is in ATF-1, it's been irradiated --
23 MEMBER REMPE: Who made the fuel that's in 24 ATF-1, is 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
52 1 MEMBER REMPE: Idaho made it, so this is 2 like a larger fabrication line?
3 MR. HAYES: Exactly.
4 MEMBER REMPE: Okay, I didn't know, I was 5 curious what was going on.
6 MR. HAYES: This fabrication line would be 7 big enough to make full-size rods --
8 MEMBER REMPE: Okay.
9 MR. HAYES: -- for a commercial reactor.
10 The fuel that's already under irradiation in ATF-1 and 11 we already have some preliminary data for silicide 12 fuel, was made at Idaho but not in the big line.
13 MEMBER REMPE: Okay.
14 CHAIRMAN CORRADINI: So can we go back a 15 slide? I'm sorry, two slides. So are these all 16 steady-state tests?
17 MR. HAYES: So ATF-3, the TREAT test 18 series are transient.
19 CHAIRMAN CORRADINI: I'm sorry, TREAT is 20 that. So are they transient with different time 21 scales? In other words --
22 MR. HAYES: TREAT is capable of quite a 23 spectrum of timescales.
24 CHAIRMAN CORRADINI: So it's in the ATF-3 25 plan to do power bursts as well as to do slow-ramp 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 ups and ramp downs? You know what I'm asking?
2 MR. HAYES: I know what you're asking, and 3 we've looked at TREAT for the slower ramps and --
4 CHAIRMAN CORRADINI: Historically, they've 5 published data on that.
6 MR. HAYES: There is some that can be done 7 in TREAT, and we'll try to make use of TREAT in that 8 area to the greatest extent possible, but for doing a 9 classical ramp test, TREAT is probably not the right 10 one. The kind of ramp testing that Halden does, TREAT 11 is probably not the right one to do that.
12 CHAIRMAN CORRADINI: Right, I'm kind of 13 sandbagging you because I was at a meeting and one of 14 the NRC staff at the meeting made this point. So then 15 what are you going to do?
16 MR. HAYES: So this is all part of our 17 Halden capability gap assessment and path forward 18 work, but for classical ramp testing, BR-2 may have 19 some capability. We think we can do something in ATR.
20 MEMBER REMPE: You have PALM cycle with --
21 MR. HAYES: Using a PALM-type device.
22 CHAIRMAN CORRADINI: I don't even, I'm 23 sorry.
24 MEMBER REMPE: Our actual locating --
25 MR. HAYES: So we're into --
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 CHAIRMAN CORRADINI: If we're into detail 2 we can wait until later.
3 MR. HAYES: Okay.
4 CHAIRMAN CORRADINI: But you answer my 5 major question as to where does it fit.
6 MR. HAYES: We haven't forgotten ramp 7 testing, but it's probably not TREAT.
8 MEMBER BALLINGER: I have question about 9 this, which one of these concepts would require 10 exceeding the 5 percent enrichment limit?
11 MR. HAYES: The one that probably 12 challenges it would be GE with the FeCrAl cladding.
13 So a lot of the work that GE's doing in the labs are 14 helping them.
15 It's how thin can you go on the cladding 16 and have enough strength and get a good weld. You 17 increase the fuel diameter to try to buy back some of 18 that neutronic penalty.
19 It's not clear you can get all the way 20 there in the same assembly design so GE, I think if 21 you talk to them, they might say they're looking at an 22 alternative assembly design to try to save a load of 23 5 percent.
24 That's the one that challenges it 25 potentially, and they're trying to make that 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
55 1 happen.
2 CHAIRMAN CORRADINI: We're a little bit 3 off topic, but it's his fault so I'll pile on. The 5 4 percent is more a licensing of the fuel fan facility 5 due to the necessary experiments versus an actual hard 6 limit because there a supplier --
7 MR. HAYES: And transportation.
8 CHAIRMAN CORRADINI: -- and 9 transportation.
10 MR. HAYES: That's right.
11 CHAIRMAN CORRADINI: Okay, fine.
12 MR. HAYES: That's right those are the 13 real concerns. It's just the activation energy to get 14 over that.
15 MEMBER REMPE: So before you switch to the 16 modeling thing when Jon Carmack was here last time, I 17 brought up a concern about what are you going to do 18 with ATF fuel if you get to a high temperature than 19 the control rod, will they liquefy?
20 And, you know, but what's the plan here?
21 Are you going to reflood and all that? Have you guys 22 done anything else about that?
23 CHAIRMAN CORRADINI: With all due respect 24 to my colleague, they went away at TMI before the 25 current fuel, so they go away.
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 MEMBER REMPE: Yes, but if we're claiming 2 we have accident tolerance now, and we have, you know, 3 reduced risk. We have increased safety. What I'm 4 kind of wondering is, do we have an increased concern 5 about an atlas?
6 MR. HAYES: So actually I would direct you 7 to EPRI who's doing a lot of studies in this regard 8 using MAP and MELCOR and from what I've seen, control 9 rods may not be the second thing to go.
10 MEMBER REMPE: May not be the second 11 thing?
12 MR. HAYES: Second thing to go, they may 13 not be.
14 MEMBER REMPE: Well, I was concerned 15 they'd be first thing to go because you're at lower 16 temperatures, so you have a reactor with intact 17 geometry.
18 MR. HAYES: EPRI is doing those analyses 19 but from what I've seen them present --
20 MEMBER REMPE: The control rods are going 21 to stay.
22 MR. HAYES: The controls rods do not go 23 before fuel.
24 MEMBER REMPE: I mean at TMI, as Mike 25 mentioned the control rods went, so I'm real curious 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 because we used to talk about after the tolerance.
2 MR. HAYES: So DOE is not doing that 3 assessment.
4 MEMBER REMPE: Okay, somebody probably 5 needs to do something on that.
6 CHAIRMAN CORRADINI: But I mean we don't 7 want to design your research program for you so we'll 8 keep on going.
9 MR. HAYES: The last one just so I don't 10 forget, Framatome is working this area as well so 11 Framatome has already had some doped UO2 fuel in 12 LaSalle for a period of time.
13 And they're actually doing some pool-side 14 exams on it this year, and next year they have plans 15 to start chrome-coated in Vogtle testing in both.
16 So legitimately by next year, 2019, 17 several years ahead of schedule, all the ATF teams 18 will have things in commercial reactors.
19 MEMBER REMPE: But what was cladding the 20 on the chromia-doped fuels?
21 MR. HAYES: It was nothing unorthodox.
22 MEMBER REMPE: So it wasn't the ATF 23 cladding.
24 MR. HAYES: It was not the ATF cladding.
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
58 1 MR. HAYES: Exactly. Okay, that's all --
2 CHAIRMAN CORRADINI: You did a great job.
3 MR. HAYES: -- the experimental background 4 to set the stage. And now we're moving into the 5 modeling area. And specifically for the next few 6 slides, we're going to be talking about DOE's approach 7 which is multiscale, mechanistic modeling of nuclear 8 fuels.
9 We use this slide a lot, you've probably 10 seen it, and we use it to try portray what we mean by 11 multiscale, mechanistic modeling of nuclear fuels.
12 The objective is to use hierarchical multiscale 13 modeling for improved mechanistic development of 14 models of fuel performance.
15 Okay, we're emphasizing mechanistic fuel 16 behavior models because we think if we have 17 mechanistic models that brings at least three 18 benefits.
19 Number one is even if some of these 20 mechanistic models require some calibration they 21 should minimize form errors because they have physics 22 built in they should provide insight where 23 experimental data may be sparse.
24 And this remains to be seen, but the 25 proposition is they may require less data for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
59 1 validation. We're not saying no data; we're saying 2 less data.
3 MEMBER REMPE: Is that because of the 4 physics-based approach?
5 MR. HAYES: Exactly.
6 MEMBER REMPE: Which you can't get data 7 from MARMOT because so far your comparisons are with 8 the integral data. Right?
9 MR. HAYES: So let me address that issue 10 right now. That's what I'll do here. So these are 11 the three scales that we typically talk about.
12 The engineering-scale is the scale at 13 which most of your classic fuel performance codes 14 operate at. That's the scale at which BISON operates.
15 So BISON is in that model of classical fuel 16 performance codes. Okay?
17 You're meshing up pellets and cladding, 18 and you're getting engineering scale predictions of 19 cladding diameter changes and fuel swelling and 20 fission gas release and perhaps a probably of cladding 21 breach. Things like that.
22 Now BISON by itself is just another fuel 23 performance code, it has some advantages, but the real 24 advantages come when you marry it to these lower-25 length scales.
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 And the two lower-length scales we talk 2 about are the mesoscale, that's what the modelers like 3 to call it. I call it the scale of microstructure.
4 You're resolving microstructure, and that tool that 5 DOE has developed to do that is MARMOT.
6 Okay, what MARMOT does is it attempts to 7 simulate the micro-structure of fuel and how it 8 changes, how it evolves under radiation.
9 CHAIRMAN CORRADINI: So let me ask you 10 this, this is, I know colleagues that do this but I 11 don't get it. So is MARMOT going to tell me fuel 12 thermo-connectivity degradation theoretically?
13 MR. HAYES: Yes.
14 CHAIRMAN CORRADINI: Therefore I need an 15 experiment to validate that. I'm just --
16 MR. HAYES: Yes and no.
17 CHAIRMAN CORRADINI: -- okay.
18 MR. HAYES: I'm about to address it.
19 CHAIRMAN CORRADINI: All right, because 20 I'm just trying to nail down some properties --
21 MR. HAYES: You're right.
22 CHAIRMAN CORRADINI: -- at the grain that 23 I get.
24 MEMBER BLEY: Before you go ahead, Steve 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
61 1 MR. HAYES: Yes.
2 MEMBER BLEY: -- I'm not, we no longer 3 have on the committee I think the person to ask the 4 question I want to poke at here a little.
5 The detailed chemistry of what's going on 6 here affects the situation well, and that's not 7 physics, that's chemistry. Is that being picked up 8 and modeled?
9 How well is that covered because as I 10 understand it, this gets very complex on the chemicals 11 side?
12 MR. HAYES: You're right, and there are 13 some areas where we look at chemistry more explicitly, 14 like in the metal fuels area.
15 MEMBER BLEY: I'm kind of hanging on where 16 we heard how good the codes are now.
17 MR. HAYES: But to be honest --
18 MEMBER BLEY: And if we miss something in 19 this area I don't know how good the codes are at all.
20 MR. HAYES: -- yes, there's not a huge 21 amount work we've done on the chemical area.
22 MEMBER BLEY: It seems an area where we 23 could go very wrong from the things I hear from 24 experiments that, I'm not an expert there at all.
25 CHAIRMAN CORRADINI: I mean, I think where NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
62 1 Dennis is kind of, I know the colleague that we're 2 missing, and he'd at least have ten questions by now.
3 But at least from my perspective, I'm just 4 looking for an experimental basis such that if I see 5 a deviation from the theoretical basis, I can then 6 back out a) an explanation or at least b) an empirical 7 correlation to modify.
8 MEMBER BLEY: Which is necessary but you 9 need more experiments than one to cover these other 10 sorts of things.
11 CHAIRMAN CORRADINI: Then you would 12 planted.
13 MEMBER BLEY: Maybe a lot yes.
14 MR. HAYES: So let me get there in about 15 a minute.
16 MEMBER BLEY: Sure, that's fine, that's 17 fine.
18 MR. HAYES: I'm not ignoring you.
19 CHAIRMAN CORRADINI: By the way, you can 20 tell us to be quiet.
21 MR. HAYES: Oh no, we're here to answer 22 your questions. Seriously, that's what we want to do.
23 Okay, so MARMOT is this tool that simulates how micro-24 structure changes under radiation.
25 And the real advance in recent years has NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
63 1 been if I know what the microstructure looks like, 2 even more so than chemistry, if I know what the 3 microstructure looks like then I can accurately tell 4 you what the properties of material are going to be 5 and what the fuel behaviors are going to be.
6 Okay, that's the proposition. Now as I 7 say that, MARMOT is a much more difficult tool to use 8 than BISON. It requires a lot of inputs that are 9 difficult to come up with experimentally.
10 And so that's the importance of these 11 atomistic simulations. Now DOE NE is not developing 12 any new tools to do atomistic simulations. We're 13 using tools that are out there and well-known to 14 people who work in this area.
15 And we're using them in a couple of 16 different ways. In some areas, they do help us 17 identify important mechanisms. But in every case they 18 allow us to calculate material parameters at the 19 microstructure scale that are important inputs to 20 MARMOT.
21 So they really let MARMOT do what it does.
22 And MARMOT tells us how micro-structure changes and 23 then based on that we assess how properties and 24 performance change.
25 And then we build new models based 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
64 1 understanding which we implement in BISON. Okay? It 2 is possible to run MARMOT and BISON in a couple 3 fashion, but that's the main way of doing business.
4 The main way of doing business is doing 5 the science and the simulations here at the lower-6 length scales, and building a mechanistic models that 7 then gets implemented in BISON that can bring some of 8 that physical understanding to the engineering scale.
9 MEMBER KIRCHNER: Now for somebody who's 10 very naive in this area, you're getting information 11 out of MARMOT that leads to modeling and BISON.
12 That's an offline process or is this somehow built 13 into the routine?
14 MR. HAYES: No, that's an offline process, 15 that's a lot of work by expert scientists, exactly.
16 MEMBER BLEY: And a lot of that --
17 CHAIRMAN CORRADINI: So Steve in MARMOT a 18 lot of it is presumed in advance, like the 19 microstructure I would assume. So that you can evolve 20 --
21 MR. HAYES: Exactly.
22 CHAIRMAN CORRADINI: -- analysis on grain 23 boundaries and --
24 MR. HAYES: Exactly, so the input 25 conditions to a MARMOT simulation are all these 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 things.
2 CHAIRMAN CORRADINI: -- What does the 3 micro-structure look like at time zero? I may have 4 more to say about that at the end.
5 So to address this issue of validation.
6 So everyone assumes that I have to do all my 7 validation on a new mechanistic model at the MARMOT 8 level.
9 First comment there is we are doing as 10 much validation at the MARMOT level as we can. Some 11 of the experiments needed to validate some of these 12 results are difficult and complicated.
13 And we're undertaking some of them, some 14 of the we just can or won't. That's doesn't mean that 15 the models developed that this scale are not 16 validation.
17 They get implemented in BISON and then 18 BISON is validated on an integral level with these new 19 models in it.
20 MEMBER REMPE: Is the empirical 21 correlation you put into FRAPCON or FRAP-T?
22 MR. HAYES: It depends on the property, 23 but for in since in the area of thermo-connectivity 24 it's nothing like it.
25 MEMBER REMPE: Nothing like it 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
66 1 MR. HAYES: In the area of gas behavior 2 it's nothing like it.
3 MEMBER MARCH-LEUBA: I'm sorry, is that 4 how you present it?
5 MR. HAYES: Pardon me?
6 MEMBER MARCH-LEUBA: Does it not reproduce 7 today?
8 MR. HAYES: No, no, it does reproduce 9 today. What I'm saying in response to Joy is the form 10 of the, it's not just a polynomial fit to a new set of 11 data that gets put into the BISON.
12 It's a more sophisticated formulation that 13 attempts to retain as much as the physics from the 14 lower-length scales as it can.
15 MEMBER MARCH-LEUBA: But again, it gives 16 you the same answer.
17 MR. HAYES: Or better.
18 MEMBER REMPE: Well, it gives you answer 19 between two data points you have. A long time ago we 20 were going to do a science-based approach, and we 21 could go and extrapolate beyond where the data is.
22 But right now, frankly, if you're saying 23 you're validated, you have the same problem that you'd 24 have with FRAPCON that you have to stay between the 25 upper and lower bounds for where you've validated 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
67 1 because it sounds to me, and maybe I'm putting words 2 in your mouth, you can't get all the data you need 3 from MARMOT to validate those models you put into 4 BISON.
5 So we shouldn't be going and extrapolating 6 beyond their data. Right?
7 CHAIRMAN CORRADINI: But I mean --
8 MR. HAYES: Not for a regulatory 9 discussion.
10 CHAIRMAN CORRADINI: So let me heretical, 11 I'm sure there's somebody back in Idaho who's going to 12 whack me good on this. If MARMOT is a stand-alone, so 13 a neutronically it's like I built all my macroscopic 14 squash sections then I go do my criticality 15 calculations.
16 MR. HAYES: It's a cross-section 17 generator.
18 CHAIRMAN CORRADINI: Right.
19 MR. HAYES: Exactly.
20 CHAIRMAN CORRADINI: Okay, so that could 21 be put into FRAP-T or FAST.
22 MR. HAYES: Exactly, so there is truth to 23 what you're saying there. The mechanistic models that 24 are developed using MARMOT and these lower-length 25 scale techniques, we are implementing them in BISON 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 and then doing the bulk of our validation at the BISON 2 level.
3 But someone else could implement those 4 models in another code.
5 CHAIRMAN CORRADINI: Okay.
6 MR. HAYES: That is true, and has happened 7 in some instances.
8 CHAIRMAN CORRADINI: All right, I'm not 9 aware. The only reason I as is the analogy in 10 neutronics seems very similar that you're going to do 11 an offline pre-calculation, what I'll call needed 12 macroscopic properties, and then do the --
13 MR. HAYES: That's absolutely correct.
14 That's a good analogy.
15 MEMBER REMPE: If that's true then, the 16 external inputs are the same, you'll look at burn-up, 17 you'll look temperature of the core, things like that 18 if I were trying take this new formulation and put it 19 into FRAPCON or FRAP-T.
20 There's nothing else you're requiring 21 that's not in FRAP-T or FRAPCON?
22 MR. HAYES: No, that's not always the 23 case. In some cases, that may be the case, but in 24 other cases it's far from the case. So, for instance, 25 thermo-connectivity and its degradation.
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 The model or correlation in a legacy code 2 may be, you know, temperature, and O-to-M, and burn-3 up, and that's all you need. And for the assessment 4 in the models, we use you need some micro-structural 5 information that the legacy codes may not have.
6 MEMBER REMPE: Such as.
7 MR. HAYES: Grain size, grain size 8 distribution, porosity, not just a value but some 9 knowledge about --
10 MEMBER REMPE: So you have knowledge about 11 that for the whole core, all the fuel in it? I mean, 12 are you just taking an average value for porosity? I 13 mean, are you doing --
14 MR. HAYES: Well, so for the MARMOT 15 simulations we have good knowledge for development of 16 the model. Now, you're right, you get to the 17 engineering scale, and you're probably not going to 18 have all that information available for every PIN, so 19 you have to make some assumptions.
20 MEMBER BALLINGER: You probably realize 21 now that you're talking to a bunch heretics but let me 22 ask an obvious question. If MARMOT can't be 23 extrapolated, if BISON can't be extrapolated then what 24 is the value of doing this other than capturing margin 25 in an accident or something like 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
70 1 So and if it's capturing margin, how much 2 margin do you get back, do you think you're going to 3 get back. Like, for the example, in the infamous 4 50.46 hit area, how much margin do you think you're 5 going to get?
6 MR. HAYES: So there may be, may well be 7 some utilities or vendors who want to go after margin.
8 We think these will be tools that will help them. I 9 can't answer quantitatively that question.
10 But let me answer the first part of your 11 question by saying I don't see these tools as being 12 accepted in the near-term making extrapolations beyond 13 experimental databases in a regulatory environment, a 14 licensing discussion.
15 But vendors could use these tools in 16 spaces like that to give them insight for things they 17 want to go after. Now they may have to go and collect 18 a data point there to have a discussion with the NRC.
19 But it gives them a tool, an informed 20 tool, maybe not a well-validated tool in these far-21 reaching areas.
22 But a tool with enough physics into it 23 they can do some meaning exploration. That's the way 24 DOE, this is the way the lab people use it. We use 25 BISON in some of our advanced fuel development areas 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 like metal fuels for fast reactors.
2 And this is the way we use the tool to 3 find places we'd want to look at more closely.
4 MEMBER BALLINGER: Again, another 5 heretical question, Company X begins with an F decides 6 that they might want to use this. They've got to make 7 a judgment on the value to them.
8 MR. HAYES: Right.
9 MEMBER BALLINGER: And that means to them 10 how much money am I going to make if I make use of 11 these tools to improve my design? So --
12 MR. HAYES: You're right, and that's an 13 evaluation they have to --
14 MEMBER BALLINGER: -- they have to be 15 convinced that there's going to be a margin.
16 MR. HAYES: Exactly.
17 CHAIRMAN CORRADINI: So Chris mentioned 18 this earlier, and I'll mention it again at the end.
19 DOE doesn't see it's role or even its ability to take 20 every one of these tools all the way across the finish 21 line and say, "This thing is validated entirely for 22 your specific fuel design.
23 You just need to go pick it up and go have 24 the conversation with the NRC. We're taking it far 25 enough to give that vendor confidence. This is really 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 a tool that can provide me some benefit.
2 They're going to have to step in and do 3 some proprietary validate of their specific concept to 4 add to the broad base that broad base validation that 5 DOE has done. To have those final conversations with 6 a regulator.
7 MEMBER KIRCHNER: Steve, to what extent 8 has MARMOT been used to --
9 MEMBER BALLINGER: There's a wealth of 10 data out there, for existing fuel. To what extent 11 have you validated MARMOT and its ability to predict 12 evolution of structure under irradiation and thermal 13 effects.
14 MR. HAYES: So MARMOT does have it's how 15 what we call assessment report that's built up every 16 year. More and more validation cases are added to it.
17 And --
18 MEMBER KIRCHNER: So how well --
19 MR. HAYES: In certain areas, it's 20 undergone a lot of validation, like grain growth which 21 is incredibly important, grain grown and 22 densification. You'll find a lot of separate effects 23 or experimental studies in the MARMOT assessment 24 support show that it can simulate things like that 25 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
73 1 CHAIRMAN CORRADINI: So we want you to get 2 through you introduction so we can go to break.
3 MR. HAYES: Thank you, okay it won't take 4 much longer probably.
5 MR. JOHNSON: I know, that's why I brought 6 it up.
7 MR. HAYES: Okay, so that's what we're 8 trying to do. So all of this stuff down here is 9 basically building models. And anyone can jump in and 10 play in this area if they want to but this is a more 11 technical area to operate in.
12 MEMBER MARCH-LEUBA: Yes, that's a concern 13 I'm having here is, if you give this to a vendor, do 14 they have anybody who knows how to use it? Or do they 15 need to hire somebody from a lab that's been working 16 on it for the last ten years.
17 MR. HAYES: Exactly, so a vendor wants to 18 use BISON and have the basic models in there that 19 already do most of what they need. And that's what 20 DOE views as its role.
21 MR. STANEK: But we're starting to see 22 that the vendors are hiring people like that.
23 MR. HAYES: Some vendors want to do this, 24 they're the minority, but some do. Okay, three or 25 four slides and then it's probably a good time for 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
74 1 break.
2 So high-level BISON overview. BISON is 3 the fuel performance code we've been talking about.
4 I'm going to have one slide on overview and then a few 5 slides discussing issues related to verification and 6 validation.
7 So the high level view of BISON. BISON is 8 a MOOSE-based application. I'll say that right up 9 front, it's built on the MOOSE platform, if you know 10 what that means it means something to you, if you 11 don't, it's not super relevant to the conversation.
12 But being a MOOSE-based application BISON 13 is a finite element-based tool for doing fuel 14 performance. It solves the fully-coupled thermo-15 mechanics and species or mass diffusion equations in 16 as many dimensions as you want to do it, one, two, or 17 three-D.
18 It's applicable both to steady-state and 19 transient, so this is a big step forward in fuel 20 performance tools, one code does both.
21 It has capability and used for L-W-R, 22 conventional L-W-R fuels, ATF, TRISO, metallic fuels.
23 Again, this is a bit of an innovation, but one fuel 24 performance code does it all.
25 CHAIRMAN CORRADINI: If we get to 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
75 1 later, then I'll stop, but I was under the impression 2 that from the standpoint of a similarity if I had 3 FRAPCON and FRAPTRAN together ,the capabilities are 4 similar.
5 MR. HAYES: Exactly.
6 CHAIRMAN CORRADINI: The end-state 7 capabilities are similar, how I got there are 8 different.
9 MR. HAYES: Right.
10 CHAIRMAN CORRADINI: The one that confuses 11 me is --
12 MR. HAYES: The BISON is one code that 13 does what the tandem of FRAPCON and FRAPTRAN does.
14 CHAIRMAN CORRADINI: -- the one thing that 15 confuses me, when you say TRISO, the kernel or the 16 actual compact? Because --
17 MR. HAYES: Vastly more has been done on 18 the particle, the coated particle.
19 CHAIRMAN CORRADINI: So we're talking 200 20 microns, 250 microns versus the compact which is 21 really the issue at hand with the --
22 MR. HAYES: But there's been plenty of 23 work, BISON can do the compact too.
24 MEMBER REMPE: Did you take the PARFUME 25 models and put them into BISON or did you do something NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
76 1 different?
2 MR. HAYES: That's always the first step, 3 you put into BISON the existing model. In TRISO's 4 case, PARFUME models, then you work on developing more 5 mechanistic models to replace the legacy, the legacy 6 models as they become available.
7 I'm going to skip these next two because 8 I'll say something about that in a minute. But not 9 just transient, we're talking about full LOCA and RIA 10 simulation capability in the case of an LWR.
11 CHAIRMAN CORRADINI: So I mean, okay, but 12 for RIA I thought BISON was limited like FRAPCON and 13 FRAPTRAN that they can't go to melt?
14 MR. HAYES: It doesn't have models in it 15 currently to progress past melt.
16 CHAIRMAN CORRADINI: And my understanding 17 was now we're getting to what I thought you were 18 saying is legacy codes was FRAPTRAN and -- ,but I 19 thought staff was developing FAST which actually can 20 go beyond that for their fuel performance.
21 MR. HAYES: So I'm not able to speak of 22 that.
23 CHAIRMAN CORRADINI: That's fine. Okay.
24 MEMBER REMPE: So you can't do melting but 25 do you, I mean, fuel starts degrading before 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
77 1 reaches melting temperatures with cladding, you can't 2 even liquefy and have cladding fuel interactions with 3 the Zircoloy. Right? Or the ARMOUR or whatever it 4 is.
5 MR. HAYES: The first models are not going 6 to track a liquid phase. And then the last point to 7 speak directly to your issue of before is BISON, MOOSE 8 which is the framework, and BISON which is the 9 application were stood up from day one to follow an 10 NQA-1 development process, and that has never changed.
11 What makes BISON different? We try to 12 illustrate that with these four little captions.
13 BISON in one sense is like any other classical fuel 14 performance code, but in certain respects, it can be 15 quite different.
16 The first, and at least three, and really 17 all four of these have some direct relevance to ATF.
18 So one thing is it truly is a code that can handle 19 arbitrary geometries with its finite element 20 formulation.
21 There's no geometry it can't handle so if 22 your ATF concept is non-cylindrical, no problem for 23 BISON. It was built from the very beginning to go 24 after multiple fuel packs. So the way the models and 25 the properties and the behavior models are implemented 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 in the code, it's very easy to introduce new models 2 that have an application to a particular fuel or not.
3 So this is where most of the ATF concepts 4 are going to fall, they're going to need new models 5 for certain types of behaviors and BISON readily 6 accommodates that.
7 A third capability and this is what comes 8 with any MOOSE-based application is coupling to other 9 codes. And this may not be a high priority thing in 10 every instance, but in some cases, you may want to 11 couple fuel performance to some other type of physics 12 to do a coupled simulation.
13 The picture here is of BISON coupled to 14 TREATS, and actually Jess, Jess will say something 15 about that in his talk later. But BISON and any 16 MOOSE-based application really is designed to 17 facilitate coupling to other codes.
18 And then lastly all MOOSE-based 19 applications, BISON, being the flagship are designed 20 to operate efficiently on a high-performance computer.
21 Not let me immediately dispel a few myths.
22 You do not need a high-performance 23 computer to run BISON. You can run it on your desktop 24 computer. Most of our developers and users do that, 25 as long as your problem's simple enough. But if 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
79 1 want to develop or need to develop complicated, 2 difficult problem and run it in 3D or something like 3 that BISON is designed to be thrown out on high-4 performance machine and operate very efficiently.
5 So if you need that capability, you've got 6 it.
7 MEMBER REMPE: What about ease of use?
8 Like, again, I know you can answer simple models, easy 9 to get going and things like that. But how difficult 10 is it to develop models to learn to use them?
11 CHAIRMAN CORRADINI: A think a regular 12 fuel, a fuels engineer that's used to you know 13 developing FRAPCON input files or what have you won't 14 find BISON terribly more challenging. There'll be a 15 few things that will have to be done differently.
16 The biggest issue is if you need a 17 complicated geometry you're going to have to generate 18 a mesh. And certainly, a BISON just automatically can 19 do a cylindrical geometry and things like that.
20 But if you want to develop a complicated 21 geometry, you have to build the mesh for it.
22 MEMBER REMPE: So it's not been keeping 23 up, do you have like user works shops and training 24 things? Do you even have an Internet?
25 MR. HAYES: Oh, 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
80 1 MEMBER REMPE: I guess you can't have an 2 international community but university -- Now there's 3 a big BISON user community, BISON training is done at 4 Idaho typically a couple of times a year. And we take 5 training on the road if someone asks for it.
6 MR. HAYES: If someone asks for it.
7 MEMBER REMPE: Is it mainly universities 8 that come? Or is it plant staff or university, or 9 vendor staff or whose?
10 MR. HAYES: There's considerable 11 universities but also a lot of lab people and vendor 12 people.
13 CHAIRMAN CORRADINI: So how are you going 14 --
15 MR. HAYES: NRC, NRC.
16 CHAIRMAN CORRADINI: So how are you going 17 with your four slides?
18 MR. HAYES: Pretty good. Verification, 19 this one won't take long, but MOOSE and certainly 20 BISON is supported by the thousands of unit and 21 regression tests that follows all the standard 22 protocols, all new code. Must be supported by 23 verification testing.
24 All tests must pass prior to merging it 25 with the official version if anything gets flagged as NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
81 1 a problem it won't be merged until though programs are 2 resolved. Again, hitting this issue of NQA-1 of BISON 3 is regularly per NQA-1 standards, and it's always 4 satisfactory.
5 We provided some documentation in advance 6 on verification, things like that. One of the things 7 we gave you was this journal articulate which does a 8 good job summarizing where we stand on BISON for 9 verification.
10 Verification in certain senses is easier 11 than validation. So validation is obviously the 12 bigger task and so a couple slides, and then I think 13 it's time for a break on the status of bison 14 validation for conventional LWR fuels.
15 So what we're talking about is for UO2 and 16 Zircaloy. You can see her 75 integral and ramp type 17 experiments are in the assessment database. A 47 18 LOCUS matching RIAs. I just give you these numbers.
19 You can go look at all the specifics in 20 the assessment report to say, "These types of numbers 21 are very similar to a FRAPCON, FRAPTRAN-type 22 assessment base.
23 Now as is already mentioned some vendor's 24 come along and they have proprietary that DOE, in 25 general, doesn't have access to. And they do some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
82 1 additional validation, so there are some assessment 2 reports out there that are proprietary that have 3 additional data. But what you'll find in here is the 4 open stuff that the DOE has done.
5 MEMBER REMPE: How does the interface work 6 because they have a version that they've used and 7 they've validated it and has their proprietary 8 reports. And then you guys change a different model, 9 and so now version X is version X+1.
10 Have you had any interference with it?
11 MR. HAYES: There's a couple of ways that 12 manifests itself. One, at the end of every fiscal 13 year there is a frozen version of BISON because some 14 people lack frozen versions of BISON, and an 15 assessment for a report role out with everything, with 16 all the results, you know, tied very specifically to 17 that frozen version.
18 That being said, all of the assessment 19 cases are rerun every night. And anything done to the 20 code the previous day that disrupts any agreement, you 21 know, comes back with some sort of a red flag.
22 And so those tend to resolved right away 23 so even though there's not going to be a frozen 24 version with a published BISON assessment report until 25 the end of the year, on any given day of that year 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
83 1 version of BISON that is operating is essentially 2 validated to that state.
3 But again, vendor has done something 4 different. Have you ever had them call up say, "Hey, 5 a year ago it worked, and now it doesn't." I mean, 6 there's always something that you didn't get with your 7 kid's case.
8 MEMBER REMPE: And you've had that happen 9 so far and figured it out?
10 MR. HAYES: There's a ticket tracking 11 thing. Any user can put in a ticket, "Hey, 12 something's wrong, and it gets resolved." Absolutely.
13 MEMBER REMPE: If you've ever used it 14 before -- okay.
15 MR. HAYES: Now a lot of these vendor 16 assessment reports through the vendor may not have 17 modified the code. In some cases, they may, but in 18 many cases, they haven't modified the code they just 19 have more experiments from Halden that were, you know, 20 proprietary, so they just have additional data to 21 compare to.
22 MEMBER REMPE: Okay.
23 CHAIRMAN CORRADINI: But if I, I know 24 we're delaying you, but if I have, I'm still trying to 25 think through the connection. So I've got BISON 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
84 1 I know BISON is maybe the source is not public but 2 I've got a public version which Company X and Company 3 Y adopt.
4 Company X and Y may change the correlation 5 --
6 MR. HAYES: They may.
7 CHAIRMAN CORRADINI: -- and that becomes 8 the proprietary thing which they must then submit to 9 the regulator for a review and analysis.
10 MR. HAYES: And then the full burden of 11 validating that version is on them.
12 CHAIRMAN CORRADINI: Right.
13 MEMBER REMPE: Right.
14 MR. HAYES: Even in that case they're 15 going to be able to take credit for things we've done.
16 CHAIRMAN CORRADINI: But to get to Joy's 17 and Jose where they're talking about modifications 18 then they have to continually do a comparison check 19 with the base version with whatever their proprietary 20 changes are.
21 MR. HAYES: That right.
22 CHAIRMAN CORRADINI: Okay.
23 MEMBER MARCH-LEUBA: And whenever they 24 approve this they will put out a specific version if 25 you don't approve an SER the most recent version 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
85 1 INL published.
2 MR. HAYES: Understood, you're right, and 3 this is why we periodically have these frozen 4 versions. We understand that in the regulatory space 5 that's the way it's going to need to work.
6 The assessment report is massive, and you 7 probably don't want to look through it but we've given 8 you also a real nice paper that kind of summarizes the 9 high-level results of BISON compared to LWR's.
10 And then this is the last one. So if you 11 look at the assessment report, I mean, it will go 12 experiment by experiment, this is how the rod 13 functioned, and these are the data collected and 14 compares to that specific experiment.
15 But also in the assessment report, you'll 16 find these higher level analyses where everything is 17 brought together. How does BISON do just overall in 18 getting fuel temperature correct or PCMI or fission 19 gas release or something related to LOCO?
20 So that's all in the assessment report as 21 well, and the punch line is I absolutely believe is 22 that you're going to find BISON for standard LWR is 23 state of the art.
24 I mean, it does as well or better than any 25 code out 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
86 1 MEMBER MARCH-LEUBA: Have you done a 2 comparison those pound lines there, the plus/minus 3 along these lines, how does this compare to FRAPCON.
4 Is it have or is it twice?
5 MR. HAYES: Jason, what would you say 6 about that?
7 MR. HALES: About the same.
8 MEMBER REMPE: You have to come up to the 9 mike and say your name. You just can't answer, sorry.
10 CHAIRMAN CORRADINI: He did it to you, go 11 on up.
12 MR. HAYES: So this is Jason Hales he 13 manages the BISON department.
14 MR. HALES: I'm Jason Hales from INL, this 15 answer's going to be pretty much a letdown because I 16 can't answer for FRAPCON. I don't know how it 17 compares honestly. What I do know is that Steve said, 18 "If you can compare the BISON results to the 19 experimental data it compares very well, and we're 20 comfortable with it.
21 MR. HALES: I'm Jason HALES from INL.
22 This answer's going to pretty much a letdown because 23 I can't answer for FRAPCON. I don't know how it 24 compares honestly, what I do know is that Steve said, 25 did he compare BISON results to the experimental 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
87 1 it compares very well and we're comfortable with that.
2 How it compares to another code is that 3 something that we've. We don't take time compare our 4 output to the put of another code. We prefer to do 5 validation and compare the data, the experimental 6 data.
7 CHAIRMAN CORRADINI: So let me ask it from 8 an uncertainly standpoint. So I've got a set of 9 experiments they have their own internal uncertainty.
10 Does it fail within the uncertainty band of the data?
11 MR. HAYES: Yes, we're very comfortable 12 with that. That's another whole issue, the 13 uncertainty with a lot of these experiments is rather 14 large, sometimes uncomfortably large. Things like the 15 power in the reactor are not known very well.
16 And so given all the uncertainty, yes, 17 we're very comfortable with where the experimental 18 data lies.
19 MEMBER REMPE: So if I look at your 20 fission gas release spot and I blow it up on my little 21 computer to see, it doesn't, I mean it looks like 22 there are more thanks that are showing higher measured 23 than I predicted. Am I right?
24 MR. HAYES: So performance is not measured 25 on the same scale for every property. So for 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
88 1 temperature, you typically look at plus or minus 10 2 percent. Okay, for fission gas release you typically 3 look at plus or minus a factor of two. But this is 4 true for all codes.
5 MEMBER REMPE: So other codes probably 6 would also be below the line on that one too. Is that 7 what you're saying?
8 CHAIRMAN CORRADINI: We're going to have 9 to separately check that out.
10 MEMBER REMPE: Yes, I just was kind of 11 curious. Also, on your material properties --
12 MR. HAYES: And there's a good discussion 13 of that in this paper.
14 MEMBER REMPE: Okay, and the material 15 proprieties did you start off with like, something 16 like map probe that has, is a function of temperature 17 and burn-ups for all those. And so when you talk 18 about the ATF later today, you'll talk about what you 19 do when you don't have some of the properties.
20 MR. HAYES: So the very first thing, 21 almost, I don't know if it's day two or three of BISON 22 development, they linked in all of the MAP properties 23 and models.
24 But over time, you know, those proprietary 25 and model and behavior models get updated with better 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 models. That doesn't mean you can't still select the 2 MAP probe data if that's what you want to do.
3 MEMBER REMPE: Oh, okay.
4 MEMBER MARCH-LEUBA: I'm just the 5 additional report, it defines well the range of public 6 ability, for example, all those red and blue elephants 7 were showing there are for PWR.
8 That doesn't mean you work for a build-up 9 like you are.
10 MEMBER MARCH-LEUBA: Is very important 11 for your evaluation.
12 MR. HAYES: Yes.
13 MR. HALES: And it is well defined?
14 MR. HAYES: Yes, in the assessment report 15 it's well-defined.
16 MEMBER MARCH-LEUBA: If it gets approved 17 now, approved SER, it will have a bunch of 18 limitations. And yes, if I were you, you can predict 19 UO2 doesn't mean you can predict the silicide.
20 MR. HAYES: Oh, of course, of course yes.
21 MEMBER MARCH-LEUBA: That's what I mean.
22 You really feel that you have to keep in mind that 23 whenever you product this you have to say, I can use 24 it for --
25 MR. HAYES: Exactly, if all this was 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
90 1 set the stage for our model enhancements for ATF to 2 say that they BISON we have, and the multiscale 3 modeling approach we've has borne out good results 4 applied to UO2.
5 So the advertisement is we're going to 6 approach these ATF concepts in the same way, use the 7 same methodology, and over tie we should get similar 8 results. That's the proposition.
9 CHAIRMAN CORRADINI: Okay, why don't we 10 take a break, and come back at 10:30 and shockingly 11 we're not so bad in time.
12 (Whereas, the above-entitled event went 13 off the record at 10:18 a.m. and resumed at 10:30 14 a.m.)
15 CHAIRMAN CORRADINI: Okay, so I think 16 Steve we gave you a break and now you're back on.
17 MR. HAYES: I appreciate that, yes, and 18 we're going to try to accelerate a few things. A lot 19 of what we have going forward for the rest of this is 20 just a status of where we are on the different 21 concepts and maybe not so much discussion is needed 22 but as much as is appropriate.
23 So model enhancements for ATFs, so all 24 that was to set the stage to allow me to say that DOE 25 believes BISON is a state of the art field 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 performance code.
2 It hasn't reached its final state even 3 for UO2 and Zircoloy we're continuing to develop new 4 and improvement mechanistic models that are going to 5 improve it.
6 But even where it is it's exceptionally 7 well verified, it's extensively validated, and we 8 would say ready for use for current LWR fuels for 9 anyone who would want to do that.
10 But all that is to say that in our 11 judgment, BISON is the right platform on which to 12 implement enhancements to simulate accident-tolerant 13 fuels.
14 And DOE started doing this back in 2015, 15 so in 2015 DOE-funded what it called a high-impact 16 problem, a HIP, on ATF. And Jason Hales who I called 17 to the microphone a minute ago, was the PI on this 18 project and if you have a lot of detailed questions 19 moving forward, he'll probably answer a lot of the 20 questions.
21 And what the HIP was was a concentrated 22 three-year where DOE spent $3 million a year for 23 three years to sort of jump start ATF modeling in the 24 MARMOT/BISON world.
25 Now the HIP ended in 2017, but 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
92 1 doesn't mean that DOE has stopped its support. ATF 2 is an integral part of any modeling programs moving 3 forward since that time.
4 And the fuel modelers working in the area 5 of BISON and MARMOT have a very close working 6 relationship with the ATF programs and personnel both 7 at the laboratories and the industry teams.
8 Particularly with a view to the test 9 programs that are going on and the data that will be 10 coming out of them because that will be critically 11 important.
12 MEMBER REMPE: Which, so they modeled all 13 of the concepts for all of the vendors that you 14 showed earlier on Slide 5. Right?
15 MR. HAYES: That's what we're talking 16 about --
17 MEMBER REMPE: For the HIP, they did 18 consider each of the vendor fuels?
19 MR. HAYES: No.
20 MEMBER REMPE: One of the vendor fuels?
21 MR. HAYES: I have a slide especially for 22 you on that.
23 MEMBER REMPE: Just for me? Okay.
24 MR. HAYES: I'm going to explain that.
25 MEMBER REMPE: Okay, I thought it would 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 be an easy answer.
2 MR. HAYES: It's better if you see it 3 than just hear it. Okay, so this slide just is meant 4 to communicate that we're not reinventing the 5 mechanistic modeling approach.
6 We're using the approach that DOE's been 7 using in the area of fuels for the last nine years.
8 And it's sort of a four-pronged approach, and that is 9 any time you start with a new fuel in the BISON 10 framework, you just stand up basic properties and 11 model for behavior using anything that you have, a 12 legacy models, or adapt UO2 models or there may be 13 existing models for some of these materials.
14 Just to get BISON up and running, all the 15 while then you're working on developing new 16 mechanistic models using these lower-length scale 17 techniques that I've discussed.
18 And this is where a ton of the work is 19 going on in ATF right now. Another important thing 20 is sensitivity analyses, when very early in the 21 development of Bison it was coupled to Sandia's 22 DAKOTA code.
23 And so we run DAKOTA with BISON regularly 24 to do sensitivity analyses especially when we're 25 starting in a new fuel. We do sensitivity analyses 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 to help us prioritize what properties or what 2 behavior models for a given fuel type have the 3 biggest effect.
4 And then we use that to prioritize which 5 mechanistic models we work on before others. And 6 then ongoing assessment and validation is this part 7 of the process.
8 Okay, Joy, this explains, this is the 9 highlights from the high-impact problem. So when the 10 HIP was originally stood up in 2015, silicide fuel 11 and FeCrAl cladding were the two big concepts getting 12 the most attention then.
13 So the way the HIP was written and the 14 way it started it focused mostly on these two 15 systems. But before the HIP was over, coded, 16 cladding concepts and doped UO2 were emerging as 17 interesting concepts as well.
18 And so they were brought into the HIP in 19 the later phases as well. So these are specifically 20 things that the HIP addressed. The program since the 21 HIP has expanded to include even more.
22 MEMBER REMPE: So again I was just trying 23 to understand what the objective of the high-impact 24 problem was and it sounds like it wasn't really an 25 analysis problem it was more trying to expand the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
95 1 capabilities of BISON to address what was perceived 2 to be an accident-tolerant fuel assisted time.
3 It was more model development --
4 MR. HAYES: That's exactly right.
5 MEMBER REMPE: It was a lot more model 6 development than a problem that you tried to analyze.
7 MR. HAYES: That's exactly right.
8 Enhancing the exhibiting tools to address emerging ATF 9 concepts. And the idea early on was to provide a tool 10 to DOE to help assess the ATF concept they were 11 funding.
12 MEMBER REMPE: So again, where I was kind 13 of going on that question eventually was that my 14 understanding is that there's not a lot of high-15 temperature data and there's not a lot irradiated 16 mater proprietary data for these materials 17 And so some of you properties you've put 18 in for ATF are guesses, right, and --
19 MR. HALES: In some cases that would be 20 true.
21 MEMBER REMPE: And this would hopefully 22 help you focus?
23 MR. HAYES: We have some tables coming 24 that try to summarize where we stand for the various 25 concepts. And not every box is checked on every 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 concept, you're right.
2 Okay, so this is the table that attempts 3 to give the 30,000-foot view of where we are on all 4 these categories of ATF features from doped UO2, 5 coated cladding, FeCrAl cladding, silicon carbide, 6 silicide fuel and metallic fuel.
7 And so this first column says we have a 8 complete set of models for what you see here. So that 9 means there's a model that allows you to do a 10 simulation of all these concepts.
11 For some, it's better than others. For 12 doped UO2 fuel, we've spent a lot of time looking at 13 the fission gas release modifications needed. Much 14 less so for the mechanical response but we expect 15 there to be some, so we're not completely there.
16 Coded cladding as you say, we can simulate 17 coded claddings now although there's very limited 18 actually there's no irradiation effects yet. FeCrAl 19 is in pretty good shape. Silicide fuel, metallic fuel 20 in good shape. Silicon carbide cladding we'll be the 21 first to say that's a work in progress. We can do 22 simulations of it but only recently has that reached 23 that level.
24 So you can see the yes or no, on the 25 evaluation, on base irradiations, so we don't have any NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
97 1 irradiation data yet on coated cladding or silicon 2 carbide as a cladding, although, there's a lot 3 irradiation data on silicon carbine.
4 CHAIRMAN CORRADINI: Was the doped UO2 5 fuel irradiations vendors supplied or you did?
6 MR. HAYES: In my anticipation, it was 7 vendor supply.
8 MR. STANEK: The answer's both, so there 9 are priorities, as far as we understand there are 10 proprietary irradiations that have been done on doped 11 UO2 as Ron was mentioning, in Europe.
12 But there are open Halden tests that have, 13 in particular, fission gas release data for doped UO2.
14 MR. HAYES: In fact, we're going to show 15 one of those.
16 MEMBER KIRCHNER: Just looking at the 17 right-hand column, it's improved fission gas, 18 diffusivity for fission gas, are you seeing much later 19 fission gas release with doped UO2 fuel?
20 MR. HAYES: We're seeing lower.
21 MEMBER KIRCHNER: I'm misreading that. So 22 you're seeing what, Steve?
23 MR. HAYES: These are the areas where we 24 are continuing to focus the work.
25 MEMBER KIRCHNER: I could see you might be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
98 1 improving your diffusivity model but are you seeing 2 higher release of fission gas and doped fuel?
3 MR. HAYES: Well, I think in general we're 4 seeing lower fission gas release and doped.
5 MEMBER KIRCHNER: So this more about your 6 model and you estimating --
7 MR. HAYES: This is about the model.
8 MEMBER KIRCHNER: -- of the fission gas 9 release not the performance of the fuel. Right?
10 MR. HAYES: Exactly. This is an 11 evaluation of where the model stands, not an 12 assessment of the concepts.
13 MEMBER REMPE: I'm curious support user 14 needs. In NRC space, if they use or need something 15 they belong to the user group. Do you have somebody 16 that has metallic fuel who's paid to join your user 17 groups and you're supporting their user needs?
18 Or is this DOE decides how much each user 19 gets?
20 MR. HAYES: So there are, the user support 21 bucket is not an unlimited bucket. But I think the 22 team does a good job of supporting needs. So you 23 mentioned metallic fuel. We actually have one vendor 24 who has a fairly aggressive way of working with us, 25 and we're trying to address needs that they have.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
99 1 A lot of these needs come from the DOE 2 program developing metallic fuels as well. Not for 3 ATF applications.
4 MEMBER REMPE: Yes, but this is an ATF 5 slide, so that's why I was curious. Okay.
6 CHAIRMAN CORRADINI: So let me, I don't 7 want to, if you're finished I have a question, kind of 8 an overall question.
9 So my interpretation at least from the yes 10 and nos that BISON is similar. I keep on doing a 11 comparison in my mind because I'm unfamiliar with 12 BISON. BISON is similar to FRAPCAN and FRAPTRAN for 13 normal operation in AOOs.
14 And where does it extend it into DBAs? In 15 other words, to put it more crudely, what does it off 16 technically that isn't already there in the current 17 code set?
18 What I hear is it might be doing it faster 19 better from an uncertainty standpoint, but in terms of 20 the current code set, it's similar.
21 Am I understanding this though?
22 MR. HAYES: Similar.
23 CHAIRMAN CORRADINI: Okay, fine.
24 MR. HAYES: Faster and better, similar.
25 CHAIRMAN CORRADINI: So if you were 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
100 1 to, well, now let me ask my second question. So maybe 2 I misheard Chris when he said it earlier. Your 3 industrial support group or NEI support group, I'm not 4 sure what it is.
5 But we'll call it industry advisory group 6 has told you you cannot tell us who's adopting BISON 7 from the industry? I'm pressing this point because 8 you took my surprise, I figured you come up here wave 9 a flag and say, "Company X, Y, and Z are adopting it, 10 and ergo it's clearly better than what they've got and 11 therefore you guys ought to consider it."
12 But I heard the exact opposite which took 13 me by surprise.
14 MR. HAYES: So let me clarify those 15 initial comments. So the industry council that I was 16 referring to is specifically focused on non-LWRs.
17 CHAIRMAN CORRADINI: Oh, okay, fine.
18 MR. HAYES: So that group does not --
19 CHAIRMAN CORRADINI: Okay, I misheard.
20 MR. HAYES: -- talk about ATF at all.
21 CHAIRMAN CORRADINI: So then I'll ask the 22 question now since we need three actors in the game --
23 MR. HAYES: Exactly.
24 CHAIRMAN CORRADINI: Framatome, 25 Westinghouse, and GE, which one of them has decided 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
101 1 take BISON as their base tool for their safety 2 analysis justification for ATF?
3 MR. HAYES: I don't think any would say 4 that.
5 CHAIRMAN CORRADINI: Why?
6 MR. HAYES: That being said, both 7 Westinghouse and Framatome have test stands that 8 they're evaluating --
9 CHAIRMAN CORRADINI: Oh, fine.
10 MR. HAYES: -- BISON with.
11 CHAIRMAN CORRADINI: So they're in the 12 process --
13 MR. HAYES: They have now announced --
14 CHAIRMAN CORRADINI: They're in the 15 process of evaluating.
16 MR. HAYES: -- the decision was made that 17 we're using BISON. But they're both, they're using 18 and testing BISON.
19 CHAIRMAN CORRADINI: So they're 20 evaluating.
21 MR. HAYES: They're evaluating BISON.
22 CHAIRMAN CORRADINI: Okay. Okay. Thank 23 you.
24 MR. HAYES: Okay. So that's the high 25 level overview. And now we're going to have a status 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 on each one of the concepts in particular.
2 So doped UO2 model summary, so the first 3 column assesses whether or not there's a model in 4 BISON. And in some of these cases for the types of 5 doping we're talking about, you're not expecting 6 significant changes in basic material properties. And 7 so a lot of times you can make use of regular UO2 8 data.
9 As I've already mentioned, we know there's 10 some work to be done in the area of mechanical 11 behavior for doped UO2. And that's not in the code 12 yet.
13 Most of the effort that has been put 14 towards this concept is towards understanding fission 15 gas release. And just as an illustration --
16 MEMBER REMPE: Could I -- I'm sorry.
17 Could you go back? That LLS-informed, that's the 18 MARMOT-informed --
19 MR. HAYES: This means we're working on --
20 so, in almost all these cases, we're working on 21 mechanistic models that will ultimately we believe 22 make it into BISON.
23 If there's a no here, that means there's 24 no model in BISON that's been developed from this 25 mechanistic process in this category.
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 But in the area of --
2 MEMBER REMPE: Is that the MARMOT?
3 MR. HAYES: -- thermal conductivity 4 degradation and fission gas release, the answer is 5 yes. There are models in BISON today that have been 6 informed by this lower length scale process.
7 MEMBER REMPE: But LLS, that's really the 8 MARMOT informing of BISON. So you don't have any 9 MARMOT models in BISON. You just have a regular 10 empirical fit for thermal conductivity degradation?
11 MR. HAYES: Thermal conductivity 12 degradation is in the area where there's --
13 MEMBER REMPE: There is a MARMOT one. But 14 for the other ones, for like mechanical properties, 15 there's no MARMOT --
16 MR. HAYES: There's no MARMOT model.
17 MEMBER REMPE: -- informed modeling.
18 MR. HAYES: Exactly.
19 MEMBER REMPE: Okay.
20 MR. HAYES: I said not all the boxes were 21 going to be checked.
22 CHAIRMAN CORRADINI: Oh, yes. That's 23 fine. That's fine. That's fine.
24 MR. HAYES: But this is a work in 25 progress.
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 CHAIRMAN CORRADINI: Understood. What is 2 a basic thermal property?
3 MR. HAYES: You know, un-irradiated 4 thermal conductivity and thermal expansion, these 5 material properties --
6 CHAIRMAN CORRADINI: Oh, thermal 7 expansion.
8 MR. HAYES: -- things like that, specific 9 heat. These basic properties, un-irradiated, you 10 know, that have to go in.
11 MEMBER REMPE: So --
12 MR. HAYES: And then you work on the 13 degradation under irradiation and things of that 14 nature.
15 MEMBER REMPE: So the data for developing 16 this model was because of the fact that they did some 17 testing of Cr2O3-doped UO2 fuel without the cladding 18 for ATF. Is that a true statement, too?
19 MR. HAYES: That's right.
20 MEMBER REMPE: I think that's what you 21 told me earlier, right?
22 MR. HAYES: That's right. So experimental 23 data means we have some data in hand to begin 24 evaluating the modeling that we're doing.
25 MEMBER REMPE: But in some of these 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 properties, I could see, it seems like some of them 2 might be influenced by the cladding. Like would the 3 creep not be influenced by the cladding and --
4 MR. HAYES: The creep of the fuel? Not 5 really.
6 MEMBER REMPE: Not in any pellet-clad 7 material and mechanical interactions?
8 MR. HAYES: That's certainly a phenomena 9 that has to be addressed. But I don't think the creep 10 model of the fuel itself is impacted by --
11 MEMBER REMPE: Okay.
12 MR. HAYES: -- those considerations.
13 And so, like Chris was saying, we have 14 this Halden experiment, this rod from a Halden 15 experiment, which was an irradiation of chromia-doped 16 UO2. It wasn't sponsored by the ATF program. But 17 this is data we're assessing against.
18 And so you've got online temperature 19 measurements that we're comparing to and fission gas 20 release.
21 And a word of caution here is this rod 22 operated at very high temperature, a higher 23 temperature than would be the norm. And so it 24 released much more fission gas than would be expected 25 in normal operations of an LWR. That being said, 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
106 1 BISON is predicting it or reproducing it fairly well.
2 And the reason we like looking at 3 experiments like this, in general, if you're familiar 4 with fuel modeling, it's much harder to get good 5 agreement on high fission gas release results than low 6 fission gas release results.
7 And so don't look at this and think that 8 chromia-doped UO2 is going to normally release 15 9 percent of its fission gas. It won't. In fact, the 10 trend is it releases less than undoped UO2. But the 11 models are working fairly well.
12 MEMBER BALLINGER: You caught me by 13 surprise. At 15 percent, it's not going to make any 14 difference. Where it really makes a difference is at 15 the low end for real operations.
16 MR. HAYES: No, I agree. But my point was 17 it's easier to find agreement with those low fission 18 gas release measurements. It's harder to get 19 agreement with the higher values. The higher values 20 are only going to show up under more extreme 21 conditions.
22 Chrome-coated cladding, so this was 23 actually added later in the process than you might 24 imagine. But there's certainly functional models now 25 in BISON to model chrome-coated cladding. We're NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
107 1 beginning to get experimental data that we need to do 2 assessments.
3 This is an area, it's a pretty simple 4 modification to BISON. We don't have any lower length 5 scale work that has fed into that. And it's not clear 6 that it's needed in the near term.
7 One thing, I thought there was an asterisk 8 here. It might have gotten deleted. In the area of 9 creep, this footnote goes with creep.
10 We don't have any irradiation performance 11 effects on chrome-coated cladding currently. So 12 that's an unknown. Although, we have chrome-coated 13 cladding in ATF-2. And we'll be getting data.
14 And so, as an example of simulations of 15 coatings using BISON, you can see them here. So we 16 don't have a lot of experimental data.
17 So what you see here is sort of a study 18 where we're comparing the mechanical response of 19 coated versus non-coated Zircaloy cladding. So this 20 is the strain results on Zircaloy for instance, and 21 then how that changes when you add various thicknesses 22 of coatings.
23 And in these two cases, the coating that 24 was added was FeCrAl. Now, none of the vendors are 25 looking at FeCrAl as a coating. Although, MIT has 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
108 1 research project where they're looking at FeCrAl as a 2 potential coating.
3 But the reason is because when these 4 analyses were done we didn't have all the data we 5 needed for the chrome coating, which we do have now.
6 We did look at chromium early on. And we 7 didn't have a chrome creep model. And that led to 8 unrealistically high stresses. So we knew we had to 9 get that in.
10 But what you see in these two cases where 11 it's a FeCrAl coating of 20 microns or 40 microns is 12 the coating is pretty stiff. It's much stiffer than 13 Zircaloy. And it adds considerable stiffness to the 14 cladding overall. And chrome follows this behavior, 15 although probably not to as great of an extent.
16 But one thing that became apparent to us 17 is someone could use BISON as a tool for optimizing 18 cladding or coating thickness, for instance.
19 FeCrAl cladding, here's the basic status 20 report, good models in BISON that allows you to do a 21 calculation, to do a simulation. We don't have yet 22 any data on irradiation creep. But FeCrAl cladding is 23 in Hatch, as we've already mentioned. It is in ATF-1.
24 And it will be going into ATF-2 this next year.
25 This is an area where we have lower length 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 scale, you know, mechanistic model development going 2 on in the area of mechanical properties and creep.
3 CHAIRMAN CORRADINI: So, just a side 4 question, there's so much other zirc in a BWR that if 5 this is really accident-tolerant, the cladding may not 6 be the driver. I could magically change out all the 7 cladding and still have a problem. I'm assuming --
8 MR. HAYES: Understood.
9 CHAIRMAN CORRADINI: -- you're aware of 10 that.
11 MR. HAYES: We are aware of it.
12 CHAIRMAN CORRADINI: Okay.
13 MR. HAYES: And there are projects looking 14 at alternative materials for channel boxes and things 15 like that. That's not part of --
16 CHAIRMAN CORRADINI: Okay, fine.
17 MR. HAYES: -- this activity.
18 CHAIRMAN CORRADINI: The reason I guess 19 I'm asking the question is, in some sense, I'm sensing 20 that it's not just, even though it's called accident-21 tolerant, I sense it has other potential normal 22 operation benefits --
23 MR. HAYES: FeCrAl cladding --
24 CHAIRMAN CORRADINI: Well, that the 25 industry --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
110 1 MR. HAYES: -- or all of these?
2 CHAIRMAN CORRADINI: All --
3 MR. HAYES: Industry hopes that a lot of 4 these concepts will bring some benefits --
5 CHAIRMAN CORRADINI: Okay.
6 MR. HAYES: -- to normal operations that 7 go beyond just the benefits in accident 8 considerations.
9 And this is just an illustration for you 10 of the FeCrAl burst model. To be frank, it's an 11 empirical model at the moment. But it's implemented 12 and working well, allowing us to do some comparisons 13 with bursting of Zircaloy.
14 And in many cases, the burst behavior 15 seems to be very similar to Zircaloy. Although, some 16 experimental results seem to indicate that the burst 17 opening may look quite a bit different than Zircaloy.
18 Silicide fuel, this is actually an area 19 where we had a jumpstart even before the high impact 20 problem. There was a university project even before 21 that that started developing some models for BISON, 22 especially in the area of creep for silicide fuel.
23 And so this is an area where BISON is 24 fairly mature. And quite a bit of lower length scale 25 model development work has been done, especially 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
111 1 the area of thermal conductivity, swelling, fission 2 gas release.
3 CHAIRMAN CORRADINI: Are there any 4 experiments on oxidation?
5 MR. HAYES: Of silicide fuel?
6 CHAIRMAN CORRADINI: Yes, I've got --
7 MR. HAYES: Yes, there are.
8 CHAIRMAN CORRADINI: I've got an awful 9 good source to make hydrogen.
10 MR. HAYES: There are oxidation studies of 11 silicide. And there are corrosion studies of 12 silicide.
13 CHAIRMAN CORRADINI: Okay.
14 MR. HAYES: The tool, the modeling tools 15 don't address those phenomena. The experimental 16 programs are looking at those.
17 CHAIRMAN CORRADINI: I mean, this is more 18 of a technical question. So I'll wait.
19 MR. HAYES: And just to give you an 20 example of the thermal conductivity modeling for 21 silicide, so this plot here on the left is thermal 22 conductivity of silicide fuel versus radial position.
23 So this would be at a center line of a 24 fuel pellet and the surface of a fuel pellet. And 25 this is what the thermal conductivity profile would 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 look like un-irradiated. And you can see degraded to 2 a certain burn-up level this is what the model 3 predicts.
4 And some of what you're seeing is also 5 reflected over here in this plot, which is thermal 6 conductivity versus temperature. The big difference 7 with silicide is silicide thermal conductivity goes up 8 with temperature more like a metal, whereas the curve 9 here for oxide fuel goes down.
10 And the models predicting degradation of 11 silicide fuel, you see it here. The silicide fuel, 12 like any fuel, will degrade under radiation. But all 13 indications are, even worst-case scenarios, it's going 14 to be far better than UO2.
15 Silicon carbide cladding, this probably 16 requires a big caveat. This is an area we started 17 working in probably much later than all the other 18 areas.
19 It says, yes, there are models in BISON 20 for pretty much everything. Even swelling could 21 probably be a yes now, although these are just 22 recently implemented. So there's not much assessment 23 of these models that have been done yet. And that's 24 just now beginning. It's not that we're not going to 25 do it. It's just now beginning.
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 There is a lot of experimental data, even 2 irradiation data, on silicon carbide. So there is a 3 wealth of data in general, not -- this is not data 4 generated as part of an integral fuel pin.
5 And so here's some, just to show you that 6 the silicon carbide cladding model in BISON is 7 working. We're not comparing to experimental data.
8 But we are part of the MIT group that has been looking 9 at silicon carbide as, and applications to cladding.
10 And they set up an early benchmarking 11 problem on stress and strain. And, you know, the 12 models in BISON are consistent with all the other 13 models in, or all the other codes in that benchmarking 14 activity.
15 And then this, then the last one is non-16 cylindrical metallic fuel. Metallic fuel is actually 17 an area that was receiving quite a bit of attention in 18 BISON even before ATF came along for FAST reactors.
19 So there was considerable metallic fuel capability in 20 BISON to begin with.
21 Now, obviously, some of that is directly 22 applicable to ATF type concepts and some needs 23 additional work. But there are good models for 24 metallic fuel in BISON for just about everything. And 25 there are some important lower length scale models 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 that are being developed as well.
2 These noes down here is radial pin power 3 distribution. Of course, in an LWR where you get flux 4 depression inside of a fuel pin, you have to account 5 for that in your temperature calculation. BISON can 6 handle that quite readily for cylindrical geometries.
7 If you go to a non-cylindrical geometry, 8 you're going to have to do something about that. And 9 nothing's been done about that in BISON yet. It's on 10 the to-do list.
11 And this gives you an example for the 12 kinds of things that we're talking about. Because of 13 its arbitrary geometry capability, BISON could 14 simulate an ATF concept that looks like this. Because 15 of its ability to couple to other physics tools, a 16 concept like this might need to look at CFD in the 17 coolant channel.
18 And so BISON can and does couple with 19 NEK5000, for instance, to do something like that.
20 You'll hear more about NEK5000 this afternoon.
21 Okay. That's a quick run-through of all 22 the concepts and sort of the status for where BISON is 23 on the various concepts.
24 Now, just a few slides on validation. And 25 I understand this is important.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
115 1 So the first point to be made is data 2 generation needed for ATF validation is underway.
3 It's early in the process, but it's underway.
4 DOE-sponsored testing of ATF concepts is 5 in progress. It has been for several years. We have 6 a very close partnership with the three industry ATF 7 teams, Framatome, GE, and Westinghouse. And all their 8 concepts are being tested. And they have a lot more 9 testing that will come.
10 Now, the DOE program, very highly, heavily 11 relies on testing in ATR and TREAT. We had envisioned 12 some testing in Halden, some important testing. And 13 as we've already discussed, that needs to be 14 redirected. And we're working on that now.
15 One thing, and, Joy, I know this is a 16 concern of yours. It's a concern of ours, too, and 17 that is the in situ instrumentation that has become a 18 real hallmark of the Halden experiments. Most of the 19 other places we're looking to move experiments don't 20 have that kind of historical legacy.
21 So we have already, DOE has taken the 22 initiative to partner with the Halden staff moving 23 forward. And they're going to help us with those 24 issues, instrumentation and how to appropriately 25 implement them wherever these experiments go. 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
116 1 don't just assume that we're going to get that right 2 without their help.
3 Here's a bullet that you should pay close 4 attention to. And that is we do foresee a need for an 5 expanded use of LTR/LTA programs in commercial 6 reactors just to generate the volume of subsequent 7 test specimens, irradiated test specimens we're going 8 to need.
9 Without Halden -- and one of the 10 advantages of Halden was instrumentation. Another 11 advantage is it was just a whole reactor dedicated to 12 the LWR fuel testing mission.
13 And most of these other reactors were now 14 going to have to go to are not going to be wholly 15 dedicated to that mission. They're going to do pieces 16 and parts of it.
17 And so we do have a need for just a larger 18 volume of irradiated fuels and materials to be 19 generated that can then be subjected to PIE or 20 possibly refabrication, reinstrumentation for 21 subsequent, more specific testing, maybe an ATR, 22 TREAT, LOCA facility. And we're hoping that materials 23 from those LTRs and LTAs can feed some of that.
24 But the bottom line here is fuel behavior 25 data needed for ATF validation is being generated. 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
117 1 mean, what you see here is silicide fuel in the hot 2 cell in Idaho. It's low burn-up. It's at about 10 3 gigawatt-days per ton. But we're beginning to get 4 data on it.
5 MEMBER REMPE: But just a caution of, I 6 know some guys from NR who used to laugh about that 7 Bruno would take the capsule out of ATR, shake it up 8 a bit, and then send it over to the hot cell.
9 And out-of-pile data is great. But you, 10 it's a lot better to get it in-pile if you can if at 11 all because --
12 MR. HAYES: We understand in situ 13 measurements are the priority. But we're not 14 discounting data generated in the hot cell as being 15 important and relevant as well.
16 Okay. So, that being said, we have some 17 serious challenges in the area of validation for ATF.
18 I've already mentioned this once in passing.
19 You know, we're at the early phase of ATF 20 fuel development. So we're not five decades or six 21 decades into looking at the same fuel system. So 22 there's, it's a given there's going to be less 23 experimental data in the near term for validating ATF 24 performance models and codes.
25 I think Shane said in his opening remarks 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 that we don't believe modeling and simulation is going 2 to replace experiments, not at all. We still need 3 experimental data. And as someone mentioned or 4 alluded to earlier, we still need experimental data 5 that's going to bound operations.
6 We're not saying in regulatory space that 7 we're going to, we envision the regulator to accept 8 extrapolations far beyond the place where data exists.
9 And these challenges really require a 10 close integration between modelers, experimenters, the 11 industry ATF teams, and the regulator to maximize the 12 quality and applicability of data. And just to 13 comment on that, this close integration is in the 14 process of developing. And we're all talking.
15 The multiscale, mechanistic modeling 16 approach to developing fuel behavior models, and these 17 are models that are informed from the level of 18 microstructure, that creates a validation challenge as 19 well. And we understand that.
20 But we want to draw attention to an 21 important distinction. We are working on separate 22 effects, so-called separate effects experiments that 23 can play a role in validating MARMOT-type models.
24 We're doing work in that area. But to be honest, that 25 is not going to be the main area where validation 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 happens.
2 The validation is still going to be 3 predominantly functioning at the integral pin level in 4 BISON using these new mechanistic models. And that's 5 really --
6 MEMBER KIRCHNER: Steve, I was --
7 MR. HAYES: -- no different than the way 8 the world works today.
9 MEMBER KIRCHNER: I was looking at your 10 tables, just glancing through them. And the pattern 11 that I think I see is that there's less LLS-informed 12 back, informed models in the cladding area than there 13 is in the fuel.
14 MR. HAYES: That's probably right, yes.
15 MEMBER KIRCHNER: I would have thought 16 that's the easier problem.
17 CHAIRMAN CORRADINI: Or is it what you 18 said about chromium cladding that it's not important?
19 MR. HAYES: Well --
20 CHAIRMAN CORRADINI: I mean, I'm trying to 21 understand --
22 MR. HAYES: Yes, so some of that comes out 23 of the sensitivity analyses which show, you know, what 24 properties or behaviors are going to make the most 25 difference. And we sort of go after those first.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
120 1 But there are probably some cladding 2 things that will fall into that area, mechanical 3 behavior of FeCrAl. But I think that's an area where 4 work is happening, for instance.
5 CHAIRMAN CORRADINI: If I might just --
6 because I think we're at the end, and I don't want to 7 delay you from ending so Jess has his time.
8 But my global interpretation is that 9 there's a lot of data to be gathered. So let me ask 10 the question. It's really not to Shane. But I'll ask 11 you, and then you'll turn to Shane.
12 Is it more a matter of DOE partnering with 13 the NRC and industry to gather the appropriate needed 14 data, identifying that as a, I'll call it a team 15 effort than it is to worry about the models, because 16 as I, unless I misinterpreted all the slides, it's the 17 data gathering to me that is the crucial element to 18 move this forward, whether I use BISON or FRAP or FAST 19 or FALCON --
20 MR. HAYES: I think that's an incredibly 21 important area where DOE and NRC needs to work closely 22 together with the vendor teams. If we want to get the 23 right data sooner rather than later, we need to all 24 agree on what the right data is early.
25 MEMBER REMPE: So, and this has been going 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 on for a while. I'm surprised. Earlier you said this 2 is under development. I'm surprised, knowing that how 3 long this ATF program has been going on, that the 4 regulator is just now coming in saying, no, you need 5 X, Y, and Z data.
6 MR. HAYES: And the regulator doesn't say 7 it quite like that. But they do help guide us. Maybe 8 DOE didn't reach out to them as early as we should 9 have. But whatever the history is, I would say now 10 it's functioning --
11 MEMBER REMPE: It's --
12 MR. HAYES: -- very well.
13 MEMBER REMPE: Okay.
14 MEMBER BLEY: Steve, I want to ask 15 something going back to something Walt asked in the 16 earlier session. And I understand most all the 17 validations done at the BISON level.
18 But for the MARMOT model in this 19 microstructure modeling, Walt asked, given all the 20 data that's out there, have you confirmed that it's 21 predicting the right microstructure, and you said, 22 yes, we've done some of that.
23 MR. HAYES: Some, yes.
24 MEMBER BLEY: Can you give me a little 25 idea about how much of that you've done? And when 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 you've done it, I can think of three possible 2 outcomes. You do it and look and you say, yes, it 3 predicted pretty well. Or you look and say 4 something's off. And you dig in and figure out what 5 it is. And maybe you change --
6 MR. HAYES: -- on the model.
7 MEMBER BLEY: -- some basic physics part 8 of the model or something that's kind of general. Or 9 you do some fine tuning to make it work, which might 10 not apply to anything else that comes along later.
11 What kind of things have you run into?
12 MR. HAYES: So we really try to avoid the 13 latter --
14 MEMBER BLEY: I hope so.
15 MR. HAYES: -- fine tuning. And at one 16 point earlier, I said something about calibrating 17 models. And there's certain people who, you know, 18 probably wish I didn't even say that. But, you know, 19 there's going to be some calibration, of course.
20 But the DOE multiscale modeling effort, it 21 really tries to avoid that as much as possible, 22 because then that's when you get --
23 MEMBER BLEY: But sort of what I'm asking 24 is when you've tried this what have you found. I 25 mean, you believe you got the right stuff in 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
123 1 And at the, over at the engineering scale, it's 2 confirmed and everything looks nice. But we could 3 have some weird stuff coming over there that just 4 happens to work.
5 MR. HAYES: Right. And the honest answer 6 is we don't know as much about that as we should. But 7 this last bullet actually speaks to this issue.
8 MEMBER BLEY: Yes.
9 MR. HAYES: So, rather than going off and 10 doing a huge amount of separate effects testing, the 11 approach DOE is taking is a little bit different. And 12 that is, you know, the mainstay of our testing is 13 still going to be integral fuel rods, miniature maybe, 14 but still integral fuel rods.
15 But we can still get at some of the 16 microstructural validation if we'll do two things, 17 one, on the front end, do a much better job of 18 characterizing the microstructure of those fuels so 19 that we have the input data for MARMOT or a BISON, a 20 MARMOT-informed BISON model that we don't always have 21 for those historical experiments that we're analyzing.
22 So we can do a better job of that.
23 And then on the back end, okay, maybe it's 24 cook and look. I don't like that term. But we 25 irradiate a fuel up to a certain burn-up. We take 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
124 1 to the hot cell.
2 And now we've stood up these new 3 facilities like the irradiation, Irradiated Materials 4 Characterization Laboratory at the INL. This is a hot 5 cell, miniature hot cell, with the ability to begin 6 characterizing irradiated fuels on a microstructural 7 level. Okay.
8 So we're going to get the microstructure 9 on the front end. We're going to get it on the back 10 end. That's not a separate effects test. But it's 11 data that can address some of these validation issues 12 for these MARMOT-generated models.
13 CHAIRMAN CORRADINI: So we're getting 14 close to the end. Can we end?
15 MR. HAYES: Yes.
16 CHAIRMAN CORRADINI: Are we there almost?
17 MR. HAYES: We're there.
18 MR. STANEK: Can I make one very quick 19 condition to Steve's comments about the lower length 20 scale modeling, which is you asked a question what 21 have we found when we do this.
22 CHAIRMAN CORRADINI: Yes.
23 MR. STANEK: And I think, generally 24 speaking, even for phenomena that we think we know 25 exceptionally well, we always seem to find something NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
125 1 very fundamental that we didn't understand.
2 And at the engineering scale for UO2 zirc, 3 there's sufficient empirical data to develop empirical 4 models that that missing physics isn't necessarily 5 important.
6 But what it's really doing is setting 7 this, allowing us as a, have a springboard into ATF 8 and advanced reactor fuel to really understand what 9 those, what that missing physics is, so now we can 10 focus on those things that we didn't fully understand 11 --
12 MEMBER BLEY: Okay. That's not surprising 13 I think. Can you give us a hint of how much of that 14 you've been able to do and how much -- you just said 15 there's more of this planned. But how much more and 16 when's it coming? And are there are any reports out 17 at that level?
18 MR. HAYES: Oh, sure, sure.
19 MEMBER BLEY: I don't know if we got those 20 or not. I didn't see --
21 MR. HAYES: We didn't give you a lot of 22 MARMOT level stuff. We gave you mostly BISON stuff.
23 But there's a MARMOT assessment report. I think we 24 provided the link. And there's tons of papers that we 25 could share with 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
126 1 It's in the pipeline. I mean, these 2 things take a few years to really --
3 MEMBER BLEY: So about ten years you'll 4 come back and tell us about that or --
5 MR. HAYES: It doesn't take ten years. I 6 mean, it took, you know, four or five years on the UO2 7 side. But we have a head start.
8 I think we're seeing some benefit already, 9 real results that are good, say, in the area of 10 fission gas release of chromia-doped UO2, major good 11 results there. And we're going to have similar 12 successes across the board in the next year or two I 13 would say.
14 MEMBER KIRCHNER: Now, those good results 15 for fission gas predictions, are those based on 16 empirical models or are they lower length scale?
17 MR. HAYES: No, the chromia-doped stuff is 18 mechanistic models.
19 MEMBER KIRCHNER: And what's the major 20 driver in fission gas release in those?
21 MR. HAYES: This is one of the scientific 22 experts in that area. I'll let him --
23 CHAIRMAN CORRADINI: Briefly.
24 MR. HAYES: Briefly.
25 MEMBER KIRCHNER: Do you 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
127 1 restructuring and cracking of UO2 or does it hold 2 together better? Is it --
3 MR. STANEK: There's no way to make this 4 brief.
5 (Laughter.)
6 MEMBER KIRCHNER: Is it a surface area 7 grain boundary effect or is it --
8 MR. STANEK: So there's a competition that 9 the -- what we found, very quickly, is that the dopant 10 in solution has an effect on not only the graining 11 structure, which is typically through grain size --
12 MEMBER KIRCHNER: Right.
13 MR. STANEK: -- but it also impacts the 14 diffusivity of the fission gas since we have competing 15 factors between grain size and diffusivity --
16 MEMBER KIRCHNER: Right.
17 MR. STANEK: -- which you need to do a 18 BISON calculation with the real power history to 19 evaluate.
20 And what you find is that even though the 21 diffusivity at higher temperature is significantly 22 greater than let's say un-doped UO2, that effect is 23 mostly at high temperature. And so the reason that 24 we're observing higher retention of fission gas is for 25 the reason that the grain sizes are larger.
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 MEMBER KIRCHNER: Thank you.
2 CHAIRMAN CORRADINI: Okay. So I want to 3 give Jess as much as possible. So we're going to 4 probably delay lunch until 12:15 to give you almost 5 what you supposedly were supposed to have.
6 MR. GEHIN: Okay. We'll pull up the 7 presentation here. I'm Jess Gehin from Idaho National 8 Laboratory. I've been there for three months, 9 formerly of Oak Ridge National Laboratory for 25 10 years. So --
11 CHAIRMAN CORRADINI: So you're still an 12 Oak Ridger at heart.
13 MR. GEHIN: I'm rapidly converting. So 14 I'm going to be talking --
15 CHAIRMAN CORRADINI: Just one of many.
16 MR. GEHIN: Yes. So I'm very happy to be 17 out of Idaho, a very nice place to be. A little 18 longer trip to Washington, D.C., but that's part of 19 the job.
20 I'm happy to be here to talk about 21 neutronics and thermal hydraulics modeling. We'll 22 spend maybe about an hour or so, as Mike mentioned, on 23 this and go through what DOE has been working on 24 there. So I appreciate the opportunity to talk about 25 what we've been doing.
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 So the presentation will follow very 2 similar to the fuels presentation. I'll give a quick 3 introduction of the codes, code descriptions, what 4 we've been doing on code validation, look at the 5 capabilities and gaps for accident-tolerant fuels, and 6 have some conclusions, so pretty straightforward.
7 In terms of introduction, DOE has been 8 developing fully coupled and resolved multi-physics 9 core simulation models for light water reactors. The 10 capability being developed under the hub is called 11 Virtual Environment for Reactor Applications, VERA.
12 This has been solely developed on light water reactor 13 development.
14 And you'll hear me talk many times about 15 the focus of this being on getting pin-by-pin detail.
16 One way we differentiate ourselves from industry 17 methods is that we directly calculate this pin detail.
18 So this is for the neutronics pin-by-pin rod powers, 19 which is shown over on the right-hand side there.
20 Thermal hydraulics, we use subchannel for 21 full core. And to get this by the rod subchannels, 22 I'll show you a picture of what that means.
23 Fuel temperature distributions, also rod-24 by-rod, either directly using a code like BISON or 25 using BISON to generate fuel temperature tables 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
130 1 can go in so we can get direct rod-by-rod.
2 One of the areas that we're working on --
3 and I should have said, these tools are being directed 4 at a set of what we call challenge problems or areas 5 that we have identified with industry to work on.
6 One of these is crud. So there is a 7 chemistry model for crud build-up. And we also do 8 very detailed isotopic depletion, so, in general, and 9 I'll go through this in more detail, more resolution 10 in detail than industry codes.
11 As I mentioned, it's primarily developed 12 for LWR's current operating plants. The emphasis, 13 therefore, has been on zirconium clad UO2 fuel in 14 PWRs, primarily for operational performance and safety 15 issues.
16 I mentioned a set of challenge problems, 17 which are things like crud-induced power shift, crud-18 induced localized corrosion, pellet clad interaction, 19 which are more operational issues.
20 And then, more on the safety issues, 21 looking at reactivity insertion accidents, LOCA, and 22 DNB are sort of the hallmarks of things that we've 23 been looking at for UO2 fuels.
24 MEMBER KIRCHNER: And in VERA now, can you 25 do LOCA?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
131 1 MR. GEHIN: The emphasis on LOCA, and I'll 2 have a slide on that, has been on the fuel performance 3 that --
4 MEMBER KIRCHNER: Yes.
5 MR. GEHIN: -- Steve has talked about.
6 So, when I get to that slide to address it, when we 7 look at thermal hydraulics at LOCA, we're really 8 planning to couple to existing system codes.
9 We're not developing as part of this 10 effort a system code capability. So, when it comes to 11 thermal hydraulics for LOCA, we're looking to 12 establish codes there.
13 There's strong industry engagement in 14 development in using these capabilities, particularly 15 for the hub. Westinghouse is a partner. So they have 16 these codes. They're applying these codes.
17 And I'll show you in a minute what the 18 codes are when it comes to neutronics and thermal 19 hydraulics. It also applies to the fuel performance 20 capabilities as well. And this includes, you know, an 21 evaluation of these codes for applicability to 22 accident-tolerant fuel.
23 Overall applicability to accident-tolerant 24 fuel, so these codes, when you look at the neutronics 25 and thermal hydraulics, they've been developed 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
132 1 demonstrated for LWRs, particularly PWR applications.
2 The BWR capability is not as matured.
3 Steady-state operation, investigation of 4 crud-induced power shift, fuel pellet-cladding 5 interaction, I've already mentioned this, operational 6 transients, startup, shutdown, power maneuver, select 7 transients, such as reactivity insertion accidents, 8 and departure from nucleate boiling.
9 So our current application set hasn't been 10 on every, you know, accident scenario and condition 11 that may exist.
12 When it comes to physics and thermal 13 hydraulics, the materials and geometry of most of the 14 concepts, particularly the cylindrical concepts, are 15 within the VERA capabilities. Some modifications in 16 development will be needed for these non-cylindrical 17 fuel geometries.
18 But the physics models, when I talked 19 about the neutronics, are fully applicable to these 20 other geometries. You just need to put those geometry 21 capabilities into the code. And I'll elaborate on 22 that.
23 When it comes to the thermal hydraulics 24 subchannel, it's generally a bit more cruder. And 25 this is where we rely more on CFD directly or CFD 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 informing the subchannel. And I'll give an example of 2 that.
3 They're validated and used for current 4 fuels, of course, zirconium-clad UO2. And we believe 5 they can be extended for ATF concepts.
6 MEMBER KIRCHNER: Just to be clear, sorry 7 to dwell on this. I mean, we're developing accident-8 tolerant fuels to withstand accidents. These are, 9 well, RIAs are one category.
10 MR. GEHIN: Right.
11 MEMBER KIRCHNER: But clearly LOCA is the 12 design basis accident for most of the current fleet.
13 So you would then switch to a different code set to do 14 the --
15 MR. GEHIN: Well, we've been supporting, 16 you know --
17 MEMBER KIRCHNER: Or you would use BISON 18 and what you've got with this --
19 MR. GEHIN: We would use --
20 MEMBER KIRCHNER: -- and then that would 21 be the initial state?
22 MR. GEHIN: No, the way we would apply 23 this capability for LOCA is we would use this 24 capability to get to the initial conditions of a LOCA, 25 whether it's getting metal into cycle in some detail.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
134 1 That would use, and I'll get to the codes in a minute, 2 you know, the neutronics, thermal hydraulics, and 3 BISON to get to that point.
4 At that point, you would switch over to an 5 accident analysis that would have to necessarily 6 involve a systems code like TRACE as an example or 7 RELAP. And that, in the case of accident-tolerant 8 fuels, we would envision using TRACE with BISON. And 9 we've made some, have done some initial work with the 10 NRC on coupling those codes for that specific 11 application.
12 When we scoped out the work for the 13 program where much of this was developed under and 14 looked at the set of problems that we were going to 15 tackle, we had decided taking on the development of an 16 advanced systems code was beyond the scope that we 17 could do. So we focused on the fuel part of that for 18 the LOCA system. Okay.
19 CHAIRMAN CORRADINI: So I'm kind of going 20 back to --
21 MEMBER BLEY: Mic.
22 CHAIRMAN CORRADINI: Oh, I'm sorry.
23 Excuse me, a little green light. Yes, my green light 24 monitor found me --
25 So, normal operation AOOs and selected 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 DBAs --
2 MR. GEHIN: Yes.
3 CHAIRMAN CORRADINI: -- is where you're 4 seeing --
5 MR. GEHIN: These are where we've chosen 6 with and --
7 CHAIRMAN CORRADINI: That's fine. I just 8 wanted to make sure I got it right.
9 MR. GEHIN: I'll just maybe give you a 10 clarification. You're right. Yes, we went through at 11 the start of the, particularly the hub program and did 12 an assessment of --
13 CHAIRMAN CORRADINI: Okay.
14 MR. GEHIN: -- of areas with industry 15 input. Those were the areas selected. It does 16 include LOCA, but it's not full scope LOCA.
17 CHAIRMAN CORRADINI: Well, I mean, just to 18 cut to a fun topic, critical heat flux, critical heat 19 flux under normal operation, you still have to do an 20 experiment.
21 MR. GEHIN: Absolutely.
22 CHAIRMAN CORRADINI: Okay, fine.
23 MR. GEHIN: And --
24 CHAIRMAN CORRADINI: Just wanted to check.
25 MR. GEHIN: Yes. Now, we'll talk a little 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 bit about that. We are seeking to develop improved 2 models. But they're all going to have to be --
3 CHAIRMAN CORRADINI: Okay.
4 MR. GEHIN: -- validated against 5 experiments. So --
6 CHAIRMAN CORRADINI: Okay, thanks.
7 Thanks, Jess.
8 MR. GEHIN: Okay. So now I'll talk a 9 little bit about the codes and code coupling for 10 simulations. So you heard about BISON extensively.
11 On neutronics and thermal hydraulics, the 12 codes, particularly neutronics, we developed a three 13 dimensional whole core neutron transport simulator.
14 It uses 51 energy groups, and I'll compare and 15 contrast these with LWR methods in a little bit, with 16 detailed cross sections.
17 And you're going to see a lot of detailed 18 pictures, details generated directly. There aren't 19 homogenization, dehomogenizations and things like 20 that.
21 For isotopic inventory, we used the ORIGEN 22 capability that provides extensive detail there. And 23 so we have the capability to run this with a different 24 number of nuclides, those important only for 25 neutronics.
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 And if you want to get into looking at 2 inventories that could feed source terms, that can be 3 supported as well.
4 And then what's very common in neutronics 5 is to support these, particularly in the verification 6 of the Monte Carlo codes. So we've been developing a 7 Monte Carlo code called SHIFT that has very good 8 parallel performance so we can run extremely high 9 fidelity models and check out physics for the 10 deterministic impact approach.
11 So those are the three codes. I'll go 12 over those in a little bit more detail.
13 The thermal hydraulics in the lower right 14 is focused at the core level on a subchannel 15 capability COBRA-TF or abbreviated CTF. This is a 16 subchannel capability. You're probably familiar with 17 it because it's been around and used significantly.
18 Transient two-fluid, three-field model, we 19 apply it at every coolant channel, you know, rod 20 channel to get the detail at the rod surface. And 21 that reflects at the core level at least, you know, 22 four regions around a fuel pin that get represented by 23 different thermal hydraulic conditions that can be 24 supplemented with CFD analysis.
25 It lists CFD here. We've used commercial 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 CFD, particularly in the hub, STAR-CCM+. We also use 2 DOE-developed capabilities, NEK5000, that you'll hear 3 more about as well.
4 CHAIRMAN CORRADINI: You're going to, 5 we're going to hear later about what you used the CFD 6 for. I'm trying to understand.
7 MR. GEHIN: Yes, I'll talk --
8 CHAIRMAN CORRADINI: Okay, fine.
9 MR. GEHIN: I'll talk about that. But 10 primarily what we've been working on there is getting 11 more detailed flow distributions to improve subchannel 12 predictions. We've also been working on improving 13 multi-phase CFD predictions. But I'm not going to go 14 into detail on that.
15 CHAIRMAN CORRADINI: So it would be, I 16 mean, just to get to a detail, it would be essentially 17 going, doing a local calculation to what I'll call 18 improve upon the correlation within CTF about the 19 crossflow resistance.
20 MR. GEHIN: That types of thing, the flow, 21 you know, the impacts of mixing vanes --
22 CHAIRMAN CORRADINI: Okay.
23 MR. GEHIN: -- inflow, things that 24 subchannel generally doesn't have models to pick up 25 that type of detail, yet you'd want to try to reflect 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 that at a whole core level.
2 CHAIRMAN CORRADINI: Thank you.
3 MEMBER REMPE: So --
4 MR. GEHIN: Yes.
5 MEMBER REMPE: -- if, you mentioned 6 earlier you'd couple to TRACE if you were going to do 7 the thermal hydraulics. If you do it like an ATWS-I, 8 would you get rid of the PARCS part of TRACE and use 9 your stuff that you've developed? And has that been 10 figured out yet or that's in the future?
11 MR. GEHIN: You can do that. It sort of 12 depends on what you believe is important and how you 13 want to do your analysis.
14 For example, the typical analysis for a 15 LOCA, you usually don't model every rod in the core.
16 You bound that with a peak rod and maybe average rod.
17 And that can be done by coupling BISON with TRACE, and 18 you choose those rods and rod towers.
19 We've been discussing but haven't really 20 pursued significantly coupling the full rod-by-rod 21 detail yet with the systems code to do that type of 22 analysis.
23 But it's something that we'd be interested 24 in, something that CTF with some, a little bit of 25 development work could be applied to that type of area 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 as well if you want to do the full detail. It's 2 really up to the user in that case to decide whether 3 they want to go after that level of detail.
4 MEMBER REMPE: So there's not a firm plan 5 on what the approach will be yet.
6 MR. GEHIN: Well, the current -- and I've 7 got a slide on this. So maybe we'll talk --
8 MEMBER REMPE: Wait till later. That's 9 fine.
10 MR. GEHIN: I'll tell you what we've been 11 doing. And I think it's correct to say there isn't a 12 decided plan on the approach for that. Whereas, I 13 think that, you know, maintaining connections with the 14 current methodology but with more advanced tools seems 15 to be what's being discussed more.
16 Okay. I'm going to talk a little bit of 17 details about each of the codes. MPACT is a 3D core 18 pin-resolved neutronics code. So this is optimized 19 for determining pin-by-pin neutron flux distribution.
20 And what I mean by pin-by-pin is that the 21 fuel rod, and if you sort of look at the diagram in 22 the upper right, each fuel rod is modeled in detail at 23 the rod level for each core. So this is modeled with 24 transport theory, method-of-characteristics, which is 25 a general geometry transport, which is why I'm saying NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
141 1 it can be extended to other fuel types so that we 2 really only applied it to cylindrical.
3 This is applied in a 2D plane across the 4 whole reactor, where most of the neutronics hydrogen-5 80 is. So, if you have differing fuel pins, you have 6 burnable absorbers, you have control rods, all that 7 hydrogen-80, if you look on a plane, you'll see that 8 hydrogen-80.
9 If you do things like put in, you know, 10 different fuel rods or different fuel assemblies, 11 those can be directly modeled with this pin-by-pin 12 capability. And it's a pretty efficient way to do 13 that.
14 Now, these pins are, of course, coupled 15 with a 3D coarse mesh solution. And so you can get a 16 full 3D rod-by-rod solution looking at all the local 17 hydrogen-80s?
18 As I mentioned, currently models 19 cylindrical fuel rods without standard approximations 20 that are made in industry codes. The industry methods 21 are typically based on nodal methods where you model 22 the full geometry detail only at the lattice level.
23 So we model it at the full core level.
24 Pin powers, we compute explicitly. We 25 don't do pin, we don't need to do pin power 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 reconstruction. We don't homogenize anything.
2 There's a direct feedback calculation.
3 There's no cross section functionalization. If you're 4 familiar with that, the typical industry approach will 5 functionalize the cross section data as temperature, 6 density, boron concentration --
7 CHAIRMAN CORRADINI: You mean a series of 8 --
9 MR. GEHIN: -- and make a table. So we 10 just do this directly. So there's no approximations 11 on the, you know, the assumed form of those functional 12 bits.
13 And all of this is to get this rod-by-rod 14 detail. Because of the problems that I mentioned 15 here, the areas that we're working on are all rod-by-16 rod phenomena that we wanted to capture. And we want 17 to do that in a way that's beyond the capabilities of 18 what already exists. Our emphasis --
19 MEMBER KIRCHNER: Can I here, Jess? But 20 I assume you use N death 7 or whatever.
21 MR. GEHIN: Yes, we use the latest cross 22 section libraries. I don't want to get too much in 23 the detail. But if you look at the processing of the 24 libraries that's done using standard tools, we use 25 standard resonance processing approaches, a subgroup 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 method that actually is applicable for these types of 2 problems with the hydrogen-80. But we keep all those 3 models up to date.
4 I had mentioned the emphasis has been on 5 PWR development. There are BWR capabilities, but they 6 have not been developed and validated to the extent of 7 the PWR capabilities. Those are in progress and 8 planned to continue.
9 And then, you know, these physics methods, 10 as I mentioned, are fully applicable to ATF. So, when 11 you look at transport with fine energy group with 12 general geometry, there really aren't limitations 13 there.
14 The limitations or the things that you'd 15 want to look at is to ensure the neutron cross section 16 data and the inputs meet your requirements. For most 17 cases, they're generally acceptable.
18 CHAIRMAN CORRADINI: I guess as I'm --
19 I'll ask the question I asked about the fuel. So what 20 industries have adopted MPACT?
21 MR. GEHIN: So it's in an assessment 22 phase. I'd say it's the same sort of situation.
23 Westinghouse has been a partner in developing this.
24 So they're --
25 CHAIRMAN CORRADINI: Well, I guess, 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
144 1 thought you were going to tell me Westinghouse.
2 MR. GEHIN: Yes.
3 CHAIRMAN CORRADINI: But they have not 4 switched over.
5 MR. GEHIN: I mean, yes, I mean, it's 6 really their call to decide when they switch over.
7 They're in an assessment phase.
8 So one of the -- you know, there are 9 existing tools, of course, they use. And they're 10 looking at these advanced tools. And one of you 11 pointed out to determine is there the value there to 12 invest the --
13 CHAIRMAN CORRADINI: Okay.
14 MR. GEHIN: -- the money into bringing 15 these tools in.
16 We've worked very hard to make these tools 17 usable and accepted by industry. But in the end, it's 18 their decision --
19 CHAIRMAN CORRADINI: Okay.
20 MR. GEHIN: -- whether they do that. And 21 we've been very fortunate to work with Westinghouse 22 and other industry organizations to get data and 23 feedback.
24 We also have an industry council. There's 25 about 20 members of that, including vendors 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
145 1 utilities, that provide strong feedback as well and 2 data as well.
3 CHAIRMAN CORRADINI: Okay.
4 MR. GEHIN: In fact, an example of the 5 data is in the lower right. You know, this is the 6 typical type of measurements you get as flux maps from 7 an operating reactor that we then can compare to. You 8 don't get the detailed pin-by-pin, of course, from the 9 reactor.
10 But these flux maps, and we've got a lot 11 of them now for a lot of reactor types, as I have 12 showed, have been very valuable to understand the 13 deployments of the codes.
14 CHAIRMAN CORRADINI: So what we're looking 15 at is a subassembly somewhere in the core and the 16 little wiggles are the axial variation?
17 MR. GEHIN: Yes. And so what this is, 18 it's a representation of a quarter core of a PWR.
19 CHAIRMAN CORRADINI: Okay.
20 MR. GEHIN: There are --
21 CHAIRMAN CORRADINI: Okay.
22 MR. GEHIN: -- detectors in various 23 locations around the core, and these move up and down.
24 CHAIRMAN CORRADINI: Okay.
25 MR. GEHIN: These are flux maps done once 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 a quarter or so. They've been mapped into one 2 quarter.
3 And this is kind of small and blurry, but 4 we've got numerous maps where we've compared our 5 calculations directly to measured results. You can 6 look at the RMS errors and show that we're getting 7 very good predictions. So we're pretty comfortable 8 with that.
9 The Monte Carlo capability, you're 10 probably familiar with Monte Carlo codes and some of 11 the more, the ones that are a little bit out, used 12 more like MCMP.
13 This shift is a Monte Carlo code in some 14 ways similar to MCMP except it's been designed to 15 scale on very large computers. And the value to that 16 is that we can then run enough particle histories to 17 get statistical uncertainties down at a finer 18 resolution to areas that are less than one percent or 19 half percent so we can ensure that it's not the 20 statistical uncertainties affecting the comparisons.
21 Some of these simulations take, you know, 22 a trillion-particle histories, which are large scale 23 simulations. And so the way we've used this code is 24 to get the best possible answer we can to help verify 25 the deterministic code.
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 And in some cases, you can see here the 2 distribution for AP1000 provided evidence that the 3 MPACT code, deterministic code, does give good local 4 details that can't be measured.
5 So Monte Carlo is the most direct physics, 6 neutronics physics simulation capable of detailed 3D 7 geometry without approximation, representation of 8 fuel, ex-core geometry. So we use this looking at 9 vessel and ex-core measurements.
10 We can get the detailed isotopics and 11 temperature distributions from MPACT to put it into 12 this so we can do verifications at state, you know, at 13 different burn-up state points as well.
14 Combined, as I mentioned, with large scale 15 computer, this provides the best available means to 16 verify more approximate physics models. And it's been 17 widely used with that. As I mentioned, we use this to 18 verify MPACT for many cases, including the AP1000 case 19 showed there.
20 MEMBER MARCH-LEUBA: What thermal-21 hydraulics have you used for the Monte Carlo?
22 MR. GEHIN: It's frozen. So those 23 conditions are frozen. So we run MPACT coupled with 24 CTF with fuel temperature either at tables or BISON.
25 At a state point, we'll take those conditions, put 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 them in Monte Carlo.
2 There are other programs that are looking 3 at coupling Monte Carlo to CFD and other things. But 4 for our purposes for applications, those are beyond 5 the timeframe.
6 MEMBER MARCH-LEUBA: So MPACT and CF use 7 the same TH.
8 MR. GEHIN: They use the same -- no, it's 9 -- so, when you run the Monte Carlo calculation, 10 you're not running a, at least for this application, 11 not running a coupled neutronics thermal hydraulics.
12 MEMBER MARCH-LEUBA: Right.
13 MR. GEHIN: You're freezing the thermal 14 hydraulics.
15 MEMBER MARCH-LEUBA: You run --
16 MR. GEHIN: You run MPACT, CTF --
17 MEMBER MARCH-LEUBA: Running, but it's 18 fragment TH and put it into Monte Carlo.
19 MR. GEHIN: That's right.
20 MEMBER MARCH-LEUBA: Okay.
21 MR. GEHIN: So, like I said, there are 22 programs looking at coupling Monte Carlo and TH, but 23 that was beyond what we felt we could achieve in our 24 program --
25 CHAIRMAN CORRADINI: From a safety NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
149 1 analysis standpoint, are you just -- no, let me ask 2 the question differently.
3 If I were a vendor or an owner/operator, 4 am I essentially identifying margin and then deciding 5 what I can economically do with the margin? I mean --
6 MR. GEHIN: You can do that.
7 CHAIRMAN CORRADINI: -- this is not safety 8 analysis I guess is what I'm kind of saying in a 9 backward way.
10 MR. GEHIN: Yes, so this, yes, and so this 11 can be used to quantify or to help quantify margin, 12 particularly those associated with physics 13 approximation.
14 CHAIRMAN CORRADINI: Okay, fine.
15 MR. GEHIN: So it won't --
16 MEMBER KIRCHNER: It's mainly 17 benchmarking.
18 MR. GEHIN: It won't cover everything.
19 CHAIRMAN CORRADINI: Well, I didn't mean 20 just this. I meant the whole shooting match, MPACT 21 with this with --
22 MR. GEHIN: Yes, so it can be. Remember, 23 some of our scope is operational challenges which 24 aren't necessarily safety-related --
25 CHAIRMAN CORRADINI: Okay.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
150 1 MR. GEHIN: -- and some are, like RIA and 2 those areas. And so they won't cover absolutely 3 everything when it comes to validating those models.
4 But --
5 CHAIRMAN CORRADINI: So, yes, but I guess, 6 so let me ask you. You brought it up. Let me just 7 ask that. So, with the expected, I'm not sure if it 8 will happen, with the expected new RIA rule, is this 9 the only way to address the issue?
10 MR. GEHIN: We've been talking with EPRI 11 and our, and vendors about the use of these tools to 12 help inform that. I don't think there's any decision 13 yet --
14 CHAIRMAN CORRADINI: Okay.
15 MR. GEHIN: -- on their point on how 16 they're going to do it. But we're very interested to 17 see our tools help on problems like that --
18 CHAIRMAN CORRADINI: Okay.
19 MR. GEHIN: -- if it's possible. But I 20 can't say whether that's going to be the solution or 21 not. Okay. All right. I think we're done with that.
22 Here's where I sort of compare and 23 contrast what we've been working on versus what's 24 available in the industry, so, again, whole-core, 25 fully coupled, steady-state, transient thermal 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 hydraulics, fuel performance with neutronics.
2 We removed many of the assumptions in 3 standard codes, particularly the homogenization, 4 dehomogenization, local lattice effects, cross section 5 functionalization, averaging of thermal hydraulics, 6 those sorts of things.
7 And it's been applied to UO2 concepts.
8 Sensitivity analysis, as I'll go through in more 9 detail, on the ATF can be used to investigate some of 10 the ATF fuel concepts to determine whether more 11 validation data is needed.
12 The table at the bottom goes through the 13 different physics model in this capability, the 14 industry practice, and the DOE code, in this case 15 VERA, where the standard practice is 3D nodal 16 diffusion with two energy groups informed by 2D fuel 17 assembly lattice transport where we do direct 3D 18 transport with detailed energy groups.
19 The power distributions, thermal 20 hydraulics, fuel temperatures are typically done at 21 nodal averages in a industry type calculation. We do 22 all that fuel pin resolved. So you can just look 23 where it says fuel pin resolved, fuel pin resolved, 24 fuel pin resolved.
25 There is a cost for this. If you 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
152 1 the bottom, target platform for the industry codes 2 runs on workstations, relatively small systems.
3 Particularly the transport is 4 computationally, more computationally intensive if you 5 want to do it at this resolution. So our target 6 system for that has been 1,000 compute cores, which is 7 a departmental size cluster.
8 Many of those are available in the 9 industry. We've been talking with -- or in DOE.
10 We've been talking with our industry partners about, 11 you know, we make these machines available. And as 12 the value of our codes becomes apparent or they decide 13 there's value, they can invest in machines like this.
14 They're achievable at this level. So -- okay.
15 COBRA-TF whole-core thermal hydraulics, 16 this is a subchannel code. This is a thermal 17 hydraulics subcommittee I believe, so you're probably 18 familiar with subchannel codes, very engineering --
19 CHAIRMAN CORRADINI: Don't give us a test.
20 MR. GEHIN: Okay. Engineering approach.
21 CTF is a two-fluid, three-field representation. So it 22 does single-phase and two-phase flow. For standard 23 PWR operation, of course, we're primarily using 24 single-phase flow with sub-cooled boiling.
25 But we've modeled some cases, if you 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
153 1 over on the lower right, main steam line break, 2 coupling CTF and MPACT that definitely are multi-phase 3 cases as well.
4 This code has been widely used and 5 validated for various reasons. We're doing our own 6 validation as well. I'll talk about that.
7 When we apply, what I mentioned earlier, 8 when we apply this code, we apply it at what I would 9 call a fuel rod subchannel, which is the, you know, 10 the channel that's at the intersection of four fuel 11 rods.
12 It's been subchannels typically applied at 13 a, either that level at an assembly or assembly 14 average conditions or quarter assembly average 15 conditions. But for the full core we model all 50,000 16 or so sub-rod channels.
17 CHAIRMAN CORRADINI: Don't go back, but I 18 should have asked. The neutronics codes that you talk 19 about SHIFT and MPACT, are they now adopted into 20 scale?
21 MR. GEHIN: No, they're separate from 22 scale. There is -- I should take that back. MPACT is 23 separate from scale. SHIFT is being incorporated into 24 scale.
25 CHAIRMAN CORRADINI: And what makes the, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
154 1 what's the decision to make it in or out? That's what 2 I don't understand.
3 MR. GEHIN: To make it what?
4 CHAIRMAN CORRADINI: To decide if it's in 5 or out?
6 MR. GEHIN: Really from a -- I mean, 7 that's a decision that would be made by the scale team 8 on --
9 CHAIRMAN CORRADINI: Oh, so it's really 10 the scale team.
11 MR. GEHIN: It's available. It's fully 12 available if there's use there. And so there's no 13 issue there.
14 CHAIRMAN CORRADINI: Okay, fine.
15 MR. GEHIN: Okay. So I was talking about 16 the rod channel resolution. Again, this resolution I 17 mentioned is applied to every rod in the core. And 18 you can see, for each quarter rod region then you get 19 a variation there. And I'll come back to that.
20 And for some problems, that's not 21 sufficient. For some of the problems that we're 22 working on that's not sufficient. I'll talk about how 23 we address that.
24 Of course, it's transient and steady-25 state. We use it, for example, coupled. If you 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
155 1 at reactivity insertion accident, we model it with 2 that level of detail as well. And it does have cross 3 flow model between channels also.
4 As I mentioned, CFD-informed models are 5 under development. This four azimuthal region around 6 the rod is okay for many cases. But for some problems 7 that we would look at, like crud, that's not enough 8 resolution if you want to resolve the crud layer where 9 you -- if you've seen crud striping, it's in more 10 detail than that.
11 In addition, the grid spacer models are 12 usually represented as losses or approximate models 13 that can be informed by CFD as well.
14 So the applications, PWR, BWR, steady-15 state, and transient, I mentioned the main steam line 16 break problem. We are applying it to reactivity 17 insertion accident as well.
18 We're not applying this right now to a 19 LOCA. Although the code could be extended and applied 20 to LOCA.
21 CHAIRMAN CORRADINI: So a lot of questions 22 come to mind. So LOCA, is there something about the 23 voiding process that then translates back to the BWR 24 also that is the limit? I'm not sure I understand.
25 MR. GEHIN: You know, and I might have 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
156 1 ask for somebody to help with the answer to this. But 2 we have not spent a lot of time validating this 3 current code version for LOCA and applying it to LOCA 4 to understand all of the issues for that application.
5 So I don't think from a fundamental point 6 the code can't model LOCA. It's --
7 CHAIRMAN CORRADINI: You just haven't 8 taken the time.
9 MR. GEHIN: Yes, maybe I'll ask Dave 10 Pointer, who works on the thermal hydraulics for us, 11 to see if I gave the right answer or not.
12 MR. POINTER: So I'm Dave Pointer from Oak 13 Ridge National Lab. And Jess gave the right answer.
14 In the course of CASL, we've actually made 15 some significant investments to improving the 16 stability of the multi-phase analysis capability in 17 CTF.
18 So, in theory, it can be applied to those 19 problems where you do generate significant void in 20 ways that we couldn't in the past. But we have not 21 gone through the next step of beginning to validate 22 those applications in CTF.
23 CHAIRMAN CORRADINI: Okay.
24 MR. GEHIN: Yes, and so, when we bring up 25 discussion of LOCA, that question of validation always 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 comes up. And it's on the list. If there's folks 2 that are interested in applying these codes that way, 3 we'd definitely be interested in looking at that.
4 CHAIRMAN CORRADINI: Okay. Thank you.
5 MR. GEHIN: So the capabilities here, 6 again, sort of the bottom line trying to tie it back 7 to ATF, is the CFD-informed rod-by-rod thermal 8 hydraulics can be used to model ATF.
9 You know, if you look at the geometry, 10 particularly the cylindrical geometry applicable, you 11 may have to take into account surface conditions, 12 things like that. But the capability, I believe, can 13 be applied and extended.
14 I'm going to go through a few examples to 15 show, to emphasize some of the things I was talking 16 about. And the first one is CFD-informed subchannel 17 modeling.
18 Some of the problems that we look at, this 19 four azimuthal regions that you get with CTF is not 20 sufficient in the models, you know, that used for heat 21 transfer aren't sufficient. So what we've done is 22 generated models, fuel assemblies, multiple grid spans 23 in CFD and calculated detailed flow distributions.
24 When you get these flow distributions, you 25 can use those to back out, if there's mapped work 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 here, around the rod, you know, the heat transfer 2 coefficient. And that's what this colored plot shows.
3 If you look along the axial, this is, you 4 know, on the left-hand side, the axial position on the 5 rod. And if you take the rod surface and roll it out 6 into a 2D plot, you see this surface, which gives you 7 a map of the detailed heat transfer coefficient around 8 the rod and up the rod.
9 It's impacted. And you can see these 10 different levels, of course, by the mixing vanes and 11 those details.
12 So, once you have these heat transfer map, 13 and it's generally done by a ratio of the actual heat 14 transfer coefficient to, say, a Dittus-Boelter heat 15 transfer coefficient that's used in CTF, you can input 16 that into CTF and get an improved simulation of the 17 details around the rod, as well as predicting the 18 overall heat transfer.
19 MEMBER MARCH-LEUBA: So what we're seeing 20 there, the horizontal lines are the spacers?
21 MR. GEHIN: Yes, these are the spacers.
22 MEMBER MARCH-LEUBA: And then what is that 23 plume where you're pointing right now?
24 MR. GEHIN: This plume right here?
25 MEMBER MARCH-LEUBA: 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
159 1 MR. GEHIN: Yes, Dave, do you have an 2 answer for that? I mean --
3 MR. POINTER: So the plume that you see 4 there is actually an opening in the configuration of 5 the spacer grid for that particular pin. This is the 6 central pin in a 5x5 bundle.
7 MEMBER MARCH-LEUBA: So it's on a smoother 8 symmetry of the --
9 MR. POINTER: It's an anomaly in the 10 dimple and spring configuration in that particular --
11 MEMBER MARCH-LEUBA: Any properties much 12 --
13 MR. POINTER: Yes.
14 MEMBER MARCH-LEUBA: -- it doesn't mix.
15 MR. GEHIN: Okay. Thanks, Dave. So we've 16 applied this to a series of case. I won't go into the 17 details.
18 But this chart here, this plot on the 19 right-hand side gives the CTF temperature prediction.
20 Solid lines are CTF, standard CTF. The dash lines are 21 with these improved heat transfer coefficients. So we 22 can a proved simulation. You can also see the error 23 as you go rod by rod is a lot more uniform.
24 CHAIRMAN CORRADINI: So say that again, 25 please. I'm sorry.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
160 1 MR. GEHIN: Okay. So this plot gives a 2 comparison of the difference between CTF and STAR. So 3 STAR is the reference. CTF is, of course, where we've 4 input these heat transfer coefficients and done a 5 calculation of CTF temperature prediction. And --
6 CHAIRMAN CORRADINI: Okay. But I guess I 7 should have asked my question more specifically. The 8 Y axis is an error in degrees Kelvin?
9 MR. GEHIN: I believe that's the case --
10 CHAIRMAN CORRADINI: Okay. So now I'm 11 going to ask the engineering question. Who cares?
12 MR. GEHIN: So, when we looked at --
13 CHAIRMAN CORRADINI: I don't mean to be --
14 MR. GEHIN: No, no, no, it's a good 15 question because we have the -- when you look at the 16 application of crud, crud is very sensitive to when 17 you get at a boron deposition threshold of when boron 18 deposits and when it doesn't deposit.
19 So, when we did our analysis looking at 20 the homogenized crud, four regions per rod, we found 21 that we did not accurately predict the crud deposition 22 and we had to calibrate that model, because there were 23 some regions that were right on the edge of this 24 threshold where temperatures like this mattered.
25 Now, it may be for your problem 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
161 1 doesn't --
2 CHAIRMAN CORRADINI: But your point is 3 you're close to a threshold. Now I need to know the 4 answer more carefully --
5 MR. GEHIN: Well, the other value to this 6 is when you looked at modeling new grid spacers or 7 incorporating grid spacer designs, certainly you're 8 going to get data from experimental measurements. But 9 you're not going to get this resolution of data where 10 you're going to have to have a model in your --
11 CHAIRMAN CORRADINI: Okay.
12 MR. GEHIN: -- subchannel code to 13 represent grid spacer.
14 So that's another application where maybe 15 you're not looking totally at improved accuracy 16 prediction, but you do need a model to represent the 17 impacts of those grid spacers. And that certainly has 18 an impact on crud, DNB, and other areas as well. So 19 those are the areas where we've really been focused on 20 this --
21 MEMBER MARCH-LEUBA: So basically that 22 figure on the right says that when you use the same 23 heat transfer coefficient in the STAR and CTF, you get 24 the same temperature.
25 MR. GEHIN: 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
162 1 MEMBER MARCH-LEUBA: And if you use the 2 wrong heat transfer coefficient, you get the wrong --
3 MR. GEHIN: I mean, it's not horrendously 4 wrong, as Mike mentioned, but it does allow you to get 5 a more uniformed prediction. If you look at the air 6 variations, they're much smaller.
7 And so we think it's one of the keys for 8 calculating crud. It may be important for DNB as 9 well. Certainly, grid spacer mixing is an effect that 10 we can capture also.
11 All right. CFD evaluation of DNB, so this 12 is something we've been working on quite a bit on the 13 hub program to be able to apply CFD directly to 14 predict when DNB conditions would occur in a PWR with 15 the idea of being able to supplement experimental 16 data.
17 So, typically what's done in industry is 18 they take a, say, 5x5 rod bundle, do DNB testing on 19 that, and check things out, particularly when there 20 are changing fuel designs like grid spacers.
21 We would like the ability to be able to 22 inform them on their grid designs that may result in 23 them having to do fewer DNB tests or find out things 24 during the DNB test that would change their grid 25 spacer designs.
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 So we've been incorporating multi-phase 2 model development to DNB, combining this with more 3 fundamental measurements on heat partitioning between 4 liquid and vapor phases, you can see some of the 5 physics that we've been looking at there, and then be 6 able to put this into a code like STAR-CCM or NEK5000 7 to be able to run, basically mimicking what an 8 experiment for electrical heated DNB test would look 9 like where you slowly ramp up the power and then you 10 detect when you actually have, you know, a temperature 11 excursion indicating DNB.
12 We've been able to get data on this 5x5, 13 bundled data --
14 CHAIRMAN CORRADINI: This is Westinghouse 15 data?
16 MR. GEHIN: This is Westinghouse data.
17 CHAIRMAN CORRADINI: Okay.
18 MR. GEHIN: Proprietary data for mixing 19 vanes and non-mixing vanes. I'm showing the non-20 mixing vane case here, which shows basically the 21 predictions are within plus or minus 16 percent or 22 better. So --
23 CHAIRMAN CORRADINI: I noticed there's no 24 axes label. So --
25 MR. GEHIN: Well, sorry about that. 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
164 1 this is a case, the value that we've had working 2 directly with Westinghouse and some of the other 3 industry organizations where we may be able to get 4 access to data. But it is proprietary.
5 MEMBER MARCH-LEUBA: So the points, the 6 sequence you're putting there are different flow 7 pressure and power.
8 MR. GEHIN: Yes, I believe that's the case 9 --
10 MEMBER MARCH-LEUBA: But the same spacer, 11 right?
12 MR. GEHIN: Yes. So, you know, the 13 vendors, of course, this is how they validated their, 14 or developed their own DNB correlations with this type 15 of data. And so they've given us some of that data.
16 And then they've actually taken these capabilities and 17 applying it themselves --
18 MEMBER MARCH-LEUBA: Right, but the value 19 of this is on the computer you can have a different 20 spacer and see how it changes.
21 MR. GEHIN: Right. And, you know, one of 22 the challenge with using CFD for this is applying 23 single-phase CFD, you know, you can only get so far in 24 that, and then when you do your test and actually have 25 two-phase, it may or may not perform --
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 MEMBER MARCH-LEUBA: By the condition, 2 yes.
3 MR. GEHIN: Yes. And so we're trying to 4 expand the application of CFD for these tests. At 5 some point, maybe we can rely on it more. But we 6 don't want to -- you know, what we want to really be 7 able to do is try to better inform these expensive 8 tests.
9 And so, of course, this is an area then 10 for ATF fuels if you're changing things that would 11 impact flow patterns to investigate impact on DNB 12 might, at this state of maturity, might be able to 13 inform where you think data is needed or if you're 14 comfortable with where things are. Okay.
15 MEMBER MARCH-LEUBA: For ATF fuels, you 16 just pick the CFD correlation, the DNB correlation to 17 change.
18 MR. GEHIN: I'm not an expert in this 19 area. In general --
20 (Simultaneous speaking.)
21 MR. GEHIN: -- experts on that side of the 22 table. It's generally driven by the mixing vane, the 23 geometry. There could be some surface condition 24 effects. But I would not --
25 MEMBER MARCH-LEUBA: If the material works 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 differently, the water level would be --
2 MR. GEHIN: Right. And then if you look 3 at non-cylindrical geometry, of course, there's --
4 MEMBER MARCH-LEUBA: Sure.
5 MR. GEHIN: -- definitely applications.
6 MEMBER MARCH-LEUBA: But for anything that 7 we are thinking in the near future, if you look from 8 the outside, the pins look the same.
9 MR. GEHIN: They look basically the same.
10 You're right. So that's why, if needed, this could be 11 applied.
12 Transient capabilities, I'll talk about 13 reactivity insertion accident. So this has been one 14 of the target problems that we've had, control rod 15 ejection for a PWR.
16 This is a capability where we've coupled 17 neutronics, thermal hydraulics with fuel temperature 18 table model to calculate, you know, the conditions in 19 the core. This is a four-loop PWR core where there's 20 a postulated control rod ejection worth $1.50, which 21 is, you know, these things are always done 22 conservatively.
23 And we can get detailed rod-by-rod power 24 distributions. This figure here shows you that's, 25 where the ejected control rod is you end up with a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
167 1 power pulse, if you can squint a little bit at that, 2 and get an axial power shape.
3 So we get all those details for every rod.
4 You can identify rods that you may want to look at, 5 use those as conditions that would go into a fuel 6 performance simulation as well.
7 We are looking at coupling and working on 8 coupling BISON indirectly. But it's usable in the 9 current state as I just mentioned. Of course, this is 10 also something that can be applied to ATF fuels as 11 well, and it's fully functional.
12 Okay. Coupling DOE codes with reactor 13 systems codes, so as I mentioned, for our work we've 14 been not relying on development of a DOE reactor 15 systems code. That may come about at some point where 16 we work on that. But in the timeframe we had, the 17 vision was to couple with existing reactor systems 18 codes.
19 So this provides a capability to analyze 20 flow regimes which CTF has not been validated. And 21 that's what I meant when we talked about the LOCA, 22 that while the capability is there, it's not been 23 validated.
24 And so we've been working with the NRC to 25 couple, you know, particularly BISON with the NRC code 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 system to look at LOCA. It's been a joint DOE/NRC 2 effort.
3 TRACE and BISON, which have been coupled 4 as a demonstration, one, you know, primary reason for 5 doing this is to show that it can be done. It can be 6 done in a relatively quick fashion that you can choose 7 some of these codes that you want to couple. And, of 8 course, TRACE is the NRC's safety analysis code.
9 The idea here is if this capability could 10 buy a means to simulate ATF simulation for LOCA as a 11 -- where other transients, TRACE has been widely 12 applied to these transients. You've heard about the 13 state of development of BISON. And so you can bring 14 those two together.
15 MEMBER MARCH-LEUBA: Does it provide two-16 way coupling?
17 MR. GEHIN: It's two-way coupling.
18 MEMBER MARCH-LEUBA: So, but --
19 MR. GEHIN: This is a --
20 MEMBER MARCH-LEUBA: -- BISON provides the 21 conductivity for TRACE --
22 MR. GEHIN: That's right.
23 CHAIRMAN CORRADINI: Can I just -- I'm 24 sorry. I didn't mean to interrupt you.
25 MR. GEHIN: No, no.
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 CHAIRMAN CORRADINI: But I'm struggling to 2 see -- I'm looking at your slides and looking at the 3 time. I'm struggling to see how this affects -- this 4 is all interesting. But I'm trying to make the bridge 5 to ATF, and I don't see a clear bridge.
6 MR. GEHIN: Okay. So the idea, in this 7 case, the idea on ATF, whether it's for NRC or 8 somebody else, if you decide BISON is a good code for 9 ATF --
10 CHAIRMAN CORRADINI: Okay, fine.
11 MR. GEHIN: -- you don't have to abandon 12 your systems code, which you put a lot of time and 13 effort in developing. You can use it with BISON as an 14 example.
15 CHAIRMAN CORRADINI: Okay.
16 MR. GEHIN: So that's the idea.
17 MEMBER REMPE: So you've totally switched 18 gears with this slide, right? You're not talking 19 about any of your MPACT stuff with this.
20 MR. GEHIN: That's exactly right.
21 CHAIRMAN CORRADINI: That was about four 22 slides ago.
23 MEMBER REMPE: Okay. Yes, well, I just 24 wanted to make sure, because earlier I had asked, 25 well, are you going to continue using PARCS 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
170 1 TRACE system. And I think even your last slide talked 2 about MPACT and CTF, right. But this particular 3 coupling did not use anything from your earlier stuff.
4 MR. GEHIN: No, no. And, you know, and 5 so, yes, just to be clear, you know, PARCS is not a 6 DOE-developed code. It's not in the --
7 MEMBER REMPE: Right.
8 MR. GEHIN: -- that I talked to you about 9 --
10 CHAIRMAN CORRADINI: But that's okay.
11 MEMBER REMPE: That's okay. But --
12 MR. GEHIN: I'm just, I didn't know where 13 you were coming from --
14 MEMBER REMPE: Right, I just was curious 15 because of my earlier question about what the vision 16 would be and stuff and --
17 MR. GEHIN: Oh, you know, LOCA simulation 18 does not use neutronics. And so, you know, you get to 19 the depleted point of the core, then you have a LOCA 20 event, and you assume your SCRAM. You have the decay 21 heat. And so this is, you know, it becomes a thermal 22 hydraulics fuel performance.
23 MEMBER REMPE: But that says, for example, 24 LOCA. And again, I'm back to ATWS-I. And again, this 25 just was focused on doing a LOCA or a station 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 blackout.
2 MR. GEHIN: Well, I knew LOCA, yes, and I 3 knew LOCA would be a question that came up. And so I 4 wanted to talk a little bit about what the thinking is 5 there.
6 MEMBER REMPE: So --
7 MR. GEHIN: That's why I introduced this 8 slide.
9 MEMBER REMPE: And on this analysis, are 10 the results documented in one of the reports we were 11 provided ahead of time?
12 MR. GEHIN: I don't believe we provided 13 this. We can provide a report.
14 But I'll emphasize, this is a 15 demonstration, primarily answering the question is can 16 you take a DOE code that's been developed in the DOE 17 system and couple it with, say in this case, an NRC 18 code. You could consider doing the same thing with an 19 industry code where the end user may want to keep 20 their own systems code. Can you efficiently take an 21 outside code and do that coupling? And so that was 22 the primary purpose of this demonstration --
23 MEMBER REMPE: Okay.
24 MR. GEHIN: -- which was successful. And 25 this figure down in the bottom is actually a movie, 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 but I think this is a PDF. So --
2 MEMBER MARCH-LEUBA: Yes, it's a PDF.
3 MR. GEHIN: Yes. But that was the point 4 of it. So, if you're not able to relatively easily 5 couple these codes, then there's no reason why you 6 would go beyond that point. So it's really to get 7 past the first phase of can you do that.
8 MEMBER REMPE: And from -- again, we 9 didn't see the results of it. But what was better 10 because you did this? I mean, you've showed you can 11 easily couple them. But did you get any results that 12 you couldn't have obtained using their own tools?
13 MR. GEHIN: Well, the, again, the idea 14 here is if you believe or if an end user would like to 15 use BISON because they believe BISON is a good tool to 16 model ATF, it would then provide those capabilities.
17 You would get the fidelity of -- in fact, 18 we can circle back to that question. You get the 19 fidelity of TRACE. You apply it to either a single 20 rod model or however you want to apply it, and then 21 apply BISON to use its ATF modeling capability is the 22 vision --
23 MEMBER MARCH-LEUBA: It is a PowerPoint 24 you can take on the --
25 MR. GEHIN: Oh, is 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
173 1 MEMBER MARCH-LEUBA: Yes.
2 MR. GEHIN: Oh, you're right. Yes. And, 3 I mean, this just shows, this is like fully coupled.
4 It's calculating temperatures. It's calculating void 5 fractions and in a two-way sense just --
6 MEMBER MARCH-LEUBA: Order of magnitude to 7 how many person-weeks or person-years or how long does 8 it take to do this?
9 MR. GEHIN: As far as the coupling?
10 MEMBER MARCH-LEUBA: Yes.
11 MR. GEHIN: I mean, the initial coupling 12 I think was done relatively quickly. I don't know if 13 there's somebody, maybe Steve Bajorek --
14 (Simultaneous speaking.)
15 MEMBER MARCH-LEUBA: Because, I mean, I 16 can do it in FORTRAN, right?
17 MR. GEHIN: Yes.
18 DR. BAJOREK: This is Steve Bajorek, 19 Office of Research. We started this work I believe in 20 last October or November. They had the essentially 21 coupling features done within a couple of days.
22 The very difficult thing is taking TRACE's 23 adaptive mesh where we look at the fine mesh re-24 nodalization, which is moving on a component basis on 25 the fly, and mapping that into a BISON code, which is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
174 1 a finite element mesh in 3D.
2 The reason we picked this is this is about 3 the most difficult coupling exercise we could think of 4 in trying to take something like TRACE and couple it 5 over to the MOOSE --
6 MEMBER MARCH-LEUBA: But you're talking a 7 week, a week of work.
8 DR. BAJOREK: A little bit longer than 9 that to get everything done with that adaptive mesh --
10 MEMBER MARCH-LEUBA: But it's not ten 11 years.
12 DR. BAJOREK: Not ten years, no.
13 MEMBER MARCH-LEUBA: Good.
14 DR. BAJOREK: No.
15 (Simultaneous speaking.)
16 MEMBER KIRCHNER: Go ahead.
17 CHAIRMAN CORRADINI: I want to be careful 18 on time. But I'm going to ask, so now for ATF, this 19 is interesting. But where is the ATF application?
20 What am I missing?
21 MR. GEHIN: Okay. So, if you decide you 22 want to use BISON as your performance code --
23 CHAIRMAN CORRADINI: But these are just 24 demonstrators of using BISON in lieu of FRAPCON and 25 FRAP-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
175 1 MR. GEHIN: That's an -- yes, that's 2 exactly right.
3 CHAIRMAN CORRADINI: Or in lieu of FAST.
4 MR. GEHIN: Yes, or if industry wanted to 5 take RETRAN coupled with BISON, they --
6 CHAIRMAN CORRADINI: Okay, fine. I get 7 it.
8 MR. GEHIN: It's a demonstration that you 9 can do the coupling, because there were questions 10 about that. And, you know --
11 CHAIRMAN CORRADINI: I know.
12 MR. GEHIN: -- can it be done? You know, 13 you got a, and the legacy code and you got a new code.
14 Can you do it?
15 CHAIRMAN CORRADINI: I got it.
16 MR. GEHIN: But you would have to buy in 17 or, you know, decide BISON is a code that you would 18 want to use, okay, and specifically focusing on the 19 thermal hydraulics of LOCA.
20 DOE code applications for source terms is 21 a question you asked. You know, the answer we had is 22 we're not working on it beyond design basis accidents.
23 But there are some things here that can provide inputs 24 for source term.
25 So we can get, you know, high resolution 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 reactor inventories. These have been well-validated.
2 These can be used as inventories that provide, you 3 know, or provide inputs to a code such as MELCOR, get 4 detailed rod-by-rod. These can be averaged or however 5 it needs to be. There are validation data from PIE 6 and reactor operation.
7 We can, since we're using ORIGEN, we can 8 provide the full detailed isotope set that's 9 consistent with NRC applications for beyond design 10 basis accidents.
11 BISON calculates fission gas release.
12 This also is information that could be useful for 13 beyond design basis accidents.
14 CHAIRMAN CORRADINI: But here I'm going 15 to, what little I know I'm going to embarrass myself.
16 It's not just the fission gas release. It's the 17 coupling of that to a series of accident scenarios and 18 the frequency and seeing if I'm going to change the 19 alternative source --
20 MR. GEHIN: And so --
21 CHAIRMAN CORRADINI: Okay.
22 MR. GEHIN: -- that part we are not doing 23 just to be clear. I wanted to be responsive to your 24 --
25 CHAIRMAN CORRADINI: No, no, that's fine.
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 I appreciate that.
2 MR. GEHIN: Yes, and then again, for 3 transient simulation, some of these, the transients 4 that I talked about could be used as information 5 informed beyond design basis accidents as well.
6 As addition, if you want to run a case at 7 end of cycle, we can deplete out the end of cycle and 8 provide all those initial boundary conditions that 9 support beyond design basis accidents. Okay.
10 Let me move on to validation. So the 11 validation approach involves single and coupled multi-12 physics. This chart here, when you look at applying 13 a coupled system or a core simulation type code are 14 the typical areas that you look at, critical 15 experiments.
16 Monte Carlo is usually used as a key part 17 of this for verifying these approximations, as well as 18 being able to map between the geometry that your codes 19 model versus what a critical experiment may be, for 20 example.
21 Fuel rod PIEs and then operating power 22 plant data are the sources of data that we have. And 23 again, Monte Carlo supplements all of this. This is 24 very typical and standard of what's done with existing 25 codes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
178 1 Subchannel validation, if you look at CTF, 2 I believe we've provided the assessment report for 3 that. There's the identified phenomenon for 4 validation. I won't go through the whole list here.
5 You're very familiar with that.
6 We've got a set of available open 7 experiments listed on the right for full channel and 8 subchannel. And in the case of like DNB, we've gotten 9 access to some proprietary data.
10 Again, this would all be supplemented if 11 a vendor were to pick up this tool with the remainder 12 of the proprietary data that they have.
13 Okay. VERA validation, on the plant 14 level, for existing fuels, there's an assessment 15 report for this. Again, this has been our emphasis 16 and, of course, where all the operating data is. So 17 we've applied it to a large number of PWR plants and 18 operating cycles.
19 Typically, the comparisons are made for 20 zero power physics tests. You've got criticality, rod 21 worth, and then flux maps during power escalation, 22 operational measurements, soluble boron, and flux 23 maps, and then our operational transients, start-up, 24 shutdown, or changes in power during operation such as 25 maybe a load-follow event where we have data as 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
179 1 These plants and cycles listed here, I 2 won't go into this detail. But they were chosen for 3 nominal conditions and some of these operational 4 occurrences like PCI and crud, which is why you'll see 5 different plants listed.
6 MEMBER MARCH-LEUBA: You said this 7 operational data proprietary?
8 MR. GEHIN: Most of it is, particularly 9 the fuel data. There are -- some of the cycles for 10 Watts Bar, the first five cycles we've been able to 11 get released. But particularly when you get to the 12 fuel data, that, the compositions and --
13 MEMBER MARCH-LEUBA: Oh, sure, the fuel 14 data but --
15 MR. GEHIN: Yes.
16 MEMBER MARCH-LEUBA: The fuel data you can 17 get it --
18 MR. GEHIN: Yes, so we've gotten this --
19 MEMBER MARCH-LEUBA: -- but I mean the 20 plant measurements.
21 MR. GEHIN: Yes, so we've gotten those 22 directly from the utilities. They're not public data 23 usually. Some of the older cycles like I mentioned at 24 Watts Bar TBA is released. And there's a public 25 document, and it's been used by others for --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
180 1 MEMBER MARCH-LEUBA: You have a DVD 2 somewhere with it just in case.
3 MR. GEHIN: Yes. So a lot of, we've been 4 able to get a lot of good data on this.
5 And there are different reactors, 6 Westinghouse four-loops, 17x17, 16x16. But, of 7 course, these are all zirconium UO2 fuels. So we've 8 got a pretty good pedigree of how this code performs 9 for operational plants that would support a good 10 validation case.
11 Transient, the data there is not as widely 12 or as much data for, if you look at reactivity 13 insertion accidents for the coupled physics largely 14 rely on the SPERT test. We use that for our 15 validation case as well.
16 I believe you're probably familiar with 17 SPERT, a PWR test reactor with a central rod that can 18 be ejected to go through a power transient.
19 And here's an example of a case, generally 20 good agreement. Again, we're modeling this in full 21 detail with the DOE codes, resolved fuel geometries 22 and thermal hydraulics with CTF with the fuel 23 temperature table, so good validation data there.
24 Now, to the ATF assessments, we've got 25 tables that are somewhat similar to what Steve had for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
181 1 BISON. Across the top we have the different fuel 2 types. And by U metal fuel, this means non-3 cylindrical geometry, just to clarify.
4 So we've got geometry, physics models, 5 materials and nuclear data, validation, have we 6 performed validation for this and is there design 7 specific validation data available. And then I've got 8 some notes below that you can refer to.
9 You can see, of course, for UO2 we have, 10 all these are yeses. Coated clad yeses, when you get 11 down to validation performed and validation data for 12 the neutronics, we don't have that data. But we 13 believe that applying a coating is well within the 14 physics capability and prediction capability of the 15 neutronics code because it's not a large impact.
16 Same is true for doped UO2 or fuel, it's 17 not a large perturbation. And if necessary, you can 18 perform sensitivity analysis to confirm that.
19 When you look at the FeCrAl or the iron-20 based clads, again the physics models and data are all 21 applicable. We've not performed specific validation 22 yet with the codes, although there is previous 23 operational data with steel clads and, as well as data 24 that could be generated through LTAs. Sensitivity 25 studies could help inform that as 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
182 1 Yes, go ahead.
2 CHAIRMAN CORRADINI: Can you expect a 3 difference, though? I'm -- it's the doped -- I'm 4 sorry. I missed that you had the U3Si. But if I look 5 down the row here about it, it's not the cladding so 6 much as the fuel constituent, right?
7 MR. GEHIN: We think the cladding, and we 8 can show this through sensitivity studies. It changed 9 the absorption properties. But that's all within our 10 --
11 CHAIRMAN CORRADINI: Okay.
12 MR. GEHIN: I think I'd be comfortable to 13 say that's within our physics prediction.
14 CHAIRMAN CORRADINI: Okay, fine.
15 MR. GEHIN: You'd want to do a sensitivity 16 study.
17 CHAIRMAN CORRADINI: Oh, no, that's fine.
18 MR. GEHIN: When you get to the fuel where 19 you're changing fuel density and things like that, 20 which I think is, you want to look at more carefully, 21 we -- and again, where it says no here, we have not 22 done the validation. The capabilities you see above 23 this line in general are yes. But we've just not done 24 --
25 CHAIRMAN CORRADINI: Okay.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
183 1 MR. GEHIN: -- the validation. And in 2 most cases, we don't have the design-specific 3 validation data.
4 Now, the one caveat on that, when you look 5 at the non-cylindrical fuel geometry, we've not 6 implemented that in the code. The physics models and 7 the data are applicable. But it's not been 8 implemented. So that's the state of where we are 9 there.
10 Thermal hydraulic capability, I think as 11 Jose mentioned, you know, for all the cylindrical fuel 12 rods, pretty much the data, you know, or I'd say the 13 geometry capabilities and the data is applicable. Of 14 course, for fuel we put not applicable because that's 15 within the, the cladding doesn't impact that.
16 We've prepared, we've performed validation 17 for these cylindrical fuel types. And there's 18 validation data generally available. And where it 19 says no here, you know, you may be able to convince 20 yourself that the current cylindrical fuel rod data is 21 acceptable.
22 For the non-cylindrical geometry, you 23 know, we've not implemented that. We've not looked at 24 whether the thermal hydraulic models would change.
25 The flow distributions would certainly be different 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
184 1 you go to that type of fuel. So that would have to be 2 investigated. And we've not done validation on that 3 --
4 CHAIRMAN CORRADINI: But again, I'm going 5 to ask this --
6 MR. GEHIN: Yes.
7 CHAIRMAN CORRADINI: -- to make sure, 8 though. And when we're thinking of this, I'm always 9 thinking of thermal hydraulic coupling back to the 10 fuels model.
11 MR. GEHIN: Yes.
12 CHAIRMAN CORRADINI: Okay.
13 MR. GEHIN: Yes. Okay. All right. So 14 conclusion, so the higher-resolution, fully coupled 15 capabilities are developed and are applicable to ATF.
16 These are generally done at a higher resolution than 17 currently available that can be used for, at a minimum 18 for investigation of impacts and insertion of LTA and 19 can be checked, you know, against higher fidelity 20 methods.
21 You can actually reduce time, in our 22 experience, in a capability like this you can reduce 23 the time to perform investigations because the current 24 industry methods take pre-generation of data. It has 25 to be tabulated, put into a code. Whereas, if you can 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 just directly simulate something, it's more computer 2 time rather than the effort spent investigating 3 challenges.
4 We have a significant validation for UO2 5 forms. Most of this could be leveraged for ATF. As 6 I mentioned, the physics we have are applicable. The 7 data is applicable. And there are sensitivity 8 approaches or Monte Carlo that can be used to 9 determine whether you need to do further, get further 10 experimental data or not. It can be used by industry 11 or NRC to support, you know, their current tools or be 12 adopted.
13 The system code, a coupling allows for 14 evaluation of broader transients. You know, as we 15 discussed in some detail, we're not developing that 16 part right now within this activity. There could be 17 some future activities in DOE that could be applied 18 there. But we've demonstrated that you can take the 19 capabilities we have and fairly readily couple those 20 with systems capability, which is, of course, 21 important for safety.
22 And I believe that's it. So are there any 23 questions?
24 CHAIRMAN CORRADINI: Do the members have 25 any more questions? No? 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
186 1 So why don't we take a lunch break? And 2 we'll come back here, reconvene at 1:15. Okay.
3 (Whereupon, the above-entitled matter went 4 off the record at 12:16 p.m. and resumed at 1:14 p.m.)
5 CHAIRMAN CORRADINI: Chris, I think you're 6 going to lead us off, right?
7 MR. STANEK: Yes. Just a few quick comments 8 to start the afternoon session. In the afternoon 9 session we will switch gears from accident-tolerant 10 fuel to non-like water advanced reactors. Although you 11 have heard this morning during the ATF presentations 12 about some of the codes that we will discuss this 13 afternoon, there are still other codes that haven't 14 yet been introduced, and so as such we've generated 15 another version of this table.
16 As was the case this morning and in the 17 interest of time I won't talk about any of the details 18 or any of the codes in this table, but hopefully it's 19 a useful guide for you as we go through the afternoon 20 presentations.
21 Also, as was the case this morning, one 22 idea of this matrix is that it serves as a detailed 23 agenda. So first this afternoon we'll hear from Tanju 24 Sofu from Argonne National Laboratory, who will 25 present neutronics code developments for non-LWRs.
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 Tanju will be followed by Rich Williamson 2 from Idaho National Laboratory, who will present fuel 3 performance. Rich will be followed by Elia Marzari 4 from Argonne National Laboratory. He'll present 5 thermal-hydraulics.
6 Then our final technical presentation of 7 the day will again be from Tanju Sofu, who will give 8 a presentation of mechanistic source terms for non-9 LWRs. If there are no questions for me, I'm happy to 10 turn it over to Tanju.
11 CHAIRMAN CORRADINI: Go ahead.
12 MR. SOFU: Good afternoon. As the first 13 technical presentation of the session today, I'm going 14 to cover recently developed neutronics analysis 15 capabilities for advanced non-water cooled reactor 16 designs. After a brief background, only one page, on 17 motivations for developments of these new advanced 18 modeling and simulation capabilities I will introduce 19 two neutronics analyses codes, PROTEUS suite and 20 RATTLESNAKE, to address their validation of the 21 verification basis and cover some example applications 22 for SFRs, high-temperature gas code reactors and 23 molten salt reactors.
24 DOE and NRC have supported development of 25 numerous neutronics analysis capabilities throughout 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 the past several decades, mainly to support deployment 2 licensing and operation of water-cooled thermoreactor 3 concepts.
4 The purpose of this presentation and 5 presentations that will follow, is not to provide a 6 comprehensive look at the entire spectrum of 7 neutronics analysis capabilities, but rather focus on 8 the recently developed capabilities that aim for 9 design and analysis of advanced reactor concepts.
10 Legacy capabilities, as well as more 11 modern Monte Carlo codes, can also support advanced 12 reactor designs through a varying degree of accuracy.
13 Therefore, the recent efforts under DOEs Advanced 14 Modeling Simulation Program focused on providing high-15 order, deterministic neutron transport solutions to 16 compensate for the limitation of these codes.
17 CHAIRMAN CORRADINI: This is an area that 18 I'm not very familiar with. When you said legacy, what 19 are you thinking of?
20 MR. SOFU: I'm thinking about existing 21 codes that could include PARCS for NRC, Scale System 22 from Oakridge, the 3-D for advanced reactors, sodium 23 task reactors, and the spectrum of other capabilities 24 that exist. That also includes Monte Carlo codes, 25 MCMP, SERPENT, SHIFT we heard about this morning.
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 CHAIRMAN CORRADINI: So those are what you 2 would term as legacy. That's what I was trying to get 3 at. You made the comment ---
4 MR. SOFU: I think I kind of mix those, 5 legacy as well as state of the art. I think in fact 6 Monte Carlo codes often provide reference solutions 7 for us to compare our results with.
8 CHAIRMAN CORRADINI: Okay.
9 MR. SOFU: It did not need to be ---
10 CHAIRMAN CORRADINI: No, I was just trying 11 to understand what you were referring to. Okay.
12 MR. SOFU: So the unique capabilities we 13 feel the PROTEUS Suite and RATTLESNAKE can offer 14 include high-fidelity solutions for complex 15 geometries, with strong heterogeneities bind with flux 16 introductions for vigorous treatment of multi-physics 17 phenomena and transient analyses with deforming mesh 18 capability, which is an important aspect of fast 19 reactor design.
20 In terms of impact, improved operational 21 safety margins through high order and high fidelity 22 modeling close-to-first principle solutions or 23 benchmarking with, benchmarking of lower fidelity, 24 lower order modeling approaches. Again, multi-level 25 interface with matching levels of fidelity, and 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
190 1 that I mean providing ten by ten power distributions 2 to match with CFD-type ten by ten analysis, subchannel 3 analysis of entire reactor core and finally, enhance 4 the impact of limited experiments to support reactor 5 design and licensing.
6 PROTEUS is the first of the two neutronics 7 codes I will cover today. It is a suite of cross-8 section generation capability, three different neutron 9 transports solvers, and a general perturbation theory 10 and sensitivity analysis capability all in one 11 package. It also comes with unstructured, finite 12 element meshing tools, or complex geometries, and 13 cross-processing and visualization capabilities.
14 MC-squared-3, the cross-section, multi-15 group cross-section generation tool can be used for 16 both fast and thermal spectrum reactors, including 17 local heterogeneity effects in a way that 18 homogenization of the units cell prior to the cross-19 section generation is not needed.
20 PROTEUS consists of two highly-scalable 21 neutron transport solvers for complex geometries.
22 Again, without any homogenization, these are solvers 23 based on the method of discrete ordinates and method 24 of characteristics. They're highly scalable in a way 25 that they could scale to very large number 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
191 1 processors for efficient parallel computing.
2 PROTEUS also has a nodal neutron transport 3 solver for simpler geometries for time-efficient 4 solutions that does not require large computing 5 platforms.
6 Finally, PERSENT is a capability used for 7 perturbation and sensitivities analyses based on 8 variational model transport methods. In fast reactor 9 applications, it is also used to determine reactivity 10 feedback coefficients.
11 The second code I will cover today to a 12 lesser extent is RATTLESNAKE. RATTLESNAKE development 13 was aimed at supporting DOE's TREAT reactor restart 14 and feeds experiment modeling efforts, but its multi-15 scheme transport solution options, based on discrete 16 ordinance and spherical harmonics methods lend 17 themselves to a broader group of advanced reactor 18 design, analysis and licensing.
19 As a MOOSE-based code, RATTLESNAKE can 20 also enable multi-physics simulations by a coupling 21 with other MOOSE-based thermal hydraulic codes such as 22 PRONGHORN for pebble bed thermal hydraulics, SAM, 23 which is systems analysis code as well as BISON.
24 You'll hear about these capabilities in the next two 25 presentations.
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 CHAIRMAN CORRADINI: You have mentioned 2 now two or three times, is there something unique 3 about some of these reactors, I think I know it for 4 sodium, but is there something unique about the other 5 reactor designs that you need the coupling 6 characteristic? Because it would strike me this would 7 make the use of it much more difficult.
8 MR. SOFU: One example would be pebble bed 9 reactor that the multi-physics coupling could enable 10 accessing the neutronics effects consistent with the 11 temperatures of the solutions reactor core.
12 CHAIRMAN CORRADINI: And you'd need the 13 coupling versus just simply, well, okay. I see your 14 point. I'll stop there.
15 MEMBER REMPE: Could you elaborate exactly 16 how you use these codes to support the TREAT restart?
17 Was it something where you did some calculations and 18 you provided them to someone from DOE, or how did 19 these codes help you get the reactor restarted?
20 MR. SOFU: Yes, I think the TREAT restart 21 effort went through a licensing process not with NRC 22 but ---
23 MEMBER REMPE: The authorization was DOE.
24 MR. SOFU: Correct. And as part of that, I 25 think RATTLESNAKE was used as one of the codes 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
193 1 supported the assessment of the reactor core, but also 2 the RATTLESNAKE's ability to analyze a particular 3 experiment where in the test channel, the test moved, 4 there's a different type of fuel, could be accident-5 tolerant fuel and coupling the overall TREAT reactor 6 core response with the experimental fuel pins tested 7 in the channel, so-called power coupling factors and 8 things like that is evaluated to a great extent in 9 very great detail with RATTLESNAKE.
10 In the past, such a power coupling was 11 done mostly based on intuition. There wasn't much of 12 a technical basis. You would actually look at the 13 temperature sensors, what you read input in the test 14 loop and then try to correlate what was measured 15 overall reactor response during a transient.
16 RATTLESNAKE did provide a mechanistic basis for that.
17 CHAIRMAN CORRADINI: But for the re-18 licensing of TREAT, for the allowance for TREAT to go 19 forward and restart.
20 MR. SOFU: Also for experiment modeling.
21 You see, I think the mission of RATTLESNAKE to support 22 TREAT operations and experiment design is not over.
23 CHAIRMAN CORRADINI: Oh, okay.
24 MR. SOFU: Its use is going to continue.
25 MEMBER REMPE: So because I have not ever 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 been personally involved with an authorization to 2 restart a reactor, did you just give them the 3 RATTLESNAKE calculations, or did you say this is what 4 we would have gotten with another code and this is 5 what we got with RATTLESNAKE, and we believe 6 RATTLESNAKE's better because of X, Y, and Z, DOE 7 looked at both, I assume it was coming from DOE, 8 looked at both of them?
9 MR. SOFU: Yes. I wonder if Jon Carmack is 10 here, would be able to answer that specific question.
11 MR. CARMACK: I'm Jon Carmack, from the 12 Idaho National Laboratory, sort of. The way I 13 understand it, Joy, we have to have a code that we 14 design the experiment packages in TREAT, but the code 15 also has to be able to model the driver core. And so 16 RATTLESNAKE is built on the MOOSE-BISON, or the MOOSE 17 framework to model the core but also provide the 18 coupling for experiment design. It's going to be used 19 in the future.
20 I don't believe there's any historical 21 code that was used to benchmark against, so they've 22 been doing a bunch of calibration experiments in the 23 reactor to correlate with that. Dan, you got input?
24 MR. FUNK: I'm Dan Funk, the former manager 25 for modeling simulation programs in NE. Joy, I 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
195 1 want to make sure we clarify here, I don't believe 2 that RATTLESNAKE was required in order to get, to 3 obtain permission and authorization to start up TREAT.
4 I want to make that clear. But it was useful in 5 confirming what some of the calculations and 6 conclusions and authorization basis was, but it was 7 really pointed more towards the development work at 8 that point.
9 The desire was to have tools that would 10 enhance the use of TREAT as we projected would be 11 needed, and so we're looking at it very critically, 12 it's proceeding not as fast as maybe some would hope, 13 but as much as funding will allow. So this is more 14 forward-looking in some ways. The capabilities are 15 there but I just wanted to clarify that so that we 16 didn't give the wrong impression.
17 MEMBER REMPE: Thank you. This helps.
18 CHAIRMAN CORRADINI: So let me make sure I 19 understand, because I'm --- So PROTEUS, I have to go 20 back a slide, PROTEUS does the reactor physics and 21 MCC-3 does the initial ---
22 MR. SOFU: Cross-section generation.
23 CHAIRMAN CORRADINI: Cross-section 24 generation? And RATTLESNAKE is, at least to a first 25 approximation, performs the equivalent function 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
196 1 PROTEUS?
2 MR. SOFU: Correct.
3 CHAIRMAN CORRADINI: So, I'm sorry to sound 4 economical, but why have two?
5 MR. SOFU: I think there are certain 6 reasons why two, the NEAMS program supported 7 development of these two capabilities.
8 CHAIRMAN CORRADING: But you're not ---
9 MR. SOFU: Historically, PROTEUS code was 10 originally supported. The NEAMS program investment in 11 the RATTLESNAKE development was mostly focused within 12 the context of TREAT restart.
13 CHAIRMAN CORRADINI: Oh. Okay. I get it.
14 MR. SOFU: I think the RATTLESNAKE has 15 proven clearly that the good capabilities to support, 16 especially for thermal reactor analyses, HTGRs and ---
17 CHAIRMAN CORRADINI: I'm jumping, but I'll 18 jump and ask the question. So if your industry council 19 comes to you and says, what would you recommend we 20 consider for our sodium fast reactor or our gas-cooled 21 thermal reactor or my liquid-fueled MSR, you'd say 22 either?
23 MR. SOFU: I would say for fast reactors I 24 would advocate use of PROTEUS, for thermal reactors 25 RATTLESNAKE. I think that's also consistent with Steve 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 Bajorek's vision of utilization of RATTLESNAKE to 2 support pebble bed type concept analyses within NRC.
3 MEMBER KIRCHNER: And could you explain why 4 one is better for one application and the other's 5 preferable for the other? You're basically using 6 transport methods, you're basically using, I would 7 guess, in-depth trials for cross-sections.
8 MR. SOFU: That's an excellent question. In 9 fact, I think you could probably do both with both 10 codes. But the level of validation basis from PROTEUS 11 is really strong, and I think those are my next few 12 slides.
13 MEMBER KIRCHNER: Okay, fine.
14 CHAIRMAN CORRADINI: Thank you.
15 MR. SOFU: So in the next pages I'll talk 16 about the verification and validation basis for these 17 two codes. It's mostly for PROTEUS, because of its 18 longer development history, but RATTLESNAKE also has 19 a very strong V&V basis, and I have a slide on that.
20 So MC-square-3 and PROTEUS solvers have 21 undergone rather extensive V&V process, and in the 22 next few slides I included only the validation cases 23 and only for advanced fast spectrum reactors. But 24 these two codes have also been applied to numerous 25 other thermal reactor V&V cases, including the TREATS 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 experiments, RPI benchmark, OECD NEAs benchmarks and 2 then ATR benchmark recently. In this slide, I'm 3 showing the critical experiments conducted at Los 4 Alamos zero-power reactor, ZPPR and EPR-II 5 experiments. Those are all US-based legacy 6 experiments.
7 The BFS experiment that you see, as well 8 as CEFR startup tests are more recent, and these are 9 leveraged through international collaborations.
10 BFS in particular is a Russian critical 11 facility, pretty much the only one that is in 12 existence today that can support fast-reactor design.
13 It's located in IPP in Obninsk and used by KRA to 14 support their PGSFR design, CEA to support their 15 Astrid design and TerraPower for their traveling wave 16 reactor design.
17 CEFR is China experimental fast reactor 18 and analysis of its physics startup test is an ongoing 19 IEA-coordinated research project. DOE and NRC are also 20 participants in this effort.
21 The figure you see on the right shows 22 ZPPR-15 tests of four measurements in comparison of 23 results for your uranium-235 fission rates, and the 24 bottom figure is for DFS 761 for uranium-238 fission 25 reaction rates.
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 This is, this page provides a bit more in-2 depth look at the comparisons for one of the tests 3 listed on previous page, zero-power reactor 6 and 7.
4 The chart on the right shows reaction rates along with 5 the radius of the core with two enrichment zones, and 6 the peak reaction rates are at the core center. This 7 is only for one of the load configurations and as 8 shown in the chart, all the results are within the 9 measurement uncertainties.
10 The table at the bottom shows eigenvalue 11 predictions for four different load configurations, 12 all within 80 pcm accuracy with experiment. This is 13 considered fairly good.
14 MD-squared-3 PROTEUS and PERSENT have also 15 been indirectly validated against integral tests 16 performed at EBR-II nf ftf. The results of two such 17 tests are shown in this and next page.
18 In this page, the comparisons of multi-19 physics analysis with data from EBR-2 inherent safety 20 demonstration tests, chart 45 is shown. This test has 21 been studied as a benchmark exercise in a recent IAEA-22 coordinated research project with participants from 12 23 countries, and the results show predictions by each 24 participant with respect to test data.
25 As an unprotected, that means un-SCRAMed 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 test, the predictions for time-dependent power relies 2 heavily on accurate assessment of the complex 3 reactivity feedback mechanisms following detailed 4 depletion analysis over several core cycles.
5 Good agreement with test data, what's 6 shown here is power, temperatures and flow rates, all 7 indicate validity, however indirect, of neutronic 8 assessments against data from a very rare integral 9 test where numerous interconnected phenomena are at 10 play all at once.
11 CHAIRMAN CORRADINI: This was just the loss 12 of, the unprotected loss of ---
13 MR. SOFU: It's a station blackout of ---
14 CHAIRMAN CORRADINI: Unprotected station 15 blackout, okay. But that's a loss of flow test for the 16 SFR, correct? That's essentially.
17 MR. SOFU: Normally loss of flow implies 18 you lose the primary pumps only. In that previous 19 test, the intermediate pumps are also, so it is loss 20 of flow plus loss of heat sink combined.
21 CHAIRMAN CORRADINI: I see. Okay.
22 MR. SOFU: The other tests that were used 23 as an indirect validation basis for these codes have 24 been, the FFTF passive safety demonstrations has so-25 called loss of flow without SCRAM test number 13. 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
201 1 test is a new benchmark exercise, is another IAA-2 coordinated research project starting this year with 3 Argonne, PNNL, TerraPower, MIT and I believe NRC as 4 U.S. participants.
5 Also as an unprotected or un-SCRAMed test, 6 the predictions for time-dependent power relies 7 heavily on accurate assessment of complex reactivity 8 feedback mechanisms. In the FFTF case, it also 9 includes the response of GEMS, gas expansion modules, 10 as novel passive reactivity shutdown device used only 11 in FFTF.
12 As this benchmark project is just 13 starting, we are showing only the preliminary results 14 in comparison to test data but already a reasonable 15 agreement has been observed for power, temperatures 16 and flow rates.
17 This space provides a summary of 18 RATTLESNAKE's verification and validation basis with 19 several TREAT validation and thermal reactor benchmark 20 verification cases that have been either completed or 21 ongoing. Strong TREAT emphasis reflects the 22 developmental focus of RATTLESNAKE and highlights its 23 strength as a capability to analyze the accidents with 24 rapid reactivity changes.
25 I'll quickly go through ---
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
202 1 MEMBER MARCH-LEUBA: Are all of these 2 benchmarks sodium, or is there any for gas-cooled 3 reactors?
4 MR. SOFU: I think there's one gas-cooled 5 reactor benchmark, HTR-10 benchmark, I had a slide but 6 Steve asked me to put it in the background.
7 MEMBER MARCH-LEUBA: You show some bias?
8 MR. SOFU: It's actually, we need to be 9 selective. I have to say that pretty much every slide 10 that you will see in my presentation, also with 11 Richard and Elia's presentations that will follow, 12 each page could be a presentation of its own. You can 13 only present, squeeze so much in short time.
14 MEMBER MARCH-LEUBA: While I have you 15 interrupted, the PROTEUS calculation shows the coolant 16 temperatures. So is PROTEUS a capital thermal-17 hydraulic and neutronic?
18 MR. SOFU: Correct. Essentially, and this 19 is really not a fully-coupled case in a way that we 20 use the electronics tools to evaluate the, start with 21 a fresh core and deplete it to the point where the 22 tests were performant to several cycles from fresh 23 core, and then do a sensitivity analysis to evaluate 24 reactivity feedback coefficients, and those reactivity 25 feedback coefficients go into codes like SAM, you'll 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 hear about, to determine the transient response of the 2 whole primary and intermediate system, and that gives 3 you the temperatures.
4 But temperatures are intimately connected 5 to power in a way that if your reactivity feedback 6 coefficients are not calculated right, then you have 7 no chance of getting temperatures right or power 8 right. So that type interconnected phenomena from 9 integral tests is kind of a unique advantage here.
10 There are no such tests outside EBR-II and FFTF.
11 There is very limited data coming from 12 Monju and Phoenix but those are all SCRAM tests. I 13 don't think this current climate, there probably won't 14 be any other opportunity to unprotected tests, 15 especially with advanced reactors. So that's all we 16 have, and I think we have a good agreement, good 17 handle on that.
18 MEMBER KIRCHNER: Just to follow up on 19 Jose's question, is PROTEUS calculating the flow, the 20 loop conditions, or is that a different code?
21 MR. SOFU: No, that's actually, that's SAM 22 that you'll hear about later.
23 MEMBER KIRCHNER: Okay. So this is a 24 combination of SAM and PROTEUS.
25 MR. SOFU: Correct. It's one-day coupling, 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 in other words you evaluate reactivity feedback 2 coefficients and core power distributions, and SAM 3 uses that to analyze the transient.
4 MEMBER MARCH-LEUBA: By reactivity you mean 5 the power reactivity feedback?
6 MR. SOFU: Correct. The reactivity 7 feedback, I think it's worth showing this slide here, 8 that net reactivity, what's shown in the top, consists 9 of several components that could be the temperature 10 coefficients for the fuel, fuel axial expansion, 11 metallic fuel as counting unique properties that way.
12 In fast reactors core can radially expand. The test 13 spectrum is very sensitive to minor geometric changes.
14 MEMBER MARCH-LEUBA: That's why you need a 15 thermal-hydraulic calculation, activity times 10.
16 MEMBER MARCH-LEUBA: I think what you're 17 referring to is spatial kinetics, but ---
18 MR. SOFU: for EDR-II that was not 19 necessary at all. It's a very small reactor. It's a 20 mega-watt thermal. Spatial kinetics is not ---
21 MEMBER MARCH-LEUBA: I'll wait until that 22 slide and then I'm asking.
23 MR. SOFU: So I'll quickly go through 24 example application of these two codes to sodium fast 25 reactors and molten salt reactors in HTGRs in next NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
205 1 seven slides. Some of these examples also highlight 2 the multi-physics use of these neutronics analysis 3 capabilities.
4 Here, one example of PROTEUS code usage in 5 combination with the NEK5000, Elia will cover that 6 that later, was assessment of hot channel factors for 7 SFRs. The traditional hot channel factor used in SFR 8 core design are largely based on several decades old 9 experimental efforts in support of EBR-II, FFTF and 10 CRBR projects. Their validity for different fuel 11 assembly designs of next-generation SFRs is 12 questionable.
13 To address this particular need, coupled 14 PROTEUS and NEK5000 calculations are used to reassess 15 a select group of hot channel factors for AFT-100, 16 which is a DOE design track, a fuel assembly designed 17 with 91 fuel pins shown on the right.
18 In the coupling scheme, the pin power 19 distributions shown on the left on a normalized scale, 20 obtained with MC-squared and PROTEUS for individual 21 fuel pins in each hexagonal ring are passed on to 22 NEK5000 for thermal assessments.
23 A preliminary comparison of the legacy and 24 newly calculated hot channel factors is provided in 25 this table. Legacy values based on EBR-2 are in red 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 font and newly-calculated indicated as SHARP in the 2 table for the name of the multi-physics interface used 3 for it are in green font.
4 CHAIRMAN CORRADINI: I don't think I 5 understand. What is SHARP?
6 MR. SOFU: SHARP is essentially coupling 7 interface between PROTEUS and NEK.
8 CHAIRMAN CORRADINI: So it's the combined 9 calculation.
10 MR. SOFU: Correct. The top row shows the 11 parameters for which the uncertainties are considered, 12 and the first column shows the sub-list of hot channel 13 factors reevaluated for AFR-100. As seen in the table, 14 these preliminary results suggest significant 15 reduction in select hot channel factors, with safety 16 significance especially since the metallic fuel 17 performance for fast reactors is often limited by peak 18 cladding temperature.
19 You can see that in some cases, for 20 example cladding interval temperature which is 21 significant parameter of interest for us, the cladding 22 thickness uncertainty is reduced from 1.03 to 1.018.
23 The cladding thermal conductivity influenced is almost 24 like a, cut down by a percentage that is small, from 25 1.09 to 1.04, and for fuel center line there'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
207 1 much of a difference but it's not particularly 2 important parameter for us.
3 It's important to note that even when a 4 newly-evaluated hot channel factor is on the same 5 order as the initially assumed value, it still 6 provides a better justification for its use based on 7 validated high-fidelity high-order multiphysics 8 computation scheme. This technique is currently being 9 leveraged to assess hot channel factors for DOE's new 10 versatile test reactor project as a new NEUP IRP.
11 CHAIRMAN CORRADINI: So what, maybe you 12 said and I missed it, what is the legacy code, legacy 13 tool you're ---
14 MR. SOFU: It's not a tool, it's an 15 experiment. It's the value used for ATR-2.
16 CHAIRMAN CORRADINI: Oh. But it was 17 evaluated using experimental data?
18 MR. SOFU: I think mostly hydraulic tests 19 and such and I'm not quite sure what else.
20 CHAIRMAN CORRADINI: So this is not, okay, 21 I misunderstood. I thought this was a calculational 22 comparison.
23 MR. SOFU: It is not. So international 24 projects, we see a lot of that too. For example, when 25 India designed their own reactors they often quote 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
208 1 values used in France or in the U.S., whether or not 2 it is really applicable for their specific fuel 3 assembly design. There's a lot of uncertainty going on 4 in hot channel factors which influence core design 5 significantly.
6 CHAIRMAN CORRADINI: Okay. But, trying to 7 figure a way to ask this question. So if I turn to 8 TerraPower, what would they use? If they wouldn't use 9 the tool that you're offering.
10 MR. SOFU: I think, I really don't know 11 what they use. I can't say. But I think generally 12 accepted hot channel factors based on EBR-II 13 operations are quoted here in red, and most of the 14 time if you don't have any other way of evaluating 15 this coefficient, you'd probably rely on that if your 16 reactor has some similarities to EBR-II, same type of 17 fuel assembly, same type of operating regime, coolant 18 outlets, inlet/outlet temperatures and so on.
19 That's usually pretty much what everybody 20 does, but TerraPower may have their own way of 21 reevaluating this using their own techniques.
22 CHAIRMAN CORRADINI: So, let me push the 23 point a little bit more. If I went and I looked at the 24 PRISM submission and Super-PRISM, or S-PRISM to the 25 NRC, what did they use? I'm trying to get a direct 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 comparison as to what tool is being replaced by this 2 calculational ---
3 MR. SOFU: I don't think there's any tool 4 that this calculation replaces. Values are mostly 5 legacy values everybody relies on. The PRISM, because 6 of its strong similarity to EBR-II, my guess is that 7 they would be using the red values that you see here.
8 CHAIRMAN CORRADINI: I'm sorry, Walt, I 9 interrupted you. You were going to say something.
10 MEMBER KIRCHNER: Was PRISM met metallic?
11 Is that the current concept?
12 MR. SOFU: Yes.
13 MEMBER KIRCHNER: Wasn't it, it was -- at 14 one time, wasn't it?
15 CHAIRMAN CORRADINI: No, that was CRBR.
16 MR. SOFU: CRBR.
17 MEMBER KIRCHNER: Also FFTF.
18 MR. SOFU: Yeah, I think there would be 19 greater similarity to FFTF, which one, PRISM? PRISM 20 would be very similar to EBR-II.
21 Another example of MC-squared-3 PROTEUS 22 tool kit used for SFR design is demonstration of its 23 multi-scale modeling capability. The idea here is a 24 select fuel assemblies from focal assembly or in 25 addition to that, surrounding six fuel assemblies can 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 be analyzed with pin-by-pin level of detail while the 2 rest of the core can be modeled with homogenized fuel 3 assemblies, which is the figure shown on the left.
4 We demonstrated this capability again for 5 the AFR-100 design using PROTEUS discrete ordinance 6 solver, achieving consistent solutions for the focal 7 fuel assemble for the second and third cases shown at 8 the bottom row.
9 Applicability of this approach, the 10 configurations with strong flux ingredients is yet to 11 be verified, but our ability to model only the focal 12 assembly with pin-by-pin heterogeneity while the rest 13 of the core is modeled homogeneously offers a 14 significant reduction in computing resources and time.
15 CHAIRMAN CORRADINI: I'm sorry, you don't 16 have to go back. I'm still back with, so I've got, I'm 17 still struggling with EBR-II. So EBR-II essentially is 18 actual data, and you're saying that most, because of 19 the metal fuel concept, most would probably default to 20 that if they didn't have a calculation. So the 21 calculation is showing that the heat challenge factors 22 are lower than what you'd actually assume or measure 23 for EBR-II.
24 MR. SOFU: I would say properly measured, 25 but at the beginning of this study, we studied 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 existing hot channel factors and tried to find 2 references for them, and references did not exist. So 3 essentially those values are taken for granted, and 4 we've been through a similar exercise for comparing 5 hot channel factors for the joint design of PGSFR 6 reactor with Korea. It was a big debating point. They 7 would use one value and our consults would advocate 8 using another value, and there was no strong basis for 9 what value would apply the most.
10 In situations like that, you usually go 11 with the higher hot channel factor because it's more 12 conservative.
13 CHAIRMAN CORRADINI: Sure. Okay.
14 MR. SOFU: Another example is application 15 of the codes to molten salt reactors for eigenvalue 16 evaluation of commercial thermospectrum concept. The 17 intent was here to use a commercial NSR design as a 18 test that for our modeling and simulation 19 capabilities. This case, both 2-D method of 20 characteristics and 3-D method transport solution 21 options were compared with Monte Carlo solutions. What 22 I'm not presenting here, but we also did compare with 23 the designer's own results, with good agreement.
24 Generally we do the comparisons with Monte 25 Carlo as a reference solution, but in this case 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 PROTEUS offers sometimes a greater, more detailed flux 2 solution of small mesh size for coupling with other 3 analytical tools as needed, such as the CFT solver. It 4 also gives you a better flux in regions with low flux.
5 These calculations were conducted for 6 stationary view. Flowing fuel treatment of PROTEUS-7 NODAL solver was also demonstrated for the graphite 8 moderated thermal NSR benchmark problem. Three cases 9 were considered to evaluate the effect of delayed 10 neutron precursor drift on neutronics. Reference case 11 with stationary field, Case A, with slow-moving fuel 12 with ten seconds transit in the core and five second 13 outside the core and Case B with fast-moving fuel with 14 one second in the core and .3 seconds outside the 15 core.
16 Calculated eigenvalue suggests about 100 17 PCM impact on eigenvalue in each case with respect to 18 previous case, due to decay of the first three four 19 delayed neutron precursor groups outside the core.
20 The impact of the delayed neutron 21 precursor drift on reactor kinetics will be greater 22 during transience, and this analysis will be performed 23 for the transient MSRE benchmark in FY19.
24 For high HTGRs, that's proof of principle.
25 PROTEUS was used for both prismatic fuel assembly 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 design shown on the left and whole-core image CRG 2 simulation shown on the right. The accurate modeling 3 of the large neutron streaming in the control hole 4 channel, control channel, the large hole at the center 5 of the left figure, when the control rod is withdrawn 6 it is a significant challenge requiring high-fidelity, 7 high-order transport solutions.
8 The whole core simulations on the right 9 were for MHTGR core assessment of computational 10 requirements and scalability of the solver. They 11 looked at all rods out, operating rods in and all rods 12 in cases. There was no data to compare with, so those 13 were more like proof of principle calculations to 14 demonstrate the ability to model complex geometries in 15 different reactors.
16 Probably the more important pebble beds 17 HTGR modeling capability comes from RATTLESNAKE, its 18 utilization in combination with pebble bed tracking 19 algorithm. Motivation for the high-resolution multi-20 physics simulation for pebble bed motion is to support 21 direct transport calculation with pebble tracking. The 22 scoot element method is used to provide time-dependent 23 position of all pebbles for establishing an 24 equilibrium course.
25 The results shown at the bottom are for 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
214 1 particle tracking transport with nine pebble model 2 with respect to reference case with SERPENT, 3 indicating good agreement, and the ultimate goal of 4 this effort is to have the ability to track the burnup 5 of individual pebbles as they move down the core 6 during the cycle.
7 The figures shown on the right are for 8 HTR-10 simulations, using RATTLESNAKE. The total, 9 almost half a million tetrahedra, almost 80,000 node 10 points ---
11 CHAIRMAN CORRADINI: So, I'm listening and 12 the question keeps on coming back to me. So, I'm back 13 to industry again. What would X Energy use if not 14 this?
15 MR. SOFU: I had that discussion with 16 Martin at a meeting at ORNL two weeks ago. They're 17 interested in this capability but I believe they have 18 their own methodology that they're pursuing. I'm not 19 quite sure what it is, or based on what code. But the 20 whole idea is understanding the depletion cycle as the 21 pebble bed core starts moving down in a very gradual 22 and slow manner is a challenge, is a significant 23 challenge for them as well.
24 MEMBER REMPE: So I vaguely remember, was 25 it South Africans, or someone did an experiment now 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 with the real pebble bed reactor but just with some 2 pebbles to try and validate that they did, were 3 capable of tracking how the pebbles would go in, and 4 it was for a core that didn't have control rods in 5 like a THTR. Is that what you're, or is this just a 6 proof of principle calculation?
7 MR. SOFU: It's a proof of principle 8 calculation at this point. Another important point 9 where we don't have a completion embedded into this 10 capability quite yet in a way, we're able to 11 understand the item value calculation, power 12 distribution in pebble bed with particles moving down, 13 but we're not able to deplete each pebble as they move 14 down. That would be the next logical step.
15 MEMBER REMPE: What did the Germans use for 16 the ABR?
17 MEMBER BALLINGER: The code that the South 18 Africans were using was called VSOP.
19 MR. SOFU: That's correct.
20 MEMBER REMPE: But what did the Germans use 21 for the ABR? Did they have a way of tracking this 22 burnup on a pebble basis?
23 MR. SOFU: I'm looking for who can help us 24 with that, but to my knowledge I'm not really sure.
25 MEMBER REMPE: I just was curious.
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 MEMBER BALLINGER: I think in Japan they 2 tried to take VSOP and turn it into a colossal Monte 3 Carlo code. I'm not sure they were successful at doing 4 that.
5 MR. SOFU: Yes, if there's a way of finding 6 out answers to those questions and getting back to 7 ACRS I'll be happy to do that.
8 MEMBER BALLINGER: There were a couple of 9 benchmarks on criticality that they had, and I forget 10 now, I'm losing track of time. There were some 11 benchmarks on pebble bed criticality height as he 12 started adding pebbles to the, pebbles to the core.
13 MEMBER REMPE: So, out of curiosity, if you 14 do have a chance to find out the answer you can send 15 it to Weidong. That's the federal designated official.
16 MR. SOFU: Absolutely. I have one more 17 slide here ---
18 CHAIRMAN CORRADINI: So I'm still back 19 with, I can't get off of EBT-II. I'm trying to 20 understand the penalty that I incur by taking these 21 conservative hot channel factors versus a more refined 22 approach, and then I have, the next question I have in 23 my mind is, if those were assumed, and I think I asked 24 you and you said you weren't sure, whether they were 25 measured or they were assumed and those were 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 considered assumed and conservative enough to proceed 2 with operation.
3 Can I get away with that with uncertainty 4 under the current situation or must I have this sort 5 of precision? Because if I demand the sort of 6 precision, I look for an experiment to validate it.
7 And I'm ---
8 MR. SOFU: I think that would be a question 9 for NRC, actually. But if I were to put regulator's 10 hat, I would say I would feel more comfortable for a 11 hot channel factor that has a sound basis as opposed 12 to a value that is being used because everybody else 13 using the same thing, and then you don't even know 14 where it comes from.
15 And if that value is greater than assumed 16 value, I don't care. That's still my preference. In 17 other words, have others reestablish the set, high 18 channel factor set and use that on a consistent basis 19 even though it turns out to be more conservative than 20 some number that everybody relies on.
21 MEMBER REMPE: So are these codes being 22 used to support the design development for the BTR?
23 MR. SOFU: Well, I think I did mention that 24 all of the NEUP projects this year, Chris' technical 25 point of contact is to ask university support to use 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 these tools to analyze VTR design to reevaluate hot 2 channel practice for BTR.
3 CHAIRMAN CORRADINI: The core itself, or 4 the experiment?
5 MR. SOFU: The core. Oh, it could be 6 experiment as well. I am sorry, I didn't mean to rule 7 that out.
8 CHAIRMAN CORRADINI: What was done, maybe 9 that was, I'm still not there. What was done, or no, 10 maybe I should ask it differently. On slide 9, when 11 you were showing the unprotected station blackout or 12 essentially loss of flow and loss of heat sink, there 13 was a whole range of calculations. What were the other 14 contributors using as their tool?
15 MR. SOFU: Oh. So ---
16 CHAIRMAN CORRADINI: I'm still struggling 17 with trying to understand how this tool relates to 18 other available tools and what the potential 19 Applicant, whether it be for sodium in this case, 20 sodium or gas, would choose to use as their depth of 21 analysis?
22 MR. SOFU: I can see that this legend is 23 kind of slow, but TerraPower was using SAS and SAS4A.
24 CHAIRMAN CORRADINI: Was using what? I'm 25 sorry.
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 MR. SOFU: SAS and SAS4A. NRG, I don't know 2 what they were using. KIT was using, I think that's 3 important because KIT was using PARCS and TRACE.
4 CHAIRMAN CORRADINI: Oh, really.
5 MR. SOFU: Yes. I'm sure, KIT, no, PSI, I'm 6 sorry. What is KIT? KIT is Germany. KIT was using 7 SIMMER, PSI was using PARCS and TRACE, KAERI was using 8 MARS-LMR, IRSN was using a code they called, I'm 9 sorry, my mind is losing. CIA was using SANS, SANS4A 10 Fukui was using their own JAEA Code.
11 CHAIRMAN CORRADINI: I guess where I was 12 going with, the only thing I could remember was SAS.
13 So, another way of saying it is PROTEUS in combination 14 with SAM is the replacement for SAS.
15 MR. SOFU: For a lot of these participants, 16 did get the reactivity to feed the coefficients they 17 used for those codes from Argonne.
18 CHAIRMAN CORRADINI: Oh.
19 MR. SOFU: They did not do the neutronics 20 part of it. Only a handful of them did. I think Fuqui 21 did and PSI did, PSI used PARCS for that, so I think 22 it turned out doing this benchmark was too much to 23 bite for a lot of the participants so they, some of 24 them, large number of them chose to only the system 25 analysis, safety analysis, transient part of it.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
220 1 CHAIRMAN CORRADINI: So from the standpoint 2 of trying to understand, what goes through my mind it, 3 they're using their legacy tools to do the analysis.
4 MR. SOFU: Correct. I think I'll just 5 briefly mention this since I am running out of time, 6 this was a multiphysics analysis of SFR core 7 deformation. It's an interesting assessment because it 8 uses coupled, fully coupled, PROTEUS, NEK5000 and 9 DIABLO for structural mechanics calculations to 10 understand how a fast reactor core would radially 11 expand.
12 It is an important reactivity effect, in 13 fact in the chart below you see that this is the 14 highest contributor to net negative reactivity in this 15 case, but typically that counts for a lot of the 16 inherent safety features that we obtained.
17 The ability to make that prediction based 18 on very detailed subassembly by subassembly 19 representation of the entire core with neutronics, CFD 20 level of thermal hydraulics and structural mechanics, 21 is the first demonstration of having a complete 22 mechanistic way of predicting this for new design.
23 MEMBER KIRCHNER: What would you venture 24 the uncertainty is on that, and how does it scale?
25 MR. SOFU: There wasn't anything for us 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
221 1 compare with on this. It was more like a proof of 2 principle type of demonstration. You know, an ideal 3 application of this would be perhaps apply this to 4 EBR-II and FFDF transients and analyze, reassess the 5 core radial extension feedback and then put that back 6 into the system analysis code to see if you're getting 7 consistently code.
8 CHAIRMAN CORRADINI: Right. So this is an 9 unfair question, but we're in the world of unfair.
10 What is the person-months of effort necessary to do 11 such an analysis?
12 MR. SOFU: Elia was involved with this 13 project.
14 MR. MERZARI: Should I answer that?
15 MR. SOFU: I'm guessing maybe a year type 16 of, maybe a little longer than a year. Go ahead.
17 MR. MERZARI: So this is the Elia Merzari 18 from Argonne National Laboratory. It's a bit of an 19 unfair question in the sense that this was a proof of 20 concept. So the first time that you do something, it 21 always takes longer. It took quite a bit to 22 demonstrate. It took about six months to the 23 calculations you see there, but I would envision that 24 once we streamline the process, this could take no 25 more than a few weeks. It wasn't done on millions of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
222 1 processors, it was done on 120 processors.
2 CHAIRMAN CORRADINI: So, modest amount 3 compared to others.
4 MR. MERZARI: Yeah. And you saw something 5 that you need to do a lot of work to find.
6 MEMBER KIRCHNER: Well, it's crucial to use 7 it as a design tool, because you're going to hit a 8 point where the size of the core isn't going to give 9 you enough, your expansion versus your leakage, etc.,
10 isn't, it's going to limit the size of the module, 11 essentially. And then you need to assess your 12 uncertainty and your margin as you push that limit. I 13 mean, that becomes a fundamental design challenge.
14 MR. MERZARI: Right. And I should mention 15 that that calculation was a demonstration again. It 16 was very little sensitivity, it was some sensitivity 17 but not near as much as we need, and there are several 18 sensitivities that go into a calculation like that, 19 means a stage lost regularly through the core. There 20 are several things that will need to be done for a to 21 think of full design tool, but I agree with you.
22 CHAIRMAN CORRADINI: Would depletion be one 23 of them?
24 MR. MERZARI: Certainly the rigid core 25 expansion depends on fluence.
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 CHAIRMAN CORRADINI: Okay. And that also 2 would have to have been done. Or needs to be done, I 3 should say. Needs to be done.
4 MR. MERZARI: We need some calculations of 5 different states, but you can assume this rate of 6 fluence as an input to the calculation.
7 MR. SOFU: In radiation swelling of fuel 8 assembly, the contribution to this expansion is 9 captured with the DIABLO models. I think the 10 importance of that particular application is there is 11 no such comprehensive multiphysics way of evaluating 12 the core radial expansion for a test reactor prior to 13 this demonstration.
14 We have codes like NOUVEAU, you may 15 remember, Mike, but they are generated during the IFR 16 program very largely on empirical basis. But you get 17 a code like that, assume applicability for a new 18 design, is always going to be an open question.
19 CHAIRMAN CORRADINI: So I know we're 20 running 15 minutes late but it's unfair, is the reason 21 we didn't see RATTLESNAKE is because it's in the 22 background slides but there are ---
23 MR. SOFU: No, I think largely I'm the one 24 to blame for it because I'm more familiar with PROTEUS 25 and I think the question about including RATTLESNAKE 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 came in the late stages of preparing for this meeting, 2 so I was able to leverage three or four slides from 3 Markie Hart, I was grateful to give you what I have.
4 But we recognize its importance for NRC, 5 especially for division, Steve Bajorek's vision to 6 utilize that capability as a component of the MOOSE 7 framework and for HTGR applications, so by no means is 8 that an indication of its importance.
9 CHAIRMAN CORRADINI: Okay.
10 MR. SOFU: Thank you.
11 CHAIRMAN CORRADINI: On to fuels. We've 12 used up about 20 minutes of your time, I'm sorry, 13 Rich.
14 MR. WILLIAMSON: Okay.
15 CHAIRMAN CORRADINI: And we'll try to give 16 it back to you, maybe.
17 MR. WILLIAMSON: I'm not sure I need the 18 full time.
19 CHAIRMAN CORRADINI: Okay.
20 MR. WILLIAMSON: Okay, good afternoon. My 21 name is Rich Williamson, I work and in fact I've 22 always worked at the Idaho National Laboratory. My 23 presentation is Fuel Performance Modeling for Advanced 24 Reactors.
25 Here's the outline of what I brought 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
225 1 present. I have a single slide to begin with. It gives 2 background information including the objective of this 3 particular presentation. I have a section of slides 4 that are the BISON fuel performance code that 5 essentially duplicate what you saw this morning in 6 Steve's presentation. I left them in there 7 intentionally so that the presentation was more 8 complete, but we'll be able to skip over that section 9 almost completely. There is one point I want to make 10 out of those slides.
11 And then really the bulk of what I want to 12 talk about is application of BISON to two fuel forms, 13 TRISO and metallic fuel. So that'll be the bulk of 14 what I have to talk about in summary, conclusions at 15 the end.
16 As you know and as you've heard already 17 this morning, DOE is developing modern fuel 18 performance, modeling tools applicable to a wide 19 variety of fuel and reactor types, operating 20 conditions, geometries, and spatial scales. This 21 capability has been developed for both advanced and 22 traditional LWR concepts with LWR fuel receiving the 23 greatest early emphasis, so that the plan has been to, 24 from the beginning has been to look at both advanced 25 and traditional fuel but to really demonstrate 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
226 1 learn by application first to LWR fuel.
2 As Steve pointed out, multi-scale modeling 3 for us, the coupling device and the norm it provides, 4 improve mechanistic material models and has delivered 5 demonstrated results for UO2 fuel and we intend to 6 follow that same path for advanced fuels.
7 The presentation objective is to provide 8 a current status of DOE fuel codes for application to 9 specific advanced reactor concepts as are identified 10 in the table, which lists the key advanced reactor 11 concepts against their fuel and coolant 12 characteristics.
13 The key takeaway from this table, and I 14 think it's simple, indicates that the DOE fuel 15 performance codes BISON and MARMOT have demonstrated 16 capabilities for metallic fuel, applicable to the SFR 17 concept, and TRISO particle fuel applicable to the 18 MHTGR and FHR concepts, but at this stage are not 19 applicable to the MSR concept which employs a liquid 20 fuel. So the focus then will be on metallic fuel and 21 TRISO fuel as I mentioned previously.
22 This is the part we skip over. You heard 23 about MOOSE, BISON, MARMOT, so you know about M.M.
24 This is almost a duplicate of what you saw before, 25 details about BISON and what makes BISON different.
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 CHAIRMAN CORRADINI: You're doing great.
2 MR. WILLIAMSON: I'm at slide 8 already.
3 So, it slows down a bit in a minute. This slide has 4 already been covered in some depth by Steve as well.
5 The only point I wanted to make is on this 6 slide, concerning BISON documentation. I wanted to 7 draw your attention to the link in the third line 8 there that is a place you can go to get a current set 9 of manuals. So although the code is, you have to have 10 a license to get the source code, the manuals and 11 documentation for the code are available, external 12 documents, so anyone can go to that site and pick up 13 those manuals.
14 The other point I wanted to make is it is 15 an enormous headache to try to keep a source code and 16 documentation and validation all in step at the same 17 time and keep everything complete, so we're working 18 and currently transitioning to a web-based 19 documentation system that will combine the theory and 20 user manuals and post more strict requirements on 21 documentation of new code, but essentially much more 22 the documentation will exist with the source code and 23 build from the source code, so when you get the code 24 you can build the documentation and that forces 25 developers to keep things much more in lockstep.
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 Moving on to TRISO particle fuel. This 2 slide is an effort to in a single slide cover the 3 capabilities that are currently in BISON for TRISO 4 particle fuel. I have a similar slide that I use for 5 LWR fuel and a similar slide that you'll see in a 6 minute on metal fuel, and the first point I wanted to 7 make is this box on the left will always be the same.
8 The general capabilities in BISON, finite 9 element based for a variety of different types of 10 geometries, fully-coupled thermomechanics, species 11 diffusion, these capabilities are the same for all 12 these fuel types and so that's one of the, in my 13 opinion, one of the real advantages to a code like 14 BISON, because much of the fundamental capabilities 15 that you need are there, and it's oftentimes more a 16 matter of enhancing and adding capabilities to include 17 in a fuel type.
18 MEMBER REMPE: Rich, I heard earlier today 19 you took the PARFUME models and put them into BISON.
20 Didn't PARFUME also have something about the silver 21 attack on the silicon carbide for accident 22 temperatures, so does the DOE approach say use BISON, 23 but then if you need to do an accident, still go use 24 PERFUME, or do you still have PERFUME that you guys 25 are using at Idaho, or what's the story?
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 MR. WILLIAMSON: We still have PERFUME. To 2 my knowledge it's no longer actively supported. The 3 intent is to move away from PERFUME but there are 4 capabilities that still exist in PERFUME that are not 5 in BISON yet, and that's part of our current gap 6 analysis. That's what we're working on now, too, to 7 identify the gaps and put our development efforts into 8 closing those gaps.
9 MEMBER REMPE: What would you use if you 10 don't start, are you going to try to have an accident 11 version, like go to MELCORE or something like that for 12 accident analysis, or what's the plan? Or you haven't 13 decided yet?
14 MR. WILLIAMSON: Yeah, I guess I would come 15 down to haven't decided yet, although we don't plan, 16 as with LWR fuel and ATF fuel that you heard this 17 morning, there's no intent of going to severe melted 18 fuel or anything like that.
19 CHAIRMAN CORRADINI: But I think that where 20 Joy was going is if you take this to different DBAs 21 and beyond-DBAs, you're going to need to do that to 22 develop source terms. To develop source terms, you're 23 going to need a release model. To have an release 24 model, you're going to have to put it in a construct, 25 so is the construct PERFUME? That's what I heard 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
230 1 kind of say.
2 MEMBER REMPE: And you said PERFUME's no 3 longer supported so I think you got to go to MELCORE 4 or something, I don't know.
5 MR. STANEK: I think a lot of this will be 6 described in Tanju's last presentation. I'm not going 7 to list source term that will go through for each of 8 the reactor types, how one could generate a mechanism 9 source term.
10 MEMBER BALLINGER: Can you do UCO as well?
11 MR. WILLIAMSON: No. Not today. We do UO2 12 only at this stage. The developments that you see here 13 actually occurred early in BISON's life. This 14 capability was available in like 2013 and there hasn't 15 been a huge amount of effort in development since 16 then.
17 MEMBER BALLINGER: Because I think that's 18 what does in UCO.
19 MR. WILLIAMSON: Again, that's a gap. And 20 you'll see when I come to the end of the triso section 21 that is one of the areas of development for next 22 fiscal year.
23 CHAIRMAN CORRADINI: So, Rich, maybe I 24 misunderstood the cartoon. Is the colored cartoon in 25 the middle the kernel or the compact?
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 MR. WILLIAMSON: In the middle is just the 2 particle.
3 CHAIRMAN CORRADINI: So, excuse me for 4 sounding holy god, you're going to nodalize that and 5 then you got to get 1,000 to 10,000 of those in a 6 compact and billions of them in a core, or are you 7 developing some sort of empirical set of calculations 8 that then would feed into some more macroscopic --- I 9 guess I misunderstood.
10 MR. WILLIAMSON: I think I have a great 11 answer to that coming.
12 CHAIRMAN CORRADINI: Okay.
13 MR. WILLIAMSON: So if I don't get to that 14 then we'll come back to it, but I think I have that 15 answer and I'll make sure to cover that.
16 So we talked about general capabilities.
17 Fuel kernel is UO2 to date, and that's principally 18 because UO2 is already a known quantity for us and so 19 the UO2 model that we use for LWRs is essentially 20 applicable to particle fuel although there are issues.
21 For example, the CO production that is important for 22 TRISO particle fuel had to be implemented or included 23 for the TRISO capability.
24 Likewise, gap behavior, with TRISO 25 particle fuel a gap can form between the inner, 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
232 1 porous pyrolytic carbon layer and the inner pyrolytic 2 carbon layer and so you have to be able to handle a 3 gap. We do gaps from the experience in the LWR world 4 and so gap behavior is essentially there, we just have 5 to worry about changing the fluid and including some 6 capability to move mass across the gap.
7 So in essence, the newness of BISON for 8 TRISO in material models for silicon carbide and 9 pyrolytic carbon. We've included an irradiation creep 10 model for silicon carbide and for pyrolytic carbon 11 anisotropic irradiation-induced strain and irradiation 12 creep.
13 And these models, Joy mentioned, come from 14 PERFUME. Our actual source for these is mentioned at 15 the bottom, but I think you'll find that's true. The 16 same models that are in this source in 2004 were in 17 PERFUME.
18 MEMBER BALLINGER: What are you defining as 19 failure? Are you modeling failure?
20 MR. WILLIAMSON: You're getting to all the 21 points that we don't do, and that's one of them.
22 MEMBER BALLINGER: Is that, the particle 23 fuel is inherently probabilistic.
24 MR. WILLIAMSON: Understood.
25 MEMBER BALLINGER: Inherently 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 probabilistic.
2 MR. WILLIAMSON: Understood. And so, again, 3 when you see my plans for even next year, that's one 4 of them. The point I need to make again is this is 5 development that has really lain dormant for a few 6 years and now with increased interest in advanced 7 reactors and TRISO fuel, has cranked back up and those 8 are areas that we will address and have begun to 9 address.
10 Now I have three examples, and I hope 11 these lead to an answer to Mike's question. I have an 12 example of an application device to a 1-D spherical 13 particle, then a 2-D and a 3-D application. I'll start 14 with the 1-D case.
15 It assumes spherically symmetric. And 16 because it's spherically symmetric, it can be done in 17 1-D. So the mesh you saw was a three-dimensional mesh 18 which is indeed complex. But if you assume it's one 19 dimensional, as is done with PERFUME, then the 20 problems run very quick, typically in under a second.
21 And so with 1-D, if one assumes spherical symmetry, 22 then one can run many of these particles and do 23 statistical analysis, failure analysis in the same 24 sense that it was done in PERFUME.
25 So this is a 1-D example. It involves 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
234 1 geometry you see here, so a UO2 kernel, porous 2 paralytic carbon layer and then the typical three 3 TRISO layers. What we've looked at as far as physics 4 in this problem is thermomechanics plus cesium 5 diffusion. We looked at a single fission product 6 diffusion. There are no issues with adding other 7 species to that, and then for this example, and this 8 is just an example calculation, we show, what we 9 looked at here was really three steps.
10 First an irradiation period that goes for 11 two and a half years to a burnup of 12 percent as a 12 step one. Then actually we moved from the core for 100 13 days in storage to decay, and then to simulate 14 accident behavior, the particles essentially were just 15 heated, furnace-heated to 1800 K for 220 hours0.00255 days <br />0.0611 hours <br />3.637566e-4 weeks <br />8.371e-5 months <br /> to see 16 the effect of an increased temperature.
17 MEMBER REMPE: This is ---
18 CHAIRMAN CORRADINI: This is an experiment?
19 MR. WILLIAMSON: This is a pure example.
20 MEMBER REMPE: This is where I'd think 21 you'd want to have the silver. Don't they start having 22 problems like 1600 C is when you start worrying about 23 the silver?
24 MEMBER BALLINGER: Silver doesn't attack, 25 it migrates through.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
235 1 MEMBER REMPE: Right, okay, so then you 2 have to worry about it ---
3 MEMBER BALLINGER: Cesium does a number on 4 the ---
5 MEMBER REMPE: But doesn't it start 6 migrating through at ---
7 MEMBER BALLINGER: Much lower temperature 8 than that.
9 MEMBER REMPE: Yeah. So it seems like 10 you're kind of missing something.
11 MR. WILLIAMSON: Yes. And I agree. I agree 12 fully. We included cesium because we knew that cesium 13 would move through the silicon carbide layer at those 14 temperatures whereas it doesn't at operating 15 temperatures. But yeah, point's well taken. Again, 16 that species can be added to this calculation.
17 This, again, is just an example 18 calculation to demonstrate capability, and there's 19 lots of results to show. I just showed two. One's the 20 tangential stress history for both the silicon carbide 21 and pyrolytic carbon layers. I would just trace the 22 red curve at the bottom which shows the silicon 23 carbide layer, and it shows the expected behavior of 24 that layer moving in a strong compression early in 25 life because of the compression of the pyrolytic 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 carbon layers but as the pressure builds in the 2 particle that stress eventually goes into tension and 3 then into significantly higher tension during the 4 accident. So again, just a demonstration of 5 capability.
6 The plot on the right shows the cesium 7 concentration, these are radial profiles from the 8 center of the particle out to the edge and two of the 9 most curves show basically the cesium distribution 10 after radiation and after storage, no change there, 11 and then when raising the temperature much higher, 12 then you start seeing the cesium diffuse out. That was 13 the point we were trying to make here.
14 Again, this is in 1-D but run times of 15 under a second permit rapid analysis if one wants to 16 look at loss of particles.
17 The second example is now two dimensional, 18 base spherical particles can occur during 19 manufacturing as is shown in this micrograph. If one 20 is okay looking at a single-faceted, aspherical 21 particle, then this problem can be analyzed in 2-D 22 axisymmetry. So it's a step up from 1-D, still 23 computationally reasonably simple but it does permit 24 one to look at a multidimensional type of effect.
25 So here we looked at the same example, 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 behavior as before, base irradiation followed by an 2 accident, but now are comparing a spherical particle 3 to a particle that has a facet, it has a flat spot on 4 it. The two results that I chose to show are shown 5 here. The red curve compared to the blue curve is the 6 stress history in the outer silicon carbide layer, and 7 you see the aspherical raising, after radiation, much 8 higher stresses, much higher tensile stresses.
9 You see the same in this contour plot.
10 This is at the end of the base irradiation and you 11 see, for this example geometry that we picked, 12 increases in the tensile stress in the order of a 13 factor of four by having this facet there. So the 14 multi-international effects are clearly important and 15 can be addressed with BISON and 2-D reasonably 16 efficiently. This problem ran in a few minutes, as I 17 recall. Six minutes, using eight processors.
18 Moving now to the 3-D. Particles can show 19 localized thinning of the silicon carbide layer due to 20 soot inclusions or because of fission product 21 interactions on that layer, which obviously will 22 degrade the capability of the silicon carbide layer to 23 handle the pressure of the particle.
24 So for this, again an example, we simply 25 added random thinning to the silicon carbide layer as NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
238 1 one might see here and simplify the case even more, 2 just looked at asymmetry of this hole in this whole 3 particle.
4 And the takeaway from this, as you can 5 see, the stresses plotted here a region that is not 6 thin compared to a thin region. You see significantly 7 higher ensile stresses, significantly higher cesium 8 fluxes in those regions. So certainly concerns that 9 come up that can really only be addressed in 3-D.
10 These cannot be done in 1-D.
11 CHAIRMAN CORRADINI: So maybe you're not 12 done. I was going to ask the question, are you 13 finished?
14 MR. WILLIAMSON: Yes.
15 CHAIRMAN CORRADINI: So, you went from 1-D 16 to 2-D to 3-D, and the real manufacturing of the fuel 17 you don't know where you are between spherical, 18 asymmetric and a flaw, so what does one assume from 19 the standpoint of prescribing the uncertainty range to 20 do the original calculation before you even do the 21 extant calc. I'm struggling here.
22 MR. WILLIAMSON: I am too. I really don't 23 have an answer to that.
24 MEMBER KIRCHNER: You have to do it 25 statistically, based on the post-production sampling 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 of the kernels.
2 MR. WILLIAMSON: I think so.
3 MEMBER KIRCHNER: Your earlier picture 4 already shows that most of them are not spherically 5 symmetric. That's the experience from manufacturing.
6 MR. WILLIAMSON: Others probably know 7 better than I do about that, you know, but that's ---
8 CHAIRMAN CORRADINI: Okay. So that would 9 lead me to my next thing. Are, you have the AGR 10 experiments. Was there pre-analysis so you knew the 11 range of geometrical conditions of those compacts so 12 you could then put that into a calculation, then see 13 what you got post-test for the AGR, I can't remember 14 which was which. I think AGR-1 was steady stage, R-2 15 was taken to failure, I don't remember. But is the, 16 are the AGR experiments the place for validation of 17 these sorts of analyses?
18 MR. WILLIAMSON: I think so.
19 CHAIRMAN CORRADINI: But they've yet to be 20 looked at?
21 MR. WILLIAMSON: Correct. I'm about to show 22 you what validation is done today. And it isn't a 23 great deal. The only other point I wanted to make on 24 this slide is now the run times have gone from a few 25 minutes, so we went from under a second to a few 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 minutes, now to a few hours on eight processors, and 2 obviously that could be reduced some with more 3 processors but now these are much more significant 4 computations.
5 MEMBER BALLINGER: There were a whole bunch 6 of benchmark runs, MPR 1, 2, 3 and all that. Are you 7 aware of those?
8 MR. WILLIAMSON: I thought you were going 9 to the benchmark, the IAEA benchmark.
10 MEMBER BALLINGER: Well, there's them too 11 but there was a whole series of benchmark codes, 12 benchmark runs in the early days that PERFUME was 13 compared to as well along with others, and they didn't 14 do too well, primarily because of the probabilistic 15 nature of the failures. And so them had to do some 16 stuff to make it work. But those, I don't see those 17 listed. They're very well characterized.
18 MR. WILLIAMSON: Point well taken. We'll 19 look at those. We haven't yet.
20 MR. STANEK: Rich, maybe this is a good 21 point to point out something that Steve went through 22 very quickly in his fuels presentation today, which 23 was, it's analogous, which was he showed a 3-D image 24 of a missing pellet surface, so that's a manufacturing 25 flaw in UO2 pellet, but the 3-D code like BISON, 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
241 1 can do an analysis on manufacturing tolerances, what 2 sort of tolerances can be, what size flaw leads to 3 what size stress concentration is essentially the 4 analysis that can be done, which would then inform the 5 manufacturing tolerance which would then have an 6 impact on your probabilistic failure assessment.
7 And so, this 1-, 2-, 3-D type of analysis, 8 I think that's sort of where we're getting at too, is 9 that yes, failures are probabilistic. Having a handle 10 on how much of a flow one can accommodate in the fuel 11 fabrication is something that can't be done with this 12 type of pressure.
13 MR. WILLIAMSON: Okay. Moving on to V&V, 14 particle fuel V&V efforts to date consist really of 15 comparisons to an IAEA benchmark which is entitled 16 CRP-6. Early 2000, I think, is when that happened.
17 They, it really was an effort to compare codes on a 18 series of problems rather than comparisons to 19 experimental data.
20 So it's not full-on validation, but it's 21 comparison to codes that are fully validated so it's 22 a good first step.
23 This IAEA CRP-6 developed a set of fuel 24 performance codes benchmarked in cases for both normal 25 operation and operational transience. They ranged 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
242 1 complexity from a simple fuel kernel having a single 2 elastic coating layer to realistic TRISO-coated 3 particles under a variety of radiation conditions.
4 As indicated in the table on the left, 5 BISON has been assessed against analytic solutions and 6 other particle fuel coats for 13 cases from that 7 benchmark. I think the total was 16, and at the time 8 we did this work we did 13 of those 16.
9 Many comparisons can be made from that 10 chart. I chose to make comparisons here to tangential 11 stress, which is important for particle integrity.
12 Comparisons of the tangential stress are shown in the 13 table on the right, but the simplest cases, cases 1, 14 2, and 3, they all have analytic solutions and so 15 BISON comparisons are excellent, as one would expect, 16 to an analytical solution.
17 But the more complex cases shown in the 18 bottom table, comparisons are made to a range of 19 solutions attained by the fuel codes and the 20 benchmark. Here the solutions are compared to the 21 range, as you see from that table, the range of values 22 that came out of the handful of codes that were 23 involved in the benchmark.
24 About all we can say from this comparison 25 is that the BISON solutions were always within 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
243 1 range of values computed by the other codes, but no 2 deeper than that.
3 Continuing with the benchmark comparisons, 4 results are shown here for three cases that have 5 increased complexity. For these cases, rather than 6 compare to all the codes, we chose to compare to three 7 well-known, reasonably well-validated codes. Those are 8 PERFUME from the U.S., STRESS the UK, and the ATLAS 9 code which is the only of those codes that is finding 10 element from France.
11 On the left, you see comparisons for Case 12 8, which was an effort to simulate the cyclic particle 13 temperature experienced during multiple passes through 14 a pebble bed reactor. You can see the passes, the 15 effort to simulate that. Predicting the tangential 16 stress in the inner pyrolytic carbon layer, which are 17 the curves at the top, essentially overlaid for the 18 four codes.
19 Then predictions of the tangential stress 20 in the silicon carbide layer, which are the four 21 curves at the bottom, nearly overlay. I want to note 22 that these are what I consider excellent comparisons 23 are really not unexpected, because this Case 8 of the 24 benchmark carefully controlled everything about the 25 calculation they compared. They controlled 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
244 1 geometry, even the material models, and so one would 2 expect very good comparisons. And I point that out as 3 we turn to Cases 10 and 11, where that wasn't the 4 case.
5 For Cases 10 and 11, on the right, these 6 were based on German fuel experiments. Here 7 comparisons are made to the tangential stress at the 8 inner silicon carbide wall, and you see substantial 9 differences in calculations, particularly for Case 11 10 which are these four curves that go to higher burnup.
11 These variations have been attributed as 12 part of the benchmark. Principally the differences in 13 the fission gas release model, so for this case they 14 didn't control the gas release and CO production 15 models and so now you're starting to see the 16 variations in codes because of the different models, 17 selections of different models.
18 The BISON fell within the, with our 19 inherent gas and CO production, this BISON is the red 20 curve which is in kind of the middle of the pack.
21 MEMBER REMPE: Explain to me they didn't 22 control the fission gas release model. It might be a 23 function of temperature or something like that?
24 MR. WILLIAMSON: They didn't specify. So 25 say for Case 8, they went to the point of specifying, 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 this is the constitutive model, this is the equation 2 you will use for irradiation creep of silicon carbide, 3 for example. And so you would expect very good 4 comparisons if you have constrained your benchmark to 5 that detail.
6 But this Case 10 and 11, they constrained 7 some of those details but they let the individual 8 users then use whatever they have in their code for 9 fission gas release and for CO production.
10 MEMBER REMPE: So then I would say that the 11 Case 11 is more typical of what would happen if a 12 licensee were to come in and say, this is what I 13 believe will happen with my reactor for this 14 particular event, and jeepers, man, there's a big 15 variation, and don't they have to have data to 16 validate any of these different models for that 17 burnup?
18 MR. WILLIAMSON: I think so. But I ---
19 MEMBER REMPE: Something we'll have to, the 20 regulator will have to testify knowing that there's so 21 much variation there.
22 MR. WILLIAMSON: It's something that we 23 will shortly learn more about as we move from these 24 benchmarks into full-on validation where we look at 25 what data are actually available and start making 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 these detail corrections.
2 MEMBER BALLINGER: Did you separate out the 3 fission gas release plus the CO production to see what 4 contributed most to the stress? I suspect it's CO 5 production.
6 MR. WILLIAMSON: And I would agree. So if 7 you look at the, there's a paper that describes all 8 this, and in there we compare the pressures and I 9 don't remember if they were separated but I would 10 agree, that's typical. I think that's because the CO 11 is the bigger component out at those higher burnups.
12 CHAIRMAN CORRADINI: So I'd like us to get 13 through your talk before we take a break, so are we on 14 our way?
15 MR. WILLIAMSON: When is break?
16 CHAIRMAN CORRADINI: I'm going to declare 17 break in 15 minutes. How we doing there?
18 MR. WILLIAMSON: I think easily.
19 CHAIRMAN CORRADINI: Okay, good.
20 MR. WILLIAMSON: I think. I hope so.
21 Because I'm ready to wrap up TRISO just by saying if 22 you want to read more detail about anything that I've 23 talked about on TRISO so far, it's in this paper, this 24 journal article, and then I want to quickly go through 25 what our plans are with regard to development. Some 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
247 1 this has already come up.
2 We started now looking at PERFUME and 3 BISON to develop this gap, capability gap analysis so 4 that we know what's in PERFUME and what is needed in 5 BISON. As Ron pointed out, UCO is an important fuel 6 for particle fuel, and so that's certainly on our list 7 as well as more mechanistic CO production. CO's a very 8 important component, as has also been pointed out, and 9 we have a very simplistic empirical model there.
10 Then particle failure probabilities have 11 already been mentioned. We also, BISON with its 12 discreet fracture capabilities, we're in a position to 13 look readily at partial de-bonding between layers and 14 actual silicon carbide fracture, and so we will begin 15 experimenting with that.
16 MEMBER BALLINGER: Again, that silicon 17 carbide fracture is also probabilistic, because the 18 Weibull modulus is way different than for metal or 19 something, and that Weibull modulus is a function of 20 how you fabricate the particles. And so there's an 21 uncertainty on the Weibull modulus which translates 22 into a big difference in stress and fracture strength.
23 MR. WILLIAMSON: I can only agree. With 24 regard to validation, talked a bit about that. Our 25 next step that we started on is to look at the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
248 1 database. We need to better understand the database, 2 review the NRC HTGR research plan to better understand 3 the issues, and then during FY19 my intent is to 4 develop a validation plan for TRISO fuel and begin to 5 learn the high priority cases from that plan.
6 MEMBER REMPE: The only designer that's 7 coming forward with a gas reactor right now is X 8 Energy, is that true? Because AREVA said, we can't 9 find a customer. They've kind of backed off. Is there 10 anybody else out there that has a gas reactor? Because 11 if that's the case my next question is, is it true 12 that X Energy is using something in scale?
13 MR. STANEK: The answer to the first 14 question is, in terms of gas reactors, I think you're 15 correct. But there is an FHR vendor that's also using 16 particle fuels.
17 MEMBER REMPE: That's true. Okay. And then 18 I guess they are interested in the BISON suite, is 19 that true? So you have some, I mean if it's true that 20 X Energy's doing something else other than BISON, I'm 21 just kind of wondering ---
22 MR. STANEK: Again, we ---
23 CHAIRMAN CORRADING: He's not going to 24 answer that question.
25 MR. STANEK: Can't answer that 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
249 1 MEMBER REMPE: Yeah, okay.
2 CHAIRMAN CORRADINI: He's been instructed 3 not to answer that question.
4 MR. WILLIAMSON: So I'm going to spend my 5 last ten minutes on metallic fuel, and I think that is 6 possible. I have less to say about metallic fuel. It's 7 at a, there's a lot of capability there to do metallic 8 fuel but very little validation to this point so not 9 as much to talk about.
10 Here's the same slide that you saw for 11 TRISO fuel. The general capabilities, as I pointed 12 out, are the same for all these codes and in fact for 13 the type of geometry here, for metallic fuel at least, 14 the gap is essentially the same. One just needs to 15 change the gap material from gas to sodium, in this 16 case.
17 So much of the capability was in place to 18 both the gap and the general capabilities. The real 19 work has been in developing material models for 20 metallic fuel. We're looking at both binary and 21 tertiary UZr and UPuZr alloys, and you see the list of 22 capabilities that have been placed in the code.
23 Temperature dependent, temperature, species and 24 porosity depending on connectivity, anisotropic 25 swelling, thermal variation creep, thermal porosity-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 based fission gas release and zirconium diffusion 2 models.
3 Cladding is different for these fuels, so 4 models have been implemented for both HT-9 and 316 5 stainless in terms of temperature-dependent 6 conductivities, thermal and radiation creep. For the 7 HT-9 there's a failure model that has been implemented 8 in BISON.
9 Coolant channel, again, borrows heavily 10 from LWR work. In fact, essentially is the same model 11 with sodium fuel fluid properties and sodium heat 12 transfer correlations.
13 I have a single fairly simple example 14 here. This is a pit from EBR-II. It's modeled assuming 15 axisymmetry, so this is a 2-D-RZ calculation. Here 16 we're looking at thermomechanics plus zirconium 17 diffusion in the fuel alloy, all fully coupled. This 18 is sodium bonded with sodium coolant channel.
19 And then for this problem to look at 20 parametric behavior and understand fuel behavior 21 better, the user looked at variations in fuel alloy 22 composition and pin power.
23 And just two fairly simple results. Here's 24 the temperature histories for different fuel 25 compositions. You see expected behavior in 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
251 1 fuel temperature rises gradually. That's mostly 2 because of increasing porosity in the fuel until that 3 reaches kind of a steady state and then the 4 temperatures become steady. Similarly, the pressures 5 as a function of different power rates show the 6 expected behavior that higher powers mean higher 7 pressures.
8 In regards to fuel metal validation, 9 there's really quite a bit of older data out there 10 from EBR-II and TREAT, and I think FFTF. We're digging 11 into those databases to better understand what's 12 available. Cases in progress are listed in the table 13 at the bottom.
14 Scheduled for completion in FY18 are three 15 cases from EBR-II and an transient case from TREAT. I 16 have results from one of those cases, from X447. These 17 are, I would still term these as preliminary because 18 we've only seen them in the last few weeks and are 19 still digesting them but this is a binary fuel at an 20 average power of 30 kilowatts per meter for about a 21 year. The multiple rods, these rods were taken out as 22 Steve mentioned at various points in time to do PIE on 23 them, in order to get these four data points. That's 24 what had to be done, is a rod had to be removed and 25 PIE had to be done to get the fission gas release.
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 CHAIRMAN CORRADINI: Just a quick question.
2 Is the manufacturer, the way you make the metal fuel, 3 going to change the properties one would then use in 4 the calculation? Because my understanding is the way 5 they made the EBR-II fuel is not the way TerraPower 6 would make their current fuel.
7 MR. WILLIAMSON: Steve, you're going to 8 have to help. Steve is a metal fuels guru, I'm not.
9 MR. HAYES: That's true.
10 CHAIRMAN CORRADINI: And you are?
11 MR. HAYES: Steve Hayes, from Idaho 12 National Lab. The answer is, metal fuels historically 13 has been manufactured in lots of different ways, and 14 you're right, TerraPower proposes doing something a 15 little bit different but one thing we've been fairly 16 successful in showing over the years is metallic fuel 17 performance is relatively insensitive to fabrication 18 routes.
19 CHAIRMAN CORRADINI: So the property, I was 20 trying to get at the property ---
21 MR. HAYES: I don't think he needs to 22 modify many properties, or any, based on fabrication 23 route.
24 CHAIRMAN CORRADINI: Okay. Thank you.
25 MR. WILLIAMSON: Thank you, Steve. 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
253 1 just drawing your attention to these three plots.
2 Cladding hoop strain is a function of position, the 3 sharp drop in both the measurements and in the 4 calculation correspond to the top of the fuel slug.
5 Fission gas release is shown in the upper 6 right, and then a comparison of the peak inter-clad 7 temperature. This is a comparison between BISON and an 8 estimate that was made by the experimenters. There was 9 no measurements for this plot.
10 So what we can say at this point is we've 11 done some very early validation of the metallic fuel 12 capabilities. The first comparisons are very 13 reasonable. There's significant work yet to be done.
14 Which brings me to the development and 15 validation plans for metallic fuel. With regard to 16 development, there's five bullets there. Our plans are 17 to improve and/or find an alternate for the fuel 18 swelling model. Early calculations have shown an issue 19 with the radial swelling. Our current Zr diffusion 20 models are applicable to tertiary but must be extended 21 to the binary. We need to investigate lanthanide 22 diffusion and fuel cooling and cladding interaction, 23 sodium infiltration and cladding swelling.
24 There's, I guess what I'd like to leave 25 the impression is there is a capability to do metal 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 fuel but it's to the point where we can begin doing 2 detailed assessments by comparison to data. We know 3 there are areas of development that remain and we are 4 focused on those. We're building the set of validation 5 cases for FY19. I list three from EBR-II but there 6 will be more than that.
7 That brings me to summary and conclusions.
8 I think I'll just let you read those. I guess I, the 9 top bullets are, I think, obvious as we walk through 10 this, and the bottom opportunities for cooperation, 11 certainly one is to implement fuel behavior models in 12 vendor or NRC codes. I just wanted to point out that 13 that effort has begun. As we learn more about material 14 models or we use MARMOT to create proof of material 15 models, we're very interested in cooperating, 16 including those in NRC codes or vendor codes if 17 there's interest. We have begun that effort and moving 18 in that direction. Obviously there's the opportunity 19 to use BISON by the industry or by NRC, 20 CHAIRMAN CORRADINI: Questions by the 21 committee? Okay, let us take a break.
22 (Whereupon, the above-entitled matter went 23 off the record at 2:56 p.m. and resumed at 3:13 p.m.)
24 CHAIRMAN CORRADINI: Okay, let's get back 25 together, and I think now we'll turn to Elia.
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 MR. MERZARI: Yes Mr. Chairman, it's a 2 pleasure to be here. It's actually an honor to be 3 presenting, and I'm grateful for the opportunity.
4 CHAIRMAN CORRADINI: I'm not so sure about 5 that.
6 MR. MERZARI: Can you hear me?
7 CHAIRMAN CORRADINI: We can hear you.
8 MR. MERZARI: Okay, thank you.
9 MEMBER REMPE: We're not sure we believe 10 you.
11 MR. MERZARI: I'm sure, I know. In fact, 12 I was sincerely not saying the word pleasure there, 13 but let me ---
14 CHAIRMAN CORRADINI: But you went further 15 and you said honored and --
16 MR. MERZARI: Yes, I know.
17 CHAIRMAN CORRADINI: -- now it's on the 18 record.
19 MR. MERZARI: Yes that's right, it's on 20 the record. So, in this presentation, I will cover 21 the capabilities for non-light water reactors for 22 three thermal hydraulic codes in development of DOE 23 and it's a lot of material so believe me, I'm aware I 24 have an accent, and I encourage you to stop me when I-25 -when you can't understand what I'm saying but please, 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 perhaps let's consider limiting the interruptions 2 otherwise you'll never get to the end of this 3 presentation but---
4 CHAIRMAN CORRADINI: I have a---
5 MR. MERZARI: --now that I've said that, 6 you're going to interrupt me more, I'm sure.
7 CHAIRMAN CORRADINI: I have a--no, I was 8 going to say we've dealt with the Spanish so we can 9 deal with the Italian.
10 MR. MERZARI: Of course. Okay, sounds 11 good. Okay, here are the codes I'm going to talk 12 about: SAM is a component-based system code based on 13 the MOOSE framework which you read about today, 14 PRONGHORN is an engineering scale akin to the code, 15 based again on the MOOSE framework, and NEK5000 is an 16 open source competition free analytics code. For each 17 code, I will briefly review the current capabilities 18 and provide the validation status for each of the 19 advanced reactor technologies of interest. So, due to 20 the massive scale separation present in nuclear 21 reactor flows, a more physical approach is often 22 desirable.
23 Akin to what you saw this morning in 24 Steve's presentation, a single-code approach is often 25 insufficient as modeling scales several numbers apart 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 requires different methods. This reason vests DOE as 2 the full investment in the development of an 3 integrated multi-scale suite of tools which spans 4 scales starting from the plant level with SAM to 5 engineering scales with PRONGHORN, and finally the 6 finer scales with NEK5000. You will notice that there 7 is some level of overlap between these codes which is 8 introduced by design to facilitate coupling based on 9 past experience. We recognize that most, if not all, 10 safety analysis would be performed at the plant level, 11 but the situation may arise when additional final 12 scale modeling is needed.
13 I want to make a point. So, we are 14 collaborating on all of these codes with industry 15 partners, so I'll follow the orders from Chris not to 16 mention specific vendors, but there is significant 17 interest from the industry and so let's not also 18 undersell that point, very significant interest. So, 19 you will see--there is a very high probability you 20 will see some of these codes in--if not in licensing 21 applications, in commercial certifications. So, what 22 I want to say here--in this presentation, we will--I 23 will discuss the capabilities, validation status, and 24 future plans.
25 We recognize that validation is a key 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 concern. The primary focus the recent year has been 2 on the validation on liquid metal reactors and high 3 temperature gas reactors, and these efforts are being 4 expanded to cover additional reactor types. Again, we 5 envision that the safety analysis will likely be 6 conducted primarily with SAM, and potentially 7 PRONGHORN for gas reactors, especially if 1-D shows 8 approximations are sufficient. There are, however, 9 several cases where these approximations may not be 10 sufficient, and coupling to PRONGHORN or NEK5000 will 11 be needed, or least desirable. Examples for this 12 could be the pool monitoring in SFRs and the modeling 13 of the core in fast spectrum molten salt reactors.
14 So, the suite offers the flexi---
15 CHAIRMAN CORRADINI: Elia, I'm sorry, say 16 that last part again please?
17 MR. MERZARI: So, what I'm saying is that 18 you might need coupling to see if the--or post medium 19 model for the modeling of--the pool modeling in--for 20 the pool modeling in SFRs, for instance---
21 CHAIRMAN CORRADINI: Oh, okay.
22 MR. MERZARI: --where thermal 23 stratification is important, and it's shown to be an 24 important factor in several transients, or of the 25 modeling of the core in fast spectrum molten salt 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 reactors---
2 CHAIRMAN CORRADINI: Okay.
3 MR. MERZARI: --where there are 4 significant effects as I will show.
5 CHAIRMAN CORRADINI: Okay.
6 MR. MERZARI: The suite offers the 7 flexibility to combine several approaches to reach the 8 desired level of resolution and accuracy. Now, I'm 9 going to dig a little bit more into SAM. SAM is a 10 component-based system co-developed on the MOOSE 11 framework. It uses a formulation particularly suited 12 for advanced reactors in the liquid form, so, 13 something where the MAC number is--oh I'm sorry, did 14 I--where the MAC number is low. It may also have some 15 applicability to gas reactors for components in 16 conditions that do not require multi-phase or fully 17 compressible solvers. Some offer significant 18 improvements over the state of the art, the leverage, 19 and the flexibility brought in using the MOOSE 20 environment, including the use of coupling to other 21 cores such as BISON, to model fuel performance during 22 transients, and three major areas of improvement that 23 I've listed here.
24 One is the modeling of large enclosures as 25 I mentioned in--like large pools where thermal 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 stratification may be an important phenomenon, the 2 model--flexible modeling of the core where--going 3 beyond the single channel approach, and improvements 4 related to molten salt reactors such as the monitoring 5 of fuel movement and delaying the precaution rift.
6 For most of its development history, SAM has been 7 focused on liquid metal reactors and the majority of 8 these improvements, as well as its validation basis, 9 pertain to this reactor type. SAM has been compared 10 against several reactor transients for EBR-II and 11 FFTF, and in general, they are shown to provide the 12 same or better accuracies than SAS4A/SASSYS-1 for 13 those tests. You also hear this sometimes called SAS, 14 but it's actually--SAS4A is combined to SASSYS-1.
15 It should be noted that SAS has benefitted 16 from decades of calibration for those same tests, so 17 we were able to reproduce that history within SAM.
18 For example, on the right-hand side of the slide, you 19 may see a comparison between SAM results for two ETF-20 II experiments. It also beat the rejection tests, and 21 in unprotected loss of flow. The PLOF represent the 22 actual temperature of one subassembly. It is possible 23 to notice that the behavior is very similar between 24 SAM and SAS, and very close to the experiment. I will 25 present a little more comprehensive table with an NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
261 1 overview of the validation status of SAM for both fast 2 reactors and other reactor types. Among the gaps 3 identified for fast reactors is one of those that I 4 guess I was listening to you before when Jose asked a 5 question--
6 CHAIRMAN CORRADINI: Uh huh.
7 MR. MERZARI: --which is the lack of 8 spatial kinetics capability, which may be required for 9 some transients, in particular, for unprotected 10 transients for very large reactors with significant 11 homogeneity.
12 MEMBER REMPE: So just out of curiosity on 13 the bottom plot, and I know nothing about the tests 14 you're trying to match against, but it's kind of 15 interesting SAM and SAS4A are just right on top of 16 each other but none of them are--SAM isn't closer to 17 the data, if I'm looking at that plot, right?
18 MR. MERZARI: But--okay, so I guess what 19 I'm trying to---
20 MEMBER REMPE: Yes, it's like why are we 21 going through this?
22 MR. MERZARI: I'm sorry, what's the 23 question?
24 MEMBER REMPE: Why didn't it come any 25 closer to the real data if you've got 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
262 1 MR. MERZARI: I mean actually, they are 2 essentially equivalent, that's the conclusion there, 3 they're essentially equivalent.
4 MEMBER REMPE: The bottom plot.
5 MR. MERZARI: The bottom plot, yes. No, 6 but they're essentially equivalent to each other. I'm 7 not trying to say they're---
8 MEMBER REMPE: Supposed to be equivalent, 9 but you didn't get any closer to the data, is why--
10 what I'm trying to ask.
11 MR. MERZARI: Right, so the point here is 12 trying to match SAS for this particular test, and on 13 the fact that we have not mentioned exactly the data 14 there, near the peak, these are largely due to very 15 large uncertainties for the measurement in that 16 particular--remember these are like--these are an 17 unprotected loss of flow.
18 MR. BAJOREK: Elia, this is Steve Bajorek.
19 I think part of the reason is when they ran this test, 20 the thermocouples were biased low at these higher 21 temperatures.
22 MR. MERZARI: Right.
23 MR. BAJOREK: So you had--even though you 24 see good agreement between SAS and SAM and they're 25 higher than the data, that's actually where they 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
263 1 they should be.
2 MR. MERZARI: Right.
3 MR. BAJOREK: It's at that high range---
4 MEMBER REMPE: Okay, that's why I was--
5 that's the answer I wanted. You want to not match the 6 data, is the answer, okay.
7 MR. BAJOREK: --there was a negative bias 8 in measurement.
9 MR. MERZARI: I was trying to get to this.
10 MEMBER REMPE: Uh huh, I got it.
11 MR. MERZARI: Thank you, Steve. Okay, 12 move on. One of the key advancements of SAM for us as 13 far as the development of advanced modeling options 14 for inlet plenum or delta pool, which is vulnerable to 15 thermal stratification. Thermal stratification has 16 been identified in several gap analyses as a key 17 future development required for these reactor types.
18 We have pursued multiple approaches in SAM, we are 19 implementing and validating a traditional 0-D model--
20 03--0-D mixing models, while also implementing more 21 advanced multi-D options including system CFD 22 coupling. The right approach will depend of course on 23 the complexity of the transient, as shown in the 24 figure on the top right. There will also be project--
25 for loss of heat rejection, a simple 0-D model 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
264 1 often sufficient. We envision the coupling to CFD to 2 be used primarily through foreman calibrated 0-D, 1-D, 3 or in general use for the models. We have 4 collaborated with several international institutions 5 in this domain, as this remains an active area of 6 research and development. This extends to evaluation 7 as well. In fact, that's where we leverage most of 8 these international collaborations, to obtain data we 9 don't have. An example is the 1995 Monju turbine 10 test, for which--which we used for our early 11 assessment of CFD system code coupling.
12 Moving to gas reactor capabilities, SAM 13 has current limited capability for gas reactors and 14 this is not currently able to handle compressible flow 15 in transients such as air and water ingress, which are 16 typically the most challenging accidents to simulate 17 for gas reactors. However, SAM can be used to simulate 18 key components such as the reactor cavity cooling 19 system. In fact--and Professor Corradini will be 20 familiar with this--SAM simulations have shown a good 21 degree of accuracy in simulating the behavior of RCCS 22 systems, and were done in collaboration with the 23 University of Wisconsin, Texas A&M, and the National 24 Shutdown Test Facility at Argonne National Laboratory.
25 You can see at the bottom right that SAM can reproduce 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 the mass flow rate as a function of the rise of 2 channel in the facility. You will also note that 3 there is a significant scatter on that data and this 4 is a function of the strong sensitivity the RCCS 5 system has on ambient conditions. That is part of 6 this work; we develop a novel model for the fitment of 7 ambient conditions and the effect of ambient condition 8 system analysis is published in a recent paper.
9 Moving on to MSRs, molten salt reactors 10 pose unique challenges to system co-modeling, and 11 several features are unique. These include the fact 12 that the vast majority of the deposition is in the 13 coolant, as well as the presence of a fuel circuit and 14 the delayed neutron precursor drift. The validation 15 basis for SAM is limited for this reactor type, as 16 little data is currently available, but an important 17 collaboration is ongoing with Louisiana State 18 University for the simulation of MSRE. In the top 19 right, you may see that SAM is able to reproduce these 20 behavior for this reactor, and the reason--this is 21 ongoing work, but these results have been published in 22 a recent 2018 ENS transaction summary. Recently, 23 we've added several MSR targeted features in SAM, 24 including salt properties, delayed neutron precursor 25 drift, generation and transport, and heat generation 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 in the coolant as well as point kinetics and coupling 2 for CFD for fast spectrum systems.
3 At the bottom right, you may see a 4 demonstration of these features for--of the novel 5 point kinetics capability in SAM that includes the 6 delayed neutron precursor drift. This--again, this is 7 just a demonstration with a very simple loss of flow 8 transient in a simplified single-channel MSR loop 9 model. One of the key gaps identified for SAM, and 10 this you will see, it's recognized as an important 11 challenge for MSRs is the lack of coupling for salt 12 chemistry models and codes.
13 CHAIRMAN CORRADINI: Can we hold off 14 there? So what is the coupling that's missing in the 15 chemistry models?
16 MR. MERZARI: So, how you get properties 17 in--as you increase the amount of fission gas 18 material, fission products inside of the key.
19 CHAIRMAN CORRADINI: Oh, okay. I always 20 thought eventually--oh no, you've got it up here, thaw 21 and freeze during overcooling events.
22 MR. MERZARI: Yes, I didn't mention that 23 because I'm going to emphasize that for FHRs.
24 CHAIRMAN CORRADINI: Okay.
25 MR. MERZARI: They're--it is a key 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 limitation for FHRs.
2 MEMBER KIRCHNER: That order, could you go 3 back? I'm still pondering about putting in air and 4 water mixtures into SAM, do you have to change your 5 basic equation sets that you're solving?
6 MR. MERZARI: For which one, sir?
7 CHAIRMAN CORRADINI: Where are you looking 8 Walt?
9 MEMBER KIRCHNER: Under SAM.
10 MR. MERZARI: Which slide?
11 MEMBER MARCH-LEUBA: He's talking about 12 air and water injection. You see---
13 MEMBER REMPE: Air ingress.
14 MR. MERZARI: Oh okay, so for the air and 15 water ingress? Okay, if you wanted to do that with 16 SAM, you would need a multi-phase model or a way to 17 track two phases.
18 MEMBER MARCH-LEUBA: Plus, you would have 19 to add the chemistry, I mean, the sodium and water 20 don't mix very well.
21 MEMBER KIRCHNER: No.
22 MR. MERZARI: No, no, I'm talking about--
23 that's--the air and water ingress are for gas 24 reactors. Those are the post-rated accidents 25 typically for gas reactors.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
268 1 MEMBER REMPE: That's not a coding 2 capability is it?
3 MR. MERZARI: Yes, that's not a coding 4 capability, we don't type that.
5 MEMBER MARCH-LEUBA: Yes.
6 MR. MERZARI: But to do that, we would 7 need to track those. I don't think you would need 8 necessarily an accurate chemistry model to do that, 9 but for air and water ingress, you must be able to 10 track two phases separately.
11 MEMBER MARCH-LEUBA: But---
12 MR. MERZARI: Well actually, I mean---
13 MEMBER MARCH-LEUBA: If you have hard 14 graphite---
15 MR. MERZARI: Oh sure---
16 MEMBER MARCH-LEUBA: --and you have an air 17 ingress---
18 MEMBER KIRCHNER: Or a water ingress?
19 MR. MERZARI: Yes, certainly there has to 20 be--graphite interaction will require some material 21 modeling for sure. So as mentioned, the next reactor 22 type is the fluoride high temperature gas--high 23 temperature reactor. This is the combined, 24 essentially features of MSRs and gas reactors. The 25 coolant is typically a molten salt while the 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
269 1 remains in solid form. The validation basis for this 2 reactor type is again limited due to lack of available 3 facilities, but several collaborations are ongoing 4 with both vendors and universities. SAM presents some 5 key attractive features for FHRs, including salt 6 properties as well as salt thaw and freeze models, 7 which are currently under development in our 8 collaborations with the University of Wisconsin, and 9 may be needed in the case of other cooling transients, 10 as well as to model the start up. I know they have a 11 nice group at the University of Wisconsin for natural 12 circulation, I've seen some very interesting thermal-13 hydraulic effects there that would be fascinating to 14 model with this. We note that SAM has also components 15 targeted specifically to FHRs, such as flow diodes.
16 As an example, on the right side you may see a 17 demonstration of FHR simulation for the P-B FHR design 18 of Berkeley, and this view states how the flow diode 19 is used to reach the desired long term cooling state, 20 so basically, to prevent the flow reversal.
21 And now to the fun part. In this table we 22 summarized the validation status of SAM for fast 23 reactors. The table represents, on the rows, the key 24 capabilities and functions in SAM, while in the 25 columns, we list a set of specific things we have NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
270 1 identified as sources of validation data, both for ITs 2 and SETs. The columns in green represent actual 3 reactors, so it's the ultimate validation basis in a 4 sense. Each cell is marked with a validation status 5 letter. C stands for Complete, O for Ongoing, and P 6 for planned. We note that a considerable amount of 7 work has been done, but far more is needed and it is 8 ongoing. We also note that for fast reactors were not 9 nearly as comparable as--were nearly comparable to 10 LWRs, a significant amount of data is available; we 11 won't be as lucky in other reactor types. Any 12 questions on this slide?
13 CHAIRMAN CORRADINI: So let me--maybe you 14 said it quickly and I missed it, SAM is not a 15 derivative of SAS4 and SASSYS, it's essentially a new 16 formulation for--and then originally for sodium 17 applications, but now single-phase system application?
18 MR. MERZARI: Right, right.
19 CHAIRMAN CORRADINI: Okay.
20 MR. MERZARI: So, the reason I guess I'm 21 expecting a question there, why the switch, or why the 22 change---
23 CHAIRMAN CORRADINI: Well I can guess, so 24 I was going to hold off, I was going to go a different 25 direction with my question, but you go ahead 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
271 1 answer that question, that sounds good.
2 MR. MERZARI: I think that's the first 3 time today that that happened. So no, the reason is 4 while certainly SAS has decades of experience with the 5 modeling of liquid metal reactors, it has shown over 6 the years it is pretty hard to modify and couple to 7 other codes and difficult to maintain. So the idea 8 there was to essentially replace at least one part of 9 SAS--let us remind ourselves that SAS does much more 10 than simple thermal hydraulics--and make it more 11 flexible and usable in the longer term. Now, I don't 12 want to make it sound as SAS and SAM are necessarily 13 one is a replacement of the other; in fact, SAM and 14 SAS can be used together. SAS has several models 15 especially for severe accidents that are needed to 16 model, for instance, in certain conditions the 17 unprotected loss of flow. So you can use SAM for the 18 thermal hydraulics and SAS for all those models that 19 come in the fuel--in the transient fuel analysis.
20 In this table we summarize the validation 21 status for SAM for FHRs and MSRs. A lot of nice work 22 has been done here for these reactor types and 23 actually you shouldn't consider this table nearly as 24 complete; we need many more tests and these will need 25 to happen in the future if these reactor types are 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
272 1 become viable. Again, we note the notable activity--
2 the notable ongoing work on MSRE.
3 CHAIRMAN CORRADINI: But the similarity 4 between all of these--well I know you're going to 5 switch now to PRONGHORN--so the similarity between all 6 of these is single-phase flow, normal operation, 7 operational occurrences, and potentially certain 8 selected design base assessments?
9 MR. MERZARI: Actually, the unprotected 10 loss of flow show certainly DBAs, but also beyond 11 design basis.
12 CHAIRMAN CORRADINI: But single-phase?
13 MR. MERZARI: Single-phase, yes.
14 CHAIRMAN CORRADINI: Okay.
15 MR. MERZARI: Well I mean in most of these 16 designs--in most of these transients, there won't be 17 a phase change for sodium, for instance.
18 CHAIRMAN CORRADINI: Well okay, but sodium 19 boiling makes life more interesting and I would expect 20 I'd see that in a beyond design basis accident 21 because---
22 MR. MERZARI: You won't see it from 23 metallic fuel. It's a strong statement, but typically 24 in--when metallic fuel forms, you will not see sodium 25 boil.
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 CHAIR CORRADINI: I know what you're 2 saying.
3 MEMBER REMPE: So back to Walt's question 4 about air ingress and water ingress, you've said well 5 for the sodium reactors, I can go to beyond design 6 basis accidents; aren't you going to beyond design 7 basis accidents for gas reactors if you're going to do 8 air ingress and water ingress events? And then, what 9 are you are going to link to, BISON, that doesn't have 10 a severe accident model for fuel?
11 MR. MERZARI: So okay, let me separate the 12 question. So for--let me answer first the sodium 13 part, so--and then the link to BISON. So right now, 14 the only code that can really some of the severe 15 accidents part is SAS4A, and the plan there is to 16 implement some of those more--the long term plan is to 17 implement some of the models into BISON and use some 18 BISON to do those severe accident parts.
19 MEMBER REMPE: Okay.
20 MR. MERZARI: Regarding the gas reactors 21 question, again, we are not planning to do water 22 ingress and air ingress with SAM, we are not. So 23 those are not part of the analysis at the moment. We 24 may plan to do those with PRONGHORN, but right now the 25 plans haven't firmed up for 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
274 1 CHAIRMAN CORRADINI: But the reason--I'm 2 still back on sodium boiling--the reason I don't get 3 it is because with the appropriate reactor design and 4 feedback, I don't get high enough temperatures or 5 enough power that will get me to the saturation 6 conditions.
7 MR. MERZARI: Uh huh.
8 CHAIRMAN CORRADINI: But--okay.
9 MR. MERZARI: No but again, this is an 10 aspect that the three tests showed in the 70s, and I 11 think Tanju can perhaps answer this question better 12 than me.
13 MR. SOFU: So you can certainly postulate 14 accidents during which sodium could get to boiling 15 temperature, but we feel modeling that particular type 16 of transient is not within our immediate plans, 17 because when sodium boiling starts, you have bigger 18 worries than just reactivity feedback or fission.
19 Those temperatures are essentially the trigger for 20 severe accident conditions.
21 CHAIRMAN CORRADINI: Okay, that's fine.
22 I figured.
23 MR. MERZARI: So that said, there are 24 boiling models in SAS, so if that became a concern, 25 that could be potentially transported to SAM, but 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
275 1 would require some development work. Those models are 2 not there in SAM today.
3 CHAIRMAN CORRADINI: Okay, I understand 4 where you're coming from.
5 MR. MERZARI: Okay, switching to 6 PRONGHORN. PRONGHORN is MOOSE-based engineering lab 7 analysis tool for advanced reactors. It targets 8 primarily pebble bed gas reactors but it is extendable 9 to other designs. So far everything that we've been 10 doing is for pebble bed reactors. It features an 11 isotropic porous media modeling formulation, as well 12 as a more advanced formulation. In its enhanced 13 formulation, it can combine porous media agents with 14 other agents to generate a full model of the primary 15 system in a significant advancement to what was 16 previously done with these porous media codes for 17 pebble beds. Moreover, distributing the existing 18 models for pebble beds developed with NEK5000 can be 19 implemented in a straightforward manner in PRONGHORN, 20 and it can also couple to SAM to model on the FCCS and 21 the overall system transient response. The validation 22 basis so far is limited; PRONGHORN is a much more 23 recent code than the other two codes I presented, but 24 as shown in a recent white paper and shown here at the 25 top right, PRONGHORN can reproduce while the pebble 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 temperature is a function of axial and radial 2 positions in the SANA experiments, which is a classic 3 validation data set from Germany for pebble beds.
4 CHAIRMAN CORRADINI: So is--I want to make 5 sure I understand. Two questions; one is what is it 6 about the physics with PRONGHORN that you've chosen to 7 use it instead of SAM where you, in your initial 8 slides, you used the terminology "applicable DOE 9 engineering-scale code," whereas SAM is useful in the 10 SFR, PRONGHORN is the choice of your interest, what's 11 the physics of it do you make the switch?
12 MR. MERZARI: So the--it's not as much of 13 the physics as a metro scale resolution, so here we're 14 solving more scales because we are actually modeling 15 three dimensionally, the old code, if it were porous 16 media.
17 CHAIRMAN CORRADINI: Oh, as where SAM is 18 a one dimensional---
19 MR. MERZARI: It's essentially one 20 dimensional.
21 CHAIR CORRADINI: It's a sodium RELAP for 22 quantify--I know that would make you upset.
23 MR. MERZARI: It's a sodium/salt RELAP.
24 I'm okay with that definition.
25 CHAIRMAN CORRADINI: Okay, so it's three 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 dimensional because of the pebbles?
2 MR. MERZARI: Because of the pebbles, and-3 -okay of course you could model, for instance, an FHR 4 with SAM, and instead of modeling 3-D through the 5 pebbles, you could replace it with a channel model.
6 CHAIRMAN CORRADINI: Okay, I see. All 7 right.
8 MR. MERZARI: But if you want to add more 9 detail, especially for the neutronics, perhaps a 10 porous media model is more effective. What also SAM 11 is missing is some of the multi-phase tracking 12 capabilities that we kind of need for gas reactors for 13 some of the transients and the compressible models.
14 CHAIRMAN CORRADINI: So is it a coupling 15 between PRONGHORN and the other tool that Tanju was 16 speaking of, the RATTLESNAKE?
17 MR. MERZARI: Yes, that's available.
18 CHAIRMAN CORRADINI: But that is the 19 logical connection?
20 MR. MERZARI: Yes, that's the logical 21 connection.
22 CHAIRMAN CORRADINI: Okay.
23 MR. MERZARI: So those two tools are both 24 based on the MOOSE framework and they can be coupled.
25 CHAIRMAN CORRADINI: Okay.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
278 1 MR. MERZARI: And they will be used for a 2 full analysis of these pebble. So--and perhaps this 3 answers also your question Mike--so this is the 4 advanced formulation in PRONGHORN--I'm showing an 5 equation, I know--and it's basically a multi-phase 6 approach in which the pebble is essentially 3-D; 7 there's another phase but it is stationary, and this 8 allows you--there is a fundamental two-phase 9 capability in PRONGHORN that can be used, for 10 instance, you can think of the operations regions that 11 have no pebbles, or the phase goes to zero, so it's 12 very easy to combine multiple regions with this, and 13 it can also, this fundamental capability, be used to 14 track some from an interphase, perhaps to here. So 15 you can see that the fundamental--the physics in 16 PRONGHORN is probably more suitable to track some of 17 the scenarios like air ingress and water ingress.
18 CHAIRMAN CORRADINI: But when you say 19 water ingress, you mean steam? That's what I 20 immediately took to be the case; or are you talking 21 liquid water?
22 MR. MERZARI: For water ingress, I think 23 what typically is assumed is not steam, it's liquid.
24 CHAIRMAN CORRADINI: Really?
25 MR. MERZARI: No, I'm sorry, steam, steam, 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 yes.
2 MEMBER REMPE: It's supposed to be steam.
3 MR. MERZARI: Yes.
4 CHAIRMAN CORRADINI: Yes, that's what I 5 thought.
6 MEMBER REMPE: Even if you---
7 CHAIRMAN CORRADINI: So it's a mixture of 8 compositions in a gas phase? Okay, fine.
9 MR. MERZARI: Yes, but there is still an 10 interphase and you still need to track it.
11 CHAIRMAN CORRADINI: Right, well a mixing.
12 MR. MERZARI: Yes, yes.
13 CHAIRMAN CORRADINI: I understand, okay.
14 MR. MERZARI: Okay sorry, do you want me 15 to--
16 CHAIRMAN CORRADINI: I'm good, no, no, no, 17 you're doing fine.
18 MR. MERZARI: Thank you, good. Okay, so 19 we present here the validation status of PRONGHORN in 20 a manner similar to what was done for SAM, you will 21 note the limited validation basis, which so far has 22 been completed only for the SANA test. Additional 23 simulations and benchmarking are ongoing. We note 24 that there is a potential issue here; some of these 25 data sets are not available.
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 MEMBER REMPE: Yet.
2 MR. MERZARI: Yet. Finally, and hopefully 3 I'm not--how much time do I have?
4 CHAIRMAN CORRADINI: You have about 15 5 minutes.
6 MR. MERZARI: Okay, that's pretty much on 7 time. Okay so finally, we introduce NEK5000, a state 8 of the art, open source computational fluid analytics 9 code that features both incompressible and 10 compressible fluid formulations. Thanks partly to 11 being open source, NEK5000 is easy to integrate and it 12 features a MOOSE interface, which is part of the 13 rationale to go toward an open source specific code.
14 NEK5000 covers a range of resolutions from direct 15 numerical simulation to large eddy simulation, to 16 Reynolds-averaged Navier-Stokes, and finally porous 17 media formulations. What sets it apart from the codes 18 is that it has a very state of the art capability 19 which allows it to run very large calculations on 20 supercomputers as well as models calculations in small 21 clusters and even laptops.
22 CHAIRMAN CORRADINI: So I want to take you 23 back to the slide that you had, but you said it 24 originally where you showed SAM in terms of length 25 scales, PRONGHORN, and NEK5000, and there was overlap; 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 and you said the overlap was good for what?
2 MR. MERZARI: To couple to different 3 codes. So I'll give you an example. So, let's take 4 a pebble bed for instance, you want to model the 5 pebble bed with a porous media approach---
6 CHAIRMAN CORRADINI: Right.
7 MR. MERZARI: --okay? You kind of want to 8 move away from the interface between the porous media 9 and the CFD codes. You don't want the interface 10 between the CFD codes and the porous media to be right 11 at the end of the pebble beds; why, because typically 12 that leads to some instabilities. So if you are able 13 to extend the PRONGHORN calculation a little bit more 14 in the open region, the coupling becomes much more 15 sound. So you couple essentially in the CFD region 16 rather than at the interface, because that interface 17 tends to be unstable very often.
18 CHAIRMAN CORRADINI: Okay.
19 MR. MERZARI: Does that make sense to you?
20 CHAIRMAN CORRADINI: I think so. I was 21 going to go a different direction; my thought was why 22 have PRONGHORN at all if NEK5000 can do porous media 23 calculations?
24 MR. MERZARI: So you can do porous media 25 calculations with commercial software even, you don't NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
282 1 even need NEK5000 really for that. The rationale 2 there is that you really want something optimized, 3 especially if you plan to couple with neutronics, and 4 having a well optimized, well established code to run 5 that particular portion does buy you something.
6 Actually, if you want a code modeling tool for pebble 7 beds, I think that's the right strategy, to include a 8 specific code that is optimized for that.
9 CHAIRMAN CORRADINI: So I'm going to ask, 10 another similarity, which I'm sure you're not going to 11 like. So, is NEK5000 just a better fluent?
12 MR. MERZARI: I'll take that; better is 13 good.
14 CHAIRMAN CORRADINI: I used the word 15 better so you would agree with me. But it's similar 16 to, it's essentially---
17 MR. MERZARI: Yes, but---
18 CHAIRMAN CORRADINI: --it's uninitiated.
19 MR. MERZARI: Yes, but you see--yes okay, 20 I'll take that, but we have access to the source code.
21 CHAIRMAN CORRADINI: Okay, fine.
22 MR. MERZARI: And that matters, it's not--
23 it's just not---
24 CHAIRMAN CORRADINI: I'm not disagreeing, 25 but in terms of similarity, there is a--that's an NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
283 1 appropriate analog.
2 MR. MERZARI: It's an appropriate analog, 3 I like the word better in there but it---
4 MR. SOFU: I think I would add that, I 5 think that you--with NEK5000, you would get reference 6 solutions that you can get validation for--RANS-based 7 approach code---
8 CHAIRMAN CORRADINI: Sure.
9 MR. SOFU: --from commercial code. Order 10 of accuracy, level of--order of the solutions is going 11 to be significantly better.
12 MR. MERZARI: Right, that's why you used 13 the word better, so I---
14 CHAIRMAN CORRADINI: So that he wouldn't 15 argue with me.
16 MR. MERZARI: So on the better part, the 17 code is ranked first in several CEA and EA blind 18 benchmarks--I mean, the blind is important--on CFD 19 applications to reactor safety. This includes the T-20 junction benchmark, which is listed there, which was 21 a model in thermal striping. Another notable test is 22 the simulation at the MAX facility, shown at the top 23 right, illustrating how NEK5000 can reproduce the 24 measurable velocity of two jets impinging on a 25 structure, again for applications in thermal striping.
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 The primary working validation basis for NEK5000 is--
2 has been focused on fast reactors. As mentioned in 3 the previous slide, there has been a significant 4 validation work performed for thermal striping, 5 including for sodium data sets, such as the PLAJEST 6 test from GE. Thermal striping is one of the limiting 7 conditions for liquid metal reactors, as the core 8 outlet temperature may have significant variations 9 between assemblies and structures; it may be subject 10 to significant thermal fatigue.
11 Additionally, I have performed extensive 12 validation tests for wire wrapped bundles, both for 13 legacy tests, and including PIV---recent PIV 14 experiments done at Texas A&M. Again, we envision 15 NEK5000 and CFD in general to be used primarily to 16 inform and calibrate SAM for safety analysis 17 purposes; but NEK5000 may also be used for component 18 level modeling, and as mentioned previously, we expect 19 that its coupling to SAM is going to play an important 20 role in the simulation of thermal stratification in 21 the upper plenum.
22 Moving to gas reactors, while not as 23 extensive as for liquid metal reactors, NEK5000 has 24 received significant validation for gas reactors as 25 well. We have two ongoing activities here I'm 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
285 1 going to briefly mention; the simulation of the upper 2 plenum tests, and the simulation of pebble bed tests 3 at Texas A&M. On the top right of the slide you may 4 see a NEK5000 simulation of jets of different Reynolds 5 numbers in the upper plenum mock-up of Texas A&M, and 6 the bottom right, you may see a NEK5000 simulation of 7 the flow through a random pebble test. This test is 8 being conducted right now, we'll have soon comparison 9 tests, so this is ongoing work. We envision that 10 NEK5000 will be primarily used to inform PRONGHORN, 11 and help define distributive resistant models for 12 pebble flows; for example, to provide closure models 13 for pressure drop in transfer and distributive 14 resistance. It mostly will be coupled to PRONGHORN 15 and SAM to simulate the upper plenum in a coupled 16 fashion if needed. The chair has disappeared. Okay, 17 well I'll go on.
18 MEMBER REMPE: You can go ahead.
19 MR. MERZARI: Okay. In the next two 20 slides I'll summarize quickly, applications for FHRs 21 and MSRs. In the interest of brevity, I do not dwell 22 long here. In general, NEK5000 will be used to inform 23 SAM and PRONGHORN. A good example in the case--is the 24 case of twisted tube heat exchangers. Available 25 correlations do not often provide good match 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
286 1 experiments in the region of interest due to the fact 2 that typically it exhibits changes for molten salt at 3 very high Prandtl and very low Reynolds, and there's 4 very little data in that region. You can see the 5 figure on the top right is the correlation, for 6 instance, provides--over estimates the loosened number 7 by over a factor of two in those--for that test, while 8 NEK5000 is relatively close to the experimental data.
9 Again, you can see the Reynolds numbers 10 included are very low, but these are actually some of 11 the Reynolds numbers that are being targeted by these 12 designs, and so nothing was changing in the NEK5000 13 model there. We could match the--we could match very 14 well the correlation in the region in the validity of 15 the correlation, but outside of the region of validity 16 of the correlation, the correlation fails miserably 17 while the same NEK model gets essentially the right 18 answer for this particular test. A possible exception 19 to the rule of using CFD models to conform with SAM or 20 PRONGHORN is the modeling of the code in fast spectrum 21 molten salt reactors, which is illustrated in this 22 figure at the top.
23 In fact, in molten salt reactors there may 24 be re-circulation regions and these may cause 25 significant local peaking. The reason for that is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
287 1 that you have very high Prandtl; the fusion doesn't 2 quite account for any effect in the heat removal, and 3 you deposit all the heat in a region where no 4 velocity--where this no velocity and suddenly you have 5 local idle position, where there is no means for the 6 heat to escape. So, of course you could take care of 7 that while removing the--by removing the re-8 circulation region, but in a real reactor that might 9 be actually quite a challenge.
10 I will now present a summary table 11 concerning the validation basis for NEK5000. NEK5000 12 is fairly mature--is a fairly mature code for this 13 class of reactors, and is considered by validation as 14 being done. As you see, there's quite a few C's 15 there, and all required features have been 16 implemented. Nonetheless, some validation work still 17 needs to be done for these reactor types. The picture 18 is far less rosy when we get to FHRs and MSRs, as the 19 NEK5000 validation basis is far more limited.
20 Significant work has been done, however, for pebble 21 bed reactors, and there's a lot of work--a lot more 22 work to be done, and we have an--we mentioned an 23 ongoing activity for MSRs.
24 CHAIR CORRADINI: Is it the fluid 25 properties that make--I mean, since NEK5000 is a--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
288 1 don't want to say standard but we'll say an open 2 source commercial tool, is it the fluid properties 3 that make it in a situation where you have to continue 4 to do validation?
5 MR. MERZARI: Well the conditions are 6 relatively different, first of all, and so the---
7 CHAIR CORRADINI: Right, I knew that. I 8 was going to say for the MSR I got it, because you're 9 essentially producing the heat within and that causes 10 the feedback you've got to check, but I'd assume that 11 for these two it would be primarily the material or 12 the fluid properties.
13 MR. MERZARI: The fluid properties is 14 relatively--it's the geometry, the complexity, and 15 certainly due to the geometry boundary conditions, and 16 in the case of FHRs in particular, the salt and 17 thermals, which we don't have and we will need to 18 validate for some of these transients. But, let me 19 also point out the fact that FHRs and MSRs have very 20 low Reynolds numbers---
21 CHAIR CORRADINI: Right.
22 MR. MERZARI: --and that poses some 23 challenges. So for instance, using standard RANS 24 models in those conditions is questionable at best.
25 CHAIR CORRADINI: But what--if I were 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
289 1 regulator, I would just say well get rid of the 2 turbulence completely, and let me look at essentially 3 a laminar calculation as a bound.
4 MR. MERZARI: Well--but, the laminar 5 calculation, I'm not sure it's a bound.
6 CHAIRMAN CORRADINI: But, what Reynolds 7 numbers are we talking about? We're talking about 8 thousands, or at the least---
9 MR. MERZARI: Sometimes hundreds, 10 sometimes hundreds and it's---
11 CHAIRMAN CORRADINI: But I'm definitely 12 laminar.
13 MR. MERZARI: Well, I mean, that's the 14 thing, you might be laminar if you consider a single 15 channel, but the moment that you add more complex 16 geometry, you might end up with instabilities and the 17 flow not actually being stationary.
18 CHAIRMAN CORRADINI: Okay.
19 MR. MERZARI: And that is definitely 20 questionable. I mean, there's no such thing as a 21 truly laminar flow, that is completely stable unless 22 you're talking about a fully developed nice pipe with 23 the flow well characterized.
24 MR. SOFU: And with that kind of a heat, 25 if you just focus on flow, the Reynolds number, 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
290 1 also have to consider the buoyancy effect--
2 MR. MERZARI: Right.
3 MR. SOFU: --associated with a tremendous 4 heat flux coming from very, very hot surfaces.
5 MR. MERZARI: That's right. So with that, 6 I conclude. Actually, I think I bought you back five 7 minutes. The DOE is developing a modern, thermal 8 hydraulic, multi-scale suite applicable to a variety 9 of reactor designs. In fact, while emphasis to date 10 has been mostly for sodium and gas reactors, 11 substantial capability exists for other reactor 12 contents, such as MSRs and FHRs. I've shown some 13 basic capabilities to simulate advanced reactor LDEs, 14 including an unprotected loss of flow, and more is 15 available in the reports and papers I've sent you and 16 I remain at your disposal for further questions. I 17 realize I haven't covered all that I--all the 18 possible--I mean, there's a lot material here and I 19 cannot possibly cover the full status of these codes, 20 but I'm available for further questions, and I hope I 21 conveyed to you the right picture, which is in my 22 opinion, that these codes have received significant 23 interest from the industry, there is a lot--there is 24 significant capability there, there is remarkable 25 progress, and there is definitely potential. 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
291 1 course, a lot more work needs to be done, especially 2 when it comes to validation. Thank you very much.
3 CHAIRMAN CORRADINI: Questions by the 4 committee? On to the world of source term. Everybody 5 wake up.
6 MEMBER BLEY: While you're waiting for it 7 Chris, I asked the question at the wrong time this 8 morning. We didn't have--well, being that we were 9 looking at micro structures, but now as we've got 10 molten fuel and liquids and aerosols being generated, 11 the chemistry's got to become extremely important.
12 How is the modeling of the chemistry in this section 13 of the work you guys have done?
14 MR. STANEK: So, I think the best answer 15 to that is that's a topic for the next time we meet.
16 MEMBER BLEY: Do you have an actinide 17 chemist who's working with you on this stuff, or is 18 that for the future?
19 MR. STANEK: Based on our interactions 20 with colleagues in the NRC and some of the vendors, we 21 now believe we have our arms around what the needs are 22 in modeling and simulation--
23 MEMBER BLEY: Uh huh.
24 MR. STANEK: --of chemistry. We're 25 working with the Advanced Reactor Technology Program NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
292 1 to hone in on what our strategy will be going forward, 2 so it's still notional at this point, but yes, we have 3 engaged the radio chemists and the right modeling and 4 simulation people to develop a needed capability which 5 frankly doesn't---
6 MEMBER BLEY: But that's off in the 7 future, so the stuff we're looking at now is without 8 that except as it shows up in some kind of 9 correlations that get picked up somewhere?
10 MR. STANEK: That's correct.
11 MEMBER BLEY: Yes. Okay, thank you.
12 CHAIRMAN CORRADINI: Tanju? If it's a 13 green you're good.
14 MR. SOFU: The last technical presentation 15 of the afternoon, so--and I'll just jump right into 16 it. The formal source term definition in the U.S 17 Regulations is provided in Part 50.2, if you read 18 there; and then this source definition is closely 19 linked with the LWR source term requirements in Part 20 50.67. The first reference there, TID-14844, makes a 21 prescriptive assumption for release of all--100 22 percent noble gases, half of halogens, and one percent 23 of the remaining solids to the containment, assuming 24 a LOCA leading to a core melt as the bonding event.
25 The other reference there you see, 1465, also assumes NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
293 1 a LOCA core melt, but it specifies unique EWR/PWR 2 releases in-vessel, ex-vessel, accounting for the 3 engineering safety features along with the uncertainty 4 analysis; and finally, the SECY-93-092 sets the stage 5 for regulatory expectations for mechanistic source 6 term evaluations for advanced reactors. And I think 7 the rationale for a mechanistic sourced term 8 assessment is that because the source term 9 requirements will significantly differ for those 10 reactors mainly because there is no single bonding 11 event like LOCA determining the source term. In fact, 12 for some concepts like those that rely on TRISO fuel, 13 very small releases can be anticipated even during 14 AOOs and DBAs, and sometimes during normal operation 15 from defective fuel particles; there's certainly a 16 circulating activity in the coolant expected.
17 CHAIRMAN CORRADINI: So the reason though 18 that you--you don't have to go back to the slide 19 before--but the reason you identified that, this is 20 for LWRs?
21 MR. SOFU: Correct.
22 CHAIRMAN CORRADINI: So there's no 23 construct except from a process standpoint for non-24 LWRs?
25 MR. SOFU: 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
294 1 CHAIRMAN CORRADINI: Okay.
2 MR. SOFU: So I'm trying to draw a contrast 3 as to why we would need a more mechanistic sourced term 4 approach for advanced reactors.
5 CHAIRMAN CORRADINI: Okay. The reason I 6 asked the question like that is that the background 7 documents you gave us, I interpreted as more process 8 discussions than tool discussions.
9 MR. SOFU: Absolutely, yes.
10 MEMBER REMPE: So this isn't really for 11 you, this is for the folks sitting around the table 12 here, but it sounds like from your first bullet when I 13 was looking at this last night, I was thinking about if 14 we had a new regulatory system and you had top level 15 regulatory criteria, it sounds to me that your 16 experience is also suggesting that you shouldn't just 17 use the--for top level regulatory criteria, 10 CFR 100, 18 you should also do--use the regulatory criteria for 19 normal operation; and so when we talked about that a 20 few weeks ago or a month ago, it sounds like there 21 might be a gap in what we heard about because they 22 might need to consider the regulatory criteria for 23 normal operating releases and AOOs; is that kind of 24 what you're thinking too?
25 MR. SOFU: I agree, that's correct, 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
295 1 think the--this licensing modernization project---
2 MEMBER REMPE: That's where I'm coming 3 from.
4 MR. SOFU: --the NRC is evaluating based on 5 DOE and NEI initiative sort of sets the stage for that 6 discussion.
7 MEMBER REMPE: Yes, but are they--in the 8 discussion we had they did not have all of like, the 10 9 CFR 20 and 50 releases--thank you--or criteria.
10 CHAIRMAN CORRADINI: I don't--I guess, I 11 don't appreciate what you're asking, I apologize.
12 MEMBER REMPE: When Former Commissioner 13 Apostolakis and Karl Fleming were here a few weeks ago 14 and we talked about that, they did not have all of the 15 normal operating regulatory criteria in there and they-16 -and George, when I brought that up actually said yes, 17 you're right, this is--basically even if we design to 18 these top level regulatory criteria that we've 19 identified, it may not meet all the regulatory 20 requirements. And so I'm just kind of emphasizing that 21 because of the guy sitting over there to the left and 22 doing his letter.
23 MEMBER BLEY: You're wrong.
24 MEMBER REMPE: I'm wrong?
25 MEMBER BLEY: 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
296 1 MEMBER REMPE: Well, years ago---
2 MEMBER BLEY: I'm not doing a letter.
3 MEMBER REMPE: Oh, okay. Well if you ever 4 do---
5 MEMBER BLEY: Not today.
6 MEMBER REMPE: --okay, if you ever do a 7 letter on that topic, okay, I just wanted to emphasize 8 that point. Again, this is off topic, but it's 9 something I wanted to emphasize. Thank you.
10 CHAIR CORRADINI: Sure.
11 MR. SOFU: So the bottom line is that for 12 advanced reactor types, you will need to focus on a 13 broader spectrum of accidents, not just local-type 14 bonding events, because there could be releases that 15 are frequent but very small, versus infrequent but 16 sizably large, but the risk factor will be comparable.
17 Also, accidents that could lead to early releases 18 versus delayed releases could have implications on 19 radionuclide discharge content as well as the emergency 20 response implications, and we need to worry not only 21 about the fuel in the reactor core but we need to 22 concern--probably consider the field and storage for 23 liquid metal coolant systems. Coolant covered gas 24 clean up systems could malfunction, they could just 25 immediately bypass the reactor pool but that could be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
297 1 right inside the containment, and any failure of that 2 system could be sourced inside the containment. And 3 for molten salt reactors--not that I know much about 4 it, but I'm assuming some chemical processing systems 5 may or may not be inside the containment--their 6 malfunction could also lead to significant releases.
7 So the--on top of all that source term 8 assessment, but also probably the mechanistic approach 9 to considering a broader spectrum of accidents for 10 advanced reactor concepts will also be needed to 11 support PRA in Level 2 and Level 3, as well as 12 emergency planning on reduction requests. So there are 13 plenty of reasons to do this mechanistically looking at 14 a broader spectrum of accidents. What I provide here 15 is a kind of--a proposed mechanistic source term 16 definition, and the general approach I present in the 17 next speech is largely based on this particular 18 definition.
19 CHAIRMAN CORRADINI: But what I--I've looked 20 at your slides ahead of time. What I see is process, 21 so it's mainly a process discussion?
22 MR. SOFU: Right.
23 CHAIRMAN CORRADINI: Okay, okay.
24 MR. SOFU: Here's the approach. The first 25 step is an inventory assessment; and then you 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
298 1 understand the release pathways depending on what 2 you're considering as the source. We need to model the 3 phenomena in volt and release pathways, we need to 4 evaluate specific scenarios, not just bonding events, 5 and of course, last step is the regulator will take a 6 look and say we agree or don't agree. So the inventory 7 step, there can be significant differences for each 8 reactor type, but even sometimes for each design. For 9 a traveling wave reactor that targets an ultra-high 10 burn-up fuel, for example, versus for a heat pipe based 11 micro-reactor with very, very low burn-up, an inventory 12 will be significantly different, even with the same 13 type of metallic fuel, for example.
14 For the release pathways shown here at the 15 second step, I have example diagrams for them for SFRs 16 and MHTGRs in coming pages. For modeling and scenario 17 evaluation, approaches that could be utilized can rely 18 on NRC and the recently developed DOE capabilities 19 you've heard this afternoon. That's something that 20 we've performed in trial mechanistic source term 21 calculations for SFRs and MHTGRs. That's the release 22 pathway diagram for a liquid metal reactor in general.
23 We've done a trial calculation based on this diagram 24 for NSFR under the DOE program. The green color shows 25 primary barriers for release of radioactivity, 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
299 1 obviously in this case we are focusing on fuel in the 2 core. We consider the fuel matrix itself as a barrier 3 because depending on the accident sequence, it could 4 hold the solid fission products within the fuel and 5 then they're not--other than small chunks that could 6 get into coolant channels--may be stayed in the matrix.
7 And also depending on the accident scenario, for 8 example, a very long loss of heat sinks type of 9 accident versus a very rapid transient overpower, what 10 radionuclides are released at what rate will differ 11 significantly.
12 So I'm showing--the first step is retention 13 of the fuel, and then once you lose the integrity of 14 the cladding, depending on the accident, you will have 15 fission gas immediately released to the coolant, but 16 some particulates and even molten fuel could get 17 released if this a rapid transient leading to fuel 18 melting. There's a complex phenomena absorption, 19 condensation, dissolution, retention in the fuel, but 20 probably more a more important unknown, which I will 21 highlight later on is along with the fission gas 22 release, some of the solid chunk fuels could be 23 scrubbed to the cover gas space with the bubbles. And 24 if the cover gas space between the sodium pool and the 25 cover gas interface, there would be vaporization, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
300 1 condensation, re-vaporization, and some vapors could 2 nucleate on, and--or condense on the particles.
3 Typically, there would be a design leakage rate from 4 cover gas space to the containment space, and then a 5 similar phenomena takes place inside the containment, 6 and then leakage from the containment. Those are sort 7 of release pathways for a typical accident--multiple-8 failure accident that could lead to a large-scale fuel 9 failures in NSFR.
10 An approach--the mechanistic source term 11 approach for liquid metal cooled reactors is therefore 12 it will involve first, inventory analysis, and then 13 transient scenario modeling. We need to understand 14 in-pin radionuclide distribution before the failure or 15 at the time of the failure; radionuclides released from 16 the failed fuel, some of those released chunks of 17 molten fuel could be carried through the bubbles--
18 fission gas bubbles--to the cover gas space, and some 19 of them could be retained inside the liquid metal pool 20 and released to the cover gas space at the surface, the 21 free surface. We need to analyze the cover gas region 22 for radionuclide tracking, containment region, and 23 finally the off-site dispersion analysis. So, these 24 are the capabilities that could be leveraged to 25 perform--to implement such an 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
301 1 An inventory analysis could be done with 2 ORIGEN, or in a radionuclide group basis using REBUS, 3 for fast spectrum reactors. Transient scenario 4 modeling could be done with SAS4A traditionally, but we 5 can also leverage our newest capabilities, SAM and 6 PERSENT to help with this process. Interim 7 radionuclide distribution are typically handled with 8 the fuel behavior codes, if you have such capabilities 9 in legacy codes, like METAL and SAS4A, but BISON, if 10 Rich gets his way, we'll be ready to provide such 11 capability in a short few months. Rich is shaking his 12 head. And for radionuclides released from the fuel, we 13 do not have a currently existing capability but we're 14 developing a module to do that type of job coupled with 15 SAS4A and SAM within the context of some international 16 collaborations; I have a slide on that.
17 And then for--really, once we know what 18 gets to the cover gas space, what goes to the 19 environment, then those implications could all can be 20 taken care of on the NRC codes like MELCOR. CONTAIN-21 LMR used to do a specific sodium-fire type of modeling 22 for liquid metal coolant and sodium coolant, but I 23 think with DOE support, CONTAIN-LMR capabilities are 24 now incorporated into MELCOR so we won't even need 25 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
302 1 MEMBER REMPE: Yes, it's--is it--it's not 2 NRC supported anymore is it?
3 MR. SOFU: CONTAIN-LMR, probably not, yes.
4 MEMBER REMPE: So--but again, you're saying 5 now we're going to put it in MELCOR?
6 MR. SOFU: It is already in MELCOR.
7 MEMBER REMPE: So--yes, okay.
8 MR. SOFU: So, as you see that I don't have 9 green boxes showing a radionuclide bubble transport and 10 liquid-metal radionuclide release; those are identified 11 as gaps. We did perform---
12 MEMBER REMPE: I'm sorry, if you're still 13 on, go ahead.
14 MR. SOFU: We did perform a trial 15 calculation to find out what the status of theses codes 16 for a simple application. We again focused on AFR-100 17 design, and performed the trial calculation. The 18 report is publicly available; there's a link here that 19 you can access. You can see the two scenarios, one is 20 a protected loss of plus, loss of flow, plus in the 21 sense that plus is very degraded decay integral 22 capacity, otherwise you won't get fuel failures with 23 those accidents; and the other one was an unprotected 24 control rod withdrawal leading to a quick rise in the 25 fuel temperatures and leading to fuel melting.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
303 1 So the consequences of these two accidents-2 -duration and consequences are quite different. What 3 we found out at the end--what we are missing is, as I 4 highlighted before, a pool bypass for bubble transport, 5 and fuel release fractions, which is the radionuclide 6 release module that I mentioned in the previous slide.
7 Professor Corradini?
8 CHAIRMAN CORRADINI: No, I was just going 9 say I do these as--well, you said examples but, one 10 would have to know the--estimate the likelihood of 11 these scenarios coupled with what would be the release 12 and look for the--I don't want to have to say the worst 13 combination--but the limiting combinations. So, these 14 are just examples; there could be a wide range.
15 MR. SOFU: Examples, and I think the sub-16 bullet here that I included based on Richard Lee's 17 comment on my slides, these sequences---
18 CHAIRMAN CORRADINI: He had a comment? I'm 19 shocked.
20 MR. SOFU: He did send a comment. So these 21 sequences are typically, normally for a design selected 22 based on a--using PRA.
23 CHAIRMAN CORRADINI: Okay, that's okay.
24 MR. SOFU: But we really didn't do PRA in 25 this case.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
304 1 CHAIRMAN CORRADINI: I was looking at your 2 slide four and I guess my interpreting of slide four as 3 you were walking down the path is you'd have to look at 4 the ensemble of things that could go wrong and their 5 ultimate result and look at that--what are the dominant 6 ones that you then would have to look is. Okay, 7 because that's what they did for the BWR and the PWR to 8 result in 1465.
9 MR. SOFU: Absolutely. Definitely.
10 CHAIRMAN CORRADINI: Okay.
11 MR. SOFU: I think you're absolutely right 12 about that.
13 MEMBER REMPE: Your trial calculation, did 14 it include using MELCOR or CONTAIN or something?
15 MR. SOFU: Yes. Trial calculation, I did--
16 I believe it did use CONTAIN--I'm sorry--MELCOR.
17 MEMBER REMPE: Okay. And it used BISON or 18 some other -- it used, like, all these other codes that 19 was on this slide earlier?
20 MR. SOFU: Trial calculation used, I believe 21 ORIGEN, SAS4A. This module is developed as part of 22 that trial calculation and MELCOR. So, for these 23 missing steps, we just assume, whatever released from 24 the field somehow reached to the cover gas.
25 So, that would be a very conservative NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
305 1 approach, because I think you would have significant 2 retention of some radionuclides in the sodium coolant 3 that we didn't take advantage of.
4 MEMBER REMPE: Was it a linked calculation, 5 where you -- or did you just get output and put it to 6 the next --
7 MR. SOFU: Pretty much, those were 8 sequential calculations. You would get the results 9 from one analysis and then, perform assessments with 10 the other codes.
11 So, these were -- and I think, at some 12 point, when you reach to this particular step, it more 13 like a spreadsheet type calculation, goes into --
14 connects to MELCOR and input, MELCOR input deck.
15 MEMBER REMPE: At this point, is this a 16 better approach than just trying to put some models and 17 have a sodium reactor MELCOR? I mean, what's the 18 benefit?
19 MR. SOFU: This question has come up during 20 our discussions with NRC and also Sandia. My 21 suggestion is trying to develop all these capabilities 22 somehow capture in MELCOR would be reinventing the 23 NEAMS program all over again.
24 It would be, in theory -- and I think one 25 suggestion, I don't remember, I think, who came 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
306 1 Argonne with, I think it was Randy. Yes. So, Randy's 2 suggestion was, can we have surrogate models that 3 capture the behavior of a particular reactor type, 4 based on running these models and understanding their 5 consequences?
6 That's certainly a possibility, but it 7 would be far from being mechanistic or general enough 8 that every time you change your reactor design or have 9 a different reactor type, molten salt, so that you need 10 to continue developing surrogate models within MELCOR 11 to achieve that goal.
12 MEMBER REMPE: I'm just remembering the old 13 days, the RELAP and things like that, and people said, 14 can't we get the important things, because you need to 15 do multiple sequences and --
16 MR. SOFU: Yes. But for light water 17 reactors, MELCOR actually does extend to scenario of 18 elevation phase. It has those capabilities. But 19 they're really very LOCA and light water reactor 20 hardwired approaches.
21 MEMBER REMPE: I thought they actually -- I 22 don't know if they've done anything for the sodium 23 reactor, but for the gas reactor, they did and try and 24 do something with PARFUME a few years ago, for the --
25 MR. SOFU: 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
307 1 MEMBER REMPE: -- NGNP stuff and so, I 2 thought they had tried to do that already. And --
3 MR. SOFU: Yes.
4 MEMBER REMPE: -- again, you've got 5 something that's actually interacting, instead of this 6 sequential thing, which might --
7 MR. SOFU: Right.
8 MEMBER REMPE: -- be more expensive to run.
9 I don't know, it's just a thought.
10 MR. SOFU: I think my takeaway from this 11 particular slide is the MELCOR has well proven 12 radionuclide tracking capabilities, within cover gas 13 and containment space.
14 If somehow, provide that link, what 15 radionuclides are reaching to that interface, cover gas 16 and containment space, MELCOR would be the best tool to 17 understand the consequences of accidents.
18 And we have the capabilities to feed that 19 information into MELCOR. I understand strong desire to 20 do everything with one code, but I would -- my personal 21 thought is that this would be too big of an effort.
22 MEMBER REMPE: Okay.
23 MR. SOFU: All right. So, this trial MST 24 calculations, mechanistic source term calculations for 25 liquid metal reactors found a lot of attraction.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
308 1 We've used that capability in our 2 collaboration with GE-Hitachi, as an art program funded 3 project as part of their PRA modernization effort. We 4 repeated that for TerraPower's TWR design and also for 5 KAERI's PGSFR design.
6 We recently received a voucher from DOE to 7 work with Fauske and Associates, as well as 8 Westinghouse, to apply this capability to a lead-cooled 9 fast reactor concept. And that also supporting the 10 development of this radionuclide release module from 11 oxide fuel.
12 And finally, we have two recent awards, 13 NEUP awards to University of Wisconsin and New Mexico, 14 to do tests with sodium and liquid lead, to assess 15 radionuclide retention characteristics of those 16 coolants, that will provide really useful information, 17 useful data, for radionuclide release module.
18 MEMBER REMPE: Why the oxide fuel thing with 19 Fauske and Associates? My understanding is, the sodium 20 fast reactors are all going with metal. Is there a 21 vendor or a design also staying in oxide?
22 MR. SOFU: This is a lead-cooled reactor.
23 MEMBER REMPE: Oh, it's a lead, okay.
24 MR. SOFU: They are focusing on oxide fuel 25 initially. I think, eventually, they want to have NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
309 1 nitrite fuel, but --
2 MEMBER REMPE: Okay.
3 MEMBER KIRCHNER: So, you probably can't 4 answer my question. What are you seeing in terms of 5 results?
6 Getting down to the bottom line, someone, 7 someplace, has to make a decision on emergency planning 8 zones. They're not going to be able to look at the 9 infinite number of combinations of scenarios.
10 So, are you finding some commonality in 11 your work that would suggest that you're seeing versus 12 the EPA protective action guidelines or what's the 13 practice in LWR industry? Are you finding some 14 scaling, based on --
15 MR. SOFU: Yes. So --
16 MEMBER KIRCHNER: -- technology and choices 17 and power level, obviously, is a big factor?
18 MR. SOFU: Correct. So, what we are seeing 19 that, first, we need to push the envelope of those 20 accidents for liquid metal applications to a point 21 where we need to have some fuel failures.
22 It's hard to get to that point with a lot 23 of those concepts, because they're inherent in passive 24 safety characteristics, passive direct gain 25 characteristics. So, we have 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
310 1 MEMBER KIRCHNER: That has not been 2 demonstrated, that a postulate --
3 MR. SOFU: Postulated.
4 MEMBER KIRCHNER: -- filling in?
5 MR. SOFU: Correct.
6 CHAIRMAN CORRADINI: The postulate being, 7 what, Walt?
8 MEMBER KIRCHNER: That these reactors do not 9 --
10 MR. SOFU: But --
11 MEMBER KIRCHNER: -- the fuel rod.
12 MR. SOFU: Correct. But we have two 13 unprotected tests coming from EBR-II and FFTF that 14 demonstrates inherent safety principles and using 15 capabilities validated with these tests, for similar 16 designs.
17 We have fairly good confidence of inherent 18 and passive response of EBR-II and FFTF designs. So, 19 that is one that these designers are taking advantage 20 of.
21 So, just to jump to your question, however, 22 oftentimes, when we analyze, really, above and beyond 23 the spectrum of accidents, we probably would normally 24 fall under the residual risk category of events, but we 25 nevertheless have to go there, because you need to have NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
311 1 fuel failures to do the source term assessment.
2 Generally, retention in fuel, retention in 3 sodium, cover gas, and containment space is sufficient 4 to minimize the dose concerns well below the regulatory 5 limits. So, that could probably support reduction of 6 those emergency planning zones, consistent with that 7 measurement.
8 But again, in my approach that I said, let 9 me just jump there, the regulatory review is the final 10 step that we're, of course, we don't want to short-11 change.
12 It all boils down to how NRC would receive 13 these calculations. These trial calculations are, 14 essentially, studies that inform the design, not 15 necessarily intended to support the license 16 application.
17 From here on, I will go really fairly 18 quickly, because I'm going to repeat a similar pathway 19 for HTGRs. And this pathway, release pathway analysis, 20 was prepared as part of NGNP project. As you see, at 21 the core of it is a TRISO fuel, much, much, much larger 22 than what it is.
23 But also, it's part of a graphite block and 24 what's shown here is the helium pressure boundary and, 25 finally, around it, surrounding it, is another barrier, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
312 1 is reactor building.
2 And several phenomenon are highlighted 3 here, steam-induced vaporization, circulating activity 4 inside the -- from defective TRISO particles. Plate-5 out, lift-off, wash-off.
6 And if you have a breach, you can have 7 either the helium leaks or helium breaks, that could 8 lead to release of this high pressure through venting 9 to the environment.
10 So, multiple barriers for HTGRs, at the 11 center is the TRISO fuel, with multiple layers. And 12 then, fuel compacts and fuel elements, the graphite 13 block is another barrier. And helium pressure 14 boundary, as long as you maintain it intact, 15 circulating activity is not necessarily a concern.
16 Finally, the reactor building, which is not 17 a traditional containment structure for modular HTGRs, 18 because if you do lose the pressure in the helium 19 pressure system, then I think it vents the helium 20 first, outside to the environment.
21 But once that initial puff is released, you 22 have the ability to seal this reactor building, to 23 retain the fission products, following that initial 24 phase of an accident.
25 So, during normal operation, relatively 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
313 1 inventory of fission products inside the helium 2 pressure boundary is expected from defective fuel 3 elements.
4 And limiting event is considered to be the 5 loss of helium pressure boundary integrity, leading to 6 a slow or sudden pressure loss and larger delayed 7 release of fission products from the fuel at elevated 8 temperatures, if, indeed, temperatures get elevated.
9 As indicated earlier, a lot of the 10 statistical analysis really don't just assume, based on 11 tests performed, I think they just assume at a certain 12 temperature, certain failure rates.
13 So, those releases from failed fuels at 14 certain temperatures is based on those correlations.
15 It doesn't have a lot of BISON modeling there, I 16 believe.
17 So, the functional containment concept 18 introduced in NRC's Reg Guide 1.232 and in the 19 criterial of MHTGR 16, it allows taking credit for 20 coated fuel particles as the primary barrier.
21 Therefore, the reactor building is not leak-tight and 22 therefore, it's not a conventional containment 23 structure.
24 And here's the MST Approach, very similar 25 steps. Inventory analysis, transient scenario NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
314 1 modeling, fuel response to scenario studied, 2 radionuclide release rates from the fuel, and then, the 3 helium pressure boundary radionuclide release, reactor 4 building analysis, and offsite dispersion analysis.
5 Codes are more or less the same. Again, 6 inventory analysis could be done with well-established 7 capabilities in ORIGEN. And in NGNP project, I believe 8 RELAP was used as the scenario, transient scenario 9 modeling. But capabilities of PRONGHORN could be 10 leveraged to make that assessment.
11 Certainly, BISON and, definitely, PARFUME 12 models will give us the fuel response, as well as 13 radionuclide release rates from the fuel at elevated 14 temperatures. I put this here, I think a lot of that 15 release rates is going to be coming from test data.
16 CHAIRMAN CORRADINI: I was going to ask 17 about that line, because I thought, I'm going to pick 18 on Rich, what I thought Rich said was that BISON is not 19 in a position to do that line.
20 That there would have to be something, 21 PARFUME, modified PARFUME, PARFUME in MELCOR, 22 something. Am I misremembering?
23 MR. WILLIAMSON: So, if I understood what 24 you just said correctly --
25 CHAIRMAN CORRADINI: The line that says NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
315 1 radionuclide release rates from fuel, under, I'll call 2 it, beyond design-basis conditions, I thought BISON 3 wasn't in that position, yet.
4 MR. SOFU: Not yet.
5 CHAIRMAN CORRADINI: At all.
6 MR. WILLIAMSON: Well, yes, I guess I'm 7 hesitating a little bit because we certainly, for a 8 given particle, can -- in fact, the prong that I showed 9 that had a cesium release, we certainly can and have 10 already demonstrated the capability to predict release 11 from an individual particle for a specific species.
12 So, we're in a position to do that, to do 13 that for a host of radionuclides. Statistically, we 14 haven't done that yet.
15 CHAIRMAN CORRADINI: Okay. But -- can you 16 go back?
17 MR. SOFU: Sure.
18 CHAIRMAN CORRADINI: So, well, okay. So, 19 let me ask the question a little differently. So, that 20 middle line is not just from the fuel, but everything 21 that kind of got dusted up in operating the reactor 22 inside the primary system, as well as what would be 23 released from the fuel, yes?
24 Because if I have any sort of transient 25 response that I have to blow down, I have to know any NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
316 1 sort of material that got accumulated from its 2 operation. And that's not in any of these, as I 3 understand it.
4 MR. SOFU: As a scenario modeling?
5 CHAIRMAN CORRADINI: But that's the flow, 6 that's not the fission product deposition that was 7 there because of --
8 MEMBER REMPE: Lift-off.
9 CHAIRMAN CORRADINI: -- because of lift-off.
10 Dust, I've got a gas reactor --
11 MR. SOFU: Oh, yes, I see --
12 CHAIRMAN CORRADINI: -- with a bunch --
13 MR. SOFU: -- what you mean, yes.
14 CHAIRMAN CORRADINI: -- of dust. And now, 15 I punch a hole in the gas reactor and the dust comes 16 out. So, I have to know, what are the radionuclides in 17 the dust. That's what I guess I'm getting at.
18 MR. SOFU: I kind of take comfort in the 19 fact that these calculations were performed for the 20 modular HTGR as part the NGNP project. There are --
21 CHAIRMAN CORRADINI: Yes, but --
22 MR. SOFU: -- two reports here.
23 CHAIRMAN CORRADINI: But you showed me these 24 two reports, but these are the ones that I thought were 25 more process than calculational. Am I misremembering NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
317 1 which ones we were sent?
2 MEMBER REMPE: No, you're correct. They did 3 peer -- or they had an expert panel on the gas reactor 4 one, to evaluate where the uncertainties were. And it 5 wasn't a calculation.
6 MR. SOFU: No, no, I mean, obviously, this 7 wasn't a kind of license application in that sense, but 8 --
9 CHAIRMAN CORRADINI: No, no, I understand.
10 MR. SOFU: But it --
11 CHAIRMAN CORRADINI: But I --
12 MR. SOFU: -- did identify --
13 CHAIRMAN CORRADINI: -- didn't disagree with 14 the process --
15 MR. SOFU: Yes.
16 CHAIRMAN CORRADINI: -- all I guess I was 17 getting at was, is that there are pieces in your 18 listing here that go beyond what --
19 MR. SOFU: Yes.
20 CHAIRMAN CORRADINI: -- the red names are.
21 MR. SOFU: I agree with you --
22 CHAIRMAN CORRADINI: Okay.
23 MR. SOFU: -- the steps could be rearranged 24 and expanded, it's just essentially -- the purpose of 25 me drafting this, I apologize if it doesn't fit your NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
318 1 vision, but, essentially, leverage some codes that 2 could fulfill certain roles, identify the codes, most 3 of them.
4 CHAIRMAN CORRADINI: No, no, it's fine, I'm 5 not worried about that, I was just trying to make my 6 point. So, if you went to, let's pick on somebody, if 7 you went to X Energy --
8 MR. SOFU: Yes.
9 CHAIRMAN CORRADINI: -- what are -- forget 10 about what the codes are, do they have the same process 11 path? In other words, how are they going to make their 12 case, if they were to come into the regulator?
13 MR. SOFU: So, I'm assuming, in their case, 14 the limiting scenario would be losing the helium 15 pressure. And at that point, they will have a 16 circulating activity already and that could be released 17 to environment directly, because they may not have a 18 containment structure that will hold that pressure.
19 So, with that accident, I think you will 20 immediately have some dose consequences of, offsite 21 dose consequences associated with helium puff, with 22 circulating activity reaching to the environment.
23 Detectors will sound and --
24 CHAIRMAN CORRADINI: Right, but --
25 MR. SOFU: -- you'll have 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
319 1 CHAIRMAN CORRADINI: -- you're going much 2 further than I am, so that's good. So, now, let me 3 push the point. Has the industry shared with you guys 4 their approach to any of this?
5 MR. SOFU: No.
6 CHAIRMAN CORRADINI: Okay.
7 MR. SOFU: Those are sort of DOE proposals, 8 based on some trial calculations are done for SFRs, and 9 what NGNP did for MHTGRs.
10 MEMBER BALLINGER: But you had to previously 11 -- the blue TRISO fuel QA, in order for you to claim 12 that that's the primary barrier and that you don't need 13 a conventional containment, you have to have previously 14 -- you have to demonstrate that your coated particles 15 meet a certain QA spec --
16 MR. SOFU: Correct.
17 MEMBER BALLINGER: -- ahead of time.
18 MR. SOFU: Correct.
19 MEMBER BALLINGER: So, that is up top.
20 MEMBER KIRCHNER: It would seem to me, 21 pragmatically, for the case of the HTGR, which is 22 rather unique, because you're going to have this 23 problem of blowing down the primary system and the 24 planned buildings for the structure can't withstand 25 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
320 1 CHAIRMAN CORRADINI: Well, they --
2 MEMBER KIRCHNER: So, they're going to vent.
3 CHAIRMAN CORRADINI: They could, but they're 4 --
5 MEMBER KIRCHNER: They could, but --
6 CHAIRMAN CORRADINI: -- they've designed --
7 MEMBER KIRCHNER: -- the cost is --
8 CHAIRMAN CORRADINI: -- it not to.
9 MEMBER KIRCHNER: No, the cost would be 10 prohibitive. So, or it could be prohibitive. So, the 11 approach currently is to vent. So, it seems to me, 12 yes, you could do all this detailed analysis to try and 13 calculate what the circulating inventory is and what 14 the plate-out is and so on.
15 Pragmatically, wouldn't you just define a 16 tech spec for what you can withstand in terms of that 17 circulating and deposited inventory?
18 CHAIRMAN CORRADINI: And demand --
19 MEMBER KIRCHNER: Ensure that the fuel is 20 manufactured so that you, based on your analysis, with 21 your detail models, you're not going to have a problem.
22 And then, from that, figure out what the offsite dose 23 is. And then, that's going to tell you what your EPZ 24 diameter is, right?
25 CHAIRMAN CORRADINI: So, your point is, work NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
321 1 the problem backwards?
2 MEMBER KIRCHNER: Work the problem 3 backwards, yes.
4 MEMBER REMPE: You need to consider a 5 spectrum.
6 MEMBER KIRCHNER: And I would think the same 7 with the LMR scenarios as well. Otherwise, to go from 8 alpha to omega, from the very beginning with every 9 single transient that you have in your PRA space, even 10 if the codes are running very efficient, it's an 11 enormous undertaking.
12 MEMBER REMPE: But don't you have --
13 MEMBER KIRCHNER: I think one is 14 underestimating where you should put the effort and 15 where mechanistic approaches are most valuable.
16 My own biases, that your codes for the 17 detailed analysis, at a smaller component level, are 18 probably much better V&V'ed than MELCOR, at the very 19 macroscopic level, where you're actually then worried 20 about release.
21 So, I would divide the labor in a way that 22 I would bound what I start with in estimating the 23 releases. And then, I would inform the design of the 24 reactors with your detailed modeling capability, to 25 convince yourself you have adequate margin.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
322 1 MR. SOFU: That is --
2 MEMBER BALLINGER: And hope --
3 MEMBER KIRCHNER: Not to get there.
4 MEMBER BALLINGER: And hope that the QA 5 requirements don't strangle you.
6 MEMBER KIRCHNER: Well, that may be unique 7 to the particle fuel.
8 MR. SOFU: I agree, that could certainly be 9 an approach, a valid one as well. But normally, we 10 don't necessarily analyze every single sequence. But 11 what you do, in LMR case for example, I analyze the 12 accidents, which you need to analyze to understand 13 their consequences.
14 MEMBER KIRCHNER: Sure.
15 MR. SOFU: You would very quickly identify 16 the bounding ones for which you would then proceed to 17 --
18 MEMBER KIRCHNER: Proceed further.
19 MR. SOFU: -- source code evaluation.
20 MEMBER KIRCHNER: Okay.
21 MR. SOFU: You have to look at the spectrum 22 for Chapter 15 purposes anyway. And in Chapter 15, if 23 you run into sequences where, with uncertainties 24 included, you will expect some fairly sizable fuel 25 failures. That's the one to look at, not necessarily NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
323 1 the whole spectrum.
2 MEMBER REMPE: Because, in your vision, 3 Walt, what about air ingress and water ingress --
4 MEMBER KIRCHNER: Well, that's another --
5 MEMBER REMPE: -- events?
6 MEMBER KIRCHNER: -- class that they would 7 have to analyze --
8 MEMBER REMPE: Yes, you've got --
9 MEMBER KIRCHNER: -- for HTGR.
10 MEMBER REMPE: -- to be able to --
11 MR. SOFU: Absolutely.
12 MEMBER REMPE: -- rule it out and I --
13 MR. SOFU: Absolutely.
14 MEMBER KIRCHNER: Yes, that's --
15 MEMBER REMPE: -- yes.
16 MEMBER KIRCHNER: The initial puff may not 17 be the more demanding problem for the HTGR.
18 MR. SOFU: So, I hope you recognize that, 19 with these questions, you're putting me in a spot where 20 I'm trying to find solutions for a specific reactor 21 type or specific company.
22 Those are just initial puff shows a pathway 23 and code capabilities that could support a real 24 application. I would have loved to be part of a 25 project like that, but this isn't my role under Shane'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
324 1 program.
2 And finally, for molten salt reactors, we 3 don't have a diagram similar, a pathway phenomena 4 diagram for MSRs. They also come with a greater 5 variety of design choices.
6 They could have solid fuel, they could have 7 dissolved fuel, they could be fast spectrum, they could 8 be thermal spectrum. So, it is anticipated that the 9 functional containment concept can also apply to MSRs 10 with dissolved fuel, as well as the TRISO fuel.
11 So, owing to high fission product retention 12 capacity of molten salt, source term may be less of a 13 concern from fuel dissolved in the coolant.
14 That's kind of counterintuitive to folks 15 who are not really immediately familiar with the 16 technology, because they consider that, well, if you're 17 worried about core melt, then here you are, you have 18 already molten fuel. But I think the fission product 19 retention capacity of the salts is significant.
20 More of a concern could come from effluence 21 of the salt chemical processing system, maybe. And 22 also, maybe the tritium generated in the core, which is 23 a kind of very elusive species, as we all know.
24 I kind of drafted this proposed mechanistic 25 source term approach for MSRs, not that I know much NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
325 1 about it, but I think I can safely say that, what's 2 immediately missing for such a source term assessment 3 would be molten salt chemistry modeling, for which, 4 even under NEAMS sphere, we don't have a whole lot to 5 offer, other than plans, currently. And also, 6 radionuclide release rates from salts, to the cover gas 7 space and such.
8 But still, we can leverage scenario 9 evaluation phase of it, because like SAM and Nek, and 10 these two codes are being already utilized to some 11 molten salt vendors, adopted as part of their design 12 process.
13 And again, another takeaway from my 14 presentation is just dependable reliance on use of 15 radionuclide tracking capabilities of MELCOR and 16 offsite dispersion analysis with codes like RASCAL and 17 WinMACCS.
18 Those are just samples, there are 19 alternatives available, but they would do the job under 20 it.
21 That's my conclusions. I know that I'm 22 running out of time, let me see if -- the takeaway from 23 this slide, perhaps, there are some gaps in our trial 24 calculations identified.
25 One of them is, radionuclide release rates NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
326 1 from the failed fuel and this bubble transport of some 2 solid and liquid phases to the cover gas space. And 3 chemistry modeling and retention in molten salt for 4 molten salt concepts and FHRs.
5 And we do believe that the emerging DOE 6 ModSim capabilities you've heard today in codes like 7 SAM, BISON, PRONGHORN, and Nek5000, in combination with 8 the radionuclide tracking capabilities of MELCOR, can 9 be leveraged to remove the empiricism embedded in 10 traditional codes, MAP, MELCOR, for advanced reactors.
11 CHAIRMAN CORRADINI: Okay.
12 MEMBER REMPE: So, as you rethink this or as 13 you continue to think along this, like, this last 14 bullet talking about you're going to support a Level 15 2/3 PRA --
16 MR. SOFU: Yes.
17 MEMBER REMPE: -- you've got to do 18 uncertainties. And so, if you do this sequential 19 thing, that's going to be really a pain to deal with, 20 if the codes aren't linked, with, like, MELCOR.
21 MR. SOFU: Correct. But I think, as I 22 answered to Walter, Dr. Kirchner, that you don't 23 necessarily follow the full source term assessments 24 text for every single accident in your PRA tree.
25 You identify the bounding events 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
327 1 follow the path, complete the source term assessments 2 for the rest of them, I think.
3 For concepts like LMRs, that's an easier 4 ordeal, because not all accidents will lead to fuel 5 failures and you don't need to do source term 6 assessment for cases where you don't have a fail 7 failures.
8 For an HTGR, this would be a bigger task to 9 tackle, because you may actually have statistical 10 releases from gazillions of TRISO particles. Even for 11 DBAs, and sometimes AOOs, I mean --
12 MEMBER KIRCHNER: Even original startup.
13 MR. SOFU: Exactly.
14 MEMBER KIRCHNER: You always have --
15 MR. SOFU: That's true. So, for that, I 16 think it's -- you're right, that's a bigger challenge 17 to do the whole sequence for every single thing you can 18 think of for an HTGR.
19 But nevertheless, for HTGR, during NGNP 20 project, DOE's approach was complete reliance on 21 mechanistic source term assessments for AOOs, DBAs, and 22 BDBAs.
23 MEMBER REMPE: It's just something to think 24 about, and I'm not sure what you'd do for the molten 25 salt 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
328 1 MR. SOFU: Yes. Correct.
2 CHAIRMAN CORRADINI: Other questions? Okay.
3 Chris, you're our cleanup.
4 MR. STANEK: Okay. Well, let me start by 5 thanking everyone for their attention today, their 6 feedback, engagement. That feedback and engagement, 7 for us, is extremely useful, as we continue to slowly 8 evaluate our code development priorities.
9 Let me maybe start by making an off-the-10 cuff observation from the day, which is that I think 11 today was semi-painful, but perhaps necessarily so, as 12 a first step, in terms of what we hope is an ongoing 13 discussion.
14 But we thought it was necessary to provide 15 a 30,000-foot view of all of the codes that are under 16 development. And so, our approach today was really to 17 present a pure informational meeting, at least that's 18 how we interpreted the guidance, maybe inaccurately so, 19 but that's how we went about today.
20 And so, what that meant was that we 21 presented the DOE codes in something like a vacuum. We 22 didn't talk about how they compared to other codes, how 23 they might interface with other codes, and we didn't 24 talk about how users are using the codes or might be 25 using the codes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
329 1 And I don't want to be presumptuous, but 2 I'm probably not being presumptuous, because one of 3 Chairman Corradini's first questions, but my assumption 4 is that our approach has left you wanting some of those 5 examples.
6 And so, now that we have all this 7 sufficient, let's say, background information on the 8 table, I think hopefully we've successfully made you 9 conversant in DOE codes, that the code names now, when 10 we say them or someone else says them, that you 11 understand what those codes are and what they can and 12 can't do.
13 Now, with that background information and 14 now that we've gotten through that, as an idea, if 15 there was to be a next briefing, perhaps it would be 16 valuable to focus on some examples, let's call them use 17 cases, of where the codes are being used or potentially 18 being used.
19 My thought here is that the timing of such 20 a next meeting, again, I'm being presumptuous, but the 21 timing might be opportune. We're, especially in the 22 advanced reactor part of this, we are working closely 23 with vendors who are having their own conversations 24 with the NRC, and so, they don't want us to be out in 25 front of their conversations with the NRC.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
330 1 But as they begin to talk to the regulator, 2 we can more naturally have conversations using very 3 specific examples that I think all of us are somewhat 4 frustrated up here that we're not able to talk about in 5 great detail, but perhaps going forward, we can begin 6 to do that.
7 In terms of a summary, a prepared summary, 8 all we wanted to do was to very quickly ascribe a 9 notional maturity to the codes you saw today. This is 10 what we interpreted as the formal request from the 11 ACRS.
12 And the maturity level of each code was 13 discussed in some detail in the previous presentations, 14 but here, we refer back to those tables we started the 15 day with, where we've tried to distill the information 16 that was presented in the presentations that you heard, 17 to give you a sense of a notional or a relative 18 maturity level of each of these codes for a specific 19 application.
20 And so, here, this is a non-rigorous way 21 of doing it, but hopefully, in an attempt to distill it 22 to a meaningful sort of way, something that leaves you 23 with a sense of where at least we think things are.
24 And so, where we've color-coded the code 25 name in green, that means that a relatively mature NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
331 1 capability exists. It doesn't mean a complete 2 capability, and the validation might be limited, but we 3 think those codes are applicable at this time for those 4 specific applications.
5 Where the code names are in yellow, there, 6 we have a basic capability, but there's some key models 7 that require additional development.
8 And finally, for codes that are in red, the 9 code is still conceptual or in its initial formative 10 phases.
11 So, what you can see from the ATF set of 12 codes, by and large, we feel that the codes that are 13 being developed are fairly mature and applicable to 14 accident tolerant fuel.
15 This is my last slide, but for non-LWR 16 reactors, the DOE codes are, let's say, reasonably less 17 mature. Reasonably meaning that we haven't spent as 18 much time working on them.
19 I think we're accelerating our maturity 20 level quickly, but compared to the maturity level for 21 the accident tolerant fuel, for the advanced reactors, 22 as you heard today, we have capabilities that are 23 applicable to each of the advanced reactor designs.
24 In each of those areas, we have interested 25 industry members, but we are well aware 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
332 1 necessary development that needs to happen and we're 2 working hard on that.
3 And you also heard, in the previous 4 presentations, mention of a chemistry code that we're 5 -- that's an urgent need that we are rapidly addressing 6 in real-time.
7 So, that was just intended to be a quick 8 snapshot of everything that you heard during the day 9 and, hopefully, sort of bubbles this up to a level that 10 is somehow digestible.
11 And I think any of us are happy to answer 12 any further questions. But in the case there are none, 13 again, I thank the ACRS for the opportunity today to 14 present this, and again, the feedback received is, for 15 us, extremely valuable. Thank you.
16 CHAIRMAN CORRADINI: Good. Well, thank you.
17 I think, we're going to go around, so here's the 18 process. We want to have public comments, first and 19 foremost. And then, we want to go around the table.
20 But I'm going to start off by thanking 21 Shane and all the myriad contractors that went through 22 this, because we started this by a phone call after I 23 got asked by the Chairman, so how are those DOE codes, 24 are they ready for prime time? My response was, we'll 25 check it out.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
333 1 So, I want to thank all of you for all the 2 effort you've put in, it was quite a lot.
3 Okay. So, why don't we first ask if there 4 are comments from the audience here, while we get the 5 outer phone line open.
6 MR. BROWN: Bridge open.
7 CHAIRMAN CORRADINI: Oh, the bridge is open?
8 Well, why don't we -- if the bridge is open, let's take 9 the public comments from the phone first, if I could do 10 that. So, hold on, everyone.
11 Is anybody out on the line that wants to 12 make a comment, please?
13 MR. WHITT: Jeff Whitt is on from Framatome, 14 just to make a brief comment, just express our support.
15 We've had an opportunity to work with both CASL and 16 NEAMS in the development of the codes and support it.
17 And I know that was a question that was 18 asked early this morning, there is an interest and a 19 desire for using of these tools for future design and 20 confirmatory work against license codes.
21 And I would say, generally, I would not 22 discount the future use of DOE codes in licensing, 23 where that seems to be the right, applicable approach 24 for some of the advanced fuels or some of the other 25 advanced concepts.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
334 1 So, we are appreciative of DOE coming and 2 making this presentation and letting us listen in and 3 be a part of it. So, we just want to express our 4 support and thank you.
5 CHAIRMAN CORRADINI: All right. Thank you.
6 Are there any other comments on the phone line? Okay.
7 Hearing none, could you close, put it on mute, please?
8 Close the line? Everett, you had a comment?
9 MR. REDMOND: Sure, thank you. Everett 10 Redmond, Nuclear Energy Institute. I also chair the 11 NEAMS Industry Council. Just wanted to give a little 12 bit of an overview there.
13 The Industry Council is comprised of the 14 chairs of the three technology working groups, molten 15 salt, high-temperature gas, and fast reactors, as well 16 as representatives from EPRI and some relationship to 17 CASL.
18 I want to thank also everybody here for 19 their interactions today and thank DOE for the work 20 putting together the presentations. There was a lot of 21 conversation about interacting and understanding who's 22 doing what with what codes.
23 Obviously, as Chris said, that's more for 24 the developers to outline that, but there will be some 25 interactions in the next few months, in terms of 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
335 1 staff from some of the developers.
2 But one thing I would highlight is that 3 NEAMS has done some training sessions in the past, with 4 some folks in industry, Molten Salt Reactor Working 5 Group, for example.
6 So, you're getting a lot of interest from 7 the industry in these codes. Exactly how much they 8 ultimately get used is up to the developers, but there 9 is great interest, I want to emphasize, on the part of 10 the community out there. Thank you.
11 CHAIRMAN CORRADINI: Thank you.
12 MR. LEE: Richard Lee from Research. Since 13 my name was invoked in his view, I wanted to tell you 14 the comments I get. First is that I said the NRC 15 source term releases to the containment and to the 16 environment.
17 So, when we do severe accident analysis or 18 source term, I do not use the FRAP code, because there 19 has no role whatsoever in my analysis. Okay? So, I 20 don't see why it should be linked to BISON whatsoever, 21 because it is not talking about steady state or has 22 anything to do with AOO or anything. So, that's one 23 thing.
24 And then, if you see the two sequence that 25 was cited, and my comment is that you really 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
336 1 have the PRA to tell us these are the risk significant 2 sequence.
3 And the same way we did when NRR asked us 4 to synthesize the high burnup fuel and the MOX fuel, 5 okay? Dana Powers went and looked at all the sequence 6 of PWR and BWR, to make sure we capture all the risk 7 significant sequence, before we synthesize analysis and 8 produce a revision to the NUREG-1465.
9 CHAIRMAN CORRADINI: Thank you. Other 10 comments? Okay. Let's go around the table. So, I 11 have two questions for the Committee. One is, are 12 there any lasting comments they want to make from 13 today's presentations?
14 And also, I'd like to take notes as to, 15 since this is an information meeting and we're clearly 16 going to need to have others, what direction would you 17 propose that we go for our next Subcommittee meeting, 18 which will also be information?
19 Because as Chris said, we really got a lot 20 of information in a relatively short amount of time 21 from a whole range of tools that DOE's developed for 22 both ATF and for advanced reactors. So, I'm look at 23 you, Ron.
24 So, question one is, any other comments 25 about today. Question two is, and now what?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
337 1 MEMBER BALLINGER: I'm sure everybody's 2 going --
3 CHAIRMAN CORRADINI: Green light.
4 (Laughter.)
5 MEMBER BALLINGER: I'm sure everybody's 6 going to say this, but I appreciate the update on 7 things. I'm familiar with some of these things, but 8 having it all in one place is a good idea. I'm 9 heartened to see that you've now expanded the animal 10 groups to include snakes, range animals, and now, fish.
11 (Laughter.)
12 MEMBER BALLINGER: I'm not sure what else 13 you can do, maybe some insects or something.
14 CHAIRMAN CORRADINI: Don't challenge him.
15 (Laughter.)
16 MEMBER BALLINGER: But with respect to going 17 forward, I think the idea of picking some really 18 important, what you consider important, of the 19 concepts, the ones that are the most likely to, with 20 divine intervention, I suppose, be going forward and do 21 some examples, like has been suggested.
22 MEMBER BLEY: Yes, sir, I do have a few 23 comments. First, same thing, it was a great day. It's 24 one of the better meetings I've sat through, I enjoyed 25 it, and the discussions. It's kind of wonderful 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
338 1 extend the state of the art and examine these 2 interesting phenomena.
3 Bottom line, for me, at the top is, I've 4 just heard bits and pieces about this whole program and 5 these large computer code developments, with some 6 worry. I gained some confidence today in hearing how 7 they're being used and still retaining the need for 8 data and experiments, that's helpful.
9 On the goal of this whole business, early 10 today, Chris pointed out that in response to 11 Congressional direction, which I guess is -- we don't 12 have any Fukushimas here, DOE issued this report on the 13 program back to Congress in 2015.
14 Early in that report, you say, with respect 15 to the goals, kind of two parts. One is, they can 16 tolerate loss of active coolant in the reactor core for 17 considerably longer time.
18 Indeed, if that ends up being true, that 19 would help us a lot, in the area of risk. I haven't 20 heard anything yet that hints at that and most of the 21 work hasn't been up at that end of the problem.
22 The kinds of things that contribute to 23 risk, though, you need minutes to hours to do anything 24 about. So, I haven't heard things that make me think 25 you're going to buy that kind of time.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
339 1 The other part is, while improving, you 2 want to do that while you're improving fuel performance 3 during normal operations, operational transience, as 4 well as design-basis accidents and beyond-design-basis.
5 Everything except beyond-design-basis, it 6 sounds like that's coming together pretty well. I 7 haven't heard much that makes me think you're covering 8 the beyond-design-basis events yet.
9 And on the mechanistic source terms, I 10 guess I'll just harp a little, and it seems like you're 11 on the right track here, getting the chemistry as right 12 as you can in that area, allowing for uncertainties, 13 because there are strange reactions going on very fast, 14 with daughter products that come and go, it's going to 15 be pretty messy.
16 So, being able to account for uncertainties 17 in that chemistry is going to be important, if people 18 are going to really buy into this and think it's 19 convincing.
20 Tanju said something I like to hear from 21 people working this area, is, every time he said 22 something, it was depending on the accident scenario, 23 barriers may or may not be useful. And that's crucial 24 and too often, people don't recognize that. But I 25 guess that's what mechanistic source term is all about.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
340 1 I hope folks consider that there might be 2 other simplifications beyond what Walt suggested 3 earlier, that might be helpful.
4 In other areas, we found that you can 5 collapse many of the PRA scenarios into classes that 6 look the same for what's coming next. And in this 7 area, I hope that's possible. Usually, there are 8 simplifications like that, there can be. You have to 9 test those. But that's kind of the whole gamut.
10 Where we go next, anything you bring will 11 be interesting to me, but I think examples would be 12 useful. And I just picked up the papers you pointed us 13 to and I look forward to looking at those. So, thanks 14 a lot for today.
15 CHAIRMAN CORRADINI: Matt?
16 MEMBER SUNSERI: So, I can't express this 17 anymore eloquently than Dennis did, so I'll just leave 18 it as, I found that the strategy that you developed is 19 a lot more developed, or farther along than I had 20 anticipated and the progress you're making on maturing 21 these codes is -- my confidence level is well raised 22 also.
23 So, I'll just leave it at that. I thought 24 all the presentations were well done and the presenters 25 were well prepared.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
341 1 And since I'm the operational guy here at 2 the table and this is kind of outside my area of 3 expertise, I'll leave it to those more qualified than 4 me to judge where we should go next. Thank you.
5 MEMBER REMPE: So, I'd like to add my thanks 6 for coming here and discussing it. It helps. I guess, 7 I'm back to the question I raised at the beginning of 8 the day.
9 Yes, I understand you're enthusiastic about 10 your research and you believe you're going the right 11 way, but I -- no one has yet, I mean, I've heard, oh, 12 they're going to be coming in, and that will be 13 interesting to see, but no one is yet willing, in the 14 ATF program, to come in and say, I want to use one of 15 these for qualifying the fuel.
16 So, I'd like to see something like that.
17 And I actually think there's a lot to be learned from 18 such a thing like that.
19 And I understand, industry says, well, I 20 can do it cheaper if I use my own code, but maybe they 21 ought to be encouraged with some funding to do that and 22 get some confidence that these codes are better, if 23 that's -- we've spent a lot of years developing these 24 codes, so let's use it.
25 And so, again, it's beyond ACRS making such NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
342 1 recommendations, but it sure seems like that that's an 2 important thing to do.
3 With respect to NRC and the next step, I 4 actually think it would be good to have a Subcommittee 5 meeting where we hear from the staff, not only the 6 folks that are doing the accident tolerant fuel or the 7 advanced reactors, but also research, see some actual 8 comparisons to see, are these codes really better?
9 Again, when we asked about that earlier, I 10 believe we were told by the person from INL, well, my 11 job isn't to look if it's better, just it matches the 12 data, but I don't know if it's better than FAST is.
13 And so, I think we need to have some folks do that, to 14 get some confidence, too. And I guess I'll leave it at 15 that.
16 MEMBER MARCH-LEUBA: I thought that was 17 Mike's job.
18 (Laughter.)
19 MEMBER MARCH-LEUBA: Okay. Yes, well, I'm 20 very impressed by the DOE team. I think you guys have 21 done a fantastic job over the years, and today of 22 course. And I'm especially impressed by the tools that 23 we have seen. I mean, something like MPACT, it's a 24 dream calculation. I mean, I see it and I love it.
25 However, I'm concerned about the complexity NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
343 1 of the new tools and the user base that -- the DOE team 2 has been working on this for the last 12 years and you 3 know how to use it. But those tools are not for you, 4 are for everybody else.
5 And I suspect that the large vendors may be 6 able to take advantage of it, but I'm worried about the 7 small companies, which are the ones that really need 8 these tools.
9 The ones that are going to design the new 10 reactors, how are they going to be able to do something 11 like this, especially if it takes 10,000 cores to run 12 the calculation like this? I can't -- wait until we 13 are through and then -- I think that the process is, we 14 talk now.
15 MR. SOFU: I think, the 2,000 -- when you 16 are trying to develop capabilities that doesn't exist 17 elsewhere, it automatically pushes you to a particular 18 domain of complexity that you're trying to do things 19 that doesn't exist. If you repeat existing capability, 20 then you're not really adding any value.
21 But it's interesting that a lot of what we 22 developed under NEAMS program is adopted earlier by 23 small companies, as opposed to large ones that already 24 have their developed capabilities. They're not 25 interested in that. So, it's the other way around. 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
344 1 just wanted to make mention of that.
2 MEMBER MARCH-LEUBA: So, for the future, I'd 3 like to see, I mean, if you know of examples where 4 companies already taking advantage of this or have 5 plans for doing it, it would be fantastic if we could 6 have some of those examples, instead of going into the 7 nitty details of how you calculate the power on node 8 27.
9 And in that line, what I would like to see 10 is a plan of application of these methods for ATF.
11 Obviously, we haven't done it, but go through how would 12 I resolve the accident tolerant fuel with this method?
13 What would I do that would be useful and that would 14 save me money on the testing?
15 CHAIRMAN CORRADINI: Walt?
16 MEMBER KIRCHNER: Thank you all for the 17 presentations. For me, it's very enjoyable. I started 18 my career at Los Alamos doing advanced code 19 development, so it's, for someone like myself, it's 20 interesting to see how much progress has been made.
21 So, I congratulate you on that.
22 Since the slide is up, I would suggest, 23 you've heard some thoughts already on source term, I 24 think this is a difficult problem, from end to end, to 25 do the source term estimates.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
345 1 So, I would have in my back pocket or my 2 thinking or my strategy, whatever you want to call it, 3 a way to bound these problems. Dennis spoke eloquently 4 to certain aspects of this. The chemistry challenges 5 can't be underestimated, especially for some of the 6 more advanced concepts.
7 So, I would have concern there and I would 8 want to have Plan B, because you're not going to have 9 the level of maturity in the PAR space that we have in 10 the LWR, backed up by all their experience. Certainly, 11 not early on. And so, I guess, a caution there, is 12 what I would put out. And thank you, again, for the 13 presentation.
14 As to the future, yes, I'd be interested in 15 hearing more. I think, hearing more attacking specific 16 problems may be useful. And I know that perhaps 17 involves vendors, then, and actual designs, but that's 18 probably where the rubber hits the road, in terms of 19 applying these codes. And with that, thank you.
20 CHAIRMAN CORRADINI: So, I'll thank 21 everybody, but you've been thanked enough. So, let's 22 just move on from there. I guess, what took me by 23 surprise today was Chris's starting comment, which the 24 industry advisory group didn't want to let us know who 25 are potential adopters.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
346 1 So, that -- because one of my first 2 questions is, okay, who are your users? So, I've got 3 to come back to that question. So, if it's still of 4 the belief that there is a potential set of users that 5 don't want to be identified, I would think the next 6 step, for me, would be user needs.
7 I'd like to hear from industry, what are 8 the user needs? So, Everett said he's got his advisory 9 group, if they don't want to say, I'll use BISON, I 10 want to hear from them, I think I want to hear from 11 them, what do they need in a tool that gets them to the 12 end game?
13 And then, if we were to talk about it that 14 way, I would flip it and go to the NRC and say, if 15 they're the users, what do they need to get it to the 16 end game? That kind of goes a little bit with what I 17 think Jose was after, which is plan for application of 18 ATF.
19 I think the one thing that Walt mentioned, 20 and Dennis mentioned, which is, simpler is better.
21 Somehow, I would like to work the problem backwards, 22 what is the simplest way to identify the limiting 23 source term?
24 And if you can do that, that simplifies all 25 the preliminary analysis and gets me to that source NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
347 1 term, rather than complexity. Because these are very 2 interesting tools, but they're complicated. And I'm 3 wondering if the industry is hesitating because they 4 view them as complicated, or maybe they're not familiar 5 with them and, therefore, that's -- so, I think, for 6 me, the next step would be to try to hear what the 7 users need. Not necessarily who the users are, because 8 I sense you're not going to be allowed to trot them out 9 in front of us to write down and certify, but at least 10 to hear what they need. What are they looking for in 11 a tool that they can use for their safety analysis?
12 I guess, that's where I would like to go in 13 the future. So, maybe we can find a way to marry that 14 together with another Subcommittee meeting. I would 15 think, since this one took four months, the next one 16 might take a few months. So, why don't we start 17 talking about, between -- I originally called Tom and 18 Tom called Chris and Chris called, and so, we can talk 19 about it and see where we go from there, okay?
20 MR. STANEK: Sounds good.
21 CHAIRMAN CORRADINI: I don't have anything 22 else. Any other members have anything else?
23 Otherwise, we're adjourned. Thank you very much.
24 (Whereupon, the above-entitled matter went 25 off the record at 5:12 p.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
Overview and Introduction to DOE Briefing to ACRS:
ATF Session Advanced Computer Models for Reactor Safety Applications August 21, 2018
Overview
- Scope of briefing: The following presentations will describe adequacy and maturity of recently developed DOE modeling and simulation tools for application to ATF (morning) and non-LWR reactors (afternoon).
- Although many other codes exist, some of which address similar phenomena as DOE codes and some which that may interface with DOE codes, discussion pertaining to non-DOE codes is beyond the scope of todays presentations.
- In order to provide a comprehensive overview, description of each capability under development is necessarily kept brief.
- Deployment: Vision is for DOE to make its codes available to NRC and US companies, from which they may create proprietary versions of the codes using data they generate.
- Software quality and validation: DOE code development efforts place high level of importance on software quality assurance and validation.
- All codes adhere to strict SQA principles.
- DOE performing sufficient validation for there to be confidence in code use. Additional validation required by users for specific applications.
2 energy.gov/ne
Outline of ATF Presentations Doped- Coated FeCrAl SiC/SiC U3Si2 Non-cylindrical UO2 cladding cladding cladding metallic fuel Fuels MARMOT MARMOT MARMOT MARMOT MARMOT MARMOT BISON BISON BISON BISON BISON BISON T-H CTF, CTF, CTF, CTF, CTF, CTF, CFD CFD CFD CFD CFD CFD Neutronics Shift, Shift, Shift, Shift, Shift, Shift, MPACT MPACT MPACT MPACT MPACT MPACT 3 energy.gov/ne
Outline of ATF Presentations Doped- Coated FeCrAl SiC/SiC U3Si2 Non-cladding cladding cladding cylindrical UO2 metallic fuel
- 1. Steve Fuels MARMOT MARMOT MARMOT MARMOT MARMOT MARMOT BISON BISON BISON BISON BISON BISON Hayes T-H CTF, CTF, CTF, CTF, CTF, CTF,
- 2. Jess CFD CFD CFD CFD CFD CFD Gehin Neutronics Shift, Shift, Shift, Shift, Shift, Shift, MPACT MPACT MPACT MPACT MPACT MPACT 4 energy.gov/ne
Fuel Performance Modeling for DOE Briefing to ACRS:
Accident Tolerant Fuels Advanced Modeling & Simulation Tools for Accident Tolerant Fuels August 21, 2018
Outline of Presentation
- Accident Tolerant Fuels Development and Testing Background
- Multiscale, Mechanistic Modeling of Nuclear Fuels
- The Bison Fuel Performance Code
- Overview
- Verification
- Validation
- Model Enhancements for Accident Tolerant Fuels
- Doped UO2 Fuel
- Cr-Coated Zirconium Cladding
- FeCrAl Cladding
- SiC Cladding
- U3Si2 Fuel
- Non-cylindrical Metallic Fuel
- Validation for Accident Tolerant Fuels
- Summary and Conclusions 2 energy.gov/ne
Accident Tolerant Fuels Development and Testing 3 energy.gov/ne
Congressional Direction and Development Plan on ATF Following the accident at Fukushima, Congress directed the Department of Energy to begin developing fuels with March 11, 2011 enhanced accident tolerance that can be used in existing light water reactors.
The Development Plan:
- Defines the general attributes of accident tolerant fuels
- Lays out an aggressive 10-year schedule starting in 2012
- Establishes the goal of inserting lead fuel rods/assemblies in an operating commercial light water reactor by 2022 4 energy.gov/ne
Industry-led Development of ATF Concepts Framatome General Electric Westinghouse
- Cr-coated M5 - Iron-based cladding - Cr-coated Zirlo cladding - ODS variants for cladding
- Doped UO2 for improved strength - SiC cladding improved thermal - Alternative fuels conductivity and with improved performance thermal conductivity and high density DOE does not currently have a formal relationship with Lightbridge 5 energy.gov/ne
DOE Irradiation Testing Program to Support ATF Halden testing must be redirected Test Series ATF-1 ATF-2 ATF-3 ATF-H-x CM-ATF-x ATF-y Commercial Test Reactor ATR ATR TREAT Halden TREAT Reactors Test Type Drop-in Loop Static/Loop Loop LTR/LTA Loop Scoping Prototypic Cladding and Focused Focused Mature concepts Mature concepts Many Compositions Integral Fuel Test Strategy Concepts Off-normal Off-normal Nominal conditions Nominal conditions Nominal conditions Nominal conditions conditions conditions Fuel UO2*, U3Si2 Rodlets Zr w/coatings, Rods conditioned conditioned in Promising near-Fe-based alloys, Promising concepts Promising concepts in LTR/LTA Cladding ATF-1 and ATF-2 term concepts advanced alloys, irradiations irradiations SiC Fuel and fuel-Key Features cladding interactions PWR conditions Integral testing BWR conditions Prototypic testing Integral testing Timeframe FY15 - FY20+ FY18 - FY22+ FY19 - FY25+ FY19 - FY22+ FY19 - ? FY22 - ?
6 energy.gov/ne
Loss-of-Coolant-Accident Test Facility (ORNL)
- Internally pressurized, irradiated fuel rods
- Flowing steam environment
- Heating rate of 5°C/sec
- Temperature up to 1200°C
- Capable of water quench LOCA furnace during operation inside the hot cell Post-burst Zircaloy-4 tubes after LOCA sequence to 1200C with internal pressurization.
7 energy.gov/ne
Plans for Lead Test Rods/Assemblies in Commercial Reactors FY18
- GE: initiate testing of FeCrAl cladding (Hatch)
- Westinghouse: establish LFR fabrication line for U3Si2 fuel (INL)
- Framatome: perform pool-side exams of chromia-doped UO2 fuel (LaSalle)
Hatch FY19
- GE: initiate LTA testing of IronClad and ARMOR fueled rods (Clinton)
- Westinghouse: initiate LTR testing of Cr-coated Zirlo and U3Si2 fuel (Byron)
- Framatome: initiate testing of Cr-coated M5 cladding (Vogtle)
Byron 8 energy.gov/ne
DOEs Approach Multiscale, Mechanistic Modeling of Nuclear Fuels 9 energy.gov/ne
Multiscale, Mechanistic Modeling of Nuclear Fuels
- Objective: Use hierarchical, multiscale modeling for improved, mechanistic, and increasingly predictive models of fuel performance
- Mechanistic fuel behavior models: 1) minimize form errors, 2) provide insight were experimental data is sparse, and 3) may require less (or different) experimental data for validation Atomistic simulations Meso-scale models Fuel performance models Atomistically-informed parameters Degrees of freedom, operating conditions MARMOT Mesoscale-informed BISON materials models
- Identify important mechanisms
- Determine material
- Predict microstructure evolution parameter values
- Determine effect of evolution on
- Predict fuel material properties and fuel performance and behaviors failure probability 10 energy.gov/ne
The Bison Fuel Performance Code
- 1) Overview
- 2) Verification
- 3) Validation 11 energy.gov/ne
Fuel Performance Code
- Finite element-based engineering scale fuel 2D axisymmetric (or 1.5D) performance code
- Solves the fully-coupled thermomechanics and species diffusion equations in 1D, 2D axisymmetric or plane-strain, or full 3D
- Applicable to both steady-state and transient operations 2D plane strain
- Readily couples to lower length-scale material models
- Designed for efficient use on parallel computers
- Includes LOCA and RIA accident capability 3D
- Development follows NQA-1 process 12 energy.gov/ne
What Makes Bison Different?
3D/Arbitrary Geometry Coupling Fuel Types Parallel Computing 13 energy.gov/ne
Bison Code Verification
- MOOSE/Bison is supported by >2000 unit and regression tests
- All new code must be supported by verification testing; all tests must pass prior to code merge
- Regularly audited per NQA-1 standards
- Documentation:
- All tests distributed with source code
- Code verification process described in journal article 14 energy.gov/ne
Bison LWR Validation Status - 1/2
- Current assessment status
- ~75 integral, normal operation and ramp fuel rod experiments
- 47 LOCA cases (43 burst tests, 4 integral rods)
- 19 RIA cases
- Some vendors have https://bison.inl.gov/SiteAssets/BISON_assessment1.4.pdf performed additional validation w/proprietary data
- Documentation:
- Assessment report updated annually and distributed with code updates
- Accessible online
- User Manual
- Theory Manual
- Assessment Report 15 energy.gov/ne
Bison LWR Validation Status - 2/2 Fuel Temperature Fission Gas Release PCMI LOCA 16 energy.gov/ne
Model Enhancements for ATF 17 energy.gov/ne
Model Enhancements for ATF
- Bison is a state-of-the-art fuel performance code
- Increasingly incorporates new and improved mechanistic models for fuel behaviors
- Exceptionally well verified using modern methods
- Extensively validated for current LWR fuels
- Bison is a sound platform on which to implement enhancements to simulate Accident Tolerant Fuels
- $3M/year for 3 years (FY2015-2017)
- ATF modeling continues in mainline DOE programs
- Close relationship with DOE (ATF laboratory and industry teams) testing programs generating new performance data to be used for validation 18 energy.gov/ne
Approach to Developing ATF Performance Code
- 1) Introduce basic ATF properties/models in 3) Use sensitivity analyses on ATF material Bison to establish simulation capability properties/behavior models to prioritize
- 2) Introduce mechanistic models into Bison LLS mechanistic modeling activities from lower length-scale (LLS) activities as 4) On-going assessment/validation they become available 19 energy.gov/ne
ATF High Impact Problem (HIP) Highlights Coated Claddings FeCrAl Capability established in BISON to model Advanced mechanical models coatings, and several case studies being developed based upon examined. cluster dynamics-informed crystal plasticity models.
Burst model developed and implemented in BISON U3Si2 Non-stoichiometry, fission gas release, swelling, thermal conductivity Doped-UO2 Detailed description of dopant solution mechanism.
Impact of dopant on fission gas behavior.
20 energy.gov/ne
Overview/Status of ATF Models in Bison Complete Evaluation: Evaluation: Activities Planned for Set of Models Base Accidents FY19 Irradiation Doped UO2 Yes for FGR, Yes Yes Improved FG Fuel No for creep diffusivity for FGR Cr-coated Yes, but no No No Investigate Cladding irradiation mechanical effects effects of coating FeCrAl Yes Yes Yes Support user needs Cladding SiC/SiC In progress No No Complete evaluation Cladding of models U3Si2 Fuel Yes Yes Yes Support user needs Metallic Fuel Yes Yes No Support user needs 21 energy.gov/ne
Cr2O3 Doped UO2 Fuel: Model Summary Model in Experimental LLS- Documented Tested Bison Data informed Basic Yes Yes No Yes Yes thermal (UO2) properties Thermal Yes Yes Yes Yes Yes conductivity (UO2) degradation Basic No (UO2) Yes No Yes Yes mechanical properties Creep No (UO2) Yes No Yes Yes Swelling No (UO2) Yes No Yes Yes Fission gas Yes Yes Yes Yes Yes release 22 energy.gov/ne
Cr2O3 Doped UO2 Model Results Bison Simulation of Halden IFA-677 Rod 5 Fuel centerline temperature Fission gas release Rod 1 Rod 5 Fuel UO2+Add. UO2+Add.
Cr2O3 content (ppm) 900 500 Average grain diameter (µm) 56 45 23 energy.gov/ne
Chromium-coated Cladding: Model Summary Model in Experimental LLS- Documented Tested Bison Data informed Basic Yes Yes No Yes Yes thermal properties Basic Yes Yes No Yes Yes mechanical properties Creep Yes Yes No Yes Yes Oxidation Yes Yes No Yes Yes Note that no models incorporate irradiation effects.
Cr-coated claddings under irradiation in ATF-2 (ATR).
24 energy.gov/ne
Cladding Coating Mechanical Behavior
- Objective: Compare the No coating:
mechanical response of coated Displacement:
vs. non-coated Zircaloy cladding 3.74e-2 mm
- Cladding-only model with representative LWR temperature, pressure 0.02 mm coating:
- Chromium coating, 0.02 mm Displacement:
thick 3.10e-2 mm
- No creep model led to unrealistically high stresses
- FeCrAl coating 0.02 mm and 0.04 mm coating:
0.04 mm thick Displacement:
- A FeCrAl coating of 0.02 mm 2.69e-2 mm carries 1/10th of the load in the hoop direction Opportunity for use as a tool for optimization 25 energy.gov/ne
FeCrAl Cladding: Model Summary Model in Experimental LLS- Documented Tested Bison Data informed Basic Yes Yes No Yes Yes thermal properties Basic Yes Yes In Yes Yes mechanical progress properties Thermal Yes Yes In Yes Yes Creep progress Irradiation Yes No* No Yes Yes Creep Oxidation Yes Yes No Yes Yes Burst Yes Yes No Yes Yes
- FeCrAl cladding under irradiation in Hatch and ATF-1 (ATR).
26 energy.gov/ne
FeCrAl Burst Model Results
- Based experimental results, a criterion for the burst hoop stress was fit as a function of temperature.
Comparison of burst predictions using a variety of cladding thicknesses indicates that the burst behavior of FeCrAl will be similar to that of Zircaloy.
Massey et al., JNM 470, 2016 27 energy.gov/ne
U3Si2 Fuel: Model Summary Model in Experimental LLS- Documented Tested Bison Data informed Basic Yes Yes No Yes Yes thermal properties Thermal Yes No Yes Yes Yes conductivity degradation Basic Yes Yes No Yes Yes mechanical (constant values for properties elasticity)
Creep Yes Yes No Yes Yes Swelling Yes Yes (very little) Yes Yes Yes Fission gas Yes Yes (very little) Yes Yes Yes release 28 energy.gov/ne
U3Si2 Thermal Conductivity Model Results Unirradiated Degraded at Buav = ~5.8% FIMA Element averaged thermal conductivity at the midplane of Degraded thermal conductivity comparisons between a 10 pellet RZ-axisymmetric rodlet. U3Si2 and UO2 for a temperatures varying from 400 to 1100 K.
29 energy.gov/ne
SiC/SiC Cladding: Model Summary Model in Experimental LLS- Documented Tested Bison Data informed Basic Yes Yes No Yes Yes thermal properties Basic Yes Yes No Yes Yes mechanical properties Creep Yes Yes No Yes Yes Swelling In Yes No In progress In progress progress Oxidation Yes Yes No Yes Yes 30 energy.gov/ne
SiC/SiC Cladding Model Results
- Verification tests of the SiC models exist, but comparison with experimental results is pending.
- An early, simple benchmark problem organized by MIT demonstrates that Bison results are consistent with those of other codes.
Koroush Shirvan (MIT) 31 energy.gov/ne
Non-Cylindrical Metallic Fuel: Model Summary Model in Experimental LLS- Documented Tested Bison Data informed Thermal Yes* Yes In Yes Yes properties progress Cladding Yes* Yes No Yes Yes mechanical Fuel Yes* Yes No Yes Yes mechanical Fission gas Yes* Yes In Yes Yes behavior progress Radial pin No No - - -
power distribution 32 energy.gov/ne
Illustration of Simulation Capabilities Simulations Performed on Model Problem, not Lightbridge Design Bison-Nek5000 Coupled Simulation 33 energy.gov/ne
Validation for ATF 34 energy.gov/ne
Data Generation for ATF Validation
- Close partnership with industry ATF teams
- Exception: Lightbridge
- Heavy reliance on ATR and TREAT TREAT
- Halden tests must be redirected (ATR, BR-2, HFIR, MITR, TREAT)
- Priority: partner w/Halden staff on instrumentation implementation
- Expanded use of LTR/LTA programs will be required
- Quantitative PIE on important LTRs/LTAs
- Re-fabrication/instrumentation of LTR segments for U3Si2 fuel in Zirlo cladding subsequent ATR, TREAT, LOCA testing
- Fuel behavior data needed for ATF validation is being generated 35 energy.gov/ne
Validation Challenges for ATF
- There will be less experimental data (near-term) for validating ATF performance models/codes
- M&S of ATF not intended to replace experiments/data
- Still need experimental data that bound operations
- Requires close integration between modelers, experimenters, industry ATF teams, and regulator to maximize quality and applicability of data
- Multiscale, mechanistic approach to developing fuel behavior models (at level of microstructure)
- Separate effects testing can play a role in model validation
- Integral fuel rod tests still priority for Bison validation
- Benefits from microstructural characterization of . . .
- As-fabricated fuels The Irradiated Materials Characterization Laboratory
- Irradiated fuels (IMCL) is a new facility at INL for the microstructural characterization of irradiated fuels.
36 energy.gov/ne
Summary and Conclusions
- Bison (world-class, well-validated for LWR fuel applications)
- Multiscale modeling approach delivered demonstrated results for UO2
- Marmot w/atom. sims building mechanistic behavior models for ATF
- Opportunity for Accelerated Fuel Development and Qualification
- 1) Use of Bison by industry and/or NRC
- 2) Implementation of fuel behavior models in vendor and/or NRC codes
- 3) Use insights obtained from Bison and mechanistic fuel models to inform experimental programs and speed licensing Bison is available for use by industry and NRC 37 energy.gov/ne
Questions?
38 energy.gov/ne
Neutronics and Thermal- DOE Briefing to ACRS:
Hydraulics Modeling for Advanced Modeling & Simulation Tools for Accident Tolerant Fuels Accident Tolerant Fuels August 21, 2018
Presentation Outline
- Introduction
- Code Descriptions
- Code Validation
- Capabilities and Gaps for Accident Tolerant Fuels
- Conclusions 2 energy.gov/ne
Introduction
- DOE has developed a fully coupled and resolved multi-physics core simulator for LWRs - Virtual Environment for Reactor Applications (VERA):
- Neutronics (pin-by-pin fuel rod powers)
- Thermal-hydraulics (subchannel two-phase density and flow distributions)
- Fuel (temperature distributions)
- Chemistry (crud build-up)
- Detailed isotopic depletion
- Developed for applications to currently operating plants
- Emphasis on zirconium-alloy clad UO2 in PWRs
- For operational performance and safety analysis
- Applied and compared to operational data from many plants
- Strong industry engagement in development and application of capabilities, including evaluation of capabilities for ATF 3 energy.gov/ne
Overall Applicability to Accident Tolerant Fuels
- The DOE neutronics and thermal-hydraulics code capabilities are well established and demonstrated for LWR, particularly PWR applications
- Steady-state operation, investigations of crud induced power shift, fuel pellet-cladding interaction
- Operational transients, startup, shutdown, power maneuver
- Select transients, such as reactivity insertion accidents, and departure from nucleate boiling
- The material and geometry of most ATF concepts are within current VERA capabilities with some modifications required for non-cylindrical fuel geometry
- The codes are validated for current fuels (Zr-clad UO2) and operations and can be extended for ATF concepts 4 energy.gov/ne
Codes and Coupling for LWR Simulations Neutronics MPACT Pin-resolved 3-D whole-core neutron transport in 51 energy groups and >5M Fuel Performance unique cross section regions ORIGEN Isotopic depletion and decay SHIFT Massively Parallel Monte Carlo transport to perform high-fidelity reference solutions to inform MPACT CTF BISON Subchannel thermal-hydraulics with Fuel performance code modeling transient two-fluid, three-field (i.e., liquid thermo-mechanics of standard and film, liquid drops, and vapor) solutions in advanced fuel and cladding 14,000 coolant channels with crossflow CFD Commercial (Star-CCM+) and DOE-developed (NEK5000) capabilities to inform subchannel and provide reference solutions Thermal Hydraulics 5 energy.gov/ne
MPACT 3D Core Pin-Resolved Neutronics Watts Bar Unit 1 Cycle 14
- Optimized neutronics for determining detailed Beginning-of-Cycle Power Distribution pin-by-pin neutron flux distribution
- In-line resonance self-shielding at local conditions
- Method-of-Characteristics (MoC) transport theory solution on exact geometry in 2D planes
- Global 3D CMFD solution for average fluxes and axial leakage
- Currently models cylindrical fuel rod geometry without standard industry code approximations
- Full geometry detail, no assembly homogenization
- Pin powers computed explicitly, no pin-power Example of Comparisons with Reactor Operating reconstruction Data (flux maps)
- Direct feedback calculation, no cross section functionalization
- Physics methods fully-applicable to ATF and elimination of approximations allows direct investigation ATF concepts 6 energy.gov/ne
SHIFT Monte Carlo Neutronics Capabilities
- Advanced Monte Carlo neutronics code designed to support a range of computers: from serial to massively parallel systems
- Most direct physics simulation of neutronics currently available
- Detailed three-dimensional geometry representation of fuel, core and ex-core geometries and materials
- Continuous energy neutron transport
- Utilizes detailed isotopic and temperature distributions from MPACT
- Combined with large-scale computers provides the best available means to verify more approximate physics models
- Used within DOE to verify MPACT reactivity and 3D pin power distributions for zero power physics experiments and reactor operations AP1000 fission rate
- Provides means to easily verify and confirm distribution physics models for ATF fuels including reactivity and detailed pin powers on a full-core scale 7 energy.gov/ne
Benefits of Advanced Neutronics Capabilities for ATF
- Whole-core, fully coupled, steady-state and transient neutronics, thermal-hydraulics, and fuel performance modeling
- Removes modeling assumptions in standard codes, reducing need to investigate and confirm approximations
- Sensitivity analysis can be used to investigate changes caused by insertion of ATF fuel concepts Physics Model Industry Practice DOE Codes (VERA)
Neutron Transport Core: 3D diffusion, 2 energy groups, 3-D transport Lattice: 2D transport, 40-70 energy 51 energy groups groups Power Distribution nodal average with pin-power fuel pin resolved reconstruction Thermal-Hydraulics 1-D assembly-averaged Fuel rod subchannel (w/crossflow)
Fuel Temperatures nodal average fuel pin resolved Xenon/Sm nodal average w/correction fuel pin resolved Depletion lattice-averaged cross sections fuel pin resolved history corrections Target Platforms workstation (single-core) 1,000 - 10,000 cores 8 energy.gov/ne
COBRA-TF (CTF): Whole-Core T/H
- Two-fluid, three-field representation of two-phase flow
- Continuous vapor (mass, momentum and energy)
- Continuous liquid (mass, momentum and energy)
- Entrained liquid drops (mass and momentum) Sub-Channel Discretization
- Non-condensable gas mixture (mass) for the entire
- Rod-channel resolution core
- Applied to every fuel rod channel in the core
- Transient and steady-state simulations
- Includes cross flow between channels
- CFD-informed models under development to improve fidelity and modeling detail
- Grid spacer grid models
- Azimuthal heat-transfer coefficients
- Applications include:
- PWR & BWR steady-state and transient
- Main steam line break analysis
- Reactivity insertion accident
- CFD-informed rod-by-rod thermal hydraulics can Main Steam Line Break Application of CTF model insertion of ATF to investigate impacts coupled with MPACT - core power distribution 9 energy.gov/ne
CFD-Informed Subchannel Modeling Example DOE has been using CFD to inform subchannel to provide more detail on clad surface heat transfer to model corrosion product (CRUD) deposition on cladding HTC rod surface map data developed by STAR-CCM+ and read by CTF Heat transfer coefficient map on Comparison of RMS error between unrolled cladding surface CFD and CTF surface temperature reconstructed from CTF based on prediction, with and without CFD-CFD calibration data informed heat transfer 10 energy.gov/ne
DNB simulations mimic electrically CFD Evaluations of DNB heated DNB experiments DNB detected by temperature excursion
- Two-fluid Eulerian-Eulerian multiphase Navier-Stokes solution
- Carrier phase evaluated using standard single phase CFD models
- Sub-grid models define dynamic behaviors of bubble/droplet populations
- Wall heat partitioning
- Single phase convection Validation for full height 5x5 bundle
- Bubble formation with nonmixing vane grid
- Sliding bubble convection +16%
- quenching
- Bubble dynamics
- Lift and drag as a consequence of bubble -16%
size and shape
- Bubble induced turbulence
- Coalescence and breakup
- Nucleation site interactions Developing capability that can be used for support DNB analysis of ATF fuels 11 energy.gov/ne
Transient Capabilities - Reactivity Insertion Accident
- A target problem for the development of the DOE coupled simulation codes is a Reactivity Insertion Accident (RIA)
- Time-dependent and coupled neutronics/thermal-hydraulics (MPACT/CTF) has been applied to a PWR control rod ejection
- Maintains fuel pin and channel resolved capabilities as previously discussed
- Fuel rod temperature model provided fuel temperature feedback
- Calculates fuel rod power profiles and histories are used as inputs for BISON fuel performance simulations Full-Core PWR Control Rod Ejection Simulation
- Capability applicable to ATF Existing commercial Westinghouse 4-loop core assessments design at End-of-Cycle 12 energy.gov/ne
Coupling DOE Codes with Reactor Systems Codes
- Provides systems code capability and analyze flow regimes for which CTF has not been validated/confirmed (for example, LOCA)
- Coupling envisioned with industry and NRC systems code
- A joint DOE-NRC effort to couple NRCs TRACE code with BISON has been performed as a demonstration of the use of both DOE and NRC codes
- Combines BISONs capabilities for ATF with NRCs systems analysis code that has been widely applied for confirmatory analysis
- Purpose of coupling is to analyze LWR ATF concepts for normal reactor operation up to Large Break LOCA events, normal operation, blowdown phase, refill phase, and reflood phase.
- An initial demonstration of systems code and fuels coupling for a typical transient has been performed (Station Black Out) 13 energy.gov/ne
DOE Code Applications for Source Terms for ATF
- Primary application of DOE Neutronics/TH codes is to provide a high-resolution reactor inventory that can be used with a severe accident code, such as MELCOR
- Can provide a detailed rod-by-rod inventory for all fuel assemblies in reactor and spent fuel pool
- Validation with data from PIE and reactor operation
- Detailed SCALE/ORIGEN isotopics, consistent with NRC applications, are available for source term inventories.
- BISON fuel performance code calculates fission gas release that can be used in source term models
- Transient simulation capability can be used to provide detailed power rod-by-rod power distributions and fuel enthalpy for comparison with coarser methods.
14 energy.gov/ne
Neutronics and Core Modeling Validation
- Validation includes single and coupled multi-physics
- Focused on existing fuels and plant operation
- Continuous energy Monte Carlo benchmark calculations supplement measured data
- Critical Boron *Criticality
- Rod Worths *BOL Pin Powers
- ITC *Temperature
- Flux Maps Worth
- T/H Feedback Operating Critical Power Experiments Plants Fuel Rod CE Monte PIEs Carlo
- 3D Core Pin
- Gamma Scans Powers
- Burnup *Intra-Pin
- Radiochemical Distributions Assays *Depleted Isotopics
- CRUD Deposition *Gamma Transport 15 energy.gov/ne
Subchannel Thermal-hydraulics Validation
- CTF Subchannel validation is against publicly-available data and some specific proprietary vendor data Phenomenon for Validation: Validation Experiments Examples
- Single phase convection
- Full channel
- Sub-cooled boiling heat transfer - PSBT (PWR)
- Single and two-phase wall shear
- Pressure drop
- Critical heat flux
- Single and two-phase grid pressure drop
- BFBT (BWR)
- Single and two-phase turbulent mixing
- Pressure drop and void
- Grid heat transfer enhancement (PWR)
- Critical power
- Nucleate boiling heat transfer - FRIGG Loop
- Vapor generation (near-wall condensation)
- Single & two-phase pressure drop
- Void drift
- Axial & radial void distribution
- Dryout
- Pressure directed cross flow
- Harwell High Pressure Two-Phase Heat Transfer Loop
- Transition boiling
- Subchannel
- Critical heat flux
- Single and two-phase heat transfer
- Flow qualities
- Mixing and void drift
- Vendor proprietary data
- Departure from nucleate boiling 16 energy.gov/ne
VERA Validation - Existing Plants and Fuels
- Integrated codes applied to a number of PWR plants and operating cycles for validation
- Typical comparisons include:
- Zero power physics tests (criticality, rod worth), flux maps during power escalation
- Operational measurements (soluble boron, flux maps)
- Operational transients
- Plant and cycles also chosen for validation of nominal and operational occurrences
- Specific plants include:
- Watts Bar Unit 1 (Cycles 1-15) (W 4-loop, 17x17 fuel)
- Watts Bar Unit 2 (Cycles 1-2) (W 4-loop, 171x17 fuel)
- Callaway 1 (Cycles 1-8) (W 4-loop, 17x17 fuel)
- Catawba 2 (Cycles 8-21) (W 4-loop, 17x17 fuel)
- Seabrook Unit 1 (Cycles 1-5) (W 4-loop, 17x17 fuel)
- Palo Verde 2 (Cycles 1-9) (CE System 80, 16x16 fuel)
- Davis-Besse Cycles 12-15 (B&W, 15x15 fuel)
- Oconee 3 (Cycle 25) (B&W, 15x15 fuel)
- TMI Cycles 1-10 (B&W, 15x15 fuel)
- Byron 1 (Cycles 17-21) (W 4-loop, 17x17 fuel)
- AP1000 startup 17 energy.gov/ne
VERA Transient Validation - SPERT-III SPERT-III Core Geometry
- Special Power Excursion Reactor Test III (SPERT III)
- PWR Test Reactor
- Purpose to investigate power excursion kinetic behavior series of transient tests provide validation for coupled multi-physics
- Widely used for validation of codes for reactivity insertion accidents (RIA)
- SPERT III Transient Experiment Modeling
- 60 assemblies radially and 20 layers axially SPERT-III Power Comparison
- Initial core inlet temperature is at 502 oF +/-4 oF.
- Initial Power is 19 +/- 1 MW.
- Fully coupled neutronics, thermal-hydraulics, fuel temperature model
- Good agreement between VERA and measured results 18 energy.gov/ne
Neutronics Capability Assessment for ATF Modeling UO2 Coated Doped FeCrAl U3Si2 SiC U Metal Fuel Clad Fuel Clad Fuel Clad Fuel Geometry Yes Yes Yes Yes Yes Yes No Physics Yes Yes Yes Yes Yes Yes Yes Models Materials & Yes Yes Yes Yes Yes Yes Yes Nuclear Data Validation Yes Yes Yes No No No No Performed (UO2) (UO2)
Design Specific Yes Yes Yes No No No No Validation data (UO2) (UO2)
Availability
- Significant validation for UO2 - zirconium clad fuel performed and underpins validation for all fuel concepts. Historical experience for steel-based cladding.
- Material perturbations can be investigated with DOE codes and sensitivity/uncertainty approaches to determine need for additional validation.
- Comparisons with Monte Carlo methods can provide confirmation of ability to codes to model non-cylindrical geometry fuels.
- Lead test rods and assemblies, startup physics testing can provide additional pre-operational data.
19 energy.gov/ne
Thermal-Hydraulic Capabilities for ATF Modeling Zirc Coated Doped FeCrAl U3Si2 SiC U Metal Clad Clad Fuel Clad Fuel Clad Fuel Geometry Yes Yes Yes Yes Yes Yes No T/H Models Yes Yes n/a Yes n/a Yes No Generic Yes Yes n/a Yes n/a Yes No Validation Performed Design Specific Yes No n/a No n/a No No Validation Data Available
- Assessment for conventional rod bundle T/H completed and is expected to be applicable to ATF.
- Design specific information for surface roughness, wettability, spacer grid location, top/bottom fuel nozzle form losses needed to verify existing validation.
- Design specific data needed for CHF and two-phase CFD validation. CFD validation of DNB underway for non-mixing and mixing vane 5x5 experiments.
- Metallic fuel with non-cylindrical geometry is the greatest challenge.
20 energy.gov/ne
Conclusions on Neutronics/Thermal Hydraulics Models
- Higher-resolution, fully coupled modeling capabilities have been developed and are applicable for ATF
- Direct whole-core simulation with explicit representation of local heterogeneities can be used for ATF, including the design and analysis of LTA test programs and reload cores
- Approaches using higher-accuracy methods (Monte Carlo, CFD) to verify and inform engineering-scale simulations provides means to accommodate ATF modeling
- Can be reduce time to perform analysis by eliminating investigation of applicability of and more directly simulate impacts of insertion of ATF
- Significant validation for existing UO2 fuels forms can be leveraged for ATF analysis
- Use of sensitivity/uncertainty methods and higher-order methods (Monte Carlo/CFD) can be used to quantify the impact of ATF in areas such as criticality and thermal-hydraulic performance
- Can be used by Industry and NRC for analysis or support confirmation of current Industry or NRC codes
- System code coupling allows for evaluation of transient licensing events and demonstrates use of NRC and DOE codes together 21 energy.gov/ne
Questions?
22 energy.gov/ne
Introduction to non-LWR DOE Briefing to ACRS:
Session Advanced Computer Models for Reactor Safety Applications August 21, 2018
Outline of non Non-LWR Presentations SFR HTGR FHR MSR Neutronics MCC-3, MCC-3, MCC-3, MCC-3, PROTEUS, PROTEUS, PROTEUS, PROTEUS, Rattlesnake Rattlesnake Rattlesnake Rattlesnake Fuels MARMOT, MARMOT, MARMOT, In progress BISON BISON BISON chemistry code T-H Nek-5000, Nek-5000, Nek-5000, Nek-5000, SAM, Pronghorn, Pronghorn, SAM SOCKEYE SAM SAM Source term SAM, Pronghorn, SAM, Nek5000, SAM, PERSENT, BISON Pronghorn, Nek5000 BISON BISON 2 energy.gov/ne
Outline of non Non-LWR Presentations SFR HTGR FHR MSR Neutronics MCC-3, MCC-3, MCC-3, MCC-3,
- 1. Tanju Sofu PROTEUS, PROTEUS, PROTEUS, PROTEUS, Rattlesnake Rattlesnake Rattlesnake Rattlesnake
- 2. Rich Fuels MARMOT, MARMOT, MARMOT, In progress BISON BISON BISON chemistry Williamson code T-H Nek-5000, Nek-5000, Nek-5000, Nek-5000,
- 4. Tanju Sofu term PERSENT, BISON Nek-5000, Nek-5000 BISON Pronghorn, BISON 3 energy.gov/ne
Neutronics Analysis Capabilities DOE Briefing to ACRS:
for Advanced Reactors Advanced Modeling & Simulation Tools for Advanced Reactors
Outline
Background
Neutronics analysis code suite Validation basis Criticality experiments Integral tests at EBR-II and FFTF Advanced reactor applications SFRs MSRs HTGRs Multiphysics analysis Conclusions 2 energy.gov/ne
=
Background===
DOEs Nuclear Energy Advanced Modeling and Simulation program supports development of high-fidelity capable neutron transport code suite for advanced reactors Deterministic tools to complement Monte Carlo techniques by compensating for their limitations Transient analyses with deforming mesh Fine-grid flux distributions for rigorous treatment of multi-physics phenomena Shielding and dose rate calculations in regions with low flux High-fidelity capability extends application regime to complex geometries and sharply heterogeneous material compositions Desired impact Improved operational and safety margins through higher-order modeling Close to first principles solutions for benchmarking lower-fidelity/order modeling approaches Multi-physics interface with matching levels of fidelity Enhance the impact of experiments to support reactor design/licensing 3 energy.gov/ne
PROTEUS Neutronics Suite Complete package with unstructured finite element meshing tools, range of multi-group cross-section generation options (for both thermal and fast spectrum), neutron/radiation transport solvers, depletion and sensitivity analyses, and post-processing capabilities
- Multi-physics interface for thermal and core deformation feedback MC2-3 and Cross Section API: For high-quality multi-group cross section generation with local heterogeneity effects
- https://www.anl.gov/technology/project/mc2-3-multigroup-cross-sections-fast-reactors PROTEUS: Two high-fidelity, highly-scalable neutron transport solver options (SN and MOC) and a nodal transport solver option (NODAL)
PERSENT: Perturbation and sensitivity analyses based on the variational nodal transport method (to determine reactivity feedback coefficients) 4 energy.gov/ne
RattleSnake A multi-scheme radiation transport application Deterministic radiation transport solver for linearized time-dependent Boltzmann radiation transport equation
- Originally developed to support TREAT reactor restart and experiment modeling, based on a finite element solver for SN and PN approximations
- Designed for tightly coupled nonlinear multiphysics simulations to capture the impact of temperature and material density changes on time-dependent flux distribution, reaction rates, and power profile
- Multi-scheme capability for a fine-scale resolution in places where interesting multiphysics phenomena take place
- Uses a lower order and/or homogenized solution for less interesting areas
- Lattice, pebble bed and hexagonal fuels, complex configurations such as Advanced Test Reactor (ATR) and the Transient Reactor Test Facility (TREAT)
- https://rattlesnake.inl.gov/SitePages/Home.aspx 5 energy.gov/ne
Validation basis 6 energy.gov/ne
PROTEUS: Criticality Experiments
- LANL Experiments ZPPR-15 (Flattop, Godiva, Jezebel, Bigten) 77 76 75 74 73 72 71 70 69 68 67 66 65 64 63 62 61 60 59 58 57 56 55 54 53 52 51 50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25 24 23 22 21 20 30 31 401 401 401 401 401 401 401 401 401 401 401 401 32 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 33 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 34 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 35 401 401 401 402 402 402 402 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 402 402 402 402 401 401 401 36 401 401 401 402 402 402 402 402 402 401 401 401 401 401 404 404 404 404 404 404 401 401 401 401 401 402 402 402 402 402 402 401 401 401 37 401 401 401 402 402 402 402 402 402 402 404 404 404 404 404 404 405 405 405 405 404 404 404 404 404 404 402 402 402 402 402 402 402 401 401 401 38 401 401 405 402 402 402 402 402 402 404 404 405 405 405 405 405 405 403 403 403 403 405 405 405 405 405 405 404 404 402 402 402 402 402 402 405 401 401 39 401 401 405 402 402 402 402 402 402 404 406 405 403 403 403 403 403 501 501 501 501 704 501 403 403 403 403 403 405 406 404 402 402 402 402 402 402 405 401 401
- Criticality, reaction rates 40 401 401 401 402 402 402 402 402 402 404 406 403 403 403 403 501 501 501 501 501 501 501 501 501 501 501 501 403 403 403 403 406 404 402 402 402 402 402 402 401 401 401 41 401 401 405 402 402 402 402 402 404 404 406 403 403 501 501 501 501 501 501 501 501 501 501 501 501 501 501 501 501 501 501 403 403 406 404 404 402 402 402 402 402 405 401 401 42 401 401 405 402 402 402 402 402 404 406 403 403 501 501 501 501 501 501 501 501 925 925 925 925 925 925 501 501 501 501 501 501 501 501 403 403 406 404 402 402 402 402 402 405 401 401 43 401 401 401 402 402 402 402 402 404 406 403 403 501 501 501 501 501 501 925 741 201 923 926 923 923 926 923 201 925 925 501 704 501 501 501 501 403 403 406 404 402 402 402 402 402 401 401 401 44 401 401 402 402 402 402 402 404 406 403 501 501 501 501 501 927 921 925 923 925 923 925 921 925 925 922 925 923 925 923 925 922 927 501 501 501 501 501 403 406 404 402 402 402 402 402 401 401 45 401 401 402 402 402 402 402 404 406 403 501 501 501 501 501 925 921 926 924 201 921 925 866 201 928 928 201 866 925 922 201 924 926 922 925 501 501 501 501 501 403 406 404 402 402 402 402 402 401 401 46 401 402 402 402 402 402 402 404 403 403 501 501 501 921 926 935 201 921 925 921 925 923 926 925 926 939 925 926 923 925 922 925 922 201 935 926 922 501 501 501 403 403 404 402 402 402 402 402 402 401 47 401 401 402 402 402 402 402 404 406 403 501 501 501 921 201 742 925 923 925 923 925 925 707 913 913 913 914 737 914 201 925 925 923 741 923 925 923 201 922 704 501 501 403 406 404 402 402 402 402 402 401 401 48 401 401 402 402 402 402 404 406 403 501 501 501 501 925 935 925 936 925 925 201 911 919 909 909 919 909 942 920 910 910 920 912 201 925 925 936 925 935 925 501 501 501 501 403 406 404 402 402 402 402 401 401 49 401 401 402 402 402 404 406 403 403 501 501 501 925 926 925 921 201 923 925 911 909 909 909 917 143 907 908 144 918 910 910 910 912 925 923 201 922 925 926 925 501 501 501 403 403 406 404 402 402 402 401 401 50 401 401 401 401 402 402 404 405 403 501 501 501 927 926 201 923 926 923 925 911 919 909 143 905 915 905 915 916 906 916 906 144 910 920 912 925 923 926 923 201 926 927 501 501 501 403 405 404 402 402 401 401 401 401
- ZPR-6 assemblies 6, 7A experiments 51 401 401 401 401 401 404 405 403 403 501 704 501 921 926 921 925 741 926 911 145 919 907 738 905 905 905 915 940 906 906 906 916 908 920 714 912 926 925 925 922 926 922 501 501 501 403 705 405 404 401 401 401 401 401 52 401 401 401 401 401 404 405 403 403 501 501 501 925 923 925 923 201 911 919 929 907 905 905 905 141 903 903 903 903 142 906 906 906 908 930 920 912 201 923 925 923 925 501 501 501 403 403 405 404 401 401 401 401 401 53 401 401 401 401 401 401 404 405 403 501 501 501 925 923 201 923 925 911 909 909 907 905 141 915 902 902 902 902 902 902 902 902 916 142 906 908 910 910 912 925 923 201 923 925 501 501 501 403 405 404 401 401 401 401 401 401 54 401 401 401 401 401 401 404 405 403 501 501 501 925 925 866 925 925 919 909 143 915 905 917 902 902 864 902 902 902 902 864 902 902 918 906 916 144 910 920 925 925 866 925 925 501 501 501 403 405 404 401 401 401 401 401 401 55 401 401 401 401 401 401 404 405 403 501 501 501 201 923 925 921 201 736 909 905 905 937 902 902 902 902 901 901 901 735 902 902 902 902 938 906 906 910 910 201 922 741 923 201 501 501 501 403 405 404 401 401 401 401 401 401 56 401 401 401 401 401 404 404 405 501 501 501 925 923 925 921 926 913 909 917 915 905 141 902 902 902 901 901 901 941 901 901 902 902 902 142 906 916 918 910 914 926 922 925 923 925 501 501 501 405 404 404 401 401 401 401 401 57 401 401 401 401 401 404 405 403 501 501 501 925 926 921 201 925 913 919 143 905 905 903 902 902 901 901 901 901 901 901 901 901 902 902 903 906 906 144 920 914 925 201 922 926 925 501 501 501 403 405 404 401 401 401 401 401 Experiment 58 401 401 401 401 401 404 405 403 501 501 501 925 923 925 928 926 913 909 907 915 915 903 902 902 901 901 901 901 901 901 901 901 902 902 903 916 916 908 910 914 926 928 925 923 925 501 501 501 403 405 404 401 401 401 401 401 59 401 401 401 401 401 404 405 403 501 501 501 925 923 925 928 926 913 909 907 915 915 903 902 902 901 901 901 901 901 901 901 901 902 902 903 916 916 908 910 914 926 928 925 923 925 501 501 501 403 405 404 401 401 401 401 401 60 401 401 401 401 401 404 405 403 501 501 501 741 926 921 201 925 913 919 143 905 905 903 902 735 901 901 901 901 901 901 901 901 902 902 903 737 906 144 920 914 925 201 922 926 925 501 501 704 403 405 404 401 401 401 401 401 61 401 401 401 401 401 404 404 405 501 501 501 925 923 925 921 926 913 909 917 915 905 141 902 902 902 901 901 901 901 901 901 902 902 902 142 906 916 918 910 914 926 922 925 923 925 501 501 501 405 404 404 401 401 401 401 401 62 401 401 401 401 401 401 404 405 403 501 501 501 201 923 925 921 201 909 909 905 905 937 902 902 902 902 901 901 901 901 902 902 902 902 938 906 906 910 910 201 922 925 923 201 501 501 501 403 405 404 401 401 401 401 401 401
- Criticality, foil measurements 63 401 401 401 401 401 401 404 405 403 501 501 501 925 925 866 925 925 919 909 143 915 905 917 902 902 864 902 902 902 902 864 902 902 918 906 916 144 910 920 925 925 866 925 925 501 501 501 403 405 404 401 401 401 401 401 401 64 401 401 401 401 401 401 404 405 403 501 501 501 925 923 201 923 925 736 909 909 907 905 141 915 902 902 902 902 902 735 902 902 916 142 906 908 910 910 912 925 923 703 923 925 501 501 501 403 405 404 401 401 401 401 401 401 65 401 401 401 401 401 404 405 403 403 501 501 501 925 923 925 923 201 911 919 929 907 905 905 905 141 903 903 903 903 142 906 906 906 908 930 920 912 201 923 925 923 925 501 501 501 403 403 405 404 401 401 401 401 401 Calculation 66 401 401 401 401 401 404 405 403 403 501 501 501 921 926 921 925 925 926 911 145 919 907 915 905 905 905 915 916 906 906 906 916 908 920 146 912 926 925 925 922 926 922 501 501 501 403 403 405 404 401 401 401 401 401 67 401 401 401 401 402 402 404 405 403 501 501 501 927 926 201 923 926 923 925 911 919 909 143 905 915 905 915 916 906 916 906 144 910 920 912 925 923 926 923 201 926 927 501 501 501 403 405 404 402 402 401 401 401 401 68 401 401 402 402 402 404 406 403 403 704 501 501 925 926 925 921 201 923 925 911 909 736 909 917 143 907 908 144 918 910 910 910 912 742 923 201 922 925 926 741 501 501 501 403 403 406 404 402 402 402 401 401 69 401 401 402 402 402 402 404 406 403 501 501 501 501 925 935 925 936 925 925 201 911 919 909 909 919 909 910 920 910 910 920 912 201 925 925 936 925 935 925 501 501 501 501 403 406 404 402 402 402 402 401 401 70 401 401 402 402 402 402 402 404 406 403 501 501 501 921 201 923 925 923 925 923 925 925 201 913 913 913 914 914 914 201 925 925 923 925 923 925 923 201 922 501 501 501 403 406 404 402 402 402 402 402 401 401 71 401 402 402 402 402 402 402 404 403 403 501 501 501 921 926 935 201 921 925 921 925 923 926 925 926 926 925 926 923 925 922 925 922 201 935 926 922 501 501 501 403 403 404 402 402 402 402 402 402 401 72 401 401 402 402 402 402 402 404 406 403 501 501 501 501 704 925 921 926 924 201 921 925 866 201 928 928 703 866 925 922 201 924 926 922 925 501 501 501 704 501 403 406 404 402 402 402 402 402 401 401 73 401 401 402 402 402 402 402 404 406 403 501 501 501 501 501 927 921 925 923 925 923 925 921 925 925 922 925 923 925 923 925 922 927 501 501 501 501 501 403 406 404 402 402 402 402 402 401 401
- ZPPR-21 Phases A - F experiments 74 401 401 401 402 402 402 402 402 404 406 403 403 501 501 501 501 501 501 925 925 201 923 926 923 923 926 923 201 925 925 501 501 501 501 501 501 403 403 406 404 402 402 402 402 402 401 401 401 75 401 401 405 402 402 402 402 402 404 406 403 403 501 501 501 501 501 501 501 501 925 925 925 925 925 925 501 501 501 501 501 501 501 501 403 403 406 404 402 402 402 402 402 405 401 401 76 401 401 405 402 402 402 402 402 404 404 406 403 403 501 501 501 501 704 501 501 501 501 501 501 501 501 501 501 501 704 501 403 403 406 404 404 402 402 402 402 402 405 401 401 77 401 401 401 402 402 402 402 402 402 404 406 403 403 403 403 501 501 501 501 501 501 501 501 501 501 501 501 403 403 403 403 406 404 402 402 402 402 402 402 401 401 401 78 401 401 405 402 402 402 402 402 402 404 406 405 403 403 403 403 403 501 501 501 501 501 501 403 403 403 403 403 405 406 404 402 402 402 402 402 402 405 401 401 79 401 401 405 402 402 402 402 402 402 404 404 405 405 405 405 405 405 403 403 403 403 405 405 405 405 405 405 404 404 402 402 402 402 402 402 405 401 401 80 401 401 401 402 402 402 402 402 402 402 404 404 404 404 404 404 405 405 405 705 404 404 404 404 404 404 402 402 402 402 402 402 402 401 401 401 81 401 401 401 402 402 402 402 402 402 401 401 401 401 401 404 404 404 404 404 404 401 401 401 401 401 402 402 402 402 402 402 401 401 401 82 401 401 401 402 402 402 402 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 402 402 402 402 401 401 401 83 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 84 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 85 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401 401
- ZPPR-15 Phases A - D experiments 86 401 401 401 401 401 401 401 401 401 401 401 401 87
- Criticality, sodium void worth, control rod worth, Doppler, axial expansion, gamma 1.1 dose, neutron spectrum BFS-76-1 1.0 0.9
- EBR-II experiments (Runs 130B - 170A)
U238 fission reaction rate 0.8
- Criticality, depletion, isotopic mass 0.7 0.6
- BFS experiments (109-2A, 76-1A, 73-1) 0.5
- Criticality, sodium void worth, control rod 0.4 worth, foil measurement 0.3 Experiment 0.2 As-built MCNP Model
- Monju startup experiments 0.1 MC2-3/VARIANT 0-D Model MC2-3/VARIANT 1-D CASE1 Model
- Criticality, temperature coefficients 0.0
-400 -200 0 200 400 600 800
- CEFR physics startup tests (ongoing) x [mm]
7 energy.gov/ne
PROTEUS: Criticality Experiments (cont.)
Detailed ZPR-6/7 Comparisons as an example 2.8
- Good agreement of core eigenvalues 2.6 EU fission, E DU capture, E EU fission, C DU capture, C with the measurements within ~80 pcm 2.4 DU fission, E Pu-239 fission, E DU fission, C Pu-239 fission, C Normalized reaction rates 2.2
- Foil reaction rates predicted within 1 2.0 sigma of experimental uncertainties 1.8 1.6 1.4 1.2 1.0 0 10 20 30 40 50 60 70 Distance from core center (cm)
Foil Reaction Rate Measurement Load PROTEUS Experiment 104 1.00147 1.00072 106 1.00134 1.00091 120 1.00127 1.00099 132 1.00016 1.00040 Flux in group 1 of 70 (10 MeV to 14 MeV)
Loading 106 8 energy.gov/ne
PROTEUS: EBR-II Inherent Safety Demonstration Test
- SHRT-45R: Unprotected (no-scram) station blackout from full power to demonstrate inherent safety
- Instrumented fuel assemblies for in-assembly temperature and flow measurements
- Neutronic benchmark for reactivity feedback coefficients
- Following detailed depletion analyses for several run cycles based on known core loading and assembly fuel compositions
- Doppler, fuel/cladding axial expansion, core radial expansion, coolant density changes, and CRDL expansion effects)
- International benchmark with 19 participating organizations from 11 countries
- Good agreement with test data confirming validity against data from a rare integral test 9 energy.gov/ne
PROTEUS: FFTF Passive Safety Demonstration Test Core
- LOFWOS #13: Unprotected loss of flow without Loading scram from 50% power at full flow
- Also with instrumented fuel assemblies for in-assembly temperature and flow measurements
- Gas Expansion Module (GEM) as a passive reactivity reduction device
- Good agreement with measured power, coolant temperatures and natural circulation flow rate Predicted and Measured Power and Decay Heat Predicted Predicted andand Measured Measured Secondary Coolant Hot-Leg Temperatures Temperatures 10 energy.gov/ne
RattleSnake V&V Completed Validation cases o TREAT Minimum Core o TREAT M8 Calibration Series (steady state and transient)
Computational benchmarks o C5G7 o LRA BWR kinetics o BEAVRS o IAEA 3-D PWR o KAIST-3A reactor quarter core o OECD 3-D MHTGR-350 Core In Progress ATR 94CIC (reactor data/ IRPhE benchmark)
GODIVA neutronic/thermal-mechanical benchmark (reactor data/IRPhE benchmark)
C5G7-TD (computational benchmark)
TREAT M2/M3 Calibration Measurements (reactor data) 2018 TREAT Transient Prescription Measurements (reactor data, steady state and transient) 11 energy.gov/ne
Application of neutronics codes to advanced reactors 12 energy.gov/ne
SFR Hot Channel Factor Evaluation High fidelity evaluation using PROTEUS-MOC and Nek5000 AFR-100 (SFR design) inner assembly - unique enrichment zoning, U-Zr binary metallic fuel Coupling of axial power distributions between PROTEUS-MOC/Nek5000 Reevaluated numerous hot channel factors to demonstrate safety margins Top: 91-pin wire wrapped bundle velocity (Nek5000)
Left: MOC power distribution 13 energy.gov/ne
SFR Hot Channel Factor Evaluation High-fidelity multi-physics evaluations with PROTEUS and Nek5000 AFR-100 (SFR design) inner core assembly
- Unique enrichment zoning with U-Zr binary metallic fuel Coupling of axial power distributions between PROTEUS-MOC/Nek5000 Reevaluated numerous hot channel factors to demonstrate safety margins Cladding Cladding Fuel thermal Coolant Cladding Wire thickness thermal conductivity density circumferenti orientation conductivity al temp.
Uncertainties (3s), % +/-3 +/-7 +/-25 +/-0.5 - -
Nominal 868.2 868.2 868.2 868.2 868.2 853.6 Cladding Perturbed - - - 869.1 866.4 856.2 Outer Wall HCF-Legacy - - - 1.016 2.19 1.01 HCF-SHARP - - - 1.001 1.002* 1.003*
Nominal 894.3 894.3 894.3 894.3 894.3 -
Cladding Perturbed 910.4 932.8 - - 896.1 -
Inner Wall HCF-Legacy 1.03 ~ 1.05 1.088 - - 1.02 -
HCF-SHARP 1.018 1.043 - - 1.002 -
Nominal 1000.5 1000.5 1000.5 1000.5 1000.5 -
Fuel Perturbed - - 1226.6 - - -
Centerline HCF-Legacy - - 1.25 - -
HCF-SHARP - - 1.226 - - -
14 energy.gov/ne
Demonstration of Multi-Scale Modeling Capability for an SFR Model detailed focal assembly at reduced computational expense Cross section processing with MC2-3 preserves heterogeneity effects in homogenized core model Can be applied for more accurate HCF calculation (includes global spectrum effects) at reduced computational expense Consistent k-eff and average power profiles across 3 cases using PROTEUS-SN solver Consistent pin-by-pin power distributions in focal assembly 15 energy.gov/ne
MSR Core Evaluations Analysis of a commercial thermal spectrum, graphite moderated design 2D stationary core calculation with heterogeneous geometry using PROTEUS-MOC 3D stationary core (378 cm high) Thermal Fast Flux (PROTEUS-MOC)
(PROTEUS-MOC) calculation with homogeneous assembly using PROTEUS-NODAL Eigenvalue Comparison Code 2D 3D OpenMC 1.01282 0.99196 PROTEUS -124 207 (k, pcm) (MOC) (NODAL)
% Difference in Assembly Flux (< 2.3%)
between PROTEUS-MOC and OpenMC 16 energy.gov/ne
Flowing Fuel Treatment for MSRs Delayed-neutron precursor drift Model Eigenvalue model in PROTEUS-NODAL Stationary Fuel 1.01458 Fuel Flowing A 1.01350 Flow path of molten salt fuel, associated channels, and transit time of inside/outside Fuel Flowing B 1.01289 core specified by user Evaluate the effect of the velocity field on neutronics Fast-spectrum MSR benchmark problem
- 2,050 MWt with PuCl3 (370x370x480 cm)
- Fuel, blanket, shield regions
- Moving fuel in the inner core and blanket regions
- Computed core eigenvalue and precursor concentrations as a function of transit time A inside/outside core Eigenvalue decreases with flow due to delayed neutron loss outside of the core Group 6
- (A) 10 sec transit in core, 5 sec outside of core B
- (B) 1 sec transit in core, 0.3 sec outside of core 17 Group 1 energy.gov/ne
High Temperature Gas-cooled Reactor Applications VHTR whole-core calculations Model for 3D prismatic fuel assemblies Hole for operating Fuel column control rod Inner reflector
- Challenges for modeling the control (replaceable)
Hole for startup Outer reflector channel leading to large neutron control rod (replaceable) streaming Hole for reserve Permanent shutdown reflector Core barrel All Rods Out Control rod hole Operating Rods In 1.2m 7.93m Inner reflector Inner reflector 1.6m All Rods In 18 energy.gov/ne
Pebble Tracking Transport (PTT) Algorithm for Pebble Bed Reactor Analysis Motivation:
- Support high resolution multi-physics simulations of pebble motion
- Enable direct transport calculations with pebble tracking Capabilities HTR-10 Pebble Bed Mesh: Total
- DEM (discrete element method) to provides time- 437,735 tetrahedra. 76,869 node points. Pebble packing region has dependent positions of all pebbles for establishing an 291,107 tetrahedra, reflection top equilibrium core and bottom.
Goal: Ability to track burnup of individual pebbles as they move down the core Preliminary Results:
- PTT can converge keff and reaction rates less than 0.1% for all cases with respect to reference case with Serpent (simple 9-sphere model, fresh fuel)
Error in keff (pcm) relative to reference model Fast (left) and thermal (right) fluxes in HTR-10 simulation 19 energy.gov/ne
Multi-physics Analysis for SFR Core Deformation Multi-physics simulation with PROTEUS-SN + Nek5000 + Diablo (SHARP)
Core deformation by thermal expansion and irradiation induced swelling is an important reactivity feedback in SFRs in both normal and transient operation Trial calculation for ULOF in ABTR ABTR core deformation (magnified 100x) computed with SHARP Fission power distribution 20 energy.gov/ne
Conclusions
- PROTEUS and RATTLESNAKE are developed to accurately and deterministically simulate various advanced reactor cores with complex geometry and compositions
- Multiple user options for improved flexibility and applicability
- MC2-3 as a well-validated cross-section generation tools against numerous fast reactor experiments (ZPR-6, ZPPR-15, BFS, EBR-II, Monju, etc.)
- Being used by DOEs VTR project, TerraPower and KAERI to support core design
- Also used by other companies, universities, national labs for the research and development purpose
- PERSENT for assessment of reactivity feedback effects
- Multiphysics simulations with PROTEUS (neutronics) and SAM, Nek5000, and Cobra-TF (thermal fluid) are in progress for SFR, HTGR, and MSR applications 21 energy.gov/ne
Questions?
22 energy.gov/ne
Fuel Performance Modeling DOE Briefing to ACRS:
for Advanced Reactors Advanced Modeling & Simulation for August 21, 2018 Advanced Reactors
Outline of Presentation
- Background and Presentation Objective
- The Bison Fuel Performance Code
- Overview
- Verification and Validation
- Documentation
- Bison for TRISO Particle Fuel
- Capabilities/Examples
- Validation
- Development/Validation Plans
- Bison for Metallic Fuel
- Capabilities/Example
- Validation
- Development/Validation Plans
- Summary and Conclusions 2 energy.gov/ne
Background and Objective
Background
- DOE is developing modern fuel performance modeling tools applicable to a wide variety of fuel and reactor types, operating conditions, geometries and spatial scales
- Capability has been developed for both advanced and traditional LWR concepts, with LWR fuel receiving greatest early emphasis
- Multiscale modeling approach (Bison/Marmot) provides improved mechanistic material models and has delivered demonstrated results for UO2 fuel Presentation Objective
- Provide current status of DOE fuel codes for application to specific advanced reactor concepts including: code capabilities, current validation status, future development and validation plans Concept SFR MHTGR FHR MSR Fuel Type Metallic Alloy TRISO particle TRISO particle Liquid Fuel Element Pin Pebble/Prism Pebble N/A Coolant Liquid Metal Gas Molten fluoride salt Molten salt Applicable DOE Bison/Marmot Bison/Marmot Bison/Marmot None Fuel Codes 3 energy.gov/ne
Bison/Marmot Fuel Performance Codes
- 1) Overview
- 2) Verification and Validation
- 3) Documentation 4 energy.gov/ne
MOOSE-Bison-Marmot (MBM)
- The MOOSE-Bison-Marmot (MBM) codes provide an advanced multidimensional, multiphysics, multiscale fuel performance capability Advanced Multidimensional Atomistic/Mesoscale Material Fuel Performance Code Model Development
- Models a wide variety of fuel
- Predicts microstructure types and geometries at an evolution in fuel and cladding engineering scale
- Used with atomistic methods
- Applicable for steady, transient to develop multiscale and accident conditions materials models Multiphysics Object-Oriented Simulation Environment
- Simulation framework allowing rapid development of FEM-based applications 5 energy.gov/ne
Bison Fuel Performance Code
- Finite element-based engineering scale fuel 2D axisymmetric (or 1.5D) performance code
- Solves the fully-coupled thermomechanics and species diffusion equations in 1D, 1.5D, 2D axisymmetric or plane-strain, or full 3D
- Applicable to both steady and transient operations and includes LOCA and RIA 2D plane strain capability for LWR fuel
- Readily coupled to lower length scale material models
- Designed for efficient use on parallel computers 3D
- Development follows NQA-1 process 6 energy.gov/ne
What Makes Bison Different?
3D FEM: Applicable to any geometry Parallel Computing: Large problems Multiple fuel/reactor applicability Readily couples to LLS or other codes 7 energy.gov/ne
Bison V&V and Statistical Analysis
- Code Verification
- MOOSE/Bison is supported by
>2000 unit and regression tests
- All new code must be accompanied with verification testing; all tests are required to pass prior to any code change
- LWR Validation
- ~75 integral, normal operation and ramp fuel rod experiments
- 47 LOCA cases (43 burst tests, 5 integral rods)
- 19 RIA cases
- More detail in ATF presentation
- Coupled to DAKOTA to enable constitutive model calibration, sensitivity analysis and uncertainty quantification 8 energy.gov/ne
Bison Documentation
- Latest externally released documentation (pdf files) available at:
https://bison.inl.gov/SitePages/Manuals.aspx
- Theory manual
- User manual
- Training workshop slides
- Link to code verification article
- Assessment report
- Currently transitioning to a web-based documentation system
- Combines theory and user manuals
- Imposes strict requirements on documentation of new code
- Much more of the documentation exists within and builds from the source code 9 energy.gov/ne
TRISO Particle Fuel 10 energy.gov/ne
Bison - Particle Fuel Capabilities General Capabilities Gap Behavior
- Finite element based 1D-Spherical, 2D-
- Gap heat transfer with kg= f (T, n)
RZ and 3D fully-coupled thermo-
- Gap mass transfer mechanics with species diffusion
- Mechanical contact (master/slave)
- Linear or quadratic elements with large deformation mechanics
- Particle pressure as a function of:
- Elasticity with thermal expansion evolving gas volume (from mechanics) gas mixture (from FGR and CO model)
- Steady and transient behavior gas temperature approximation
- Parallel computation Silicon Carbide Fuel Kernel (UO2)
- irradiation creep
- Temperature/burnup/porosity dependent thermal conductivity
- Solid and gaseous fission product swelling Pyrolytic Carbon
- Anisotropic irradiation-
- Densification induced strain
- Thermal and irradiation creep
- Irradiation creep
- Fission gas release (two stage)
- CO production Empirical models for SiC and PyC from:
D. Petti, P. Martin, M. Phelip, R. Ballinger,
- Radioactive decay Development of improved models and designs for coated-particle gas reactor fuels. Technical Report INL/EXT Tangential Stress 02615, December 2004.
11 energy.gov/ne
Example:1D Spherical Particle
- Spherically symmetric
- Thermo-mechanics plus cesium diffusion
- Analysis of normal irradiation, storage, and accident testing periods
- Run times of ~1 s permit rapid analysis of large numbers of particles accident testing storage Tangential stress histories Cesium radial profiles 12 energy.gov/ne
Example: 2D RZ - Aspherical Particle
- Aspherical TRISO particles can occur during manufacturing
- Single facet particle simulated assuming 2D axisymmetry
- During simulated accident, asphericity raised peak tensile stress in SiC layer by ~4x
- Typical run times of a few minutes using 8 processors accident testing storage 175 MPa 650 MPa Tangential stress comparison for a spherical and Stress history at the outer SiC surface aspherical particle following simulated accident 13 energy.gov/ne
Example: 3D - Defective SiC layer
- Localized thinning of SiC layer can occur due to soot inclusions or fission product interaction
- BISON 3D capability demonstrated on an eighth-particle with localized thinning of the SiC layer at random locations SiC SiC layer layer
- Thinned SiC regions experience significantly higher tensile stress and greater cesium release; impossible to predict with 1D analysis
- Typical run times of a few hours on 8 processors 14 energy.gov/ne
Comparison to IAEA Benchmarks (1/2)
- IAEA CRP-6 on HTGR technology developed a set of fuel performance code benchmarking cases for normal operation and operational transients
- Ranged in complexity from a simple fuel kernel having a single elastic coating layer, to realistic TRISO-coated particles under a variety of irradiation conditions
- Bison assessed against analytical solutions and other particle fuel codes for 13 cases Comparisons of tangential stress Case Descriptions at end of simulation 15 energy.gov/ne
Comparison to IAEA Benchmarks (2/2)
- Bison comparisons to PARFUME (US), STRESS3 (UK) and ATLAS (France)
Inner pyrolytic carbon layer Silicon carbide layer Case 8: Code comparisons for the cyclic Cases 10 and 11: Code comparisons of the particle temperature in pebble bed reactor tangential stress at the inner SiC wall.
Variations attributed to differences in fission gas release and CO production models.
16 energy.gov/ne
TRISO Particle Fuel - Documentation TRISO particle capability development, examples and benchmarking are documented in 2013 journal article 17 energy.gov/ne
Particle Fuel Development/Validation Plan Development
- PARFUME/Bison capability gap analysis (FY18)
- UCO fuel models (FY19)
- Mechanistic CO production (FY19)
- Particle failure probabilities (FY20)
- Partial debonding and SiC fracture with XFEM (FY20)
Validation
- Investigate data base and approach for existing particle fuel codes (FY18)
- Review NRC HTGR research plan (FY19)
- Develop Bison particle fuel validation plan (FY19)
- Prepare highest priority validation cases (FY19 forward) 18 energy.gov/ne
Metallic Fuel 19 energy.gov/ne
Bison Metallic Fuel Capabilities General Capabilities Gap/Plenum Behavior
- Gap heat transfer (sodium conductivity)
- Finite element 1.5D, 2D-RZ axisymmetric and Cartesian and 3D fully-coupled thermo-
- Mechanical contact (master/slave) mechanics with species diffusion
- Plenum pressure as a function of:
- Linear or quadratic elements with large evolving gas volume (from mechanics) deformation mechanics gas mixture (from FGR model)
- Elasticity with thermal expansion gas temperature approximation
- Steady and transient operation
- Parallel computation Cladding Behavior HT-9 and 316 Stainless Steel Metallic Fuel Behavior
- Temperature dependent conductivity UPuZr and UZr
- Thermal and Irradiation creep
- Temperature/species dependent conductivity
- Failure model for HT-9
- Anisotropic Swelling
- Thermal and irradiation creep
- Fission gas release (porosity based)
Coolant Channel
- Zr diffusion
- Closed channel thermal hydraulics with heat transfer coefficients
- Sodium fluid properties Temperature 20 energy.gov/ne
Example: EBR-II Fuel Pin
- Representative EBR-II pin geometry modelled assuming axisymmetry (2D-RZ)
- Thermo-mechanics plus Zr diffusion in the fuel alloy
- Variations considered in fuel alloy composition and pin power Typical temperature field fuel cladding (geometry scaled 10x radially) 21 energy.gov/ne
Metallic Fuel Validation
- With basic capability established, validation efforts have begun
- Data bases include experiments from EBR-II and TREAT EBR-II TREAT Cases in progress and scheduled for completion in FY18 Experiment Fuel type Cladding Measurements X447 (EBR-II) U10Zr HT9/D9 Pin profilometry, fission gas release X441 (EBR-II) U10Zr HT9 Pin profilometry, cladding strain, U19PuxZr plenum pressure, fission gas release X423 (EBR-II) U10Zr 316 SS Fuel slug dimensions UxPu10Zr M7-IFR (TREAT) U10Zr HT9/D9 Cladding temperature history, cladding U19Pu10Zr failure 22 energy.gov/ne
Metallic Fuel Validation: X447 Experiment X447 experiment in EBR-II
- U10Zr fuel at average power of 30 kW/m for ~360 days
- Pins examined at multiple burnups
- First comparisons are very reasonable 23 energy.gov/ne
Metallic Fuel Development/Validation Plans Development
- Improvements and/or alternates to fuel swelling model - radial swelling is currently under-predicted (highest priority)
- Current Zr diffusion models are applicable to UPuZr; must be extended to UZr
- Lanthanide diffusion and Fuel Coolant Cladding Interaction (FCCI)
- Sodium infiltration
- Cladding swelling Validation cases planned for FY19 Experiment Fuel type Cladding Measurements X429 U10Zr HT9/316 Plenum volume, pressure, and gas UxPu10Zr analysis X496 U10Zr HT9 Bow and length measurements X501 U10Zr HT9 Minor actinide bearing fuel, plenum UxPu10Zr pressure and void volume 24 energy.gov/ne
Summary and Conclusions DOE is developing modern fuel performance modeling tools
- Applicable to a wide variety of fuel types, geometries, spatial scales and operating conditions
- Basic capabilities in place and demonstrated for high-priority advanced reactor fuel concepts including:
- TRISO particle fuel (MHTGR and FHR)
- Metallic fuel (SFR)
- Multiscale modeling approach (Bison/Marmot) has delivered demonstrated results for LWR fuel and will be extended to advanced reactor fuel
- While early comparisons to experimental data for particle and metallic fuel are very encouraging, additional validation is clearly needed
- In collaboration with the NRC and industrial collaborators, capability gaps are being identified and will be given high priority in future code development Opportunities for cooperation
- 1) Implementation of fuel behavior models in vendor or NRC codes
- 2) Use of Bison by industry and/or NRC 25 energy.gov/ne
Questions?
26 energy.gov/ne
Thermal-Hydraulic Capabilities for DOE Briefing to ACRS:
Advanced Reactors Advanced Modeling & Simulation Tools for August 21st, 2018 Advanced Reactors
Outline of Presentation
- Background and Objective
- Capabilities/Examples
- Validation
- Pronghorn
- Capabilities/Examples
- Validation
- Nek5000
- Capabilities/Examples
- Validation
- Summary and Conclusions 2 energy.gov/ne
Overview of Thermal-hydraulic capabilities
Background
- DOE is developing modern multiscale thermal-hydraulic (T/H) tools applicable to a variety of advanced reactor concepts
- While validation focus has been primarily on Sodium Fast Reactors and Gas Reactors to date, the validation basis is being extended to other designs.
- Due to massive scale separation in nuclear systems a multiscale approach is desirable.
Plant Engineering Fine Scales Scales Scales System Sub-channel Reduced- Porous Momentum RANS Hybrid Codes order models media Sources CFD CFD LES/DNS modeling (Coarse CFD)
System Analysis Module (SAM)
Pronghorn Nek5000 3 energy.gov/ne
Objective Objective
- Provide current status of DOE T/H codes for application to specific advanced reactor concepts including:
- Code capabilities
- Validation status
- Future plans
- It can be combined as desired with other DOE T/H codes when 1D approximations are inadequate.
- Engineering and Fine scale codes can be used also to provide closure information to SAM.
Concept SFR HTGR FHR MSR Applicable DOE SAM SAM/ SAM SAM system code Pronghorn Applicable DOE SAM Pronghorn Pronghorn SAM Engineering scale code Applicable DOE Nek5000 Nek5000 Nek5000 Nek5000 fine scale code 4 energy.gov/ne
SAM: Overview
- A modern plant-level system analysis tool for advanced reactors in liquid form (SFR, LFR, MSR/FHR) safety analysis. Some application to Gas reactors.
- Advances in software environments and design (MOOSE), numerical methods, and physical models.
- Focused on system T/H.
Mass flow rate in EBR-II SHRT-45R Benchmark
- Enhancements in large volume modeling:
0D, 1D stratification models and full 3D modeling (porous media).
- Enhancements in core modeling: Single-channel, Multiple-Channel and Intermediate fidelity (targeted toward SFR) core modeling.
- Enhancements related to MSRs: delayed neutron precursors transport, freeze and thaw models.
- Flexible multi-scale multi-physics. Representation of a PLOF in ABTR 5 energy.gov/ne
SAM for Fast Reactors: Challenges & Status
- Unique T/H challenges:
- Decay heat removal through passive heat removal mechanisms
- Natural convection heat transfer for low Prandtl Fluids
- Thermal stratification (impedes natural circulation in upper plenum)
- SAM validation basis:
- SAM can reproduce behavior of several EBR-II and FFTF benchmarks
- SAM improvements :
- Enhanced Multiphysics integration to connect separate phenomena (e.g., coupling of SAM to BISON)
- Advanced Pool and stratification models in SAM
- SAM Gaps: EBRII - Subassembly 6C4 outlet temperature during unprotected loss of heat rejection tests
- Spatial kinetics modeling to be (BOP-302R - top) and unprotected loss of integrated. flow (SHRT-45R - bottom) 6 energy.gov/ne
SAM for Fast Reactors:
Options for stratification models
- Computationally efficient coarse-grid multi-D flow mixing and heat transfer model in SAM
- Using a primitive variable based FEM formulation
- Multiple options for shear stress modeling
- 1-D models and multiple 0-D volume modeling are also being pursued
- Work in progress: closure model developments and V&V of reduced-order models for thermal-stratification SAM EBR-II BOP-302R Simulation Results using multiple 0-D volume model
- Coupling to CFD codes for additional accuracy and flexibility
- Ideal usage is to calibrate 1D or reduced order models
- Several verification and validation tests have been completed or planned
- Includes international collaborations (i.e., TALL-3D) Coupled CFD calculation in the upper plenum of Monju (1995 turbine trip test) - Temperature predictions 7 energy.gov/ne
SAM for Gas Reactors: RCCS applications
- Unique T/H challenges:
- Decay heat removal through Reactor Cavity Cooling systems
>> Need to account for radiation effects
>> Environment effects
- Unique accident scenarios (Air ingress, water ingress) not suited for current SAM formulation
- SAM validation basis:
- SAM can reproduce RCCS tests at NSTF facility (Argonne)
- SAM improvements :
- Novel environment condition models for RCCS
- Coupling to Pronghorn and Nek5000 for detailed modeling of system.
- SAM Gaps:
- Additional components and models National Shutdown Test Facility validation.
Picture of the facility (top), comparison between would be needed for Air ingress and experiment RELAP5 and SAM results.
Water ingress transients.
(Acknowledgement: B. Hollrah) 8 energy.gov/ne
SAM for MSR Reactors: Challenges and Status
- Unique T/H challenges:
- Heat is produced directly in the coolant
- Unique design with fuel circuit
- Unique core configurations with potential recirculation and stagnation zones
- Reactivity management challenges
- Delayed neutron precursor drift
- Quick reconfiguration of the core geometry (gravitational draining) for passive safety
- Salt thaw and freeze in overcooling transients.
- SAM validation basis: Ongoing validation effort against MSRE data in collaboration MSRE - Comparisons between available experimental with PSU. data and SAM model. Top - overall reactor data, Bottom -
Head loss comparison
- SAM improvements:
- Delayed neutron precursor generation and transport
- Direct heat generation in coolant
- Point Kinetics with DNP drift
- Coupling to Nek5000 for 3D modeling (fast spectrum systems)
- Coupling to chemistry models for salt simplified SAM MSR loop model for verification purposes.
9 energy.gov/ne
SAM for Flouride High Temperature Reactors:
Challenges and Status
- Unique T/H challenges:
- FHRs (i.e., coolant salt reactors) combine features of molten salt and high temperature gas reactors.
- High temperatures and associated material modeling issues
- Salt may freeze in certain reactor transients. Freezing and thaw modeling are necessary.
- SAM validation basis: Ongoing validation effort against separate effect tests and MSRE.
- Collaboration with vendors and universities.
- SAM improvements:
- Salt thaw and freeze models.
- Development of specific components needed for FHRs (flow diodes)
- SAM Gaps:
- Need of data for validation.
PB-FHR- LOFC Transient with flow reversal in the primary side of DHX (K. Ahmed, R. Scarlat, R. Hu, Proceedings of NURETH-17) 10 energy.gov/ne
Fast Reactors: SAM Validation Status and Plans
- Extensive validation performed or planned. C- Complete, O-Ongoing, P- Planned.
11 energy.gov/ne
FHR/MSR: SAM Validation Status and Plans
- Ongoing validation and plans identified. C- Complete, O- Ongoing, P-Planned.
12 energy.gov/ne
Pronghorn: Overview
- A modern engineering-level system analysis tool for advanced reactors.
- Advances in software environments and design (MOOSE), numerical methods, and physical models.
- Targeting primarily pebble bed reactors (FHR/HTGR) but extendable to other designs.
- Anisotropic porous media modeling as we well as more advanced formulations. Comparison between Pronghorn and SANA dataset.
- Enhanced formulation: It can combine Pebble temperature at various axial locations.
porous media regions with open regions.
Flexible multi-scale multi-physics.
- It can be described as homogenized conjugate heat transfer (CHT), where each finite element may contain a mixture of coolant, fuel, moderator, or other core internals.
- Correlations for anisotropic resistance from Nek5000 can be implemented in a straightforward manner.
B. Stocker and H. Nieben. Data Sets
Technical report, Forschungszentrum Julich, 1996.
13 energy.gov/ne
Pronghorn: Formulation Think of the model as a two-phase flow problem where the second phase (pebble stack) is stationary.
Conservation of Mass, Momentum, Energy Assumptions:
Nodalization for a gas reactor
- 1. No bed motion with Pronghorn
- 2. No phase change
- 3. Ensemble-averaged turbulence effects 14 energy.gov/ne
Gas Reactors: Pronghorn Validation Status and Plans
- Validation on SANA dataset for Pronghorn, other simulations are planned. C- Complete, O- Ongoing, P- Planned.
15 energy.gov/ne
Nek5000: Overview
- A state-of-the-art high-order computational fluid dynamics code.
- Incompressible and compressible formulations.
- Extensive user base.
- Easy to integrate: Open-source, MOOSE-interface.
- Range of fidelity: from Direct Numerical Simulation to Large Eddy Simulation, Reynolds Comparison of near-wall velocity of two impinging Averaged Navier-Stokes and Porous Media. jets predicted by Nek5000 for the MAX thermal-striping and mixing experiment.
- Scaling Performance: it can be run on laptop as well as a supercomputer. On supercomputers it 4.6 allows for state-of-the art scaling (which enables 3.6 otherwise impossible large scale calculations). x/D 2.6
- Flexible multi-scale multi-physics. 1.6
- Extensive general validation basis: Code has Lo-res Re=40K Med-res Re=40K ranked #1 in several OECD/NEA blind Expt Re=90K benchmarks on CFD applications to reactor OECD/NEA blind benchmark on T-junction flow.
safety. Nek5000 ranked first in temperature predictions 16 energy.gov/ne
Nek5000 for Fast Reactors: Challenges and Status
- Unique T/H challenges:
- Turbulence modeling for low Prandtl Fluids in natural convection
- Thermal striping is a a major concern for structures in the upper plenum
- Mixing in Large enclosures
- Complex flow structures in the fuel assembly require the evaluation of mixing coefficients.
- Nek5000 validation Basis:
- Nek5000 has extensive validation for thermal striping
- Evaluation of Nek5000 against legacy wire-wrapped Wire-wrapped fuel assembly calculation (61 pins).
fuel assembly experiments and new experimental Instantaneous velocity magnitude campaigns.
- Nek5000 improvements:
- Coupling to SAM for modeling of upper plenum
- Coupling to Diablo (structural mechanics) for evaluation of thermal-striping in structures.
- Accurate predictions through Large Eddy Simulation modeling
- Nek5000 Gaps:
- Advanced turbulence modeling options for low Prandtl Fluids (in development)
Comparison between Nek5000 and experimental data at TAMU (central subchannel of a 61 pin bundle) 17 energy.gov/ne
Nek5000 for Gas Reactors: Challenges and Status
- Unique T/H challenges:
- Modeling of Decay heat removal through Reactor Cavity Cooling systems to drive system code model development.
- Simulation of onset of unique accident scenarios (Air ingress, water ingress).
- Mixing in large enclosures.
- Pebble bed flow and heat transfer models
- Nek5000 validation basis:
Nek5000 Simulations in the upper plenum of a gas reactor at
- Ongoing validation effort against TAMU various Reynolds numbers, increasing from 100 (i) to 1000 (v).
experiments for upper plenum. Results will be compared against ongoing TAMU experiments.
- Ongoing validation against random pebble bed data at TAMU (useful also for FHRs). Verification against available DNS data.
- Collaboration with vendors.
- Nek5000 improvements:
- Anisotropic distributed resistance and heat transfer models for Pronghorn.
Nek5000 simulations in the TAMU random pebble experiment.
Left - PIV experimental results, Right - Snapshot of the velocity field. Comparisons are ongoing.
19 energy.gov/ne
Nek5000 for FHRs : Challenges and Status
- Unique T/H challenges:
- Low Reynolds number flows encountered in several components.
- Potential for excessive stress in high temperature components (upper plenum).
- Salt may freeze in certain reactor transients. Freezing and thaw modeling are necessary.
- Nek5000 validation basis:
- Validation on available experiments for twisted tubes HX (e.g., DRACS) and pebble beds
- Collaboration with vendors.
- Nek5000 improvements:
- Large Eddy Simulation in Nek5000 provides an efficient and accurate mean to evaluate friction and heat transfer.
- Coupling to SAM for full system analysis (upper plena simulations)
- Nek5000 Gaps: Twisted tube performance evaluation using Nek5000. Top -
- Need of data for validation. comparison with an available experiment and correlation.
Bottom - Fluid flow in various lattice configurations.
- Salt thaw and freeze models.
20 energy.gov/ne
Nek5000 for Molten Salt Reactors: Challenges and Status
- Unique T/H challenges:
- Unique core configurations with potential recirculation and stagnation zones (exacerbated by the high Prandtl number)
- Delayed neutron precursor drift and associated transport
- Uncertainty of thermo-physical properties of salt.
- Potentially very Low Reynolds numbers in some designs.
- Nek5000 validation Basis:
- Limited MSR-specific validation basis. Ongoing collaborations for validation on MSRE data. Demonstration of flow in a fast spectrum MSR core (Nek5000). Highlighting the importance of 3D effects on the
- Nek5000 improvements: prediction of the highest structure temperature.
- Coupling to SAM for modeling of full system when 3D effects are important (fast MSR core)
- Modeling of mass transport for DNP and fission products.
- Efficient Large Eddy Simulation approach provides accurate results at low Reynolds number conditions.
- Nek5000 Gaps:
- Coupling to neutronics needs to be updated to handle DNP drift
- Coupling to chemistry models for salts
- Need to extend validation basis Demonstration - Large Eddy simulation in a representative thermal-spectrum MSR. The cross section may present local laminarization and strong inhomogeneity of the flow field.
21 energy.gov/ne
Nek5000 for Fast Reactors: Validation Status and Plans
- Substantial validation has already been performed in Nek5000. C- Complete, O-Ongoing, P- Planned.
22 energy.gov/ne
Nek5000 for FHR/MSR: Validation Status and Plans
- Limited validation already performed in Nek5000. Several Ongoing efforts. C-Complete, O- Ongoing, P- Planned.
23 energy.gov/ne
Conclusions
- DOE is developing modern T/H multiscale modeling tools applicable to a wide variety of reactor designs.
- While greatest emphasis to date has been for sodium and gas cooled reactors, substantial capability exists for other reactor concepts
- Basic capabilities have been demonstrated to simulate advanced reactor LBEs.
- Existing capability gaps have been identified and will be given high priority in future code development work
- Past and Ongoing comparisons to experimental data are very encouraging however substantial validation work remains 25 energy.gov/ne
Questions?
26 energy.gov/ne
Source Term Assessment DOE Briefing to ACRS:
Approaches and Codes for Advanced Modeling & Simulation Tools for Advanced Reactors Advanced Reactors
=
Background===
- Source term definition in 10 CFR 50.2:
- The magnitude and mix of the radionuclides released from the fuel, expressed as fractions of the fission product inventory in the fuel, as well as their physical and chemical form, and the timing of their release.
- Deterministic accident source terms requirements for LWRs in 10 CFR 50.67
- TID-14844 makes prescriptive radionuclide release assumptions for a LOCA leading to core melt as the bounding event
- Instant release of 100% of noble gases, 50% of halogens, 1% of remaining solids to the containment
- NUREG-1465 specifies unique BWR and PWR releases based on scenarios considered in NUREG-1150 and supplemental analyses
- Prescribed timed (early/late) in/ex-vessel releases accounting for engineered safety features along with uncertainty analyses
- SECY-93-092 sets the regulatory expectations for advanced reactors to rely on more realistic source term evaluations 2 energy.gov/ne
Rationale for MST for Advanced Reactors
- For advanced reactors, source term is not always limited to bounding events
- For some advanced reactor types, very small releases can be anticipated during even AOOs/DBAs
- Broader spectrum of accidents needs to be considered
- Potential for frequent but small vs. infrequent but large releases
- Accidents that could lead to early vs. delayed releases with different radionuclide discharge and emergency response implications
- Radionuclide sources other than the fuel in the reactor core
- Fuel storage, coolant/cover-gas cleanup and chemical processing systems
- Also to support Levels 2/3 PRA and EPZ reduction requests
- Mechanistic Source Term (MST) Assessments:
- Analysis of radionuclide release, in terms of quantities, timing, and other characteristics, resulting from the specific event sequence being evaluated using best-estimate phenomenological models for the transport of radionuclides from their source to the environment through all holdup volumes and barriers, taking into account the mitigation features.
3 energy.gov/ne
General MST Approach
- Identify radionuclide sources and characterize the potential inventory Inventory associated with each source (core, spent fuel, cleanup system, )
- Identify relevant pathways and phenomena associated with Release radionuclide transport from their source to the environment pathways
- Modeling of the phenomena involved in transport of radionuclides Modeling through identified pathways and barriers
- Application of models to a spectrum of accidents (AOOs, DBAs, and Scenario BDBAs) for evaluation of release rates and frequency evaluation
- Interactions with the regulator as part of pre-application and license Regulatory review process review 4 energy.gov/ne
Release pathways for LMRs Vapors nucleate or condense on particles Mechanical Condensation Deposition Leak Leakage Plugging Leakage 4.Primary Circuit Condensation Cover Gas Mechanical Revaporization Resuspension Deposition Vapors nucleate or Vaporization condense on particles Vaporization Gas/Vapor Transport in Bubble Gas/Vapor Transport Bubble Redissolve 3.Sodium Redissolve Resuspend Resuspend 1.Fuel & 2.Clad Adsorption &
Adsorption Mechanical Deposition Condensation Fission Gas Dissolution
& Dissolution Plenum & Entrainment Legend Vapors &
Retention Particulates Barrier to Release Gases Particulate Radionuclide Path in Fuel Particulate Radionuclide Inhibitor Vapor and Gas Radionuclide Path Vapor and Gas Radionuclide Inhibitor 5 energy.gov/ne
Existing codes MST Approach for LMRs NEAMS Program Tools Developmental Capabilities NRC-supported codes Inventory analysis ORIGEN, REBUS Transient scenario modeling SAS4A, SAM, PERSENT In-pin radionuclide distribution LIFE-M, SAS4A, BISON Radionuclide Release Radionuclide release from failed fuel Module (RRM)
Radionuclide bubble transport Liquid-metal pool radionuclide release Cover gas region analysis MELCOR, CONTAIN-LMR Containment region analysis MELCOR, CONTAIN-LMR Offsite dispersion analysis RASCAL, WinMACCS 6 energy.gov/ne
Trial LMR MST Application
- AFR-100 design
- Trial MST calculations for a 100 MWe pool type small modular SFR with metallic fuel
- ANL-ART-49: http://www.ipd.anl.gov/anlpubs/2016/11/131283.pdf
- Two transient scenarios considered as example cases
- PLOF+ Long, slow heat-up of core and primary system with fuel failures but no melting
- UTOP+ Quick rise in fuel temperatures leading to melting, but with primary system at near-nominal conditions
- Sequences are typically selected based on a risk-informed approach using PRA
Conclusions:
7 energy.gov/ne
Other LMR MST Applications
- GE-Hitachi
- MST findings as a major part of PRISM PRA update/modernization effort (2-year collaboration between GEH and Argonne)
- TerraPower
- Company-funded work at Argonne to repeat trial MST calculation for TWR design
- Korean Atomic Energy Research Institute (KAERI)
- KAERI-funded effort for preliminary source assessments and experiments are Argonne to support PGSFR licensing
- Supports development of RRM for metallic fuel
- Fauske & Associates
- GAIN voucher to Argonne for coupling SAS4A with FATE for LMR source term assessments (initial application to WEC LFRs)
- Supports development of RRM for oxide fuel
- Two 2018 DOE-NE NEUP awards to UWM and UNM
- Radionuclide retention tests in liquid sodium and liquid lead 8 energy.gov/ne
Release Pathways for HGTRs 9 energy.gov/ne
Multiple Barriers for HGTRs Pyrolytic Carbon (Inner and Outer)
Silicon Carbide Porous Carbon Buffer Fuel Kernel Multicell reinforced concrete reactor building based on IBC, ACI, AISC standards COMPACTS FUEL ELEMENTS Helium Pressure Boundary fabricated based on ASME B&PV Code Section III 10 energy.gov/ne
- Radionuclide retention within fuel during normal operation
- Relatively low inventory inside HPB from defective fuel particles
- Limiting off-normal events characterized by
- an initial release from the HPB depending on the size of leak/break/pressure-relief
- a larger, delayed release from the fuel at elevated temperatures
- Functional containment concept to meet 10 CFR 50.34 requirements and EPA PAGs with wide margin for spectrum of off-normal events
- Coated fuel particle is the primary barrier to radionuclide release during normal operation (due to initially defective particles) and off-normal events (through diffusion and recoil of radionuclides at elevated temperatures for long transients) 11 energy.gov/ne
Existing codes MST Approach for HTGRs NEAMS Program Tools Developmental Capabilities NRC-supported codes Inventory analysis ORIGEN Transient scenario modeling RELAP, PRONGHORN Fuel response to scenario studied PARFUME, BISON Radionuclide release rates from fuel BISON, TRISO fuel QA HPB radionuclide release MELCOR Reactor building analysis MELCOR Offsite dispersion analysis RASCAL, WinMACCS 12 energy.gov/ne
- HTGR Mechanistic Source Term White Paper (INL/EXT-10-17997)
- https://www.osti.gov/biblio/989901-htgr-mechanistic-source-terms-white-paper
- Scoping Analysis of Source Term and Functional Containment Attenuation Factors (INL/EXT-11-24034)
- https://www.osti.gov/biblio/1037782 13 energy.gov/ne
- Functional containment concept can also apply to MSRs (with dissolved fuel) and FHRs (with solid fuel)
- For FHRs, coated fuel particle is also the primary barrier to radionuclide release during normal operation and off-normal events
- For MSRs, molten salt can retain almost all non-gaseous fission products
- Source term will likely be dominated by minor releases during incidents more likely than BDBAs
- Leaks from molten salt chemical processing system (may be inside or outside the containment)
- Tritium generated during irradiation of molten salts (especially those containing Li) will require approaches to prevent its uncontrolled release to the environment
- assess its rate of generation in the core,
- trace its movement in and out of the core with moving molten salt,
- monitor its accumulation in critical locations (i.e., cover gas space),
- establish mechanisms for collection and purge
Existing codes MST Approach for MSRs NEAMS Program Tools Developmental Capabilities NRC-supported codes Inventory analysis ORIGEN Transient scenario modeling SAM, Nek5000 Molten-salt chemistry modeling Radionuclide release rates from salt Cover gas region analysis MELCOR Reactor building analysis MELCOR Offsite dispersion analysis RASCAL, WinMACCS 15 energy.gov/ne
Summary and Conclusions
- Regulatory expectations for mechanistic source term evaluations for advanced reactors
- For some advanced reactor designs without cliff-edges, source term will include minor release from incidents more likely than BDBAs
- Also to support EPZ reductions
- Trial LMR and HTGR MST calculations based on a combination of legacy DOE and NRC codes with identified gaps (data and codes):
- Radionuclide release rates from failed fuel
- Bubble transport (scrubbing) in LMRs
- Retention in molten salt (MSRs, FHRs) and liquid metal coolants (LMRs)
- Emerging capabilities can play significant role in analysis of phenomena leading to fuel failure and radionuclide release
- SAM, BISON, PRONGHORN, Nek5000, and new MSR chemistry modeling tools
- Can remove the empiricism embedded in the legacy codes
- These capabilities interface with well-established radionuclide tracking capabilities of NRC codes like MELCOR and CONTAIN to support mechanistic source term assessments and Level 2/3 PRA 16 energy.gov/ne
Questions?
17 energy.gov/ne
Concluding Remarks DOE Briefing to ACRS:
(ATF and Advanced Reactors) Advanced Computer Models for Reactor Safety Applications August 21, 2018
Notional ATF Code Maturity Doped- Coated FeCrAl SiC/SiC U3Si2 Non-cylindrical UO2 cladding cladding cladding metallic fuel Fuels BISON BISON BISON BISON BISON BISON T-H CTF, CTF, CTF, CTF, CTF, CTF, CFD CFD CFD CFD CFD CFD Neutronics Shift, Shift, Shift, Shift, Shift, Shift, MPACT MPACT MPACT MPACT MPACT MPACT Where Green = Mature capability exists, with limited validation Where Yellow = basic capability exists, but key development required Where Red = conceptional capability 2 energy.gov/ne
Notional Non-LWR Code Maturity SFR HTGR FHR MSR Neutronics MCC-3, MCC-3, MCC-3, MCC-3, PROTEUS, PROTEUS, PROTEUS, PROTEUS, Rattlesnake Rattlesnake Rattlesnake Rattlesnake Fuels BISON BISON BISON In progress chemistry code T-H Nek-5000, Nek-5000, Nek-5000, Nek-5000, SAM, Pronghorn, Pronghorn, SAM SOCKEYE SAM SAM Source term SAM, Pronghorn, SAM, Nek-5000, SAM, BISON BISON Pronghorn, Nek5000 BISON 3 energy.gov/ne