ML20147A301
| ML20147A301 | |
| Person / Time | |
|---|---|
| Issue date: | 05/05/2020 |
| From: | Zena Abdullahi Advisory Committee on Reactor Safeguards |
| To: | |
| Abdullahi, Z, ACRS | |
| References | |
| NRC-0902 | |
| Download: ML20147A301 (165) | |
Text
Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION
Title:
Advisory Committee on Reactor Safeguards Thermal Hydraulic Phenomena Subcommittee Docket Number:
(n/a)
Location:
teleconference Date:
Tuesday, May 5, 2020 Work Order No.:
NRC-0902 Pages 1-112 NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers 1323 Rhode Island Avenue, N.W.
Washington, D.C. 20005 (202) 234-4433
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 2
3 DISCLAIMER 4
5 6
UNITED STATES NUCLEAR REGULATORY COMMISSIONS 7
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 8
9 10 The contents of this transcript of the 11 proceeding of the United States Nuclear Regulatory 12 Commission Advisory Committee on Reactor Safeguards, 13 as reported herein, is a record of the discussions 14 recorded at the meeting.
15 16 This transcript has not been reviewed, 17 corrected, and edited, and it may contain 18 inaccuracies.
19 20 21 22 23
1 UNITED STATES OF AMERICA 1
NUCLEAR REGULATORY COMMISSION 2
+ + + + +
3 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4
(ACRS) 5
+ + + + +
6 THERMAL HYDRAULIC PHENOMENA SUBCOMMITTEE 7
+ + + + +
8 TUESDAY 9
MAY 5, 2020 10
+ + + + +
11 The Subcommittee met via Teleconference, 12 at 1:00 p.m.
- EST, Walter
- Kirchner, Chairman, 13 presiding.
14 COMMITTEE MEMBERS:
15 WALTER KIRCHNER, Chairman 16 RONALD G. BALLINGER, Member 17 DENNIS BLEY, Member 18 JOSE MARCH-LEUBA, Member 19 DAVID A. PETTI, Member 20 JOY REMPE, Member 21 22 ACRS CONSULTANT:
23 MICHAEL CORRADINI 24 STEVE SCHULZ 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
2 DESIGNATED FEDERAL OFFICIAL:
1 ZENA ABDULLAHI 2
3 ALSO PRESENT:
4 PAUL CLIFFORD, NRR 5
TOM EICHENBERG, Tennessee Valley Authority 6
7 8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
3 T-A-B-L-E O-F C-O-N-T-E-N-T-S 1
PAGE 2
Opening Remarks...............
4 3
Regulatory Requirements...........
7 4
Timeline and Stakeholder Comments...... 25 5
Revise Guidance and Analytical Limits 6
Reactor Coolant System Pressure
.... 39 7
Damaged Core Coolability
....... 47 8
Radiological Consequences
....... 51 9
Cladding Hydrogen Uptake Models....... 84 10 Burnup Extension
.............. 88 11 Public Comment 107 12 Discussion 109 13 Adjourn..................
112 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
4 P R O C E E D I N G S 1
1:04 p.m.
2 CHAIR KIRCHNER:
Okay.
Let me 3
ceremoniously bang the table with a gavel. The 4
meeting will now come to order. This is a meeting of 5
the ACRS Metallurgy and Reactor Fuels Subcommittee and 6
Thermal Hydraulics Subcommittee of the Advisory 7
Committee on Reactor Safeguards. I am Walter 8
Kirchner, Chairman of today's subcommittee meeting.
9 ACRS members in attendance are -- and at this point, 10 I'll just ask for an affirmation if you're there. Ron 11 Ballinger?
12 MEMBER BALLINGER: Yep, I'm here.
13 CHAIR KIRCHNER: Dennis Bley?
14 (No response.)
15 CHAIR KIRCHNER: Jose March-Leuba?
16 MEMBER MARCH-LEUBA: Yes, I'm here.
17 CHAIR KIRCHNER: David Petti?
18 MEMBER PETTI: Here.
19 CHAIR KIRCHNER: Joy Rempe?
20 MEMBER REMPE: Here.
21 CHAIR KIRCHNER: Matt Sunseri?
22 PARTICIPANT: Walt, Matt said that he was 23 not going to be able to participate.
24 CHAIR KIRCHNER: Okay. Thank you. And I 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
5 also note that Pete Riccardella would not be able --
1 he had a conflict as well. Vesna Dimitrijevic?
2 MEMBER BALLINGER: Vesna said that she was 3
not going to be able to attend.
4 CHAIR KIRCHNER: Okay. Thank you. Now 5
have I left anyone out? Joy Rempe?
6 MEMBER REMPE: You already asked, and yes, 7
I'm here.
8 CHAIR KIRCHNER: Okay. Sorry, Joy. All 9
right.
10 MEMBER REMPE: Not a problem.
11 CHAIR KIRCHNER:
All right.
And 12 consultants, I believe we have Mike Corradini?
13 MR. CORRADINI: Yes.
14 CHAIR KIRCHNER: And Steve Schulz?
15 MR. SCHULZ: I'm here.
16 CHAIR KIRCHNER: Thank you. Okay. Zena 17 Abdullahi is the Designated Federal Official for this 18 meeting. During today's meeting, the subcommittee 19 will hear presentations and will hold discussions with 20 the NRC staff regarding Regulatory Guide 1.236, 21 Pressurized Water Reactor Control Rod Ejection and 22 Boiling Water Reactor Control Rod Drop. There's a 23 long history related to this Regulatory Guide. The 24 staff was -- the staff briefed the ACRS Committee on 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
6 the draft Regulatory Guide 1327, I believe back as far 1
as 2007 and again in October 2016.
2 In today's meeting, the staff will present 3
the finalized draft of DG-1327 which is now Regulatory 4
Guide 1.236. This meeting is open to the public. The 5
ACRS section of the U.S. NRC public website provides 6
our charter, bylaws, agendas, letter reports, and full 7
transcripts of all open and full and subcommittee 8
meetings, including slides presented there.
9 The meeting notice and agenda for this 10 meeting are posted there. We have received no written 11 statements or requests to make an oral statement from 12 the public. At this point, the Committee will gather 13 information, analyze relevant issues, facts, and 14 formulate proposed positions and actions as 15 appropriate for deliberation by the full committee, 16 and that would be in June.
17 A transcript of the meeting is being kept 18 and will be made available at the NRC website. This 19 meeting is being held virtually as part of the COVID-20 19 preventative measures. We ask everyone except the 21 subcommittee members and presenters to mute their 22 microphones until the public session starts as shown 23 in the meeting agenda.
24 The meeting participants should first 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
7 identify themselves and speak with clarity and volume 1
so that they may be readily heard. We will now 2
proceed with the meeting, and I will start by calling 3
on Paul Clifford of NRR. Good afternoon, Paul. Are 4
you there?
5 MR. CLIFFORD: Good afternoon. I am here 6
and ready.
7 CHAIR KIRCHNER: Okay. Did anyone from 8
NRR management also wish to make a statement before we 9
begin?
10 MR. CLIFFORD: I do not believe that is 11 the case.
12 CHAIR KIRCHNER: Okay. Thank you, Paul.
13 Why don't you go ahead then, Paul. We look forward to 14 your presentation. I might note that this is very 15 timely.
16 MR. CLIFFORD: Well, thank you very much.
17 Okay. So hello, everybody. Good afternoon. Welcome 18 to virtual ACRS. My name is Paul Clifford, and I am 19 the Senior Level Advisor for Reactor Fuel in the 20 Office of Nuclear Reactor Regulation. I've been with 21 the NRC in the Division of Safety System since 2003.
22 Prior to that, I worked in the commercial nuclear 23 industry for 16 years, starting in Combustion 24 Engineering in Connecticut, then moving to Arizona to 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
8 work at the Palo Verde Nuclear Generating Station.
1 And finally, I moved to Maryland and worked at Calvert 2
Cliffs.
3 The focus of my 30-plus years of 4
experience has been nuclear fuel design and 5
performance, poor reload safety analyses, and plant 6
operations. Today, I will be presenting Reg Guide 7
1.236 which provides guidance for evaluating a nuclear 8
reactor's initial response to a postulated control rod 9
ejection or control rod drop accident. So let's 10 begin.
11 Today, we will start off with an overview 12 of the postulated reactivity insertion accidents and 13 applicable regulatory requirements, then a timeline of 14 the staff's guidance and how it evolved with an 15 expanding empirical database. I will describe 16 stakeholder comments received by two public comment 17 periods along with major changes to the guidance 18 prompted by those comments.
19 Next, I will walk through the guidance 20 with emphasis on what has changed since 2016 which is 21 the last time I briefed the ACRS, then I will describe 22 cladding hydrogen uptake models which were developed 23 to aid in the implementation of this guidance. And 24 finally, I will describe staff efforts to support a 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
9 relatively new industry initiative on fuel rod burnup 1
extension.
2 CHAIR KIRCHNER: Paul, this is Walt. Just 3
to show you that these virtual meetings work, I can 4
interrupt you. May I ask very simply why you went 5
from the previous reactivity insertion accidents title 6
for the Reg Guide to the rather more prescriptive PWR 7
rod ejection, BWR rod drop title?
8 MR. CLIFFORD: Well, I think there's 9
always been some confusion --
10 CHAIR KIRCHNER: Go ahead.
11 MR. CLIFFORD: Okay, sorry. I think 12 there's always been some confusion with the title and 13 classification of these types of accidents. People 14 referred RIAs as reactivity insertion accidents. Some 15 referred to them as reactivity initiated accidents.
16 I was never a big fan because there are many accidents 17 which involve reactivity insertion, and the accident 18 is actually not initiated from reactivity perspective.
19 It's initiated from the failure or the movement of a 20 particular blade or a control element.
21 And as I mentioned, the guidance is 22 stylized to these particular classes of events and not 23 all accidents involved in reactivity. So that's kind 24 of why I moved back because we're really dealing with 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
10 how to properly analyze these accidents and to assess 1
the impacts and to show compliance with the 2
regulations.
3 CHAIR KIRCHNER: Thank you. I would 4
submit that the material that's in the Reg Guide has 5
broad applicability beyond just the ejection and drop 6
accidents.
7 MR. CLIFFORD: Yeah, I understand.
8 MEMBER MARCH-LEUBA: Yeah. Paul, this is 9
Jose. I wanted to agree again with my colleague, 10 Walt, that almost all of this Reg Guide applies to 11 reactivity excursion events, whether it is because of 12 a rod or something else. And the question I have is 13 by changing the
- title, are we changing the 14 requirements a plant would have to satisfy if they 15 have a reactivity excursion by any other means?
16 I realize this is a Reg Guide. It's not 17 a rule. But are we saying if you have a reactivity 18 excursion, say, for example, you are injecting cold 19 unborated water into the core and producing a critical 20 event, would you generalize that differently, that rod 21 ejection?
22 MR. CLIFFORD: Yes, you would. So there 23 are different classifications of events, I'm sure 24 you're aware. There are reactivity events that are 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
11 classified as AOOs. There are others that are 1
classified as postulated accidents. And each one of 2
the accidents can have a different set of acceptance 3
criteria. I understand that the reactor kinetics are 4
the same. However, the acceptance criteria and the 5
guidance and the GDCs for which you're judging 6
compliance are different for the different accidents.
7 MEMBER MARCH-LEUBA: So it would be based 8
on the frequency for the excursion event that would 9
make it more like a severe accident in which you can 10 have more relaxed acceptance criteria?
11 MR. CLIFFORD: That's correct. Generally, 12 the lower the probability of the initiating event, the 13 higher allowable consequences.
14 MEMBER MARCH-LEUBA: But the consequences 15 are more severe. The consequences are pretty severe, 16 even on the control rod. But okay, we'll go through 17 this during your presentation.
18 MR. CLIFFORD: Right, okay. So sounds 19 like I'm in trouble. We haven't even gotten off the 20 title page yet. Okay. Moving on to Slide 3. The 21 reason for concern with this type of accident, the 22 safety significance of a reactivity insertion accident 23 is evident from the fatal accident that occurred at 24 the U.S. Army's prototype modular reactor, SL-1.
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
12 On January 3rd, 1961, SL-1 experienced a 1
prompt critical power excursion felt almost 2
instantaneously by a steam explosion as a result of an 3
improper central control rod withdrawal. All three 4
reactor operators were killed as a result of the 5
physical trauma received in the violent explosion.
6 And even if the operators had shielded themselves from 7
the explosion, lethal amounts of radiation was 8
released into the building. This old INL safety 9
poster shows what remained of the SL-1 reactor. It is 10 a good reminder.
11 MEMBER MARCH-LEUBA: Yes, Paul. I know 12 you will talk over all of Slide 3. I am very glad you 13 presented this picture because this is a visual of how 14 important this Reg Guide is. We are trying to 15 protect. We're making sure that the plants have the 16 proper power-dependent and pressure limits or what 17 their limits you -- so this doesn't happen.
18 Now in a different time this week, we've 19 been talking about a different reactivity event. Is 20 there anything special in this event that was because 21 it was a rod? You have put up prompt criticality 22 source of reactivity in the core. It would have 23 behaved differently, say, for example, a cold 24 unborated water slug.
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
13 MR. CLIFFORD: Well, certainly, yes. So 1
the width of the pulse has a significant effect on the 2
performance and response of the fuel through that 3
pulse. I mean, this type of prompt reactivity occurs 4
as a result of the quick withdrawal of a single rod 5
element results in a very localized prompt critical 6
excursion. And it's the height and the width of the 7
response in a local reactor power that influences how 8
the rod responds, whether the rod fails, and whether 9
you melt fuel or fail cladding or in a resulting --
10 MEMBER MARCH-LEUBA: Is it whether you 11 have enough reactivity? If you have 0.9 dollars of 12 reactivity, this will happen. So you have 1.2, 1.3, 13 this will happen. If you have 1.05, it may not 14 happen.
15 MR. CLIFFORD: That is correct. The width 16 of the prompt pulse is directly related to the 17 amplitude. So --
18 MEMBER MARCH-LEUBA: So the insertion 19 reactivity or positive reactivity, the insertion rate 20 is important on the development. But it's not because 21 it is a rod. It is because rods can move fast where 22 other types of reactivity essentially may or may not 23 move that fast.
24 MR. CLIFFORD: I agree. I agree.
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
14 Fundamentally, you're designing your core. And if you 1
think about it, you have other accidents where you 2
rely upon the actions of a lot of safety-related 3
systems. Your reactor trips or ESFAS or ECCS or 4
something to mitigate the consequences.
5 This event, this accident -- this family 6
of accidents I should say is really developed to limit 7
the design of the fuel, as you mentioned, how much 8
work you can have in a single blade, how much 9
enrichment you can put in the core. You're designing 10 the core so that the initial response, the prompt 11 critical response would be limited and that you'll 12 also have inherent feedback from Doppler that'll turn 13 the power around before you even get to the point 14 where you're relying on engineered safety functions to 15 act.
16 MEMBER MARCH-LEUBA: Yeah, which you 17 probably don't have time to react if you're in prompt 18 critical.
19 MR. CLIFFORD: Absolutely.
20 MEMBER MARCH-LEUBA: Okay. I just wanted 21 to emphasize that this is a really good visual of what 22 we're trying to achieve today. We're going to prevent 23 this.
24 MEMBER BLEY: This is Dennis Bley. Since 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
15 you brought this one up, SL-1 as I recall was very 1
highly enriched, like, over 90 percent. So we had no 2
Doppler in this plant. It was all thermal response 3
that shut it down. Is that true?
4 MR. CLIFFORD: I'm not 100 percent sure.
5 I mean --
6 MEMBER MARCH-LEUBA: I am looking at the 7
size of the core, and I agree with Dennis. This 8
cannot be other than highly enriched core if you were 9
to make it critical.
10 MR. CLIFFORD: Okay. So as this poster 11 shows, it's important that we protect against this 12 class of accidents. And I believe that this nuclear 13 accident was likely in the minds of the Atomic Energy 14 Commission when they drafted the General Design 15 Criteria in 10 CFR Part 50 just a few years later.
16 Let's get to the specific regulatory requirements.
17 Appendix A, GDC 28 limits the amount and 18 the rate of reactivity insertion to protect the 19 reactor coolant pressure boundary and ensure a 20 coolable geometry. Essentially, it looks like it's 21 written to protect against this. In addition to the 22 reactivity requirements, there are dose requirements 23 to the general public located in 10 CFR Part 50 and 24 Part 100. Reg Guide 1.236 provides and acceptable 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
16 means to meet these regulations.
1 Reactivity insertion accidents are safety 2
significant because of their potential ability to 3
challenge fuel rod integrity, fuel bundle geometry, 4
and the integrity of the reactor coolant boundary.
5 The uncontrolled movement of a single control rod out 6
of the core results in a positive reactivity insertion 7
that promptly increases local power and is considered 8
the limiting reactivity insertion accident. Of the 9
various postulated single failures to the control rod 10 drive system which would lead to an uncontrolled 11 movement of a single control rod, the PWR control rod 12 ejection and BWR and control drop are considered the 13 most limiting scenarios for the current operating 14 fleet.
15 MEMBER MARCH-LEUBA: Hey, Paul. Sorry.
16 This is Jose again. Sorry to interrupt, but this is 17 how real meetings go. We are always interrupting you.
18 A year or two ago over several review cycles, we have 19 wondered if you have to assume an additional single 20 failure.
21 So if you initiate the event, it's a 22 control rod ejection. Do you need to assume that 23 another rod doesn't scram, if it fails to scram? And 24 we worked this with the staff. And at the end, I'm 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
17 not sure how we resolved it. But they convinced us 1
that no, you don't have to. What's your view on that?
2 MR. CLIFFORD: I think when you get into 3
the assume single failure, you're in the GDC 27. And 4
this Reg Guide does not address GDC 27. But from my 5
experience working in the industry, we would always 6
assume that instead of an N-1 scram worth, it would be 7
an N-2 scram worth for the rod ejection event.
8
- However, in the current plant 9
configurations of the operating fleet, N-1, N-2 10 doesn't make a difference because this is not a 11 limiting shutdown margin accident. And as I 12 mentioned, it's initially turned around by Doppler.
13 So the assumption of an N-1 or N-2 scram configuration 14 does not affect this accident.
15 MEMBER MARCH-LEUBA: It doesn't affect the 16 first ten seconds of the transient, but it makes an 17 impact 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> behind. So basically you're telling 18 me that this Regulatory Guide is silent on the single 19 failure criteria and other GDCs would control that?
20 MR. CLIFFORD: Correct. We focus -- the 21 stated purpose of this Reg Guide is that it provides 22 guidance on how to evaluate the initial response of 23 the reactor. In other words, the first five seconds 24 and how you show compliance with those applicable 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
18 regulations in 28 and in the dose requirements which 1
all are occurring in the first five seconds.
2 MEMBER MARCH-LEUBA: Yeah, our job as ACRS 3
members is to try to look a little bit in the future 4
or what is changing instead of just blindly applying 5
the rule. And one thing that's changing is reactors 6
are becoming smaller and they have less control rods.
7 So when it's in a very large PWR, N-1 or N-2 makes no 8
difference.
9 In these small reactors, N-2 is a lot 10 minus 2. When N is only 10, that minus 2 makes a big 11 difference, but I get your meaning. This Reg Guide 12 does not address the issue, and this has to be address 13 by other Reg Guides or GDCs.
14 MR. CLIFFORD: Yes, 100 percent.
15 MEMBER MARCH-LEUBA: Yeah, okay. Thank 16 you.
17 MR. CLIFFORD: Okay. Now getting into the 18 accident sequence for these postulated accidents. The 19 control rod ejection is postulated to occur because of 20 the mechanical failure which causes an instantaneous 21 circumferential rupture of the control element drive 22 mechanism housing or its associated nozzle. This 23 results in the reactor coolant pressure ejecting the 24 control rod and drive shaft to the fully withdrawn 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
19 position.
1 BWR sequence of events is a little 2
different. A control blade is inserted into the core 3
and becomes decoupled from the drive mechanism. The 4
drive mechanism is subsequently withdrawn. The 5
control blade is assumed to be stuck in place, and at 6
a later moment, the control rod suddenly falls free 7
and drops to the control rod drive position.
8 MEMBER MARCH-LEUBA: This is Jose again.
9 Sorry to be disrupting what you were planning to say.
10 But we have also been talking in our meetings about 11 especially on the PWR event what we call a missile.
12 So if the control rod housing breaks and the control 13 rod gets ejected, I think this new Reg Guide is mostly 14 worried or concerned about the thermodynamics inside 15 the core.
16 We were concerned, does that ejected 17 control rod has to be treated as a missile and see 18 what it hits? We know that some reactors have missile 19 shields for such events. I'm going to say again that 20 your Reg Guide is silent on this issue.
21 MR. CLIFFORD: Yes, this Reg Guide is 22 silent on that issue because that involves the 23 mechanical design of the reactor vessel itself. And 24 that responsibility resides in the Division of 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
20 Engineering, and that's a different section. I 1
believe that's SRP Section 394 or 395, and there's 2
guidance in that Reg Guide on the -- what's the word 3
I'm looking for -- the integrity of the vessel and the 4
pedigree at which the vessel is manufactured. And it 5
involves GDC 14 compliance. So once again, this Reg 6
Guide doesn't get into GDC 14.
7 MEMBER MARCH-LEUBA: Yeah, the question we 8
were -- I'm just bringing up things we've talked about 9
in our previous meetings over the last couple of 10 years. And we were not very sure whether -- who's in 11 charge of reviewing this missile or even if you have 12 to review it as a missile or not? This guy -- what 13 we're reviewing today is not covered by that. That 14 should be some other topic of responsibility.
15 MR. CLIFFORD: Yes, it needs to reside in 16 the Division of Engineering because they are 17 responsible for reviewing the reactor vessel.
18 MEMBER MARCH-LEUBA: And changing the 19 topic but it's related. We also talked about when you 20 open the hole on the vessel, you have the first five 21 seconds are the most important ones when you have the 22 power excursion and the Doppler turned it around. But 23 you are also creating a LOCA.
24 Is this supposed to be covered by the 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
21 small-break LOCA analysis that was performed somewhere 1
else in Chapter 15? I'm going to say again you're 2
silent on it because you can have the first five 3
seconds, the big power pulses. Those are the 4
important ones. But then as you -- if you have a 5
LOCA, you have to make sure you don't uncover the 6
core, right?
7 MR. CLIFFORD: Right. So I would say that 8
this reactivity initiated accident class of events was 9
stylized from a point of limiting fuel rod design and 10 control element design so that you would avoid the 11 catastrophic failure and show compliance with GDC 28.
12 A break in the reactor vessel that would result in a 13 slow depressurization of the RCS would then require 14 engineered safety functions and systems to limit the 15 consequences of that type of very slow evolving loss 16 of coolant accident. So ultimately, your ECCS would 17 be actuated and your ECCS would be relied upon to 18 ensure long-term cooling (telephonic interference) and 19 the removal of decay heat.
20 MEMBER MARCH-LEUBA: I understand your 21 point of view. What I'm worried -- I've been worried 22 for several -- I've been a member for now three years, 23 and I worry there is some discontinuity, some 24 compartmentalization of the review. My job is to look 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
22 at the first five seconds. Somebody else will look at 1
the LOCA.
2 I would have loved to see this RG point or 3
assess and verify that the small-break LOCA spectrum 4
analysis covers the break that you discussed today.
5 And just a simple -- ensure that you -- because the 6
guys that do the small-break LOCA -- look at pipes 7
that connect to the vessel. That's where things 8
break. And now we're postulating that a control rod 9
housing is breaking and it's a different size, 10 different location. How do we know it's covered?
11 And what I'm worried -- and I'm not saying 12 this happened -- is Paul is telling me, I made sure 13 the electronics were okay. My -- temperature didn't 14 melt. I didn't violate any of my calorie per grams 15 limit. And the small-break LOCA is on Chapter 15 and 16 say, well, we look at all the pipes and when they 17 break, nothing happens, and nobody looked at the 18 housing breaking.
19 MR. CLIFFORD: I agree with what you're 20 saying, and I understand where you're going. And that 21 really comes from the discontinuity in 50.46 because 22 50.46 which is a regulation on -- which governs the 23 design of your emergency core cooling system, 24 specifically limits breaks to those in piping and not 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
23 in penetrations in the reactor vessel.
1 So there could be multiple penetrations in 2
the reactor vessel. And if you were to break those, 3
then you would have a loss of coolant accident. But 4
it's outside of the regulatory requirements for the 5
ECCS system. So in other words, you're not designing 6
your ECCS to mitigate the consequences of a vessel 7
break. And I don't know the history of all of that 8
because, once again, that really resides in the 9
Division of Engineering. But I believe it's because 10 the penetrations are all designed to GDC 14 standards.
11 And GDC 14, to meet those standards, you have to show 12 that there's an extremely low probability of gross 13 failure of those vessel penetrations.
14 MEMBER MARCH-LEUBA: By this, you are 15 assuming it broke because you're saying my rod ejects.
16 CHAIR KIRCHNER: Again, Jose, this is --
17 because as Paul used the operative words, this is the 18 stylized event specifically targeted at GDC 28 and 19 hence sets a limit on things like control rod work or 20 reactivity in the core, et cetera, for purposes of GDC 21 28.
22 MEMBER MARCH-LEUBA: Yeah. I'm not saying 23 I see any problem anywhere. I'm just trying to make 24 people think if you see there are holes, and I see 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
24 some discontinuities. I'm not seeing a problem. I 1
mean, it depends on what people think how we're doing 2
this stuff. Please do continue.
3 MR. CLIFFORD: Okay. So next, we will 4
discuss a timeline and stakeholder comments. Now onto 5
Slide 8. This timeline shows the evolution in 6
regulatory guidance beginning in 1974 with the release 7
of Reg Guide 1.77 and ending with today's briefing.
8 Note the two public comment periods for DG-1327 which 9
was subsequently designed Reg Guide 1.236. And now 10 we'll go on to Slide 9 and talk about each of these.
11 Before we begin with the requirements or 12 the guidance, it's important to know that in-pile 13 prompt critical power excursion type testing is 14 required to understand the phenomena, to identify 15 damage mechanisms and failure modes, to calibrate 16 analytical
- models, and
- finally, to establish 17 analytical limits to ensure acceptable fuel 18 performance. In 1974, the year Reg Guide 1.77 was 19 issued, the empirical database consisted of ten in-20 pile property activity insertion tests conducted 21 within the capsule driver core of the Special Power 22 Excursion Reactor Test program, referred to as SPERT-23 CDC. These tests were conducted in 1969 and 1970.
24 Reg Guide 1.77 used the empirical 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
25 database, although limited at the time, to provide 1
acceptable analytical methods and models and 2
assumptions for modeling a PWR control rod ejection.
3 In addition, it provided coolable geometry limits, 4
limits on reactor coolant system pressure, and limits 5
on radiological consequences, and these are provided 6
here on the slide. Two hundred and eighty calories 7
per gram was the peak radial average fuel enthalpy 8
allowed. Reactor system pressure was limited to 9
Service Level C, and offsite dose consequences were 10 limited to well within the guidance in Part 100.
11 Now well within translates to 25 percent.
12 So for the maximum hypothetical accident which is of 13 a lower frequency, the allowance would be 100 percent 14 of 10 CFR Part 100. But for this class of accidents 15 which is considered more likely, the radiological 16 consequences are limited to 25 percent of the limits.
17 MR. CORRADINI: Paul?
18 MR. CLIFFORD: Yes.
19 MR. CORRADINI: Can I just get a 20 clarification? I think I know what peak radial 21 average fuel enthalpy means. That means you're 22 looking for the peak in the radial direction, the 23 average within the pellet. Is that correct?
24 MR. CLIFFORD: Correct.
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
26 MR. CORRADINI: Okay. Which means you get 1
some sort of spatial distribution in the pellet which 2
has to be then -- I don't want to use the word back 3
calculated but evaluated because there might be 4
regions which are solid and there might be regions 5
where there's melt.
6 MR. CLIFFORD: Correct. But as we'll get 7
into, the allowable melt is very limited.
8 MR. CORRADINI: Okay, understood. And 9
then the second part of this is with UO2, I'm just 10 trying to remind myself of the numbers. With UO2, 280 11 calories per gram average over the pellet is a small 12 amount of melt is in existence if I do just a 13 thermodynamic balance, right?
14 MR. CLIFFORD: Correct. And that portion 15 of molten fuel would increase with burnup as melting 16 point diminishes.
17 MR. CORRADINI: Correct, right. Sorry, 18 that's right because of the change of property.
19 Excuse me. Got it. Thank you.
20 MR. CLIFFORD: Okay. So moving to Slide 21 10, we're going to 1980. So from 1978 to 1980, a 22 series of in-pile tests were conducted at the Power 23 Burst Facility in Idaho. This led to additional data, 24 and this data suggested a need for new analytical 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
27 limits.
1 A paper was published which recommended 2
changes to the existing criteria. The coolability 3
criteria was reduced from 280 calories per gram to 230 4
calories per gram. Cladding failure thresholds were 5
reduced from 170 to 140 for irradiated fuel. It was 6
noted that the failure mode strongly dependent on 7
prior irradiation history. And based upon advanced 8
analytical methods and evaluations, it was concluded 9
that there was no eminent safety concern for the 10 operating fleet. The guidance in Reg Guide 1.77 was 11 not revised. However, many plants voluntarily 12 implemented reduced limits.
13 Moving on to 2004. By 2004, the empirical 14 database had increased significantly with in-pile 15 testing being performed in France, Japan, and Russia.
16 In response to reported PCMI cladding failures at 17 lower than expected deposition energies, the NRC 18 completed a detailed safety assessment. Based upon 19 the research which was conducted, the expanded 20 empirical database, and advanced analytical methods, 21 RIL-0401 concluded that there was no eminent safety 22 concern in the current fleet. However, different from 23 in the past, it was decided that the guidance would be 24 updated based on the latest and greatest information.
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
28 So building upon the expanded empirical 1
database, the NRC issued interim guidance in 2007.
2 There were ACRS briefings before the interim went out.
3 There was a public comment period before the interim 4
guidance went out. The guidance is documented in 5
Appendix B, the Standard Review Plan 4.2.
6 The guidance provides hydrogen-dependent 7
PWR and BWR PCMI cladding failure thresholds, cladding 8
delta P-dependent failure thresholds for high 9
temperature failure
- modes, burnup-dependent 10 coolability criteria, and finally, transient fission 11 gas release. This guidance was classified as interim 12 because the staff was awaiting further test results to 13 help resolve some scaling issues with the data.
14 Moving on to today, Slide 13 presents the 15 current empirical database. Tests conducted since the 16 interim guidance in 2007 are highlighted in red. This 17 data collectively was used in the development of the 18 final guidance within Reg Guide 1.236.
19 MEMBER MARCH-LEUBA: Paul, this is Jose 20 again. Can you spend a little more time on the 21 difference between the black dots and the open dots?
22 What are we looking at here?
23 MR. CLIFFORD: I'm glad you said that. I 24 actually had a note to say something about that, but 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
29 I missed it. Okay. So the open dots -- well, first 1
of all, there's different symbols based on where the 2
tests were conducted. There's a different symbol for 3
SPERT versus PBF versus BIGR and IGR which is -- let's 4
see what that -- those are the test reactors in 5
Russia. Impulse Graphite Reactor, I believe that's 6
what IGR stands for, the French reactor CABRI and the 7
Japanese reactor NSRR.
8 The close symbols are those that failed.
9 The open symbols are those that did not fail. It's 10 important to note that not everybody reported the same 11 information for CABRI and for NSRR. They reported the 12 enthalpy at the time of failure. But for IGR and 13 BIGR, they just reported the maximum enthalpy. So you 14 don't know when it failed. You only know that it 15 failed, whereas NSRR and CABRI, you know when it 16 failed. So it's a little more -- it's much more 17 valuable from that perspective.
18 MR. CORRADINI: Paul, can you -- I guess 19 I wanted you to talk through the red since some of the 20 red failures are at low values, and I guess I want to 21 understand the new data that you emphasized.
22 MR. CLIFFORD: Well, the red was just 23 emphasized as being the most current data. We'll get 24 through how I used this data because the data is all 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
30 used differently, depending on what the failure mode 1
is. And we'll get into that in the coming slides.
2 MR. CORRADINI: Thank you.
3 MEMBER MARCH-LEUBA: So you have -- this 4
is Jose again. You have looked at it, I mean, all 5
your life and understand it. Do you see any 6
correlation with new fuels, new materials, better 7
cladding, or it's all the same?
8 MR. CLIFFORD: We can get into that.
9 Certainly, there are burnup effects and there are 10 corrosion effects and there are also fabrication 11 effects. And all those are considered in this Reg 12 Guide.
13 MR. CORRADINI: But I think what Jose is 14 asking is if it were M5 versus ZIRLO versus Zirc-4 15 versus Zirc-2, is there a difference?
16 MR. CLIFFORD: Yes, and there are 17 different analytical limits for each one of those 18 cladding types.
19 MEMBER BALLINGER: This is Ron. In fact, 20 most of the Zircaloy-2 and Zircaloy-4, you don't see 21 those in cores anymore at all, right?
22 MR. CLIFFORD: Right.
23 MEMBER BALLINGER: The last batch of 24 Zircaloy in a core I think has probably already been 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
31 discharged. So it's either all ZIRLO or M5, right?
1 MR. CLIFFORD: So for the PWRs, I don't 2
believe there's any reactors that are taking fresh 3
batches with Zirc-4. I think that's --
4 MEMBER BALLINGER: Yeah, that's what I was 5
6 MR. CLIFFORD: -- the last one.
7 MEMBER BALLINGER: -- thinking. So the --
8 rates for M5, they're much, much lower than for Zirc-9
- 4. And the hydrogen pickup rate is also much lower 10 than for Zirc-4. Am I right?
11 MR.
CLIFFORD:
- Yes, and that is 12 specifically accounted for in this guidance.
13 MEMBER BALLINGER: Yeah.
14 MR. CLIFFORD: And I'll get to that.
15 CHAIR KIRCHNER: Just a quick break here, 16 Paul. Members, when you ask a question or make a 17 statement, please identify yourself. That will help 18 for people who are not on Skype but are on a public 19 line. Thank you.
20 MR. CLIFFORD: Okay. As we go through the 21 guidance, I'll try to keep that in my mind to make 22 sure I get into the alloy specific aspects of it and 23 the fuel specific aspects of it. If I don't, just ask 24 me some questions. So we're moving on to Slide 14, 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
32 and this is the first public comment period in 2017.
1 So there was an ACRS briefing in 2016.
2 Shortly afterwards, DG-1327 went out for public 3
comment. Comments were received from 12 stakeholders 4
with a total of 124 comments. Over 100 comments were 5
accepted and prompted changes to the guidance. This 6
pie chart below shows the distribution of comments to 7
the various sections of the Reg Guide.
8 And if you quickly look at it, you see the 9
lion's share of comments were in the analytical 10 methods section, the range of applicability and in the 11 failure threshold curves. However, there were 12 comments received in really all aspects of the 13 guidance. And as I mentioned, 100 of the comments 14 were accepted and the NRC made changes to the guidance 15 in response to those comments.
16 In 2017, DG-1327 went out for a second 17 round of public comments. There was comments received 18 from seven stakeholders with a total of 54 comments.
19 Over 30 comments were accepted and prompted changes to 20 the guidance. Here's the distribution of the comments 21 received in 2019.
22 It's worth noting that there a large 23 portion of the comments were on the radiological 24 source term information, and those comments were 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
33 mostly directed at removing that information from this 1
Reg Guide and putting it back into Reg Guide 1.183.
2 There were also comments on the failure threshold 3
curves.
4 It's kind of worth noting that every time 5
we've gone out for
- comments, we received 6
recommendations on how to redraw the figures, one way 7
or another. And each time we've gone out, we have 8
redrawn the figures. And I'll get to that later on, 9
but I think that leads to a little regulatory 10 instability when you start continuous -- when you 11 continuously redraw figures based on the same data 12 set. And I'll get to what the staff's recommendations 13 on that is.
14 MR. CORRADINI: Paul, this is Corradini.
15 One thing that you don't have to answer now but 16 whatever is logical to put in. When you identify peak 17 radial average enthalpy -- fuel enthalpy, is there a 18 spatial area over which this is analyzed? That is, 19 are we looking at a single pin, at a fuel assembly, at 20 a group of fuel assemblies, because I think that has 21 evolved over time given the ability to do better 22 resolution of reactor physics. And I can't remember 23 where that sits now.
24 MR. CLIFFORD: No, you're 100 percent 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
34 correct. It really depends on the granularity of the 1
physics model that's used to calculate that. Going 2
back in time, you would see a physics node would be 3
approximately six inches. But since then, they've 4
grown shorter because the computational abilities are 5
much stronger. So you can afford more detailed 6
resolution when you're solving your equations. So I 7
think it depends on what's approved and when it was 8
approved.
9 MR. CORRADINI: In terms of the analysis.
10 But what I was trying to get at is -- well, I'll ask 11 it again later when you get to loss of coolable 12 geometry. I'm trying to understand the spatial area 13 over which loss of coolable geometry is considered.
14 If I had a single pin or a single two or three pins 15 for whatever that would undergo failure, cladding 16 failure, that's different than would be a fuel 17 assembly. I'm just trying to understand how that's 18 done now with the new guidance.
19 MR. CLIFFORD: Okay. Yeah, we'll get to 20 it in the new guidance. But to answer your question 21 now, it is based essentially down to the pellet. You 22 can now see the coolability criteria down to a pellet 23 or whatever node that the limiting node is. That node 24 might be three or four pellets that are you're 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
35 calculating, but it's down to that one single point in 1
the most limiting rod.
2 MR. CORRADINI: Thank you.
3 MR. CLIFFORD: Okay. So --
4 MR. SCHULZ: Paul, excuse me. This is 5
Steve Schulz.
6 MR. CLIFFORD: Yes.
7 MR. SCHULZ: You mentioned a radiological 8
source terms and fission product release fractions 9
that the number of comments there were disposition by 10 referring them to Reg Guide 1.183, 1.195. Does that 11 mean that those are in process of being redeveloped, 12 and then therefore the information that was developed 13 here that affected those guides is going to be 14 addressed in the future?
15 MR. CLIFFORD: Yes, we are actively 16 updating those Reg Guides now. As you may now, Reg 17 Guide 1.183 has kind of a long history. It went out 18 for public comment in DG-1199, I believe, many years.
19 Lots of comments were received and resolved. However, 20 the Reg Guide was never issued final. That was one of 21 the reasons that prompted me to take the relevant 22 information that was going to be in that revision to 23 that Reg Guide 1.183 and move it into Reg Guide 1.236.
24 Knowing that, it really didn't -- it shouldn't reside 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
36 there but just a means to get it out. But the 1
comments we received were that it shouldn't reside 2
here if it's confusing to the industry because you're 3
now going to have a duplication of information in two 4
different Reg Guides and they're not identical. So --
5 MR. SCHULZ: Right.
6 MR. CLIFFORD: -- the comments were move 7
it back where it belongs to Reg Guide 1.182, and 8
that's what we plan on doing. We're actively updating 9
that and we're pushing to move that forward this year.
10 MR. SCHULZ: This year, okay. That was --
11 the second part of my question was on schedule.
12 MR. CLIFFORD: So we have a major action 13 involving burnup extension. So we need to update that 14 Reg Guide, not just to bring it to the current state 15 of knowledge but also to update it to support the 16 higher burnups.
17 MR. SCHULZ: That's right. And the 18 elements that are associated with the newer fuel 19 designs are also affecting that Reg Guide and that 20 effort?
21 MR. CLIFFORD: Well, certainly, the burnup 22 has a direct impact on the source terms.
23 MR. SCHULZ: Okay. Thank you.
24 MR. CLIFFORD: But the source term was not 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
37 just burnup specific. We needed a new source term 1
that would have a higher power level at extended 2
burnups so that it would support even current fuel 3
management techniques. Right now, people can't meet 4
the Reg Guide 1.183 applicability limit. I believe 5
you can't be higher than 6.3 kilowatts a foot beyond 6
54 gigawatt-days. They can't meet that with modern 7
fuel management.
8 So we needed to expand that power burnup 9
window, and we did that in the early revision to Table 10 3 which is the gap fractions. But now we're going to 11 extend that even further to encompass the extended 12 burnup. So it's both extended burnup and extended 13 power.
14 MR. SCHULZ: Good. Thank you.
15 MR. CLIFFORD: Okay. So in response to 16 two rounds of public comments, we made many changes 17 and I'm just going to go over the major ones. And 18 we'll get into more detail as we get into the specific 19 guidance. But we expanded the allowable fuel burnup 20 or the applicability range of the guidance out to 68 21 gigawatt-days rod average burnup. That's in Section 22 C-1.
23 We improved the analytical requirements.
24 As I mentioned, there were lots of comments to on to 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
38 provide clarity to the requirements or the 1
organization. For instance, we lumped them all 2
together. And as a result of comments, we tried to 3
put the PWR-specific guidance in one section and the 4
BWR-specific guidance in a different section just to 5
make it clear.
6 We revised the PCMI cladding failure 7
threshold curves in Section C-3. We removed the 8
radiological source term information which was put in 9
for the second round, now being taken out for the 10 final. And that included the analytical requirements, 11 the fission product gap release fractions, and an 12 acceptable analytical procedure for calculating design 13 specific gap fractions which has not existed. We 14 amended the implementation section to reflect the 15 Commission's revised backfit guidance, and we added 16 hydrogen uptake models to aid in the implementation.
17 MEMBER REMPE: Paul, this is Joy Rempe, 18 and I have a question that's more related to 19 knowledge-management of the agency. We followed the 20 Reg Guide. We've been given these slides which have 21 more of a technical basis for changes, and there was 22 a memo that was passed to us yesterday.
23 But in some cases when I look at the 24 comments you received and responses back, you made 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
39 some changes. Is there some sort of Technical Basis 1
Document that underlies this Reg Guide for knowledge-2 management within the agency so they can understand 3
more about what's in the Reg Guide and the basis for 4
it?
5 MR. CLIFFORD: Right. So there was a memo 6
made 40 or 50 pages long that provided the technical 7
and regulatory basis for what was in the Reg Guide.
8 And that was developed to support the interim 9
criteria. And then that was revised to then support 10 the changes in the first -- should I say the 2017 11 version of DG-1327.
12 And then further changes to the Reg Guides 13 are document as part of a public comment response 14 table which is also captured in ADAMS. And then we 15 would have then, of course, the second round of public 16 comment responses. We would show the changes to the 17 documents there.
18 So you're right. It is kind of a 19 piecemeal -- if you went back in history, it is a 20 little piecemeal that you would have three sources 21 defining the technical basis for the Reg Guide. It 22 would be the initial memo, then it would be the 23 revisions and the public comment response table from 24 2017, and then it'd be revisions from the public 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
40 comment response table in 2019.
1 MEMBER REMPE: So is that trail at least 2
documented somewhere internally within the agency? I 3
know it's in ADAMS, and I just don't know all the 4
links and things. But if a staffer wants to figure 5
this out, to at least know how to get to the piecemeal 6
trail.
7 MR. CLIFFORD: Yeah, if you -- it's 8
probably the introduction section of the Reg Guide, 9
one of the first pages. There's kind of a history 10 there, and it describes the basis, exactly what I 11 said. And all three of those documents are in ADAMS 12 and they're referenced in the Reg Guide.
13 MEMBER REMPE: I didn't see, like, your 14 memo provided us references, but perhaps I've missed 15 that somewhere in the introduction.
16 MEMBER BALLINGER: This is Ron Ballinger.
17 Talking about the memo related to the 68 gigawatt-days 18 for time?
19 MR. CLIFFORD: No.
20 MEMBER REMPE: That came yesterday --
21 (Simultaneous speaking.)
22 MEMBER BALLINGER: That's yesterday.
23 MR. CLIFFORD: That's yesterday.
24 MEMBER BALLINGER: But in a lot of these 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
41 Reg Guides, there is a technical basis that's in one 1
place. And might we not consider instead of forcing 2
the reader, because this is a fairly important Reg 3
Guide, to go and have sort of ASME codes fingered in 4
five different locations pages and have it in one 5
place that kind of unifies everything. It's 6
especially good for this knowledge-management thing, 7
but just for people's sanity when they want to go look 8
this up.
9 MR. CLIFFORD: No, I mean, that's a good 10 recommendation. That could be a good recommendation 11 for the ACRS to update that.
12 MEMBER REMPE: Thank you.
13 MR. CLIFFORD: So we're at a natural break 14 here. Before we move on to the next topic which is 15 walking through the guidance, are there any more 16 questions on the timeline or the comments we received 17 to date?
18 MEMBER REMPE: I'm not sure if this is a 19 good place or later, and I'm willing to wait till 20 later. But I had a question about the applicability 21 for this Reg Guide. Is this a good place to ask this 22 and your logic for it and why you have what you have 23 there, or should I wait?
24 MR. CLIFFORD: I think if we wait and hold 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
42 that question because I'm going to be walking through 1
the different failure modes and how the different 2
analytical limits are applied to different fields. So 3
maybe that'll answer it naturally as we go through it.
4 MEMBER REMPE: Okay. So this is, like, 5
Section 1.1 of your Reg Guide is where I -- like, you 6
ruled out greater than five percent.
7 MR. CLIFFORD: And so that's the question, 8
why we ruled out greater than five percent?
9 MEMBER REMPE: There's a couple of 10 different ones like that that I'm just wondering what 11 the logic will be. Is everything going to just be 12 case by case in the future? You've talked about the 13 new fuel. Well, they may go to higher enrichment.
14 What about if you have a modular reactor with PWR fuel 15 operating at BWR pressures? Is that something that --
16 are these all just going to have to be case by case?
17 MR. CLIFFORD: Yeah. I mean, if you read 18 the introduction paragraph of the applicability 19 section, it does kind of state that this is applicable 20 to the current operating fleet for the PWRs and BWRs.
21 And the design-specific changes to modular reactors or 22 advanced reactors would need to be addressed on a 23 case-by-case basis. I guess the applicant would need 24 to come in and say why the Reg Guide continues to be 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
43 applicable to their design given whatever changes 1
they're making to that design.
2 It's difficult to try to make any blanket 3
statements on how you could apply this to a different 4
design because I just don't understand what that 5
design is. I mean they could have an entirely 6
different pulse width profile or different blade 7
widths, or there could be a lot of different aspects 8
of it.
9 MEMBER BALLINGER: This is Ron Ballinger 10 again. I mean I think the limits, the way they're 11 posed in terms of differential pressures and burnups 12 and hydrogen content and those kinds of things kind of 13 lay out the field. In an SMR, as long as the SMR fits 14 in those fields, the fact that if one is operating PWR 15 fuel with BWR differential pressure is naturally taken 16 care of, right?
17 MR. CLIFFORD: Yes, in that particular --
18 MEMBER BALLINGER: Anything that would 19 related to the source term is now going to be shifted 20 to Reg Guide 1.183, right?
21 MR. CLIFFORD: Correct. As we go through 22 the presentation, if we haven't addressed it, we can 23 circle back.
24 CHAIR KIRCHNER: Okay. Paul, this is --
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
44 is this a logical break point?
1 MR. CLIFFORD: Yeah, sure.
2 CHAIR KIRCHNER: Okay. Then why don't we 3
take a -- on my NRC computer, it says it's 2:00 p.m.
4 Eastern Time -- Daylight Time. Let's take a 10-minute 5
break and reconvene at 2:10, 10 minutes from now.
6 (Whereupon, the above-entitled matter went 7
off the record at 2:00 p.m. and resumed at 2:10 p.m.)
8 CHAIR KIRCHNER: Okay. Let's reconvene 9
our meeting. This is Walt Kirchner from the ACRS, and 10 this is a meeting of the Thermal Hydraulics and 11 Metallurgy and Reactor Fuel Subcommittee. We are 12 addressing Reg Guide 1.236, and Paul Clifford of the 13 staff is the presenter. So Paul, if you're ready, 14 please continue.
15 MR. CLIFFORD: Okay. Welcome back, 16 everybody. Next, we're going to walk through the 17 guidance with emphasis on what has changed relative to 18 2016 version of DG-1327, the version that was 19 presented to the ACRS. So we're not going to get into 20 intermittent changes between the two public comment 21 periods. We're just going to say this was a line in 22 the sand in 2016, and here's the final.
23 So we begin with the RCS pressure 24 analytical limit. The changes to the text are in 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
45 blue. We modified the text based on comments 1
received, noting that not all existing plants use the 2
same pressure criterion. It's consistent with the 3
current guidance and the practice, and it protects the 4
reactor coolant pressure boundary, hence satisfying a 5
portion of GDC 28.
6 CHAIR KIRCHNER: Now Paul, just for the 7
record, when one does your pressure calculations, you 8
take no credit for the fact that you actually have a 9
LOCA, in the case of the PWR of sizable LOCA. Does it 10 have actual failure of the control rod drive housing?
11 So that is not credited. Is that correct?
12 MR. CLIFFORD: That is true. Any leakage 13 through the postulated break in the control rod drive 14 mechanism housing is not credited. However, it's 15 probably worth noting in the first few seconds when 16 you get your peak pressure even if you did credit it, 17 it probably would not have a big impact since it's --
18 I forget the size. I want to say 0.04 square feet is 19 something that pops in my mind.
20 CHAIR KIRCHNER: Yeah, I think that it's 21 probably in the order of a six-inch flange.
22 MR. CLIFFORD: I think it's smaller than 23 that.
24 CHAIR KIRCHNER: Smaller than that? The 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
46 nozzles would be smaller. I wasn't sure about the 1
flange size.
2 MR. CLIFFORD: I don't know. 0.04 square 3
feet is something that just pops in my head.
4 CHAIR KIRCHNER: Okay. Thank you.
5 MR. CLIFFORD: Okay. Moving on to the 6
next guidance, this is damaged core coolability. So 7
here are some photographs of fuel rod test specimens 8
from the SPERT-CDC test program which I mentioned took 9
place in 1969 and 1970. When we define an analytical 10 limit for damaged core coolability, we are trying to 11 preserve the fuel pellet stack within the fuel rod 12 cladding within a fuel assembly bundle array, 13 essentially a known configuration which limits fuel-14 coolant interaction and also preserves a geometry 15 that's amenable to coolant.
16 MR. CORRADINI: This is -- Paul, this is 17 fresh fuel, Paul? This is Corradini.
18 MR. CLIFFORD: So if we went back and 19 looked at it, I believe this was all fresh fuel --
20 MR. CORRADINI: Thank you.
21 MR. CLIFFORD: -- or really low burnup, 22 within five or ten -- probably five gigawatt-days.
23 MR. CORRADINI: Thank you.
24 MEMBER PETTI: And Paul, the energy 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
47 deposition is an average number? This is Dave. It's 1
not the peak is it, or --
2 MR. CLIFFORD: I believe it was the peak 3
deposited in the fuel.
4 MEMBER PETTI: Okay. Thank you.
5 MR. CLIFFORD: You could almost see the 6
line, how they got to 280 at this point in time. At 7
240, the cladding was failed but it was still a rod 8
structure. It still had the same geometry which could 9
be shown to be coolable. It wasn't -- it released its 10 fission gas, but it didn't melt and the cladding 11 maintained its structure.
12 So onto the guidance itself. Here's the 13 wording that's in the guidance document. There's only 14 editorial changes relative to 2016. The coolability 15 limits remain unchanged. A loss of fuel rod geometry 16 is limited to 230, and that's based on the earlier 17 tests, not just SPERT but also reinforced with PBF and 18 NSRR.
19 The limited centerline melt which is 20 consistent with many plants FSAR design basis. Even 21 though it allows melt, it's limited to 10 percent 22 volume in the centerline region, hence avoiding molten 23 fuel-coolant interaction. It should be noted that 24 fuel melt criteria becomes more limiting at about 30 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
48 gigawatt-days per metric ton of uranium.
1 In other words, you'll stop melting fuel 2
below 230 calories per gram as you increase burnup.
3 So the higher burnup fuel will be limited by the melt 4
criteria. Note that these limits are designed to 5
preserve coolability and satisfy GDC 28. And as I 6
mentioned --
7 MEMBER REMPE: Paul?
8 MR. CLIFFORD: Yes.
9 MEMBER REMPE: This is Joy. I'm sorry.
10 I thought you were done. If you have to finish a 11 sentence, go ahead and finish it first.
12 MR. CLIFFORD: I was just going to remind 13 it's just editorial changes made since 2016.
14 MEMBER REMPE: Okay. When I was looking 15 through this -- and I apologize if I asked you this at 16 the meeting whenever last time we met on it. But why 17 is there so much focus on melting versus liquefaction, 18 because in the severe accident world, we know that 19 fuel and cladding materials can become liquid at lower 20 temperatures than melting.
21 MR. CLIFFORD: Right. So what we're 22 trying to do is really limit the fuel-coolant 23 interaction. So if we're limiting it to just 24 centerline melt, then we're not going to have that 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
49 interaction with the coolant, even if the cladding 1
fails.
2 CHAIR KIRCHNER: This is Walt. Joy, I 3
think the other thing here is the time. This is 4
almost instantaneous whereas in the severe accident 5
world, you're setting time and temperature and you get 6
changes in the phase diagrams, et cetera, et cetera.
7 But this is -- here, we're dealing with a rather short 8
time frame, and I think --
9 (Simultaneous speaking.)
10 MEMBER REMPE: So I think there's --
11 CHAIR KIRCHNER: -- focusing on melt 12 versus other phase -- parts of the phase diagram.
13 MEMBER REMPE: I figured timing would be 14 one of the big ones, but I just was wondering. I 15 don't see any sort of lower temperature phenomena 16 occurring because of the cladding and the fuel 17 becoming liquid. I realize your point about timing is 18 what I suspected the answer would be. But I never 19 noticed that in any of these tests, I guess is what 20 you're saying.
21 MEMBER PETTI: Joy, this is Dave. I can 22 remember being in Idaho when --
23 (Simultaneous speaking.)
24 MEMBER REMPE: Dave, you're going to have 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
50 to talk louder. I just cannot hear you.
1 MEMBER PETTI: Oh, okay. I remember 2
talking to the guys that were doing the fuel-coolant 3
interaction analysis on one of the PBF tests that was 4
particularly Korea. And the severe accident loss was 5
just starting, so they knew about that stuff at the 6
same time. But it never came in. It must have been 7
the time to get an interaction on a phase diagram. It 8
was the kinetics of, frankly, it's got to take some 9
time to occur, and it's just not occurring in the 10 millisecond time scale here.
11 MEMBER REMPE: Yeah, that's what I was 12 wondering, if they never seen it at all. Thank you.
13 MEMBER PETTI: An interesting question, 14 for sure.
15 MEMBER REMPE: I wish you could find a way 16 to get the volume up, Dave. I can kind of hear you.
17 But boy, it's hard compared to other speakers.
18 MEMBER PETTI: I'm talking directly into 19 cell phone. Even on the computer, I can't -- nobody 20 can hear me through my headphones for some reason.
21 CHAIR KIRCHNER: For those of us in Idaho, 22 just yell out the back window.
23 (Laughter.)
24 MEMBER REMPE: I didn't think of that.
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
51 You're right.
1 MR. CLIFFORD: Okay. Moving on to Slide 2
22, we're going to get into fuel failure estimates and 3
radiological consequences. So to perform a proper 4
radiological consequence assessment, first you need a 5
conservative estimate of the number of rods that fell, 6
and then you need an estimate of how much fission gas 7
is released from each of the rods that fell. In the 8
Reg Guide, we provide guidance on estimating the total 9
number of failed fuel rods from several different 10 failure modes.
11 First, there's a high temperature cladding 12 failure threshold which will be applied to prompt 13 critical scenarios. And there's the traditional 14 assumption of cladding failure if you violate your DNB 15 critical power ratio design limits which are applied 16 to non-prompt power excursions, then there's PCMI 17 cladding failure. And finally there's a presumption 18 of cladding failure upon centerline melt. We'll be 19 getting into each one of these.
20 The guidance for estimating or providing 21 a conservative estimate of the amount of fission gas 22 which is released from each failed rod includes 23 steady-state gap inventories, releases during the 24 transient as a result of separation of grain 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
52 boundaries, and there's also additional releases if 1
there's fuel melt. That guidance is being moved to 2
Reg Guide 1.183 as we talked about.
3 So let's get into the guts of the guidance 4
with respect to estimating fuel rod failures. We 5
begin with high-temperature failure modes, and I'm 6
just showing you the guidance. We're going to get 7
into more detail when we get to Slide 25.
8 So the empirical database supporting this 9
failure mode is comprised mostly of tests conducted in 10 Japan NSRR test facility and Russia's BIGR and IGR 11 test facilities. Tests reported which had failures 12 reported due to PCMI were removed from the database.
13 Also, tests that were conducted in the CABRI sodium 14 loop were also removed since the cladding would not 15 have experienced any high temperature excursion.
16 There are two failure modes which are reported. The 17 first is brittle failure due to post-DNB oxygen-18 induced embrittlement, and the second is ductile 19 failure due to post-DNB balloon and rupture.
20 So sensitivity studies. Here's the 21 database plotted not as a function of burnup but as a 22 function of cladding differential pressure. Note that 23 the tests reported peak fuel enthalpy and not enthalpy 24 of failure. So you're looking for the intersection 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
53 between the failed solid symbols and the non-failed 1
hollow symbols. And this red dotted line attempts to 2
draw that kind of a line of demarcation between the 3
failed and the non-failed.
4 As I mentioned in one of the earlier 5
slides, sometimes the reported enthalpy at failure was 6
provided. Sometimes it was just the peak enthalpy.
7 And these particular events, it was peak enthalpy. So 8
you don't know when it failed. You just know that it 9
failed. So you have to look at the intersection 10 between the two, the failed and the non-failed test 11 segments.
12 The data exhibits a clear trend with 13 decreasing failure enthalpy with increasing cladding 14 differential pressure. It's also worth noting that 15 any fission gas released during the transient would 16 contribute to rod internal pressure and hence cladding 17 DP, and the failure trend does not continue 18 indefinitely, below 100 calories per gram. Reported 19 cladding temperatures remained below 800 degrees F and 20 is no longer sensitive to failure.
21 MEMBER MARCH-LEUBA: Hey, Paul, this is 22 Jose. Reinforcing what you just said a moment ago, 23 and I know you understand this. But I wanted to see 24 if I understand it. Let's look at the one, two, 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
54 three, four, five block squares at three megapascal.
1 It tells me if I read it correct that one of them 2
failed at 140, the other one failed at 160, the other 3
one 200, 240, 300. But what you said is that most 4
likely, they all failed at 140. It's only that the 5
transient was so large that even though it failed at 6
140, it continued to go all the way to 300. Is that 7
what you said?
8 MR. CLIFFORD: Yes, that's exactly what 9
I'm saying.
10 MEMBER MARCH-LEUBA: Okay. Thank you.
11 MR. CLIFFORD: Okay. So you guys can 12 actually see my pointer on the screen when I move it?
13 CHAIR KIRCHNER: Yes, we can see it.
14 MEMBER MARCH-LEUBA: Yes, we can see it.
15 MR. CLIFFORD: Okay. I guess I didn't 16 know that. Good, okay. Now we're moving on to Slide 17
- 27. Okay. This figure shows the proposed cladding 18 failure threshold along with the supporting empirical 19 database. This failure threshold has remained 20 unchanged since 2016. So at the beginning of this 21 curve, you're protecting against brittle failure, then 22 you move into ductile failure. And then at the end, 23 there'd be no failures because the enthalpy is so low 24 that you don't experience a significant transient.
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
55 MEMBER BALLINGER: Paul, this is Ron 1
Ballinger. Quite a long time ago now, there was a 2
study done where they did burst tests. Admittedly, I 3
believe it was on unirradiated fuel cladding. But 4
depending on the temperature, the differential 5
pressure, and that's what I'm talking about here, 6
differential pressure, the actual failure geometry of 7
the cladding was quite different. It ranged from just 8
a little bit of a perforation kind of thing to a case 9
where they had a fair amount of uniform expansion, 10 almost superplastic behavior, and that would have a 11 big difference, a big effect on coolable geometry. Is 12 any of that built into this?
13 MR. CLIFFORD: No, we're not trying to 14 define -- like in LOCA, for instance, you have balloon 15 and burst models which predict the extent of the 16 balloon because you're interested in the effect of the 17 balloon on kind of the long term thermal hydraulics.
18 MEMBER BALLINGER: But you are trying to 19 factor in coolable geometry. That's one of the 20 requirements, right?
21 MR. CLIFFORD: Correct. With respect to 22 GDC 28, you're just trying to limit the rate and the 23 amount of reactivity insertion and how that 24 corresponds to basically violently failing the rod.
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
56 Ballooning and bursting the rod, you still preserve a 1
bundle geometry that's still felt to be amenable to 2
cooling. And also, this event -- this class of event 3
is very localized. It's not going to be a core-wide 4
phenomenon like a LOCA would be where you're worried 5
about ballooning a lot of rods and blocking flow for 6
a large portion of the core. These are very localized 7
activity events.
8 MEMBER BALLINGER: I remember photographs 9
and things where the balloon section was basically six 10 inches long.
11 MR. CLIFFORD: And realistically, if you 12 look at the database for ballooning, there is a 13 tremendous amount of spread in the size and shape of 14 the balloon. And also those events are so slow 15 compared to this prompt critical. You don't have a 16 long time to heat up the cladding.
17 MEMBER BALLINGER: Okay. Yeah, I remember 18 that. Okay, thanks.
19 MR. CLIFFORD: So as I mentioned, this is 20 the basis for the high temperature cladding failure 21 mechanisms. Here's the analytical limit, and it 22 hasn't changed since 2016. One important item as I 23 mentioned is the influence of transient fission gas 24 release.
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
57 Now during normal operation, fission gas 1
is released. And depending on fuel design and 2
operating history, rod internal pressures will 3
continue to increase as a result of this fission gas 4
release. It may even exceed reactor system pressure.
5 During the transient, additional fission gas may be 6
released which would also contribute to rod internal 7
pressure. Hence, this is a contribution which must be 8
accounted for when applying this curve. Looking at 9
the available database, staff came up with burnup-10 dependent transient fission gas release correlations 11 which are provided in the guidance to help estimate 12 rod internal pressure. These correlations are 13 unchanged from 2016.
14 MEMBER MARCH-LEUBA: It took me a while to 15 unmute. Can you explain the figure a little bit? Can 16 you tell us what the figure says?
17 MR. CLIFFORD: Okay. So the figure 18 provides measured transient fission gas release for 19 each of the tests as a function of the peak enthalpy 20 increase during the event. And the symbols are color 21 coordinated based on the burnup. So the green symbols 22 are the lower burnup, below 30 gigawatt-days. The 23 blue symbols are intermediate burnup, between 30 and 24
- 50. And the red symbols are greater than 50 gigawatt-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
58 days, so higher burnup.
1 And as with the other curves, the open 2
symbols did not fail. So after the test, they 3
would've been punctured and they would have measured 4
the fission gas in the rod. And then the closed 5
symbols are ones that failed, and they used different 6
means to then measure the fission gas that was 7
released into the test capsule.
8 MEMBER MARCH-LEUBA: So then I understand 9
the red solid line is trying to bound all the failed 10 rods with the red dots which are high burnup.
11 MR. CLIFFORD: Correct. Right. So when 12 looking at the data, once again, there's a lot of 13 spread in the data as there always is spread in the 14 data when you're talking fission gas release. So 15 transient gas release, not surprising to see this 16 spread. But just from looking at the data, I guess we 17 felt that there was an observation of higher fission 18 gas release for the higher burnup rods. So at the 19 same energy deposition, the higher burnup rods, about 20 50 gigawatt-days, would release more fission gas --
21 more relevant fission gas release.
22 MEMBER MARCH-LEUBA: And the blue solid 23 line is trying to bound all of the green and blue dots 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
59 MR. CLIFFORD: Correct, right.
1 MEMBER MARCH-LEUBA: -- which are the ones 2
below 50? And fission gas release percent, is that a 3
percent of the theoretical calculated maximum, 4
calculated production?
5 MR. CLIFFORD: Correct. So that --
6 MEMBER MARCH-LEUBA: What evidence?
7 MR. CLIFFORD: So that is a percent of 8
fission gas that was created as a result of the burnup 9
history. So of course, the moles of gas increases 10 with time and with irradiation. So it's not the moles 11 of gas. It's the percentage of the gas that's 12 released. So that would be multiplied by the moles of 13 gas at --
14 (Simultaneous speaking.)
15 MEMBER MARCH-LEUBA: So if I read 25 16 percent FGR, I mean, 75 percent of the gas is the 17 fission gases where it's still retaining, say, inside 18 the crystalline structure, inside a pellet?
19 MR. CLIFFORD: Correct.
20 MEMBER BALLINGER: Though we have to be --
21 this is Ron again. We have to be a little bit careful 22 when we look at this because it's really -- it doesn't 23 say it in the figure, but it's 25 percent of the gas 24 released in the area for which the power excursion 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
60 occurs. It's not the full rod, right?
1 MR. CLIFFORD: And that's the reason if 2
you look at the text right above it is 2.3.3. We say 3
you should calculate it along the length of the rod.
4 MEMBER BALLINGER: Right.
5 MR. CLIFFORD: So --
6 MEMBER BALLINGER: Along the length of the 7
rod. Okay, okay.
8 (Simultaneous speaking.)
9 MEMBER MARCH-LEUBA: Go ahead, Paul.
10 MR. CLIFFORD: -- calculated it so like in 11 the area where the rod was ejected which would be the 12 top of the core, you might have 20 percent release.
13 But six inches below that, you'd have 0 percent 14 release because the change in fuel enthalpy would be 15 16 MEMBER MARCH-LEUBA: Yeah.
17 MR. CLIFFORD: You wouldn't see a 18 transient. So it's only a high percentage of those 19 pellets that actually experience the release -- I mean 20 the transient.
21 MEMBER BALLINGER: So the actually impact 22 of the differential pressure is probably not that 23 large.
24 MR. CLIFFORD: It's not that large.
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 However, there was a least one test where it ballooned 1
and failed when it shouldn't have, and that's really 2
what focused the world's attention -- the nuclear 3
world's attention on the potential for the transient 4
fission gas release.
5 MEMBER MARCH-LEUBA:
Just for my 6
education, Paul. Is this the internal of the whole 7
cycle burnup of production of gas, or is it only the 8
gas that is produced during the peak during excursion 9
as the rod is ejected? This is a transient meaning 10 during those two minutes -- two seconds where the 11 power increased, you produce some gas. And this is 12 the fraction of that two second production that gets 13 released, or is it the fraction of the gas that was 14 produced since I started up?
15 MR. CLIFFORD: It's the fraction of gas 16 since you started up, but --
17 MR. SCHULZ: It's the latter, yeah.
18 MR. CLIFFORD: Yeah, it's not -- if you 19 were using this information for calculating doses, of 20 course you would have to take into account the decay 21 of each of the radionuclides and stuff like that that 22 occur and how much activity was released. But here, 23 we're just talking moles of gas going into rod 24 internal pressure.
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
62 MEMBER MARCH-LEUBA: So it's mostly 1
helium, but there is some --
2 MR. CLIFFORD: Mm-hmm.
3 MEMBER MARCH-LEUBA: Okay. Thank you.
4 Somebody else had a question. Im over and out.
5 MR. SCHULZ: Paul, this is Steve Schulz.
6 It looks this is what again is being transferred to 7
8 MR. CLIFFORD: So this was maintained in 9
this Reg Guide because it feeds into the mechanical 10 analysis of rod internal pressure. There would also 11 be this information along with a breakdown of how to 12 treat this for each of the radionuclide groups when 13 calculating activity releases as a result of transient 14 fission gas release. But that would be slightly 15 different.
16 MR. SCHULZ: Would you not expect in terms 17 of the fission gas release calculation from the event?
18 Just in terms of the mechanics of the calculations, 19 it's not trivial to implement the analysis of the 20 steady state and now the transient fission gas release 21 creating that pressure. So wouldn't you want to make 22 sure that you have a consistent approach in this 23 evaluation of gas release and the one that's used in 24 1.183?
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
63 MR. CLIFFORD: It will be consistent. The 1
only difference will be this would be concerned mostly 2
just about the moles of gas released, not the activity 3
of each, because, for instance, if you have a long-4 term stable nuclide, krypton-85, it's not decaying.
5 So whether it was born on day one or it was born a 6
second before the transient, it's still going to be 7
there and it's going to be released and you'll have to 8
account for the dose count.
9 But if it was iodine-131 which has a much 10 shorter half-life, the iodine that was created three 11 weeks before the accident isn't active anymore. So it 12 doesn't really affect it. So there has to be guidance 13 on how to treat each of the radionuclides with respect 14 to calculating the activity that's released or the 15 additional activity that's released during the 16 transient. But it's still using the same database to 17 come up with a baseline. So you would see this figure 18 in Reg Guide 1.183, but it would be focused more on 19 how to use this information to determine the activity 20 released from each of the radionuclide groups.
21 MR. SCHULZ: Yeah, that's post-accident, 22 if you will, post-event in terms of the radiological 23 release.
24 MR. CLIFFORD: Mm-hmm.
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
64 MR. SCHULZ: Okay. Thank you.
1 CHAIR KIRCHNER: Paul, I think -- this is 2
Walt. I think you described it well. This is for the 3
structural, the pressure loading calculation, not for 4
the source term. But they would be consistent. You 5
would start with this and then have to look at the 6
speciation, half-lives, et cetera, et cetera. But for 7
purposes of the PV=nRT, this is what is added to the 8
existing rod internal pressure.
9 MR. CLIFFORD: That is correct. Okay. So 10 let's move on to calculating additional failure modes, 11 or using additional failure modes to calculate the 12 total number of failed rods. So we get into the 13 requirements associated with DNB and CPR.
14 Based upon comments
- received, the 15 application of the high temperature cladding failure 16 curve which we just talked about was changed from all 17 zero power scenarios to all prompt critical power 18 scenarios to be more consistent with the empirical 19 database. So the way this is now worded was if you're 20 at prompt critical excursion, you use Figure 1. If 21 you're non-prompt, then you use your DNB, CPR thermal 22 design limits to estimate cladding failure. Any 23 questions?
24 (No response.)
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
65 MR. CLIFFORD: And the use of DNB as a 1
cladding failure metric, that's the way it's always 2
been done. Okay. Over on to the next failure 3
mechanism which is PCMI, Pellet-Clad Mechanical 4
Interaction. Here's the guidance that's in the Reg 5
Guide for estimating the number of fuel rod failures 6
due to PCMI. We'll get into a lot more discussion.
7 What we're trying to protect is the failure of the 8
cladding due to the mechanical interaction of the 9
expanding pellet. So onto the details, Slide 31.
10 The PCMI empirical database as a function 11 of fuel enthalpy rise versus cladding excess hydrogen.
12 Hydrogen which is absorbed during normal operation 13 forms zirconium hydrides which reduce the cladding's 14 ductility. Hydrogen uptake depends on several 15 factors, time and temperature, power history and 16 fluence, alloy-specific corrosion and hydrogen pickup 17 kinetics, proximity to dissimilar metals, and RCS 18 chemistry. Low burnup and low corrosion fuel rods 19 retain sufficient cladding ductility and will likely 20 fail by the high temperature mechanisms we previously 21 discussed.
22 Besides overall cladding hydrogen content, 23 PCMI database exhibits sensitivity to hydride 24 distribution and orientation. Hydride distribution is 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
66 influenced by thermal and mechanical treatment during 1
manufacturing and the stress state prevailing during 2
hydride precipitation. This force micrograph shows 3
the cross section of a high burnup PWR cladding 4
manufactured in a stress-relieve annealed state.
5 This OD oxide layer is clearly visible at 6
the top. Hydrides preferentially precipitate in a 7
circumferential direction in the SRA state. Hydrides 8
preferentially reside along the cooler outside 9
surfaces as shown in the micrograph.
10 MEMBER BALLINGER: This is Ron Ballinger 11 again. The PWR cladding sample would be Zircaloy-4, 12 and the BWR cladding sample would be Zircaloy-2?
13 MR. CLIFFORD: In this micrograph?
14 MEMBER BALLINGER: Yeah.
15 MR. CLIFFORD: I believe VA-2 was the 16 Japanese cladding MDA which is --
17 MEMBER BALLINGER: Okay. So it's not 18 Zircaloy-2 or 4?
19 MR. CLIFFORD: VA-2 on here. Now this 20 figure here is Zirc-2.
21 MEMBER BALLINGER: Okay. The one on the 22 right is Zirc-2? Okay.
23 MR. CLIFFORD: And all BWRs currently use 24 Zirc-2.
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
67 MEMBER BALLINGER: Right. The one on the 1
left is not Zircaloy-4?
2 MR. CLIFFORD: No, I believe it's more 3
like a ZIRLO alloy.
4 MEMBER BALLINGER: Okay, okay.
5 MR. CLIFFORD: But it's not the alloying 6
that affects the hydride preferential orientation 7
distribution.
8 MEMBER BALLINGER: But it does affect the 9
amount?
10 MR. CLIFFORD: Correct, it affects the 11 amount of oxidation and the amount of hydrogen uptake.
12 But it doesn't affect the morphology of the zirconium 13 hydride which is influenced by the manufacturing.
14 MEMBER BALLINGER: Yeah.
15 MR. CLIFFORD: So if you look to the 16 right, the second micrograph shows a cross section 17 from a high burnup BWR fuel which is manufactured in 18 a fully recrystalized state. RXA cladding exhibits 19 randomly oriented hydrides. Note that the radial 20 hydrides provides the pathway for crack propagation.
21 As a result, the RXA cladding is more sensitive to 22 hydrogen content.
23 MEMBER BALLINGER: This is a texture 24 effect?
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
68 MR. CLIFFORD: Correct. So if you look at 1
the sample on the left, the failure begins on the OD 2
in a brittle manner through the oxide and the hydride 3
rim on the OD of the cladding and then exhibits more 4
of a ductile failure as it moves to the lower hydrogen 5
region towards the ID, whereas if you look at the 6
figure on the right where you have basically this 7
pathway of radial oriented hydrides, you see it behave 8
in both the brittle and a ductile fashion. It begins 9
as brittle failure with a straight line, and then you 10 see the classical 45 degree tearing in a ductile 11 fashion as you move away from the hydrides.
12 Okay. So because of the sensitivity to 13 hydrogen, separate PCMI failure threshold lines were 14 established which account for the impacts associated 15 with not only the amount of hydrogen or the texture or 16 the cladding which affects hydride precipitation but 17 also initial RCS temperature. So you'll see four PCMI 18 curves, one for SRA material at high temperatures, one 19 for SRA cladding materials at low temperature, and the 20 same is true for RXA material.
21 So here we're accounting for -- and also 22 it's a function of the initial hydrogen content of the 23 cladding at the time of the accident. So you're 24 accounting for alloy effects and the hydrogen pickup 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
69 fraction and the corrosion rate. You're accounting 1
for fabrication effects for SRA versus RXA, and then 2
you're accounting for temperature effects, whether the 3
effect begins at cold zero power conditions or hot 4
operating conditions.
5 So let's get into a comparison of these 6
four curves. This plot shows the SRA high temperature 7
cladding failure from the DG-1327 along with a revised 8
curve from Reg Guide 1.236 and the supporting 9
database. In response to comments, the algebraic form 10 of the failure correlation was changed. The new 11 correlation exhibits a steeper decline showing a 12 higher sensitivity to initial hydrogen content but 13 then a saturation effect.
14 To provide regulatory stability, the staff 15 elected to use a more conservative fit as opposed to 16 the prior best estimate fit of the failure data. Best 17 estimate curves are susceptible to constant change as 18 new data becomes available. And as we started in 19 2007, as more data became available, the staff found 20 themselves redrawing the curves, trying to reestablish 21 a best estimate fit.
22 And it's noting that in 2007, the rate at 23 which data became available was very low. CABRI had 24 been shut down for many years. Following Fukushima, 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
70 the Japanese were conducting less tests at NSRR. But 1
now today's environment, we're getting more data from 2
JEA. They're back up and running. CABRI just 3
finished their qualifications and starting to run 4
tests. We expect multiple tests a year. And now, we 5
spent a lot of money starting up the TREAT reactor at 6
INL.
7 So we expect more data to be available as 8
we move forward, and we wanted to design these curves 9
and draw these curves so it was more likely that the 10 new data would confirm the continued applicability of 11 the curve as opposed to necessitating changes to the 12 curve because it was found to be non-conservative. So 13 with regulatory stability in mind, we move from a best 14 estimate fit to more of a lower bound drawing of the 15 curve.
16 MEMBER PETTI: So Paul, I have a question.
17 MR. CLIFFORD: Sure thing.
18 MEMBER PETTI: At the high hydrogen pickup 19 numbers, there's a bunch of solid squares that are 20 below the curve. Why would you have drawn your curve 21 to make one be above it --
22 MR. CLIFFORD: Right. I mean --
23 (Simultaneous speaking.)
24 MR. CLIFFORD: Right. There's a certain 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
71 amount of engineering judgment that goes into drawing 1
these curves. We weren't statistically trying to come 2
up with a 95/95 only because it was such a limited 3
data set. So I think we're looking for the overall 4
trends with an emphasis on trying to bound most of the 5
data but not all of the data.
6 There are certain points on the curve 7
which I show that were considered to be faulted either 8
due to there was an incident where one of the more 9
recent tests, the water was initially water-logged 10 which meant it had failed before the event occurred.
11 So the results aren't really applicable. And some of 12 the earlier CAPRI tests had a lot of spallation and 13 hence hydride lenses on the cladding which would fail 14 at a much lower enthalpy and aren't allowed by the 15 current design of fuel. So we didn't intend to bound 16 all the data. We just bounded most of the data.
17 MEMBER BALLINGER: This is Ron Ballinger 18 again. On Slide -- just pick one, Slide 34 which is 19 what you're on and pick -- I don't know. I don't 20 know. Pick 250 ppm hydrogen. This is all Zircaloy-4 21 data, or is there any M5 or ZIRLO data here? And is 22 there a difference?
23 MR. CLIFFORD: Yes, yes. There is a 24 difference. M5 is RXA. So the Zirc-2 -- the database 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
72 for the RXA consists of RXA Zirc-2 for BWRs and M5 for 1
PWRs.
2 MEMBER BALLINGER: Right.
3 MR. CLIFFORD: This database consists of 4
mostly Zirc-4 which is SRA.
5 MEMBER BALLINGER: Right. And so what I'm 6
trying to be mindful of is that there is no more Zirc-7 4.
8 MR. CLIFFORD: That is true, but we don't 9
expect that the failure enthalpy -- once you remove 10 the corrosion dependence of the alloy which is 11 accounted for in the x-axis here, the excess cladding 12 hydrogen, that it's going to be the morphology or the 13 orientation of the hydrides that has a first-order 14 effect. And that's dependent on the manufacturing.
15 So whether it's SRA ZIRLO or whether it's SRA Zirc-4, 16 both of which are represented here, it doesn't have a 17 first-order effect. They're both SRA.
18 MEMBER BALLINGER: But it's multivariate.
19 For the same hydrogen concentration in, say, ZIRLO or 20 M5 or whatever versus Zircaloy-4, you'd have a much 21 higher burnup for Zirc-4.
22 MR. CLIFFORD: And that would be accounted 23 on how you apply these curves. That is correct. Like 24 for instance, if this was an SRA version of a good 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
73 niobium-based alloy like optimized ZIRLO, you wouldn't 1
have more than 200 ppm at the end of life, so --
2 MEMBER BALLINGER: That's why I picked 3
250.
4 MR. CLIFFORD: So you wouldn't have to 5
fall down this curve. You wouldn't --
6 MEMBER BALLINGER: Right.
7 MR. CLIFFORD: -- come down this slope.
8 You would be up at 150 delta calories per gram for the 9
first two cycles.
10 MEMBER BALLINGER: Right.
11 MR. CLIFFORD: So you're considering that 12 in how you apply the curves.
13 MEMBER BALLINGER: Yeah.
14 CHAIR KIRCHNER: Yeah, Paul. Yeah, this 15 is Walt. Thank you for that observation because I was 16 going to ask you to just point out what a typical 17 excess cladding hydrogen level would be at full 18 burnup.
19 MR. CLIFFORD: Okay. So --
20 CHAIR KIRCHNER: It's well below 500.
21 Isn't that correct?
22 MR. CLIFFORD: Well not for the legacy 23 ones. Zirc-4 could get out to 700, and ZIRLO could 24 get out to 500 or 600. But as I mentioned, we're not 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
74 really loading ZIRLO and Zirc-4 anymore. We're using 1
optimized ZIRLO and M5 and other advanced cladding 2
alloys. So you're starting to see end of life at 3
about 200 or lower, so --
4 MEMBER BALLINGER: And that's why -- I 5
mean a lot of this curve is irrelevant.
6 CHAIR KIRCHNER: Yeah, the reason I wanted 7
to bring it up, though, is because Dave had pointed 8
out some of the points that aren't necessarily well 9
captured by the red line that's drawn here out at 600 10 or 700 ppm. But that probably, going forward, is 11 irrelevant.
12 MR. CLIFFORD: That's true. So here's the 13 first of the four curves. This is the SRA at hot, and 14 there's an adjustment to account for the temperature 15 scaling. So there's a slight change in the SRA. The 16 RXA because I mentioned the radial hydrides that can 17 form, the random orientation of the RXA hydrides, some 18 of which would be radial. It shows that it's much 19 more sensitive. So you're going to see a steeper, 20 earlier decrease in ductility with excess hydrogen.
21 So here's the curve.
22 What you'll notice here is in DG-1327. We 23 didn't have any data beyond 300 ppm, so the curve was 24 truncated there. We got some more data and noting the 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
75 saturation effect. We just drew the curve through 1
that last data point that we had there at 700. And 2
this is for the hot -- so this would be for -- hot 3
would be, for instance, M5 in a PWR and cold would be 4
RXA Zirc-2 and in a BWR for reactor startup 5
conditions. So those are the changes in the curves.
6 MR. SCHULZ: Excuse me, Paul. I've got a 7
-- this is Steve Schultz -- a general question, and it 8
comes from your comment that experimental data now has 9
the opportunity to come in, that there's more work 10 that's going to be done. Can you help with 11 understanding of, if you will, the purpose of this 12 experimental work? When you look at what's been done 13 here and all the experimental data that is available 14 so far, in a way, you kind of narrowed in on the right 15 parameters.
16 You've got some data that is being well 17 captured in a conservative way by the curves that have 18 been drawn. What do you see as the future benefit of 19 this new data? You indicated that you wanted to have 20 curves that aren't going to change in spite of it.
21 That is to say, I presume you're suggesting that a Reg 22 Guide review in the future might be five or 10 years 23 out depending on data that comes in. But I mean what 24 is the value of the new data that's going to be 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
76 created --
1 MR. CLIFFORD: So --
2 MR. SCHULZ: -- that you see?
3 MR. CLIFFORD: -- fuel designs don't stand 4
still. They're continuously evolving. We've seen a 5
lot of different zirconium-based alloys that have been 6
introduced in the last 10 years, and each one of those 7
is going to have different oxidation kinetics and 8
potentially a different formation of hydrides along a 9
different preferential orientation that have to be 10 accounted for.
11 But we're starting to see tests that are 12 being done for ATF designs. Doped fuel is a big thing 13 that's coming our way. So there's a lot of tests 14 that'll be run on doped fuel, and there's various 15 doping agents. There's going to be coated cladding, 16 whether they put chromium on the outside of the 17 cladding or some other the metallic barrier. We could 18 see that, and there would be testing done kind of to 19 confirm the applicability.
20 Higher enrichment, we expect to see 21 licensing actions for higher enrichment and higher 22 burnups, first to 68 gigawatt-days. That's why we 23 went to the trouble of evaluating the database, and I 24 have a presentation on that upcoming. But then 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
77 further to 75 gigawatt-days, so you might see more of 1
an intrinsic or inherent burnup phenomena that occurs 2
that causes failure at lower enthalpies, maybe due to 3
higher gaseous swelling, due to the availability of 4
more fission gas along grain boundaries or within the 5
grains. So there are changes that are occurring, and 6
there are tests that are being done to try to capture 7
those for either demonstrating that the existing 8
failure limits are applicable or for doing new ones.
9 MR. SCHULZ: That's what I want to do to 10 get to because -- and it's stated in the Reg Guide 11 clearly that the expectation or the desire would be 12 for the new cladding to be demonstrated to either 13 match or be modified to this curve as a result of data 14 being provided.
15 MEMBER BALLINGER: But the newer -- and 16 this is Ron again. The newer cladding -- any newer 17 cladding, in order for it to be used, would push 18 things to the left in terms of hydrogen pickup and 19 corrosion rates and things like that. And the only 20 thing that would push things out to the right, which 21 would be to higher hydrogen, would be burnup. And 22 until we raise the enrichment limit considerably, 23 we're not going to get burnup much higher, certainly 24 not more than 100.
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
78 MR. SCHULZ: That goes back to Joy's 1
earlier question too which we might come to later.
2 MR. CLIFFORD: Right. So when we -- I 3
think that memo was distributed maybe yesterday or the 4
day before, staff's memo on the applicability to 5
higher burnup for Reg Guide 1.236. In that memo, we 6
not only assess the sensitivity with burnup and the 7
extent of the database. But we also identified data 8
gaps.
9 So we've identified several important data 10 gaps if we want to go to higher burnup to try to 11 confirm whether or not there's intrinsic burnup 12 effects or fill in the gaps that are in the existing 13 database. For example, looking at this Slide 37, not 14 a whole lot of data here, right? So it would be -- I 15 think we're very limited in RXA data. And since a lot 16 of reactors are moving towards say M5 which is an RXA 17 material, it would be good to get more information.
18 So maybe we move on to -- also, we're 19 sticking with PCMI but moving into a different topic.
20 We're getting into the impact or the potential impact 21 of barrier lining on the application of PCMI failure 22 thresholds. So in response to numerous fuel rod 23 failures during normal operation in the BWR fleet in 24 the '70s and '80s, the industry developed barrier 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
79 cladding.
1 The root cause of these fuel failures was 2
PCI stress corrosion and cracking. Barrier cladding 3
consists of a natural or low alloy zirconium liner on 4
the cladding ID of standard Zirc-2 cladding. It has 5
been shown to be less susceptible to PCI SCC. One 6
unexpected benefit of the liner was its higher 7
affinity for hydrogen.
8 Micrograph on the bottom right shows a 9
cross-section of a radiated Zirc-2 barrier cladding.
10 Zirconium hydrides are clearly visible. Note the 11 concentration of hydrides within the barrier liner 12 along the ID. Hence, the liner depletes the base 13 metal of the detrimental effects of hydrogen. And 14 also, studies have shown that the liner will remain 15 intact even at higher concentrations of hydrides.
16 Next, you'll see -- here's the plot for 17 the RXA failure threshold at hot zero power and the 18 supporting database. The blue symbols represent test 19 segments consisting of Zirc-2 cladding with a barrier 20 liner. Note that these barrier tests are the most 21 important test results since they define the shape of 22 the curve.
23 So the test results shown in blue are 24 shifted to account for the presence of up to 30 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
80 percent of the hydrogen in the barrier lining. Using 1
the scale test results, the staff developed RXA PCMI 2
cladding failures for both fuel rod types with and 3
without barrier cladding. Are there any questions?
4 CHAIR KIRCHNER: Five-second rule, Paul.
5 (Laughter.)
6 MR. CLIFFORD: Okay. So we're moving on 7
to a different topic. So let me note something here, 8
and I think it's important. So by having the barrier 9
liner, not only were you protecting against PCI stress 10 corrosion cracking and allowing plants to maneuver 11 more aggressively, you were also basically absorbing 12 all the hydrogen, hence forming less hydrides in the 13 base material. So it was giving you the double 14 benefit.
15 And if plants were to move away from 16 zirconium liner, then they wouldn't have that second 17 benefit, and thus at the same excess hydrogen level 18 would have a higher impact on the cladding ductility, 19 hence the shift in the curve. Okay. We'll go to the 20 next slide. Now we're getting into a different agenda 21 topic which is the hydrogen models.
22 In support of the now-dormant 50.46(c) 23 rulemaking, the staff developed acceptable 24 conservative cladding hydrogen uptake models. These 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
81 models were part of the public comment period for DG-1 1263 in 2014 and were published in a draft Reg Guide 2
1.224 in 2015. These models have been part of 3
previous ACRS briefings on the 50.46(c) rulemaking.
4 CHAIR KIRCHNER: You picked up on that 5
now-dormant wording.
6 (Laughter.)
7 MR. CLIFFORD: Given the status of 8
50.46(c), the staff decided to include these hydrogen 9
models in Reg Guide to help facilitate implementation 10 of the hydrogen dependent PCMI failure curves.
11 MEMBER BALLINGER: This is Ron again. I 12 keep coming back to the difference between M5 and 13 Zircaloy-4. The lower curve is for Zircaloy-4, and 14 the upper curve is probably for Zircaloy-2. But 15 there's no M5 data there on the lower one.
16 MR. CLIFFORD: Hold on. So you're looking 17 at -- hold on.
18 MEMBER BALLINGER: Over at Slide 40.
19 MR. CLIFFORD: Right, okay. So the top 20 curve is the BWR cladding, so that was Zirc-2. Below 21 that is just the empirical database for Zirc-4. So 22 for the BWR Zirc-2 where there is a noticeable impact 23 of fluence --
24 MEMBER BALLINGER: Yeah.
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
82 MR. CLIFFORD: -- we developed kind of a 1
fluence-dependent correlation for hydrogen pickup 2
fraction. For the PWR alloys where we don't see that, 3
we developed conservative bounding pickup fractions 4
which would then be combined with an approved fuel 5
mechanical design code which calculates oxide 6
thickness to then translate into hydrogen content 7
which would then --
8 MEMBER BALLINGER: But the M5 slope would 9
be shallower, right?
10 MR. CLIFFORD: Well, correct. M5 has the 11 benefit of having both lower corrosion, a much lower 12 corrosion rate, but also a lower pickup fraction.
13 MEMBER BALLINGER: Right.
14 MR. CLIFFORD: So yeah, the data to the 15 bottom right is just Zirc-4 data. And what I'm 16 showing here is just if you use a prediction of oxide 17 and you use a 20 percent pickup fraction, you bound a 18 majority of the data for measured hydrogen content.
19 So it's just one acceptable model for translating on 20 approved fuel mechanical design into an estimate of 21 cladding hydrogen content just to implement those 22 failure curves.
23 MEMBER BALLINGER: Are we being sure that 24 we're not compounding conservatism, especially with 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
83 M5, to negate some of the advantages of M5?
1 MR. CLIFFORD: Compounding conservatism?
2 MEMBER BALLINGER: In other words, if you 3
got a conservative hydrogen pickup curve where M5 is 4
considerably below that, does that, in effect, 5
penalize M5, because now that translates back into the 6
allowable enthalpy in an insertion accident. Am I 7
missing something here?
8 MR. CLIFFORD: So it just happened if you 9
look at the last paragraph in that C-1, it says, the 10 hydrogen pickup fractions should be used along with a 11 best estimate prediction of peak oxide thickness from 12 an approved model. So --
13 MEMBER BALLINGER: Right.
14 MR.
CLIFFORD:
it's using a
15 conservative hydrogen pickup model but with a best 16 estimate prediction of corrosion. So I don't think 17 it's overly conservative. It's certainly more 18 conservative than if the industry was to measure a lot 19 of data, come in and get an approved hydrogen model.
20 I think it would be less conservative than this, but 21 22 (Simultaneous speaking.)
23 MEMBER BALLINGER: I keep -- if I go back 24 to Slide -- where am I -- Slide 37 or 36 where the 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
84 knee in that curve is in some cases around 150 ppm.
1 That conservatism can shift where you are in the knee 2
in that curve, right?
3 MR. CLIFFORD: Correct.
4 MEMBER BALLINGER: Okay.
5 MR. CLIFFORD: I would say, though, this 6
is excess cladding -- excess hydrogen. So you figure 7
hot full conditions, you may have 70 ppm in solution 8
and the remainder in being excessed in the form of 9
hydride. So you would need -- see the curves falls at 10 about -- and just eyeballing it -- falls at about 80.
11 MEMBER BALLINGER: Yeah.
12 MR. CLIFFORD: So you would need more than 13 150 ppm before you started to see the detrimental 14 effects. And I don't believe for M5 you would see 15 more than 150 ppm at end of life. So it's not being 16 overly conservative for the advanced cladding alloys.
17 MEMBER BALLINGER: Okay. All right. I've 18 got to think about it. Okay.
19 MR. CLIFFORD: That's the problem with 20 animating stuff. You've got to walk through each of 21 these changes. Okay. So once again, these hydrogen 22 models are there to aid in the implementation of the 23 failure curves. You've seen them before in 2014 and 24 2015. Anymore questions on this topic?
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
85 Okay. Hearing none, good. We're moving 1
on to the last topic which is the burnup extension.
2 In support of the near-term licensing actions to 3
extend allowable fuel rod burnup out from 62 gigawatt-4 days per metric ton of uranium to 68 gigawatt-days per 5
metric ton of uranium, the staff completed a critical 6
assessment of the empirical database supporting each 7
portion of Reg Guide 1.236.
8 And there's the ML number for the -- so we 9
investigated the sensitivity of the phenomena to 10 burnup, and then we assessed the extent of the 11 empirical database. And here's just a graph shown 12 earlier that just shows all of the data as a function 13 of burnup so you can see the extent of the data beyond 14 the current burnup limits. It's important to note and 15 this is kind of a source of confusion. When they 16 report data -- sorry. When they report burnup, 17 they're reporting a burnup -- the average burnup on 18 the segment -- the test segment.
19 And the test segments can be four inches 20 or they can be -- I think the largest were close to 21 two feet or three feet in length. So when you're 22 looking at the reported burnup, it's closer to a nodal 23 burnup or even closer to a pellet burnup. But it is 24 a rod average burnup, so you've got to kind of adjust 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
86 it based on that.
1 So what I'm saying here is the empirical 2
data for the test segments, if you consider that a rod 3
average 68 gigawatt-days which is the target we're 4
trying to achieve here is approximately 75 gigawatt-5 days on a test segment. So when we're looking at the 6
extent, we're saying, well, how much data is there out 7
to 75 on the local fuel burnup? I just wanted to 8
introduce that.
9 So we'll start with the high-temperature 10 cladding failure threshold. Here's the database as a 11 function of burnup with failure enthalpy. Or once 12 again, this isn't failure enthalpy. It's the enthalpy 13 and then the closed symbols -- solid symbols show 14 which tests failed.
15 As a function of burnup, you'll notice if 16 you go across at 150 here, you don't see a sensitivity 17 to burnup. And it was determined that the cladding 18 failure threshold was more sensitive to differential 19 pressure. And of course, there is a relationship 20 between differential pressure and burnup because 21 higher burnup has more fission gas release, more 22 fission gas release, higher delta P. So there is 23 somewhat of a
connection between burnup and 24 differential pressure.
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
87 And if you look at the extent of the 1
database of these higher burnups, you see these points 2
here. You guys can see my thing here and I'll make it 3
real fancy. Hold on a sec. Look at this, huh? My 4
simulated laser pointer, it shows that these points 5
survived above 100 calories per gram which is towards 6
the end of life here. In fact, you didn't have 7
failures up to 150 calories per gram, even in the 8
higher burnup segments. So it was felt that the 9
phenomena, in fact, was not sensitive to burnup and 10 there was data out past higher burnups to support 11 expansion in the applicability. Now --
12 MEMBER BALLINGER: This is Ron again.
13 Back up a slide. What's the uncertainty? The 14 differential pressure is a calculated number, right?
15 MR. CLIFFORD: Well, it would be on how 16 you apply it. But in this database, it was something 17 that was designed. It was measured.
18 MEMBER BALLINGER: Okay.
19 MR. CLIFFORD: They set the initial 20 pressure to achieve a certain target.
21 MEMBER BALLINGER: Right. But in real 22 life, it would be a calculated number?
23 MR. CLIFFORD: Correct.
24 MEMBER BALLINGER: Okay, which has 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
88 uncertainty, and I'm looking at the numbers. We're 1
talking about three or four MPA. Well, okay.
2 MR. CLIFFORD: And realistically, this 3
event, because you have more reactivity in a fresher 4
bundle, by the time you get to high burnup, the 5
reactivity in that bundle, hence the worth of the 6
blade next to that bundle would be a lot lower. So 7
this event is more limiting for fresher fuel, and 8
fresher fuel probably don't have a DP because the RCS 9
pressure is not decreasing during the event. It's 10 just barely changing in the first couple seconds. So 11 you're not seeing -- it's not like a LOCA where you 12 also have pressure is dropping. So you have to have 13 an initial pressure beyond system pressure to even 14 move to the right on this scale over here. So --
15 MEMBER BALLINGER: Yeah.
16 MR. CLIFFORD: -- chances are for the rods 17 that are more susceptible, you're going to be in this 18 realm right here.
19 MEMBER BALLINGER: But at the end of, say, 20 third cycle, the fuel rod is actually probably being 21 driven --
22 MR. CLIFFORD: Correct.
23 MEMBER BALLINGER:
by adjacent 24 assemblies, right?
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
89 MR. CLIFFORD: Right. But if it resides 1
out in the core peripheral, there's going to be no 2
worth --
3 CHAIR KIRCHNER: Actually, it's running on 4
plutonium, Ron.
5 MEMBER BALLINGER: It's running on 6
plutonium, yes.
7 MR. CLIFFORD: So your concern is the 8
pedigree of the calculation of rod internal pressure 9
that gets used to implement this?
10 MEMBER BALLINGER: Yeah. I mean, those 11 burnups, especially with fission gas release, the 12 starting differential pressure is going to be very, 13 very low, right, because you've got system pressure on 14 one side and you've got almost liftoff on the other 15 side, right?
16 MR. CLIFFORD: Yeah, right. The limit on 17 rod internal pressure from the design -- the 18 mechanical design calculation is that you can't lift 19 off. So right, you could be 800 pounds per square 20 inch above system pressure and a very high burnup rod 21 at end of life.
22 MEMBER BALLINGER: Right. Okay. I just 23 am saying that things get a little bit muddled when 24 these calculations become tenuous.
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
90 MR. CLIFFORD: Potentially. It depends on 1
how fine they want to do the calculations. I've 2
always felt that if you wanted to try to predict the 3
exact number of rods that fail by each mechanism for 4
each core loading, and since you have a differential 5
pressure, a dependent failure threshold, and you have 6
a cladding hydrogen dependent failure threshold, then 7
that could -- it could be very difficult to find the 8
worst possible scenario because you're no longer 9
looking for the highest-worth ejection. So you just 10 can't do a quick physics calculation saying, what's 11 the maximum ejected rod worth, and do an analysis 12 there, because there may be a rod worth that's of 13 lower magnitude but is next to a cladding that has 14 higher hydrogen content, hence a lower failure 15 threshold.
16 MEMBER BALLINGER: And those calculations 17 are basically almost full core calculations nowadays, 18 right, rod by rod?
19 MR. CLIFFORD: I don't know. I don't know 20 off the top of my head. I mean, there's been a lot of 21 evolution in the analytical methods over the last 22 couple of years as we're moving towards 3D space-time 23 kind of kinetics as opposed to the 1D 1970s versions.
24 MEMBER BALLINGER: Yeah.
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
91 MR. CLIFFORD: But where I was going with 1
it, you want to do a very, very detailed consensus 2
evaluation where you figure out how many rods fail for 3
every mechanism and every point during a cycle for 4
each reload. It could be very difficult, but 5
remember, what we're trying to do here is estimate 6
fuel failures that goes into a dose calc. This is not 7
very limiting from a dose calc.
8 Even though you're only allowed 25 9
percent, you could still fail. You could survive even 10 if you failed 20 percent of the core. You could meet 11 your dose requirements, and there's no way to get to 12 20 percent fuel failure because the event is very 13 localized.
14 MEMBER BALLINGER: Right.
15 MR. CLIFFORD: So you could do a bounding 16 assessment once and say, look, I'm going to look at 17 these failure thresholds and show that. I'm never 18 going to get to this point because I've done a 19 bounding dose assessment and I just can't fail any 20 rods and then break down to some reload checklist 21 item. But we're just providing the level of detail 22 should somebody want to get into the weeds.
23 Okay. So where was I? Okay, PCMI. Now 24 we're moving on to the burnup dependence of PCMI.
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
92 Okay. This plot shows the SRA PCMI failure threshold 1
along the empirical database. The data suggests that 2
PCMI is more sensitive to and better represented by a 3
function of cladding hydrogen content as opposed to 4
burnup.
5 Here's the same data, both rise and peak 6
enthalpy but as a function of burnup instead of excess 7
cladding hydrogen. It's important to note that you 8
would expect some burnup effects on cladding PCMI 9
performance under RIA conditions. For example, 10 increased exposure leads to a closure of the initial 11 fuel pellet to cladding gap which would have an effect 12 on PCI performance. And also, higher exposure leads 13 to enhanced gaseous swelling, would affect PCMI 14 loading.
15 So you would expect some inherent burnup-16 dependence in PCMI. However, when you start looking 17 at the data, at least up to 68 gigawatt-days, you 18 don't really see it. Here's another plot. Here is 19 the empirical database with cladding excess hydrogen 20 as a function of fuel burnup. So it gives you an idea 21 of the extent of the database.
22 So at least as far as this database is 23 concerned, as you get into higher burnup, your failure 24 mechanism is clearly related to hydrogen and not 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
93 related to burnup. You don't have a lot of data here.
1 But when you start looking across -- this point right 2
here is suspect. That's SIBQ (phonetic) which was 3
monologued.
4 When you start looking here at about 125 5
calories per gram, you don't see failures at 40 and 6
you don't see failures at 65. So you don't see a 7
strong burnup dependence, and you have to consider 8
this because you could have an M5 as we talked about 9
that doesn't have the hydrogen uptake. So it doesn't 10 fall down that curve, all the way down to 65 calories 11 per gram failure threshold. It stays up at 150 even 12 at end of life.
13 So that's why you need to try to 14 understand the sensitivity of the failure with burnup.
15 And here, going across the data, there's no obvious 16 trend in decreasing. If you just looked at the right, 17 you would think there would be a trend like you could 18 draw a line like this. But the only reason that these 19 are failing is because they're at high hydrogen 20 content which we're capturing with the failure 21 criteria.
22 MEMBER BALLINGER: This is Ron again. Do 23 we have Slide 43, your Slide 43, that one? Well, wait 24 a minute. Okay.
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
94 MR. CLIFFORD: So the other one, they're 1
animated, so --
2 MEMBER BALLINGER: Okay. All right. I'm 3
just trying to --
4 MR. CLIFFORD: You start here. This is 5
44, and you add this additional plot, then you change 6
one of the plots out.
7 MEMBER BALLINGER: Okay. I think we're a 8
little bit -- I don't think -- maybe I have the wrong 9
set of slides, but I don't have that sequence.
10 MR. CLIFFORD: If you have it, just -- if 11 you have the PDF version, the PDF version doesn't 12 capture the --
13 MEMBER BALLINGER: Oh, okay, okay.
14 MR. CLIFFORD: -- information.
15 MEMBER BALLINGER: Yeah.
16 MR. CLIFFORD: So back, if you're looking 17 at the data, there's not a pronounced sensitivity with 18 burnup, at least to this burnup level. It's certainly 19 more sensitive to cladding hydrogen content. And if 20 you start looking at the database, remember that the 21 kind of saturation effect on higher hydrogen contents.
22 So there's a kind of asymptotic limit for SRA at 65 23 and RXA at 50 calories per gram.
24 And then when you start looking at both 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
95 burnup effects and failure of really high burnup test 1
specimens, they're not failing below those limits 2
here. So 65, 50 is right here. So you're not failing 3
in these data points. They failed above those limits.
4 So based on that information, we felt it was 5
acceptable to extend the range of applicability out to 6
68 gigawatt-days for the PCMI. And we look at both 7
RXA and SRA. I'm just showing you the SRA right here.
8 Next part is the damaged core coolability 9
analytical limit. And here it is right here, the 230 10 calories per gram and the limited amount of melt. And 11 here's the supporting database for these limits. It's 12 really the entire. But realistically, you're looking 13 at this cloud of data over here at very low burnups.
14 We identified a need for more data out 15 here because if -- let's walk through the logic here.
16 So while there's no data out here at very high 17 burnups, the criterion itself since there's two parts 18 of the criterion and the first one being the 230.
19 It's kind of the upper -- the ceiling. But then the 20 next part being the limited melt which decreases with 21 increasing burnup.
22 You are capturing a burnup effect using 23 the melt criteria. But we identified this as a data 24 gap, that we needed to have tests run on higher burnup 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
96 rods, especially if you started to go to higher 1
enrichment, because as you go to higher enrichment, 2
you're going to have more reactive worth at higher 3
burnup.
So in that
- scenario, (telephonic 4
interference) to here so that we could verify that the 5
ten percent melt limit would be conservative because 6
we just don't have data in this region here.
7 MR. CORRADINI: Paul, this is Corradini.
8 I don't completely follow what you just said. Are you 9
telling me the black dots below your little red star 10 or your red pointer are ten percent melt fuel volume 11 limits and you're worried that with higher enrichment 12 and higher burnup, you'd actually fall below those?
13 Is that what you're trying to get at? I'm still not 14 clear about what you're saying.
15 MR. CLIFFORD: Okay. So right now, we 16 have the two criteria. The first one is the 230 which 17 is really based on fresh fuel.
18 MR. CORRADINI: Right.
19 MR. CLIFFORD: And the second part of it 20 is no melting which is going to capture burnup effects 21 on how you predict melting, both from a radio power 22 profile and also a melting threshold which are both 23 burnup dependent. But what we don't have is, for 24 instance, what if you had a loss of coolable geometry 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
97 before you melted because of some intrinsic effect and 1
really high burnup? Right now, you would expect that 2
fresh fuel is going to be about 260 calories per gram 3
is when you're going to melt. And high burnup fuel is 4
going to drop down to below 200, somewhere down here 5
6 MR. CORRADINI: Right.
7 MR. CLIFFORD: -- just based on radio 8
power profile, plutonium production, and melting 9
temperature. So it's going to drop down to about 10 here. But what if you had a loss of coolable geometry 11 up here? For instance, your gaseous swelling would be 12 enhanced at high burnup. So maybe the pellet would 13 essentially push its way out. Not melt but it would 14 swell so violently at the higher burnup that it 15 dispersed into the coolant. So we don't have the data 16 to account for that phenomena. So --
17 MR. CORRADINI: Okay. I think I get it.
18 MR. CLIFFORD: Right. These dark symbols 19 are just when the cladding fails, not when it melts 20 and not when you lose the integrity of the cladding.
21 MR. SCHULZ: Paul, this is Steve Schulz.
22 You're talking about a combination of phenomena that 23 may not have been incorporated or considered at lower 24 burnup.
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
98 MR. CLIFFORD: Well, let me rephrase this.
1 So at low burnup, you're not going to run into the 2
melt criteria until you're up above 260. But you have 3
an empirically based threshold at 230 based on the 4
destruction of the rod geometry. So at 230, it didn't 5
melt, but you still violently failed the rod so that 6
you couldn't guarantee coolability because it was no 7
longer one or two pieces of rod. It was multiple 8
sections of rod.
9 MR. CORRADINI: Even though you didn't 10 melt, you would have a loss of coolable geometry in 11 that region?
12 MR. CLIFFORD: Correct.
13 MR. CORRADINI: Okay.
14 MR. CLIFFORD: So as you go up and burnup, 15 the melt criteria becomes more limiting and eventually 16 drops down to 200 calories per gram at end of life.
17 So it's much smaller than the 230. But who's to say 18 that the other phenomena -- not the melt phenomena but 19 just the phenomena of the expansion of the pellet and 20 the response of the pellet and how it failed the rod 21 in a very complex manner. Who is to say that doesn't 22 become more limiting at higher burnup.
23 MR. SCHULZ: Or in combination. That's 24 what I was getting to.
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
99 MR. CLIFFORD: Yeah, so that's just a data 1
gap right now. But once again, for today's fuel, the 2
way enrichments are, by the time you get out to high 3
burnup, you have very low worth. The blade or the rod 4
next to that or within that high burnup assembly is 5
very limited in worth. So you just can't get enough 6
power into that fuel assembly. So you're not going to 7
go above 200 calories per gram in an end of life fuel 8
rod. I know you're worried about this, but if you 9
were to go to eight percent enrichment, you could.
10 MR. CORRADINI: Okay. I see your point 11 now.
12 MR. CLIFFORD: So next section was the 13 transient fission gas release and a burnup assessment.
14 Here's the transient fission gas release database.
15 This is measured fission gas release versus fuel 16 burnup just to give you a feel for the extent of the 17 database. Here is the extent of fuel rod burnup 18 versus peak fuel enthalpy.
19 So it's important to look at both, how 20 much high burnup fuel was included in the database and 21 how hard they pushed that fuel. So here, you can see 22 that they pushed the high burnup fuel up above 100 23 calories per gram which is good. In other words, if 24 all this data for supporting high burnup was down 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
100 here, then you really wouldn't get a good accurate 1
measure. But since you're pushing the high burnup 2
fuel, it gives you -- it has more pedigree to it.
3 So the empirical database does show a 4
sensitivity with burnup as we described before, and 5
also the amount of deposited energy. And that's why 6
the staff developed these burnup dependent 7
correlations which is shown right here. So it's 8
important to note that these four data points that are 9
at or above 70 are these four data points here, one, 10 two, three, and four. So all four of the high burnup 11 specimens remained below the correlation. So based on 12 the extent of the database and the fact that the 13 correlations bounds the high burnup test specimens, we 14 felt we could extend the range of applicability out to 15 68.
16 In conclusion for the burnup extension, 17 based on the extent and the sensitivity of important 18 parameters and phenomena of burnup, the staff found 19 that Reg Guide 1.236 to be applicable up to a fuel rod 20 average burnup of 68. This assessment is predicated 21 on the use of approved core neutronics models and fuel 22 rod thermal-mechanical models which are validated up 23 to at least 68 gigawatt-days. In the course of the 24 evaluation, we also identified two limitations.
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
101 The first is as you go march down towards 1
higher burnup, you're going to have a longer residence 2
time. A longer residence time could mean higher 3
corrosion, cladding corrosion levels. If you have 4
excess cladding corrosion, you could localize defects 5
in the cladding such as a spallation or hydride 6
blisters.
7 And tests conducted on cladding with these 8
localized defects have shown significantly reduced 9
cladding failure threshold. So therefore, the 10 applicability of the failure thresholds is limited to 11 fuel rod designs and cladding alloys and plants which 12 control and limit oxide thickness to prevent these 13 localized effects. And right now, when we approve a 14 cladding alloy, we approve the alloy up to a specified 15 oxide thickness, and that thickness is based upon 16 poolside examinations which have shown that there's no 17 localized effects, no spallation up to a certain 18 thickness. So we've accounted for it, but we just 19 wanted to reinforce it here in the Reg Guide.
20 And the second limitation was that fuel 21 fragmentation, relocation, and dispersal is an 22 important phenomena. It affects the ability to 23 demonstrate coolable geometry, and it has been shown 24 to be sensitive to burnup. This Reg Guide does not 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
102 provide an acceptable means of addressing fuel 1
fragmentation and dispersal at higher burnups. So are 2
there any questions on the burnup extension portion?
3 Hearing none, I will move on to the final 4
slide, 3:30. So based upon the latest research data, 5
revised research data and new analyses and 6
international perspectives, the NRC has developed the 7
guidance that's within Reg Guide 1.236. It represents 8
a significant advancement in guidance which separately 9
captures fabrication, burnup, and corrosion effects on 10 fuel rod performance under RIA conditions.
11 We have been actively involved with the 12 industry and the ACRS beginning well before 2007.
13 There have been several previous ACRS briefings 14 beginning with the interim criteria. Actually, 15 beginning with RIL-0401 and then the interim criteria 16 and then the draft guidance and now the final 17 guidance. In addition, there's been numerous public 18 workshops and three rounds of public comments if you 19 consider the standard review plan which was the 20 interim to be the first round.
That's my 21 presentation.
22 CHAIR KIRCHNER: Thank you, Paul. Very 23 good. Excellent piece of work. So before we turn to 24 the public, Members, have you any specific questions 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
103 to ask of Paul at this juncture, and that includes 1
consultants?
2 Okay. Hearing none, when can we open the 3
public line so that we can request any comments from 4
the public?
5 MS. ABDULLAHI: Is anyone here?
6 CHAIR KIRCHNER: Go ahead. Please state 7
your name and then make your comment, please.
8 MS. ABDULLAHI: Maybe your line is muted 9
from last time unless it's a bridge line.
10 CHAIR KIRCHNER: I'm sorry. I couldn't 11 understand the comment.
12 MEMBER MARCH-LEUBA: That was Zena trying 13 to figure out if the line is open.
14 CHAIR KIRCHNER: Oh, okay. Thank you.
15 MR. EICHENBERG: Hello. Can you hear me?
16 CHAIR KIRCHNER: Yes, go ahead.
17 MR. EICHENBERG: Yes, this is Tom 18 Eichenberg, Tennessee Valley Authority, and --
19 CHAIR KIRCHNER: Go ahead.
20 MR. EICHENBERG: -- I just wanted to ask 21 a simple question, if there is any anticipated 22 timeline for cleanup for SRP 4.2.
23 (Simultaneous speaking.)
24 CHAIR KIRCHNER: Paul, you don't have to 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
104 answer that. Hold on, Jose. Paul, we -- Tom from 1
TVA, we normally take comments. If you have a 2
question about a schedule or such, direct that to Zena 3
who is the Designated Federal Official and we'll go 4
through proper channels and give you an answer to your 5
question.
6 MR. EICHENBERG: Okay. Thank you.
7 CHAIR KIRCHNER: Any other members of the 8
public wishing to make a comment?
9 Hearing none, I think we can close the 10 public line, and then I would like to turn to members.
11 Right now, our plan would be for a letter at the June 12 Committee meeting. Are there any other comments from 13 members about that?
14 MEMBER MARCH-LEUBA: Hi, this is Jose. I 15 just wanted to say how impressed with the whole 16 presentation. I mean, this was very thorough, very 17 informative, and it's impressive whenever -- Paul, 18 whenever we ask you a question, you know what we're 19 asking before we finish talking. So I wanted to thank 20 you for a fantastic presentation, and I think this is 21 a very good RG and I support writing a letter in June.
22 MR. CLIFFORD: Thank you very much.
23 MEMBER BALLINGER: This is Ron. I mean, 24 I think this along with unfortunately the, what did 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
105 you say, now dormant 10 CFR 46(c) have taken so long 1
that with the evolution of cladding, much of this 2
stuff -- much of the information, while extremely 3
valuable, is moot.
4 MR. CORRADINI: This is Corradini. I 5
guess I do want to ask Paul something besides, as 6
usual, telling him he did an excellent job of trying 7
to explain to us when we were confused. Paul, if we 8
go back to the slide -- now I'm going to get it wrong 9
-- Slide 40 where you basically point in that the 10 hydrogen uptake models now are included consistent 11 with 50.46(c). Was that required, or does that just 12 make it easier for the applicants to use it and clear 13 to have it as part of this RG?
14 MR. CLIFFORD: It's not required, no.
15 We're just providing it to aid in the implementation 16 because not all vendors have approved hydrogen uptake 17 models for each one of alloys, so no.
18 MR. CORRADINI: So you're basically making 19 it consistent is the way I understood.
20 MR. CLIFFORD: Right. We're just giving 21 them the option of using something that we've provided 22 and that we already find acceptable.
23 MR. CORRADINI: Okay, fine. But other 24 than that, as usual, I thought it was an excellent 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
106 presentation and you helped us a lot when we were 1
confused -- when I was confused.
2 MEMBER REMPE: Paul, this is Joy, and I 3
also find your presentations, as
- usual, very 4
enlightening. But that's what motivated my question 5
about knowledge management because I think you have --
6 your background did reference this earlier memo. But 7
the later memo, I still could not find referenced in 8
the Reg Guide. I also think that as you went through 9
the presentation, you provided us some background that 10 I'm not sure is easily found by other staff members in 11 the future as they start trying to accommodate new 12 fuels. And I would hope that our letter would mention 13 the need for something that's a little more 14 comprehensive that's a background document to support 15 this Reg Guide.
16 (Simultaneous speaking.)
17 MEMBER PETTI: This is Dave. I have a 18 question for Paul. I think about all of this, and I 19 think about accident-tolerant fuels. And you look at 20 how much time and effort it took to get this database.
21 And even today, some of it has some holes in it. Have 22 you given any thought to what you think it's going to 23 take to get enough data from an accident-tolerant fuel 24 perspective, whether it'd be the cladding or the new 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
107 fuels?
1 What I worry about is it could be a two-2 part answer. It could be, well, we just need some 3
data to confirm. That leads you down one path. But 4
what if something else shows up? When you make a 5
change over here and you find out it causes a 6
deleterious effect over there that you didn't think 7
of, that could really take a lot of effort. Have you 8
given any thought as you following the accident-9 tolerant fuel program from this perspective?
10 MR. CLIFFORD: Okay. So I think accident-11 tolerant fuel is a good example because there's both 12 cladding -- design changes in the cladding and design 13 changes in the pellet itself. And of course, the 14 further you migrate away from the empirical database, 15 the more difficult it will be to license. I mean, 16 something like silicon carbide which does not possess 17 significant ductility relative to a new metal, that's 18 going to be a challenge and it's going to likely fail 19 at a lower threshold.
20 So that's something that would have to be 21 evaluated. But given -- if it's paired with the 22 existing UO2 design, you understand the behavior of 23 the UO2 for a given deposit of energy. So maybe 24 there's separate effects that could be used to help 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
108 speed up the licensing process and limit how many in-1 pile tests that have to be done on an irradiated 2
cladding material that's not Zircaloy.
3 But once you get to changes in the fuel 4
pellet where you just don't understand the response of 5
the fuel pellet to a certain energy deposition, 6
uranium silicide and uranium nitride, there's going to 7
have to be a lot of RAI tests done, in-pile testing 8
done at TREAT or CABRI or NSRR to establish the 9
failure modes for that fuel design because you've got 10 significantly different compositions of the pellet, 11 both from an alloy perspective and the microstructure 12 of the pellet and the grain sizes.
13 And the melting temperature of the pellet 14 could be significantly different.
Thermal 15 conductivity of the pellet is different. Swelling 16 rates are different. So if you start changing from 17 the UO2 pellet to something else, there's a lot of 18 work that has to be done.
19 CHAIR KIRCHNER: Along those lines, Paul, 20 there are a number of concepts in terms of changing 21 the cladding or coatings. And so I was thinking back 22 in your presentation the positive effect of the liner 23 for the BWR fuel cladding. Has there been any 24 preliminary look at what some of these coatings and 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
109 such might to in terms of changing either ductile or 1
brittle fracture properties for the cladding under 2
these kind of accident conditions?
3 MR. CLIFFORD: So for the coatings, well, 4
certainly, the coating is going to ensure that there's 5
almost no hydrogen present on the cladding at the 6
beginning of the transient because the oxidation rates 7
of chromium are so low compared to zirconium that you 8
would not even begin to fall down that PCMI failure 9
curve. You would always be at the very top which 10 gives you tremendous benefit.
11 With respect to the actual performance, 12 the strain capability of the cladding with the 13 coating, we don't expect there to be much because the 14 coatings are generally two to three percent thickness 15 of the underlying substrate. So it's not going to 16 have a
significant impact on the mechanical 17 performance of the bare cladding if you want to think 18 of it as bare.
19 CHAIR KIRCHNER: Whether it's brittle or 20 ductile. Okay.
21 MEMBER BALLINGER: This is Ron. The 22 argument that is being made was made for putting these 23 things in lead test assemblies or the like that are 24 already in core.
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
110 CHAIR KIRCHNER: Yes.
1 MR. CLIFFORD: Yes, so I don't -- overall, 2
I don't think coated cladding is a significant 3
challenge from a licensing perspective. I think the 4
challenges are going to be something that's a 5
nonmetallic cladding or a change, ceramic pellet to a 6
metallic pellet or a metallic bore or something else.
7 CHAIR KIRCHNER: Yes.
8 MR. SCHULZ: Paul, this comment follows 9
Ron Ballinger's comment and your introduction today 10 and that is that, as you stated, some of us have been 11 involved with this issue for either near or more than 12 three decades. And so as a result, a lot of the 13 information and data that has gone into this 14 technology is from cladding materials which are no 15 longer used.
16 At the same time, I did want to comment 17 that the NRC team has done a very excellent job in the 18 handling and disposition and analysis of the public 19 comments that were put in. And as you stated, there 20 were many. But the thoroughness by which you address 21 those
- comments, including some very specific 22 evaluations that have been done of different and new 23 types of cladding, are very helpful in both the issue 24 of moving forward with how this can be applied to 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
111 current design, how it can be applied to future 1
design, and also did contribute to an element of 2
knowledge management and technical capture.
3 But I do agree with Joy that it would be 4
good to somehow figure out or make sure that the 5
information that's been developed from the responses 6
to those questions is captured in the archives of the 7
NRC. I did try to access one or two -- one of the 8
referenced documents from the questions, and I 9
couldn't get it in ADAMS. It said it wasn't 10 available. So that's just an example. It's not a 11 complaint, but it's an example that we need to make 12 sure that the information is held for future use.
13 CHAIR KIRCHNER: Okay. Members, any 14 further questions or comments at this point?
15 (No response.)
16 CHAIR KIRCHNER: I'm using a ten-second 17 rule because often we have trouble getting our cursor 18 back on the unmute the microphone.
19 (Laughter.)
20 CHAIR KIRCHNER: Okay. That's ten 21 seconds. Well, then at this point, I want to thank 22 you again, Paul, and thank all the members for 23 participating. I ask Steve and Mike, our consultants, 24 if you have any specific comments, would you please 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
112 get those to me within the next couple of weeks. And 1
also, Members, I will be putting together a draft 2
letter again for the June meeting. So with that, I 3
think we are complete with our work today. Thank you 4
all, also in the public attending. And with that, we 5
are adjourned.
6 (Whereupon, the above-entitled matter went 7
off the record at 3:45 p.m.)
8 9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
SUBCOMMITTEE ON METALLURGY AND REACTOR FUELS REGULATORY GUIDE 1.236, PWR CONTROL ROD EJECTION AND BWR CONTROL ROD DROP ACCIDENT.
May 5, 2020 Subcommittee Meeting ATTENDEES ACRS SUBCOMMITTEE MEMBERS Walter Kirchner (Chairman)
Vesna Dimitrijevic Ronald Ballinger Dennis Bley Jose March-Leuba Joy Rempe David Petti ACRS Consultants Michael Corradini Stephen Schultz ACRS Staff Zena Abdullahi (Designated Federal Official)
Scott Moore (ACRS Director)
Lawrence Burkhart (Technical Branch Chief)
Christina Lui Alesha Ballinger (Branch Chief, PMDA)
Thomas Dashiell Makeeka Compton Paula Dorm Joanne Johnson Shandeth Montgomery Hossein Nourbakhsh Janet Riner
SUBCOMMITTEE ON METALLURGY AND REACTOR FUELS REGULATORY GUIDE 1.236, PWR CONTROL ROD EJECTION AND BWR CONTROL ROD DROP ACCIDENT.
Tammy Skov Derek Widmayer Weidong Wang Quynh Nguyen Christopher Brown NRC staff Paul Clifford (presenter)
Edward ODonnell (RG 1.236 Senior Staff)
Donald Agama Andrew Bielen Andrew Proffitt Ngola Otto External Attendees Alex (Guest)
Gregory Broadbent Stephen Geier Kent Halac (GEH)
Charles Heck (GEH)
Nathanael Hudson Shawn Lamb (GEH)
Christan McElory (GEH)
David Mitchell Brian Mount Kurshad Mufruoglu (Guest)
Nathan Peck (GEH)
Scott Pfeffer (GEH)
Eric S. Scott Eric Thomas Ken Yueh
SUBCOMMITTEE ON METALLURGY AND REACTOR FUELS REGULATORY GUIDE 1.236, PWR CONTROL ROD EJECTION AND BWR CONTROL ROD DROP ACCIDENT.
Telephone Attendees Electric Power (Guest)
GE Nuclear Energy (Guest)
RNT G Wireless Caller (Guest)
+467XXXXXXXXX (Guest) 201XXXXXXXXX (Guest) 571XXXXXXXXXX (Guest) 910XXXXXXXXXXX*Guest)
1 Draft Regulatory Guide 1.236 PWR Control Rod Ejection and BWR Control Rod Drop Accidents ACRS Metallurgy and Reactor Fuels Subcommittee May 5, 2020 Paul M. Clifford Division of Safety Systems Nuclear Reactor Regulation
2 Agenda
- 1. Regulatory Requirements
- 2. Timeline and Stakeholder Comments
- 3. Revised Guidance and Analytical Limits
- a. Reactor Coolant System Pressure
- b. Damaged Core Coolability
- c. Radiological Consequences
- 4. Cladding Hydrogen Models
- 5. BU Extension
3 Reason for Concern Idaho National Engineering and Environmental Laboratory, INEEL 81-3966
- Fatal accident at Armys prototype modular reactor -
Stationary Low Power Reactor (SL-1)
- Improper withdrawal of central control rod resulted in prompt critical power excursion and steam explosion Protection Against Violent Explosion and Loss of Pressure Boundary
4 Regulatory Requirements
- 10 CFR 50, Appendix A GDC 28 requires reactivity control systems to be designed with appropriate limits on potential amount and rate of reactivity increase to assure that the effects of postulated reactivity accidents can neither (1) result in damage to the reactor coolant pressure boundary greater than local yielding nor (2) sufficiently disturb the core, its support structures, or other reactor pressure vessel internals to impair significantly the capability to cool the core.
- 10 CFR 100.11 and 10 CFR 50.67 establish radiation dose limits for individuals at the boundary of the exclusion area and at the outer boundary of the low population zone.
RG 1.236 provides an acceptable means to meet requirements
5 Reactivity Insertion Accidents
- Reactivity insertion accidents are safety significant because of their potential ability to challenge fuel rod integrity, fuel bundle geometry, and the integrity of the reactor pressure boundary
- The uncontrolled movement of a single control rod out of the core results in a positive reactivity insertion that promptly increases local core power
- Considered the limiting reactivity insertion accident
- Of the various postulated single failures of the CRD system which may initiate an uncontrolled movement of a single control rod, PWR CRE and BWR CRD are considered the most limiting scenarios for the current operating fleet
6 Postulated Accidents
- A PWR CRE event is postulated to occur because of a mechanical failure that causes an instantaneous circumferential rupture of the control element drive mechanism housing or its associated nozzle. This results in the reactor coolant system pressure ejecting the control rod and drive shaft to the fully withdrawn position.
Power / Temperature Time Rod Power Fuel Temperature Clad Temperature
- A BWR CRD event is postulated to occur because of the following sequence of events: a control rod (blade) inserted into the core becomes decoupled from its drive mechanism, the drive mechanism is subsequently withdrawn, the control blade is assumed to be stuck in place, and at a later moment, the control rod suddenly falls free and drops to the control rod drive position.
7 Agenda
- 1. Regulatory Requirements
- 2. Timeline and Stakeholder Comments
- 3. Revised Guidance and Analytical Limits
- a. Reactor Coolant System Pressure
- b. Damaged Core Coolability
- c. Radiological Consequences
- 4. Cladding Hydrogen Models
- 5. BU Extension
8 Timeline 1974 RG 1.77, Assumptions used for Evaluating a Control Rod Ejection Accident for Pressurized Water Reactors 1980 Nuclear Safety article (MacDonald et.al.) suggests need for new analytical limits for coolable geometry and failure threshold 2004 RIL-0401, An Assessment of Postulated Reactivity-Initiated Accidents (RIAs) for Operating Reactors in the U.S.
2007 SRP 4.2, Appendix B, Interim Acceptance Criteria and Guidance for the Reactivity Initiated Accidents 2017 DG-1327 1st public comment period 2019 DG-1327 2nd public comment period
9 Timeline - 1974 1974 RG 1.77, Assumptions used for Evaluating a Control Rod Ejection Accident for Pressurized Water Reactors
- Acceptable PWR analytical methods and assumptions
- Fuel radial average energy density limited to 280 cal/g at any axial node
- Offsite dose consequences limited to well within the guidelines in 10CFR Part 100
10 Timeline - 1980 1980 Nuclear Safety article (MacDonald et.al.) suggests need for new analytical limits for coolable geometry and failure threshold
- Coolability criteria reduced to 230 cal/g at any axial node
- Cladding failure threshold reduced from 170 cal/g to 140 cal/g for irradiated fuel rods
- Failure mode strongly dependent on prior irradiation history
- Based on advanced analytics, no imminent safety concern
11 Timeline - 2004 2004 RIL-0401, An Assessment of Postulated Reactivity-Initiated Accidents (RIAs) for Operating Reactors in the U.S.
- PCMI cladding failure at much lower fuel enthalpy
- Based on advanced analytics, no imminent safety concern
12 Timeline - 2007 2007 SRP 4.2, Appendix B, Interim Acceptance Criteria and Guidance for the Reactivity Initiated Accidents
- Cladding P-dependent cladding failure thresholds
- BU-dependent coolability criteria
- Transient FGR
13 Empirical Database Circa 2019
14 Public Comment - 2017 Comment submissions received from 12 stakeholders with a total 124 comments Over 100 comments were accepted
15 Public Comment - 2019 Comment submissions received from 7 stakeholders with a total 54 comments Over 30 comments were accepted
16 Major Changes Expanded fuel burnup range to 68 GWd/MTU (Section C.1)
Improved analytical requirements (Section C.2)
Revised PCMI cladding failure threshold curves (Section C.3)
Removed radiological source term information (Section C.4)
Analytical requirements Fission product gap release fractions Future revision to RG 1.183 Analytical procedure Amended implementation to reflect revised Backfit guidance (Section D)
Added cladding hydrogen uptake models (Appendix C)
17 Agenda
- 1. Regulatory Requirements
- 2. Timeline and Stakeholder Comments
- 3. Revised Guidance and Analytical Limits
- a. Reactor Coolant System Pressure
- b. Damaged Core Coolability
- c. Radiological Consequences
- 4. Cladding Hydrogen Models
- 5. BU Extension
18 RCS Pressure Analytical Limit 5.
Allowable Limits on Reactor Coolant System Pressure For new license applications, the maximum reactor coolant system pressure should be limited to the value that will prevent stresses from exceeding Emergency Condition (Service Level C), as defined in Section III of the American Society of Mechanical Engineers (ASME) Boiler and Pressure Vessel Code (Ref. 14). For existing plants, the allowable limits for the reactor pressure boundary specified in the plants updated final safety analysis report should be maintained.
Consistent with current guidance and practice Protects reactor pressure boundary - satisfies GDC-28
19 Agenda
- 1. Regulatory Requirements
- 2. Timeline and Stakeholder Comments
- 3. Revised Guidance and Analytical Limits
- a. Reactor Coolant System Pressure
- b. Damaged Core Coolability
- c. Radiological Consequences
- 4. Cladding Hydrogen Models
- 5. BU Extension
20 Protect Against Loss of Fuel Bundle Geometry
21 Coolability Analytical Limits 6.
Allowable Limits on Damaged Core Coolability Limiting peak radial average fuel enthalpy to prevent catastrophic fuel rod failure and avoiding molten fuel-coolant interaction is an acceptable metric to demonstrate that there is limited damage to core geometry and that the core remains amenable to cooling. The following restrictions should be met:
a.
Peak radial average fuel enthalpy should remain below 230 cal/g.
b.
A limited amount of fuel melting is acceptable provided that it is less than 10 percent of fuel volume. If fuel melting occurs, the peak fuel temperature in the outer 90 percent of the fuel volume should remain below incipient fuel melting conditions.
Loss of fuel rod geometry limit (230 cal/g) based on earlier SPERT, PBF and NSRR prompt critical experiments Limited centerline melt (current license bases) avoids molten FCI Fuel melt becomes more limiting at ~30 GWd/MTU Preserves coolable geometry - satisfies GDC-28
22 Agenda
- 1. Regulatory Requirements
- 2. Timeline and Stakeholder Comments
- 3. Revised Guidance and Analytical Limits
- a. Reactor Coolant System Pressure
- b. Damaged Core Coolability
- c. Radiological Consequences
- 4. Cladding Hydrogen Models
- 5. BU Extension
23 Radiological Consequences
- 1. Requires a conservative estimate of the total number of failed fuel pins from all failure modes Prompt critical high temperature cladding failure Non-prompt DNB/CPR cladding failure PCMI cladding failure Centerline fuel melt cladding failure
- 2. Requires a conservative estimate of fission product release fractions Steady-state gap inventories Transient fission gas release Future revision to RG 1.183 Fuel melt fission gas release
24 HT Cladding Failure Threshold
25 HT Cladding Failure Mechanisms
- High temperature cladding failure mechanisms Brittle Failure: High-temperature post-DNB (film-boiling) oxygen-induced embrittlement and fragmentation Ductile Failure: High-temperature cladding creep (balloon/rupture)
26 Sensitivity Study
- Database suggests ductile failure sensitive to cladding P
- Transient FGR contributes to rod internal pressure
- Below 100 cal/g fuel enthalpy, cladding remains below 800 F and is no longer susceptible to HT failure
- Minimal BU effects
27 HT Cladding Failure Mechanisms
28 Impact of Transient FGR on RIP Unchanged from DG-1327
29 DNB/CPR Cladding Failure Consistent with current guidance and practice
30 PCMI Cladding Failure
31 Hydrogen-Enhanced PCMI
- Hydrogen, absorbed during normal operation waterside corrosion, forms zirconium hydrides which reduce the claddings ductility
- Hydrogen uptake depends on several factors:
Time-at-temperature (residence time)
Power history and fluence Alloy-specific corrosion and hydrogen pickup kinetics Proximity to dissimilar metals RCS chemistry
- Low burnup and low corrosion fuel rods retain sufficient cladding ductility and will likely fail by high temperature mechanisms before PCMI
32 Hydride Orientation
- Besides overall concentration, PCMI sensitive to hydride distribution and orientation which are influenced by:
Thermal and mechanical treatment during manufacturing Stress state prevailing during hydride precipitation T. Sugiyama, JAEA, PCMI failure of high burnup fuels under RAI conditions, Fuel Safety Research Meeting, Tokai, Japan, May 2007
33 Separate PCMI Failure Thresholds
- Separate PCMI cladding failure threshold lines established which account for impacts associated with initial RCS temperature, excess hydrogen, and hydride sensitivity 1.
SRA cladding at high RCS coolant temperature 2.
SRA cladding at low RCS coolant temperature 3.
RXA cladding at high RCS coolant temperature 4.
RXA cladding at low RCS coolant temperature
34 SRA Hot PCMI Failure Threshold
35 SRA Cold PCMI Failure Threshold
36 RXA Hot PCMI Failure Threshold
37 RXA Cold PCMI Failure Threshold
38 Impact of Barrier Liner
- In BWR liner (i.e., barrier) fuel, the natural or low alloy zirconium liner acts as a sponge for hydrogen.
- Depletes base metal of detrimental effects of hydrides
- Liner remains ductile even with high concentration of hydrides
39 Agenda
- 1. Regulatory Requirements
- 2. Timeline and Stakeholder Comments
- 3. Revised Guidance and Analytical Limits
- a. Reactor Coolant System Pressure
- b. Damaged Core Coolability
- c. Radiological Consequences
- 4. Cladding Hydrogen Models
- 5. BU Extension
40 Hydrogen Uptake Models
- Originally published in draft RG 1.224 (2015) in support of 50.46c rule
- DG-1263 public comment period in 2014
41 Agenda
- 1. Regulatory Requirements
- 2. Timeline and Stakeholder Comments
- 3. Revised Guidance and Analytical Limits
- a. Reactor Coolant System Pressure
- b. Damaged Core Coolability
- c. Radiological Consequences
- 4. Cladding Hydrogen Models
- 5. BU Extension
42 Burnup Extension
- In support of near-term licensing actions to extend allowable fuel rod average burnup to 68 GWd/MTU*, staff completed a critical assessment of the empirical database supporting RG 1.236 guidance (ML20090A308)
- For each portion of the guidance, the staff (1) investigated sensitivity with burnup and (2) assessed the extent of empirical database
- Empirical database reports test segment (i.e., local) burnup. Rod average 68 GWd/MTU equivalent to approximately 75 GWd/MTU test segment burnup.
43 HT Cladding Failure Thresholds
- Empirical database does not exhibit sensitivity with burnup
- Burnup effects captured in rod internal pressure (cladding P)
- Extended burnup test specimens survived above 100 cal/g
44 PCMI Cladding Failure Thresholds
- Empirical database does not exhibit pronounced sensitivity with burnup
- PCMI failure more sensitive to cladding hydrogen content
- Extended burnup test specimens survived above asymptotic limit (SRA -65 cal/g, RXA 50 cal/g)
45 Damaged Core Coolability
- While the damaged core coolable geometry empirical database does not include any high-burnup fuel rod segments, the detrimental effects of higher burnup must be accounted for to satisfy the limited fuel melt restriction.
Peak radial average fuel enthalpy should remain below 230 cal/g.
A limited amount of fuel melting is acceptable provided that it is less than 10 percent of fuel volume. If fuel melting occurs, the peak fuel temperature in the outer 90 percent of the fuel volume should remain below incipient fuel melting conditions.
46 Transient FGR
- Empirical database shows sensitivity with burnup and deposited energy
- BU-dependent FGR correlations
- Extended burnup test specimens fall below high burnup correlation
47 Burnup Extension Based upon the extent of the empirical database and sensitivity of important parameters and phenomena to burnup, the staff found RG 1.236 to be applicable up to a fuel rod average burnup of 68 GWd/MTU.
This assessment was predicated on the use of approved core neutronics and fuel rod thermal-mechanical models. The following limitations were identified:
Excess cladding corrosion will promote localized effects (e.g., spallation, hydrogen blisters) which have been shown to significantly reduce the cladding failure threshold.
Hence, the applicability of the high temperature and PCMI cladding failure thresholds to any fuel burnup, including extended burnup up to 68 GWd/MTU rod average, is limited to fuel rod designs, cladding alloys, and plants which control and limit oxide thickness to prevent these localized effects.
Fuel fragmentation, relocation, and dispersal (FFRD) is a phenomenon which challenges coolable geometry and has been shown to be sensitive to fuel burnup. The susceptibility of fuel pellets to fragment into fine particles increases with burnup. RG 1.236 does not provide guidance related to an acceptable treatment of FFRD.
48 Conclusions
- Based upon latest research data, revised research data, new analysis, and international perspectives, the NRC staff has developed CRE/CRD guidance in RG 1.236
- Represents a significant advancement in guidance
- Separately captures fabrication-, burnup-, and corrosion-effects on fuel rod performance under RIA conditions
- ACRS and stakeholder involvement starting prior to 2007
- Several ACRS briefings beginning prior to Interim Guidance
- Numerous public workshops and 3 rounds of public comments