ML21323A021
| ML21323A021 | |
| Person / Time | |
|---|---|
| Issue date: | 11/15/2021 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| Brown, C., ACRS | |
| References | |
| NRC-1748 | |
| Download: ML21323A021 (126) | |
Text
Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION
Title:
Advisory Committee on Reactor Safeguards Metallurgy & Reactor Fuels Subcommittee Docket Number:
(n/a)
Location:
teleconference Date:
Monday, November 15, 2021 Work Order No.:
NRC-1748 Pages 1-79 NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers 1716 14th Street, N.W., Suite 200 Washington, D.C. 20009 (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 METALLURGY & REACTOR FUELS SUBCOMMITTEE 7
+ + + + +
8 MONDAY 9
NOVEMBER 15, 2021 10
+ + + + +
11 The Subcommittee met via Video 12 Teleconference, at 2:00 P.m. EST, Ron Ballinger, 13 Chairman, presiding.
14 COMMITTEE MEMBERS:
15 RONALD G. BALLINGER, Chair 16 CHARLES H. BROWN, JR. Member 17 VESNA B. DIMITRIJEVIC, Member 18 GREG HALNON, Member 19 JOSE MARCH-LEUBA, Member 20 DAVID PETTI, Member 21 JOY L. REMPE, Member 22 MATTHEW W. SUNSERI, Member 23 24 ACRS CONSULTANT:
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
2 STEVE SCHULTZ 1
2 DESIGNATED FEDERAL OFFICIAL:
5 ALSO PRESENT:
6 BOB CALDWELL, NRR 7
DAN DENIS, Structural Integrity Associates 8
ALLEN HISER, NRR 9
ELLIOT J. LONG, EPRI 10 SCOTT MOORE, Executive Director, ACRS 11 JEFFREY POEHLER, NRR 12 DAVID RUDLAND, NRR 13 R O B E R T T A Y L O R,
N R R 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
3 T-A-B-L-E O-F C-O-N-T-E-N-T-S 1
PAGE 2
3 Presentation by David Rudland 11 4
5 Presentation by Elliot Long 58 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 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
4 P R O C E E D I N G S 1
2:00 p.m.
2 CHAIR BALLINGER: Good afternoon, the 3
meeting will now come to order. This is a meeting of 4
the Advisory Committee on Reactor Safeguards 5
Subcommittee on Metallurgy and Reactor Fuels.
6 I'm Ron Ballinger chairing this 7
Subcommittee meeting. The ACRS Members present are 8
Charles Brown, Dave Petti, Dennis Bley, Greg Halnon, 9
Jose March-Leuba, Joy Rempe, Matt Sunseri, Vesna 10 Dmitrijevic, and our consultant Steven Schultz.
11 The purpose of today's meeting is discuss 12 reactor pressure vessel monitoring and prediction in 13 long-term operations, it's actually an update. The 14 Subcommittee will gather information, analyze relevant 15 issues and facts, and formulate proposed positions and 16 actions as appropriate.
17 This matter will be scheduled for a future 18 full Committee meeting at which time the Committee may 19 develop a letter report on this topic, something some 20 time in the future.
21 By way of background, the Committee has 22 been following this issue for several years and has 23 had several briefings and has written at least two 24 letters on the related topic.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
5 In that letter, one of them, Reg Guide 1
1.99 letter that we wrote, we cautioned that the high 2
fluids issue is important and suggested that the newer 3
more accurate correlation, ASME 900-15, which had been 4
underdeveloped for quite some time and contained an 5
enormous amount of data compared to the original 6
database, will be considered for use in an updated 7
revision of Reg Guide 1.99.
8 The Staff disagreed with this suggestion 9
and provided an explanatory document, Reg Guide 1.9 on 10 Revision 2 evaluation status March 19, 2021. Also a 11 document where the favor code was used to evaluate the 12 effective high fluids was developed.
13 Additionally, a petition for rulemaking 14 has been submitted by NuScale Power LLC docketed by 15 the NRC in September 11, 2019 and has been signed 16 Docket PRM50-120. To allow the use of the trend curve 17 suggested in the ASME 900-15 as the basis for RPV 18 embrittlement.
19 Based on more recent analysis by the 20 Staff, a rulemaking effort may be one path forward to 21 ensure embrittlement is adequately addressed and high 22 fluence, especially the uncertainty.
23 ACRS was established by statute and is 24 governed by the Federal Agents Advisory Committee, 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
6 FACA. The NRC implements FACA in accordance with its 1
regulations found in Title 10 of the Code of Federal 2
Regulations Part 7.
3 The Committee can only speak to its 4
published letter reports. We hold meetings to gather 5
information and perform preparatory work that will 6
support our deliberations at a full Committee meeting.
7 The rules for participating in all ACRS 8
meetings including today's were announced in the 9
Federal Register on June 13, 2019.
10 The ACRS Section of the NRC public website 11 provides our charter bylaws, agendas, letter reports, 12 and full transcripts of all full and Subcommittee 13 meetings including slides presented at these meetings.
14 The meeting notice and agenda for this 15 meeting were posted there.
16 As stated in the Federal Register notice 17 and in the public meeting notice posted to the 18 website, members of the public who desire to provide 19 written or oral input to the Subcommittee may do so 20 and should contact the designated official five days 21 of the meeting.
22 I should say that the designated federal 23 official for this meeting is Mr. Christopher Brown.
24 Today's meeting is open to public attendance and we 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
7 have received one request to make an oral statement.
1 I think that's correct. Time is provided 2
in the agenda after presentations are completed for 3
this oral statement and for spontaneous comments for 4
members of the public attending or listening to our 5
meetings.
6 Today's meeting is being held over 7
Microsoft Teams, which includes a telephone bridge 8
line allowing participation to the public over their 9
computer using Teams or by phone.
10 A transcript for today's meeting is being 11 kept, therefore, we request that meeting participants 12 on Teams and on the Teams call-in line identify 13 themselves when they speak and to speak with 14 sufficient clarity and volume that they can be readily 15 heard.
16 Likewise, we request that the meeting 17 participants keep their computer and/or telephone 18 lines on mute when not speaking to minimize 19 disruptions. I also ask that the Teams attendees make 20 sure that they are all muted so we may commence the 21 meeting.
22 I should say that when it comes time to 23 make comments, the way to do that is to use star 6, I 24 think that's correct. We'll now proceed with the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
8 meeting and I think it's Rob Taylor. Rob Taylor, are 1
you there?
2 MR. TAYLOR: I am, can you hear me?
3 CHAIR BALLINGER: Yes, very well. If 4
you'd like to make some comments, please do so.
5 MR. TAYLOR: I'd like to make just a few 6
opening remarks and thank you very much for the 7
opportunity. Good afternoon, for those who don't know 8
me my name is Rob Taylor, I'm the Deputy Office 9
Director for New Reactors at NRR.
10 I don't just do new reactors, my portfolio 11 includes materials, issues, and subsequent license 12 renewals for the operating fleet as well so the staff 13 that are working on this topic work for me.
14 And this is an issue that I've been 15 following very closely and working with them on here 16 for the last year or so as we've been working up the 17 issue in more and greater detail.
18 So, we appreciate the opportunity to 19 discuss this topic which we are working on to assess 20 its safety significance and the potential need for 21 regulatory action.
22 Specifically, today you'll hear from NRC 23 Staff about their effort associated with monitoring 24 and prediction of reactor pressure vessel 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
9 embrittlement during long-term operation of nuclear 1
power-plants.
2 As you indicated, this is part of a 3
continuing discussion of issues that were first 4
presented in a May 2020 public meeting. These issues 5
related to the embrittlement trend curves in Reg Guide 6
1.99 Rev 2 and 10 CFR 50.61 and the surveillance 7
testing in Appendix H.
8 During the meeting, the Staff will 9
describe a holistic risk-informed analysis performed 10 of these issues and its potential impact on reactor 11 pressure vessel integrity.
12 I want to assure everyone that the NRC has 13 high confidence that operating plants remain safe and 14 currently the NRC regulations provide reasonable 15 assurance of adequate protection against brittle 16 fracture of the reactor pressure vessel.
17 The issue we are going to hear about today 18 has the potential down the road to impact the Staff's 19 confidence in the integrity of the reactor pressure 20 vessel during long-term operation for some plants in 21 the United States.
22 The NRC Staff has further work to do to 23 determine which specific plants are impacted and to 24 what extent. The NRC Staff is proactively considering 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
10 risk-informed operations to address the combined 1
effects of both issues to ensure continued reasonable 2
assurance protection agenda brittle fracture in the 3
reactor pressure vessel during long-term operation.
4 The Staff is very interested in receiving 5
feedback from ACRS regarding the NRC Staff's approach 6
taken and its holistic risk-informed analysis being 7
presented today.
8 Other potential adverse impacts to plant 9
operations some of that should be considered and is 10 now the appropriate time to pursue these issues?
11 We sincerely appreciate the interest in 12 these topics and we're expecting a very interesting 13 and productive meeting.
14 So, that concludes my opening remarks.
15 Should I turn it over to Dave or do you want me to 16 hand it back to you?
17 CHAIR BALLINGER: I think we'll turn it 18 over to Dave but let's ask if any Members have any 19 comments that they would like to make before we pick 20 this up? We have a five-second rule for Members and 21 donuts so I think we're okay.
22 Dave, over to you.
23 MR. RUDLAND: Thanks, I think everybody 24 can see my slides, right? I'm going to talk a little 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
11 bit about this that Rob just gave us background on and 1
the reactor pressure embrittlement monitoring 2
prediction, especially long-term operations.
3 My name is Dave Rodman and I'm a senior 4
technical advisor for materials in the Division of New 5
and Renewed Licenses in NRR.
6 And just for bandwidth, I'm probably going 7
to keep my camera off while I'm making presentations, 8
and then I may turn it back on during questions and 9
answers if needed.
10 So, the purpose today is to go over some 11 of these issues. As Ron pointed out in his opening 12 comments, we have talked to the ACRS Subcommittee 13 several times about some of these issues so I'm not 14 going to go into too much detail about the individual 15 issues.
16 But I will touch on the embrittlement 17 trend curve that's in Regulatory Guide 199 Rev 2 and 18 the same trend curve that's in 10 CFR 5061 and the 19 issues there, as well as some issues with Appendix H 20 surveillance testing.
21 I want to spend a little bit more time 22 talking about a holistic risk-informed analyses that 23 the Staff did and its thoughts on the potential impact 24 to reactor pressure vessel integrity.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
12 And then I'll discuss some path forward 1
options. Even though we've had a couple of briefings 2
in the past I thought I'd go a little bit over 3
background of what it is that we're looking at and how 4
the uncertainties may come from these issues.
5 Building a trend curve is an estimate of 6
the change in fracture of toughness as a function of 7
fluents. If you look at this plot that's in the lower 8
lefthand corner, on the Y axis is the embrittlement 9
with a measurement of temperature on the Y axis.
10 The X axis is operating time and/or 11 fluents. The solid red line represents a schematic of 12 some embrittlement trend curve that has some initial 13 value of this transition temperature from brittle 14 fracture to ductal fracture that increases as the 15 operating time influence increases.
16 In addition, there's requirements for 17 surveillance capsule testing that provides monitoring 18 to ensure that this embrittlement trend curve properly 19 predicts plant-specific behavior.
20 In that same stage, the orange data-points 21 represent a fictitious schematic of that particular 22 data. Typically, the way that the current regulations 23 work is that embrittlement trend curve, there is a 24 margin added to that.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
13 That gives us our adjusted reference 1
temperature, or ART in this particular illustration 2
and then at different times those ART values are used 3
with the formulations that are in 10 CFR 50 Appendix 4
G to calculate pressure temperature limits for heat-up 5
and cool-down, which are represented by the blue, 6
green, and red lines on the slide on the right.
7 Those lines are the limits for operation 8
that will move or shift to the right as material 9
becomes more embrittled, so to warmer and warmer 10 temperatures.
11 So, ideally, we want an embrittlement 12 trend curve that provides conservative predictions of 13 embrittlement and surveillance data that covers all of 14 the operating periods for any one particular reactor.
15 If, for instance, the embrittlement trend 16 curve under-predicts the data we could have a level of 17 uncertainty which is illustrated in this left figure.
18 Again, in this particular figure which is the same 19 from the previous slide, you have a couple of 20 data-points.
21 And we have future data-points that have 22 not been pulled yet and if the particular trend curve 23 that we have underpredicts that, we can give an 24 uncertainty in what the actual embrittlement is from 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
14 that relationship.
1 Also, if the surveillance data is limited 2
or if periodic embrittlement measurements are not 3
made, that of course increases the uncertainties since 4
we don't know if the plant-specific behavior is 5
matching the embrittlement trend curve.
6 I can state this in a slightly different 7
way and maybe a little bit more clear way.
8 If you look at this top behavior on this 9
slide where we have periodic regular data that's 10 pulled from particular plants, embrittlement trend 11 curve, which is represented again by the red line, 12 that matches that behavior relatively well.
13 Then we have known embrittlement behavior 14 with a known uncertainty. And from that then are 15 regulations and margins that we put on that would be 16 perfect to keep the predictions conservative.
17 However, if we have gaps in the data or 18 not future data plus an embrittlement trend that is 19 underpredicting the experimental data from the plant 20 surveillance, then we end up with very large 21 uncertainties.
22 And so what we have in the regulations 23 may not properly account for that amount of 24 uncertainty and we don't know exactly what 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
15 embrittlement we have at a high fluence long-term 1
state.
2 And that uncertainty and what those 3
impacts are we felt needed to be studies holistically 4
to really understand what the impacts of those 5
uncertainties are on the behavior of the vessel.
6 I'm going to start a little bit with our 7
conclusions and then I'll talk with her about the 8
details afterwards.
9 And so from this the Staff still has very 10 high confidence that the current operating plants 11 remain safe and that our licensing actions that we've 12 made to date remain valid.
13
- However, there is insufficient 14 embrittlement monitoring under prediction that occurs 15 of this reactor embrittlement and that will eventually 16 impact the Staff confidence in the integrity of the 17 vessel.
18 MEMBER BLEY: Can I interrupt you at this 19 point?
20 MR. RUDLAND: Sure.
21 MEMBER BLEY: This is Dennis Bley. Two 22 questions for Ron and then something for the Staff.
23 Ron, are we expecting to write a letter on this next 24 month or soon?
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
16 CHAIR BALLINGER: To answer your question, 1
I would like to write a letter on this topic but not 2
right now. We need to decide on where there's a 3
convenient milestone point, if you will, for this to 4
happen.
5 The Staff is in the process of doing a lot 6
of analysis and evaluating and should a rule be the 7
path that they want to take, then we go from there.
8 I suspect that would be the best time, 9
unless the Subcommittee wants something different, to 10 write a letter because then we'll have a complete idea 11 of what the path forward is.
12 MEMBER BLEY: Thanks, that helps me a lot 13 to put perspective on this. I just wanted to make a 14 comment for the Member first but then I'd be happy to 15 hear what the Staff has to say.
16 When I think about uncertainty of the kind 17 I think we're going to hear more about soon, good 18 treatment of uncertainty gives us and is a reasonable 19 approach, in fact, it's the only approach I think that 20 allows us to have confidence in our decision-making.
21 On the other hand, using an uncertainty 22 based on limited data when there's new data available 23 just seems like we're dodging the issue of keeping 24 current with the science and that sort of thing I find 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
17 uncomfortable and not part of best practice.
1 So, I'm done, if anybody wants to say on 2
that, that's great, if not that's okay too.
3 CHAIR BALLINGER: This is Ron again. I 4
would expect that the trend curve that was based on 5
the latest data would certainly factor into and has 6
been used in this analysis to determine the 7
uncertainty.
8 Am I right, Dave?
9 MR. RUDLAND: That's correct.
10 CHAIR BALLINGER: So, also, there's a bit 11 of a subtlety that may have been missed on Slide 2 and 12 that is that we have been in the past time, much to my 13 dismay, dealing with PTSand embrittlement in different 14 ways, in effect.
15 And the analysis that the Staff is doing 16 I think would seek eventually to rationalize and come 17 up with a single trend curve for both PTS and 18 embrittlement. Am I right again, Dave?
19 MR. RUDLAND: I'll be touching on that in 20 the presentation.
21 CHAIR BALLINGER: That's kind of an 22 important thing.
23 MR. RUDLAND: To be clear, right now the 24 embrittlement trend curve that's in 199 Rev 2 and 5061 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
18 is the same trend curve, just to be clear. So, as we 1
move forward and as I talk about this, I want you to 2
keep that in mind.
3 CHAIR BALLINGER: But 61A allows a 4
probabilistic approach, right?
5 MR. RUDLAND: Yes, 61A uses a different 6
trend curve.
7 CHAIR BALLINGER: That's what I was 8
referring to.
9 MEMBER HALNON: This is Greg. Could you 10 just briefly tell us why you have high confidence that 11 we're safe now, just that first bullet on that slide 12 you were just on?
13 MR. RUDLAND: Again, I'll get to that as 14 I'm discussing it but I think the reason why is 15 because at the current state of where the plants are 16 in terms of the fluence levels, the trend curve does 17 a pretty good job of predicting the embrittlement.
18 And so we're not as concerned with the 19 behavior now because, again, the trend curves are 20 predicting things. It's out in the future that they 21 begin to underpredict the behavior.
22 And it's out in the future where we also 23 don't have surveillance data to make sure that the 24 trend curves are being accurate.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
19 MEMBER HALNON: I think at some point 1
you're going to talk about the year 2025. If I 2
haven't figured it out by then I'll get back into your 3
questioning.
4 MR. RUDLAND: Is there anything else? Can 5
I go on?
6 MEMBER HALNON: Sure.
7 MR. RUDLAND: So, as I was talking about, 8
it will eventually impact our confidence.
9 Right now we're kind of estimating that it 10 would be about 10 years before we would start to have 11 an issue and I'll get into the details here in a 12 second for the PTS.
13 And it's about 23 years for PT limits, and 14 why those are different and stuff like that I'll get 15 into in a second.
16 But the Staff needs to do continued work, 17 especially to determine which plants are impacted by 18 this potential issue because it becomes very plant-19 specific issue as you'll see in a second.
20 I'm going to start with the trend curve 21 issues before I move on to the surveillance issues.
22 In May of 1988 the NRC published Reg Guide 199 which 23 contained an
- improved, and this is Rev 2,
24 embrittlement trend curve that was fit to 177 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
20 data-points.
1 And in 1991, the NRC updated 5061 to 2
include that same EPC from Reg Guide 199. Since then, 3
we haven't really touched it all that much but we did 4
a reevaluation for the continued adequacy in 2014 and 5
in more detail in 2019, which kicked off basically 6
this effort that I'm talking about today.
7 It was both those documents that are in 8
Adams and publicly available. So, the issue is shown 9
here.
10 This plot gives you on the Y axis is the 11 difference between the predicted embrittlement and the 12 measured embrittlement, and the measure from 13 surveillance data predicted in this case by the trend 14 curve in Reg Guide 199 where a zero value would be a 15 perfect prediction against the fluence, which is on 16 the X axis.
17 As you can see, the data is relatively 18 scant. The red data here represents the U.S.
19 surveillance data and the grey data-points represent 20 international surveillance data.
21 What we see is that when we get to about 22 3e to the 19 neutrons per centimeters squared, the 23 trend curve begins to deviate from that zero or 24 perfect prediction.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
21 By the time you get to 6e to the 19 it 1
becomes statistically significant, which means the 2
trends themselves are below the 2 sigma scatter within 3
the data. The dashed heavy lines represent plus and 4
minus 2 sigma on the data.
5 Those values are in Reg Guide 199. By 6
the time we get to 1e to the 20th about 180 degree 7
Fahrenheit under prediction in embrittlement. This 8
particular plot is for base metals and again, we'll 9
talk about what these values mean in terms of our 10 current operating fleet in a little bit.
11 For weld materials we see similar trends 12 but not as apparent, same plot but the data now 13 represents surveillance data from welds and what you 14 see is not as much data in high fluence.
15 Data seems to be more scattered but 16 doesn't seem to have that large drop-off and that may 17 just be due to the limited data of high fluence.
18 MEMBER REMPE: This is Joy. On the prior 19 plot, I just looked at this, what was the reason that 20 the non-U.S. seems to have more variance?
21 MR. RUDLAND: In terms of this data down 22 in the --
23 MEMBER REMPE: Yes.
24 MR. RUDLAND: I think what's happening is 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
22 that we just don't have enough data at high fluence.
1 So, we have some data here which represents these red 2
data-points that are in this region.
3 I don't know if you can see my cursor or 4
not.
5 MEMBER REMPE: I can.
6 MR. RUDLAND: This data in here is U.S.
7 data. For international, we just don't have any data 8
right now that's in this range.
9 MEMBER REMPE: That helps, thanks.
10 MR. RUDLAND: I think the thought at least 11 from the Staff is that the data that we do have seems 12 to be following the same trend. And so as more and 13 more data is gathered, we suspect it will follow in 14 this trend but we don't have that evidence.
15 And actually, Elliot Long will be talking 16 about some programs that are currently underway to try 17 to develop some of that data in a more efficient 18 manner and he'll give some timeframes in when he 19 expects to have that data.
20 MEMBER REMPE: Thank you.
21 MR. RUDLAND: You're welcome.
22 So, within Reg Guide 199, the prediction 23 of the change in embrittlement due to radiation is 24 represented by this functional form that's shown here 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
23 that's a function of something called a chemistry 1
factor, which is a function of nickel and copper, and 2
a function that is just a function of fluence itself.
3 So, if we were to plot that particular 4
fluence function against fluence itself, we can see 5
the trend and what we see is that once we get above 3 6
to the 19, the slope of that particular fluence 7
function begins to change and begins to decrease and 8
actually reaches a maximum of about 2e to the 20 and 9
begins to drop off.
10 This is the same point at which we saw 11 that the material in two slides above began to deviate 12 from accurate predictions.
13 So, that seems to suggest that the fluence 14 function itself is probably part of the issue of why 15 we're seeing this under-prediction in embrittlement 16 trend because the fluence function that's the base of 17 those calculations begins to decrease at that point.
18 Whether the actual trend follows this 19 light blue line or doesn't or is slightly higher is 20 unknown at this point but the behavior seems to be 21 related to the actual fluence function that is part of 22 the embrittlement trend curve in Reg Guide 199.
23 I'm going to move on to a little 24 background on surveillance capsules. Appendix H to 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
24 Part 50 requires periodic monitoring of the changes in 1
fracture toughness caused by neutron embrittlement.
2 And typically, these programs are three to 3
five surveillance capsules that are placed inside the 4
vessel closer to the core than the vessel wall itself 5
and they're pulled periodically.
6 The surveillance capsules contain sharpie 7
specimens and pencil specimens and other specimens, 8
and they're tested to make a prediction of what the 9
embrittlement in the vessel would be.
10 Like I said, there's usually about three 11 to five in there and Appendix H incorporates by 12 reference ASTM Standard E185-82.
13 In that particular standard which was 14 developed for a 40-year life originally, it allows the 15 final capsule to be pulled at a fluence of about 2 16 times the design fluence, which over its life would be 17 about 40 years.
18 However, plants have been changing that 19 design fluence to the current license length of 60 or 20 80 years depending on if their license had been 21 extended.
22 That allows the last capsule to be held 23 without testing and the reason being that for a 24 typical program with a 40-year life, the second to 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
25 last capsule typically had sufficient fluence to be 1
able to represent the vessel wall at 40 years.
2 However, as plants began to apply for and 3
receive extended licenses, these last capsules had 4
been delayed a bit to represent the fluences at 60 or 5
for subsequent license renewal 80 years.
6 And in
- 1997, the NRC wrote an 7
administrative letter that stated that staff review of 8
these requests of surveillance capsule withdrawal 9
scheduled changes has stayed limited to verification 10 of performance with the NSTN standard and cannot be 11 based on a technical or safety review.
12 And because of
- that, the capsule 13 withdrawals and testings all repeatedly delayed in 14 some cases to achieve higher fluence.
15 In license renewal the regulations were 16 unchanged when it came to surveillance testing.
17 It was that way and the recommendations 18 were put into guidance and it was done that way for 19 flexibility, to allow the plants to be flexible in how 20 they demonstrated adequate aging management of DRPV 21 due to embrittlement and different specific 22 circumstances.
23 I'm sorry, is there a question? And in 24 that guidance, which was both in the GALL Report which 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
26 is NUREG 1801 for license renewal and in SOR GALL 1
which is NUREG 21 and 91, there were several 2
statements about it.
3 In 1801 Rev 1, there was a statement that 4
at least one capsule needed to be tested with the 5
projected fluence equal to exceeding 60 years of life 6
and in GALL SOR a very similar statement was put in 7
that said a neutron fluence capsule between one and 8
two times the peak fluence of interest at the end of 9
the subsequent period of operation should be tested.
10 And it also specified that it's not 11 acceptable to redirect or postpone the withdrawal of 12 testing to achieve a higher neutron fluence.
13 In practice, what's happening is that many 14 licensee are changing those capsule withdrawal 15 schedules prior to the application of license renewal 16 and receiving a review that is basically a conformance 17 verification review.
18 And then when they submit their license 19 for renewal, they are consistent with the GALL 20 program. It's still allowing the plants to possibly 21 delay their final capsule, leaving large gaps of time 22 between data that's being pulled for embrittlement.
23 Here's a particular example of that. This 24 plot shows neutron fluence, again the Y axis and the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
27 date on which the capsule is pulled and tested on the 1
X axis. The last data-points represent four capsules 2
that were pulled and tested.
3 The orange symbols represent the fifth 4
capsule that will delay four times finally for their 5
subsequent license renewal of capsule that would be 6
tested where you see the red symbols.
7 That's a lot of years, that's 20 years 8
between the last capsule and the next, and many 9
licensees have done this. Here are some examples that 10 are shown here in this plot of plants that have 11 delayed their final capsules.
12 But not all have done it. Many have 13 actually tested capsules very periodically and not 14 have these large gaps.
15 CHAIR BALLINGER: This is Ron. These 16 capsules are forward-looking so if you look at the 17 slide that says the 80 years point, how long is that 18 in calendar year? Do you see what I'm getting at?
19 How much of a forward --
20 (Simultaneous Speaking.)
21 MR. RUDLAND: This is when the plant will 22 get to 80 years and that's the 80-year fluence. So, 23 that's the difference between when the plan is 24 actually 80 years and when it reaches 80-year fluence.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
28 When that capsule reaches 80 fluence it 1
will be in 2020 whatever that is, 2024 or 2025.
2 Compared to when it actually reaches 80 years which 3
will be in 2051 or something like that.
4 CHAIR BALLINGER: So, it's considerably 5
forward-looking?
6 MR. RUDLAND: Yes, and that's typical 7
because again, remember these capsules are closer to 8
the core than the wall so they have a higher lead 9
factor so they're getting fluence at a much higher 10 rate than the wall.
11 CHAIR BALLINGER: Thanks.
12 MR. RUDLAND: I'll show the same kind of 13 data but in another plot and I think this goes to 14 answering one of the questions earlier.
15 This is the same plot that shows the 16 difference between the predicted and measured values 17 of embrittlement as the functional fluence.
18 The four data-points here in green 19 represent four of the capsules pulled and you can see 20 when those four capsules were pulled at that 21 particular fluence. The embrittlement trend curve 22 does a
pretty good job of predicting that 23 embrittlement.
24 It reaches 60 years here and 80 years here 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
29 and this is where the fifth capsule will be pulled.
1 So, it may be that particular plant has values of 150 2
to 180 degrees under-predicted. It may be, I don't 3
know for sure until they actually pull the capsule.
4 But you can see there's that large gap 5
there which, again, adds to uncertainty. I don't 6
really know what the embrittlement state is beyond 7
that 2020 value until they've actually pulled and 8
tested that capsule.
9 And again, the way the rules are now, that 10 particular plant could get to that time and decide 11 they want to delay that capsule even further and we 12 would have to approve that with the same performance 13 check as we've done in prior approvals.
14 I show this in a slightly different space 15 but it also shows the behavior of the embrittlement 16 trend curve to predict some of these. These are the 17 same four data-point, embrittlement on the Y axis 18 versus fluence on the X again.
19 If you use just the chemistry for that 20 material to predict the trend curve I would get this 21 orange line.
22 Reg Guide 199 allows the user to fix the 23 data-points they
- have, fit the embrittlement 24 data-points they have, basically adjust the chemistry 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
30 factor to give them a new trend curve that fits their 1
data.
2 That would be the blue line here. So, you 3
can see that drops their prediction of embrittlement 4
by about 25 to 50 degrees Fahrenheit.
5 If this particular data follows the trends 6
that were expected, I could expect to have about 150 7
degrees Fahrenheit underpredicted for this data if it 8
follows the trends of the database data that was shown 9
earlier.
10 And again, if I fix that data with the 11 trend curve as it is, I would get this orange yellow 12 line and what's happening is because of that fluence 13 function that kind of levels off, it does not accuracy 14 predict this behavior.
15 So, I would still have even when I would 16 fit that particular data about a
75 decree 17 underprediction in the data.
18 CHAIR BALLINGER: This is Ron again. I've 19 been looking at this for so long, sometimes you don't 20 see the forest from the trees. That fluence factor is 21 nothing more than an empirical fifth.
22 There's no physics associated with that.
23 MR. RUDLAND: That's correct.
24 CHAIR BALLINGER: And so the fact that it 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
31 all of the sudden goes a little strange after a 1
certain fluence, well, it's just an empirical fit.
2 MR. RUDLAND: That's right, and realize it 3
was only an empirical fit to about 170 data-points at 4
the time. Now we have 1000 data-points or something 5
like that, so it was fit well to the data they had at 6
the time.
7 And it was not expected to be used for 8
this high fluence values because there was no data at 9
that point to help calibrate the fifth.
10 CHAIR BALLINGER: It's a good lesson, 11 don't extrapolate.
12 MR. RUDLAND: Exactly right. What are the 13 impacts of those two things? Any one can be explained 14 away with conservatisms or whatever but together they 15 may actually have a much larger impact.
16 So, the Staff wanted to look at these 17 combined effects and they wanted to use risk-informed 18 thinking in order to do that, which includes not just 19 looking at how these changes affect the actual 20 calculation of risk but also how do they affect 21 performance margins, safety monitoring, defense 22 in-depth in order to help us make a decision.
23 To do that, we used a targeted sample of 24 plants and we used the data that we had available to 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
32 us. I realize that not much plant-specific 1
information is available. We only have limited data 2
when we did this particular analysis.
3 The assumptions that we made were that we 4
wanted to do some comparisons with AASME 915, I think 5
Ron talked about this a little bit earlier. A report 6
was written in 2020 where the Staff found that this 7
particular embrittlement trend curve provided the most 8
accurate characterization of the database.
9 So, the database is what I'm talking about 10 here. This is data that was used by ASPM to basically 11 fit this embrittlement trend curve to this data. So, 12 you would expect then that these trends would show 13 that trend curve could accuracy predict through the 14 third range of fluence those data.
15 And that's what it shows. And so we 16 wanted to use that as kind of our comparison tool 17 since we knew it was the most accurate to the data in 18 that database.
19 We did a targeted sample, a targeted 20 study, we looked at 21 plants, we emphasized high 21 fluence plants but we threw in a few that had low 22 copper and BWRs to round out the study.
23 And we looked at changes in these adjusted 24 reference temperatures I talked about earlier, from 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
33 switching embrittlement trend curves.
1 How much would the adjusted temperature 2
change if I switched from Reg Guide 199 to E915? And 3
we called that the embrittlement shift delta. And the 4
embrittlement shift delta results were used to help us 5
benchmark our risk analysis.
6 So, what we found out from the data that 7
we were able to develop for these 21 plants was that 8
there's a tendency for the material reference 9
temperatures to increase when switching from Reg Guide 10 199 to ASME 900.
11 And you've got to realize that there's 12 other parameters in E900 besides what I talked about 13 for Reg Guide 199, which is just nickel, copper, and 14 fluence.
15 There are other things that E900 also 16 uses. The trends were to basically increase that 17 reference temperature when moving to ASME 10 E900. We 18 saw that trend more in base metals and in weld metals, 19 I'm not too sure why that is, but that was the trend 20 we saw.
21 And only really a small handful of plants 22 had changes in this reference temperature greater than 23 50 degrees Fahrenheit and those tended to be plants 24 that were at higher fluences greater than the 6e to 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
34 the 19.
1 We assumed that this range that we got of 2
the shift deltas bounded the three and so we used 3
those in our risk studies. Staff did a probabilistic 4
fracture mechanics calculations.
5 We looked at a bunch of different 6
transients, we looked at a bunch of different flaw 7
sizes. We didn't have a lot of information on fluence 8
maps and things like that so we kept those kinds of 9
things constant, and we ran these analyses.
10 What I'm showing here in this particular 11 plot is one particular place. I had conditional 12 probability of initiation and conditional probability 13 of failure on the Y axis and this embrittlement shift 14 delta on the X axis.
15 And remember, this embrittlement shift 16 delta basically represents the under-prediction in the 17 transition temperature in the delta RT and DT.
18 If you look at the conditional probability 19 of failure just as a comparison for this point, a 50-20 degree shift in this is a 50 degree underprediction 21 basically from Reg Guide 199.
22 This would give us about three orders of 23 magnitude increase in the conditional probability of 24 failure. And a 150-degree shift would give us about 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
35 six orders of magnitude, a change in additional 1
probability of failure.
2 But please realize these values are 3
conditional on the frequency of the transients. For 4
instance, this particular one is for a cool-down that 5
follows the PT curve, which is a very unlikely event.
6 We don't know really what the frequency of 7
some of these transients are. We did do some actual 8
plant transients, which I'll talk about in a second, 9
but a lot of the frequency of some of these more rare 10 transients is unknown.
11 There's also a lot of plant variations 12 that we did not account for in these analyses that add 13 to the uncertainty.
14 We don't know if these analyses are 15 bounding and we really don't know how much protection 16 administrative and other limits provide against 17 violating PT curves and I'll get into that when we 18 talk about safety margins a little bit.
19 All of the analyses are summarized in this 20 report that I have here that was a 2021 report that's 21 publicly available on ADAMS. So, the risk seems to be 22 low because of the frequency of the transient but the 23 uncertainties are very large.
24 This was again for heat-ups and cool-25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
36 downs.
1 MEMBER PETTI: Can I ask a question? What 2
are the yellow curves?
3 MR. RUDLAND: The yellow curve is the 4
conditional probability of initiation. So, the way 5
these probabilistic fracture mechanics work is they 6
look at when a crack may initiate versus when it may 7
actually go through a wall and fail the vessel.
8 Those are always going to be lower because 9
in some cases the cracks are rushed depending on the 10 embrittlement.
11 CHAIR BALLINGER: This is Ron. The report 12 has a lot more information in it but whenever I see a 13 curve that has conditional probability of initiation 14 or failure, 10 to the -7 or -8, I immediately ask the 15 question what's the uncertainty?
16 And I look for air bars to be plotted on 17 these things and I never seem to see them.
18 MR. RUDLAND: The air bars are difficult 19 to determine what they are because there's so much 20 unknown right now.
21 So, I'll talk about this in a second but 22 that's why we decided to look at some other things 23 besides just these risk numbers because the 24 uncertainties are so large.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
37 I need to make a correction to what I 1
said. The yellow lines down here are actually a 2
different flaw size so the difference between the dark 3
purple and the light purple is the difference between 4
the initiation and failure probability.
5 The difference between the light and dark 6
lines are different flaw sizes. The Appendix G 7
calculations assume a quarter T flaw where the flaw is 8
a quarter way through the wall thickness.
9 So, that's the higher curves. The lower 10 curves are smaller flaw, that's only 3 percent through 11 the wall.
12 MEMBER BLEY: I guess that last discussion 13 left me a little lost. If we can't characterize that 14 uncertainty well because it's so large, we don't have 15 a clue if these things are means, medians, or 90th 16 percentiles or 20th percentiles.
17 They're just numbers. It's back to that 18 question, how are you getting confidence? And if you 19 don't have any confidence in the uncertainty, you 20 can't have any confidence in what these numbers really 21 represent.
22 MR. RUDLAND: Right, I think we tried to 23 pick as many bounding results as we could to not be 24 caught by the uncertainties so much. So, the fluence 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
38 maps that we picked, the geometries that we picked, 1
the flaw sizes that we picked were all meant to try to 2
bound the results.
3 MEMBER BLEY: Same thing for upper bounds?
4 MR. RUDLAND: I think so, yes, and as I 5
talk about it in this particular plot, where we get 6
worried is we don't know exactly what the event 7
frequencies are.
8 When we calculate conditional probabilities of 9
failure that are less than 1e to the -6, we know that 10 the frequency of the transients doesn't matter all 11 that much because the impact to core damage 12 frequencies will be very low.
13 Because the two wall clash frequencies 14 will be way lower than 1e to the -6. Remember the two 15 wall clash frequency is conditional probability of 16 failure times the transient frequency.
17 And in a risk analysis, we equate the two 18 wall clash frequency to core damage frequency for 19 conservatisms. So, any conditional probability of 20 failures that are less than 1e to the -6 are not 21 concerning to us.
22 The ones that are the most concerning are 23 the ones where the conditional probability of failures 24 are greater than 1e to the -6 but I don't really quite 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
39 know what the event frequencies are.
1 That's where I start to get concerned. I 2
know they're probably low because following the PT 3
limit is a very low frequency event, I just don't 4
really know what it is and so I'm uncertain about 5
that.
6 For the leak test, for instance, what 7
drives the leak test are not the frequencies so much 8
as the cool-down rates and I need additional 9
information to determine if the high cool-down rates 10 really are an issue.
11 That's the added uncertainty there. And 12 because it was difficult for us in this study to 13 answer these questions, that's why we looked at some 14 of the other things like safety margins and 15 performance monitoring before making the decision that 16 we need to consider change.
17 CHAIR BALLINGER: The key here is I think 18 that the embrittlement trend is slowly varying and as 19 long as some other phenomenon doesn't rear its ugly 20 head that we don't know about, which I suppose is 21 possible, we're in a good way.
22 MR. RUDLAND: Right now we're in a good 23 way and I think our concern comes in the future when 24 we begin to have this underprediction.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
40 Because what happens is that will continue 1
to get worse, it will continue to underpredict at 2
higher and higher
- fluences, which drives the 3
uncertainty even higher.
4 In addition, we looked at 5061, which is 5
the thermal shock rule and again, in that particular 6
rule, a calculation of RPPS is done using the 7
embrittlement trend curve. That might be impacted.
8 And what you need to remember is there's 9
a screening criteria where if the RPPS exceeds 270 10 axel welds or 300 for circ welds, then something needs 11 to be done for PTS.
12 However, for the sample plants that we 13 did, we ran those through the analysis as part of 14 5061A and demonstrated that the wall crack frequency 15 for PTS were all less than 1e to the -6 for all the 16 cases investigated.
17 And so the impact on risk for a single 18 shock was shown to be low.
19 CHAIR BALLINGER: But 61A uses a different 20 correlation. How does that work?
21 MR. RUDLAND: We actually looked at the 22 correlation using Reg Guide 199, the correlation which 23 is in 5061A as well as E900-15. We looked at all 24 three of those trend curves in doing these 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
41 calculations.
1 CHAIR BALLINGER: Thanks.
2 MR. RUDLAND: So, because of this whole 3
discussion of uncertainty and our feelings that we 4
were very unsure about the behavior, especially at 5
high fluence, we wanted to look at other things.
6 We knew we needed to resolve these issues 7
to maintain safety and be comfortable with the 8
uncertainty so we decided to look at both the impacts 9
on the safety margins as well as on performance 10 monitoring.
11 I illustrate these safety margins with 12 this particular schematic. The Y axis again is 13 pressure and the X axis is temperature, and the shaded 14 area is a typical operating window where a plant may 15 cool down or heat up.
16 So, it starts up here at a high 17 temperature and high pressure and decreases pressure 18 and temperature to stay within this window.
19 There's a limit that sits out here, a 20 structural limit, where if the plant was to be in this 21 region up here for the high temperature and pressure 22 and decrease temperature without decreasing pressure, 23 they may cross this structural limit and increase the 24 chances of a brittle fracture.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
42 On the regulations with an accurate PT 1
curve, we have a regulated PT curve that lays here 2
that restricts them from getting close to that 3
structural limit by an adequate amount of margin.
4 So, now that when they're cooling down 5
they need to stay to the right of this orange curve 6
and the amount of margin that is chosen here is 7
directly proportional to that uncertainty in the 8
problem.
9 This margin was chosen knowing the 10 analysis methodologies that went into calculating 11 that, the amount of conservatisms, the amount of 12
- unknowns, and that margin was chosen to be 13 appropriate.
14 However, if your PT curve actually is here 15 because we're using a trend curve that underpredicts, 16 we can get a false sense for what the actual operating 17 margins are and a reduced margin to brittle fracture.
18 What that's showing is that actually, even 19 as the uncertainties are increasing, the margins to 20 brittle fracture are actually decreasing. So, that's 21 going opposite to the trends that we want to stay 22 consistent with in our definition of margin.
23 So, with a reduced PT curve we're reducing 24 that particular margin but the uncertainties are 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
43 increasing. That's something that we think we 1
probably need to fix because the margins that we have 2
need to be represented by the amount of uncertainty 3
that we have.
4 Same with performance monitoring. The 5
performance monitoring that we have ensures a couple 6
of things. It ensures that analysis results remain 7
valid with time and that there's nothing unexpected, 8
there's no adverse safety issues or changes that we 9
were not prepared for.
10 And delaying those capsules for a long 11 period of time with the possibility of no future data 12 because, again, remember ASTM E185-82 allows the last 13 capsule not to be tested. That really represents a 14 lack of performance monitoring.
15 Improved dictations were aware for 16 particular plants we have no data at those high 17 fluence levels. With our current state of knowledge, 18 the general analyses that we did suggests that the 19 overall risk of brittle fracture is low.
20 And again, we tried to be as bounding with 21 those analyses as we could based on our knowledge and 22 the data that we had, but the uncertainty is high 23 because of that and it increases with time.
24 A lot of the plant-specific details that 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
44 are needed for individual analyses weren't considered 1
because it wasn't available at the time for the Staff.
2 Under certain conditions, safety margins may be 3
impacted and are probably decreasing as that 4
uncertainty increases.
5 And delaying capsules at high fluence 6
really does represent the lack of sufficient 7
performance monitoring. But most of these issues of 8
the plants with the fluences are greater than 60 to 9
the 19.
10 Who is impacted by this? If we look at 11 the embrittlement underpredictions within the fleet 12 for both 1680, we have about 34 percent of the PODRs 13 will surpass that fluence level on the ID of their 14 vessel by 80 years.
15 They will surpass 15 percent by 80 years.
16 Realize that plant-specific details such as limiting 17 material and other things may contribute to which 18 plants are impacted.
19 So, the Staff needs to have more time to 20 be able determine which individual plants may or may 21 not be impacted. About a third of the PWR fleet could 22 be impacted by 80 years.
23 Of course, any plant that renews its 24 license and chooses to renew the last capsule would be 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
45 impacted by these issues.
1 So, because of the time-dependent 2
behavior, the Staff feel the regulations right now are 3
sufficient for a reasonable show-in for adequate 4
protection against brittle fracture for the plants as 5
they are right now.
6 But the Staff wants to ensure continued 7
reasonable assurance in long-term operations and come 8
up with fixes for these identified issues of those 9
embrittlement and surveillance, and use a
10 risk-informed performance-based solution to provide 11 those remedies.
12 We also want to make sure that whatever 13 solutions we come up with are focused and doesn't 14 impact plants that are not adversely affected by the 15 issues. So, for instance, there are plants that have 16 surveillance data that cover the end of their license 17 already.
18 And there are plants such as BWRs, where 19 they project improvements at the end of the license is 20 going to be less than 3 to 19 so they're not going to 21 be impacted by this.
22 To do that the Staff has been considering 23 and doing continued study to determine what their 24 options should be to find the remedies for this, 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
46 whether it's a blanket specific action or a focused 1
regulatory action, or possibly some other tools to 2
properly fix this for the long term.
3 But again, in summary, like I stated at 4
the beginning, the Staff really does have high 5
confidence right now that the operating plants remain 6
safe and that the recent licensing actions are valid 7
but the issue is a long-term issue.
8 It could be 10 years for PTS or 23 years 9
for PT limits but these issues will eventually impact 10 the Staff's confidence in the integrity of the vessel.
11 I focused mainly on performance margins and safety 12 monitoring because those seemed to be most highly 13 impacted by these issues.
14 And further work needs to be done. But we 15 want to come up with and we want to be proactive in 16 coming up with a solution that provides continued 17 reasonable assurance and use risk-informed 18 performance-based solutions to do that.
19 And like I say here, we decided on a 20 really focused solution and don't want to overburden 21 any plant that's not adversely impacted by this issue.
22 I think that's my last slide.
23 CHAIR BALLINGER: We are way ahead of 24 schedule.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
47 MEMBER REMPE: To help you slow down, Ron, 1
I've got a question.
2 CHAIR BALLINGER: That was my next comment 3
was now we have plenty of time for questions.
4 MEMBER REMPE: I have to go back again to 5
that plot and the non-U.S. data, if you look at even 6
before 10 to the 19 fluences, you can look at Slide 15 7
or several slides that have it there.
8 We talked about there isn't much data as 9
you get near 10 to the 20th but what about earlier?
10 Is it statistically significant, some of the 11 data-points that are non-U.S. that are a lot larger in 12 the difference between predicted and measured?
13 MR. RUDLAND: Are you talking about in the 14 region that is --
15 MEMBER REMPE: Starting where the green 16 line is on the left and you see one greyed off, and 17 then you see several between there and 10 to the 18 20th.
19 MR. RUDLAND: I don't think that's a lot 20 of data so it's not really significant.
21 Sometimes there are outliers and actually, 22 when it comes to determining if material is credible 23 or not, there are statistical rules you can do to 24 check whether or not they are statistically 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
48 significant or not.
1 But these kinds of data-points that are 2
that far away from the mean, they're outside of the 3
signal from the mean. So, they're probably not 4
significant.
5 MEMBER REMPE: It's just interesting that 6
U.S. data seems to be closer to what's considered the 7
mean value or zero variance than the non-U.S. data and 8
I just wondered if there was anything to be learned.
9 But you just think all of those 10 data-points are not statistically significant?
11 MR. RUDLAND: I think what I learned from 12 this, Joy, is that the data that we have now 13 demonstrates that the standard deviation is actually 14 probably greater than what came from the Reg Guide 199 15 analysis.
16 So, remember these dashed lines are the 17 standard deviation that came out of Reg Guide 199 Rev 18 2 from the 170 odd data-points that they originally 19 calibrated the curve to.
20 And again, it's clear, and we've done 21 these deviations from the database, it's slightly 22 larger than that which was predicted back when Reg 23 Guide 199 was developed.
24 So, standard deviation is bigger so the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
49 scatter in data is more so than what was originally 1
thought.
2 (Simultaneous Speaking.)
3 MEMBER REMPE: I'm just wondering if 4
there's something to be learned about why the U.S.
5 data seems to be closer. There's only a few 6
data-points, I just thought it was curious.
7 MR. RUDLAND: Again, I don't know about 8
this data-point here at 1e to the 19 but it may be 9
that certain conditions may not be fully as 10 representative to our data as the data-points because 11 they actually come from our plants.
12 I don't know about those particular 13 data-points, we'd have to go back and take a look at 14 them.
15 MEMBER REMPE: It will be interesting as 16 you go forward if you see the same thing is still 17 occurring.
18 For example, if it stays around -50, 19 whereas the data-points on the far right down low are 20 a bit lower, if there's something going on with the 21 U.S. versus non-U.S. data.
22 Just a question.
23 MR.
RUDLAND:
Priors look at the 24 embrittlement trend curve, we tried to focus only on 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
50 the U.S. data in terms of whether the fix to the trend 1
curves are good or not.
2 And only use the international data to 3
help us understand the overall trends because there 4
are some differences, of course, between the way 5
international plants operate their plants versus how 6
we operate ours in terms of the design and stuff like 7
that also.
8 So, in developing the trend curve we would 9
try to only use the U.S. data as best we could.
10 MEMBER REMPE: Thanks.
11 MEMBER HALNON: This is Greg Halnon. One 12 of the public comments was to go after the 13 decommissioned reactors to try to get more data. Was 14 that useful?
15 MR. RUDLAND: I think the problem with 16 that is that a lot of the decommissioned reactors 17 don't have high enough fluence to really be useful for 18 this particular study.
19 MEMBER HALNON: Yes, I was wondering about 20 that because a lot of them are shutting down early --
21 (Simultaneous Speaking.)
22 MR. RUDLAND: And while it would be good 23 data to have, I'm not sure it would be that useful for 24 this particular issue.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
51 CHAIR BALLINGER: Other questions from 1
Members?
2 MEMBER PETTI: I have one, this data is 3
all from actual plants but in the early days there was 4
a ton of work done just in the test reactors and the 5
like. And I recall there being some bias with that 6
data.
7 But in terms of trend, is there data out 8
there from accelerated testing and material test 9
reactors with these higher fluence? Does it confirm 10 what you see in the green data-points?
11 MR. RUDLAND: Yes, there's data from test 12 reactors and there appears to be a flux effects.
13 MEMBER PETTI: That's what I remembered, 14 yes.
15 MR. RUDLAND: And actually, ASME Section 16 11 is putting together a code case and one of the 17 things they're looking at is how to correct the test 18 reactor data for the flux effect so that it could be 19 used and prepared directly then to the data from 20 operating reactors and power reactors.
21 MEMBER PETTI: But this decrease, this 22 drifting away from the correlation seems to be in that 23 data set too.
24 MR. RUDLAND: I'm not sure if it is or 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
52 not, to be honest. I don't know if one of my 1
colleagues knows the answer to that. Allen, do you 2
know the answer to that? I can't remember off the top 3
of my head.
4 MR. HISER: I'm sorry, what was the 5
question, Dave?
6 MR. RUDLAND: The question was whether or 7
not this behavior of underprediction shows up in the 8
test reactor data.
9 MR. HISER: I don't think that anybody has 10 really looked at the test reactor data with the number 11 of surveillance data-points from operating reactors.
12 We've pretty much stuck with that.
13 MR. RUDLAND: I know there's a bias that 14 occurs due to the flux but I don't recall there being 15 the same trends. But I'd have to go back and look.
16 CHAIR BALLINGER: We need to remember this 17 correlation is exactly that, it's not physics-based so 18 you're outside of the range of its applicability out 19 there.
20 MR. RUDLAND: That's absolutely right.
21 MR. HISER: For the record, this is Allen 22 Hiser, the NRC Staff.
23 MR. RUDLAND: There was some chemistry 24 correlation they used but Ron's right, it's not 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
53 mechanistic.
1 MEMBER HALNON: Just one more question.
2 I saw somewhere, and I'm not sure where, that the PWRs 3
will be effective until 2025 and it's just kind of a 4
continuation of an earlier question.
5 I assume that's using the highest amount 6
of uncertainty in margin and everything else that you 7
add in there that's the earliest that you would be 8
concerned, is that fair?
9 MR. RUDLAND: Yes, now where that number 10 comes from is that's when the first plants' ID of 11 their vessel reaches the 6e to the 19. That's where 12 that number comes from.
13 MEMBER HALNON: So, it's an approximate 14 number?
15 MR. RUDLAND: Yes, it's based on the 16 plant's fluence predictions at this point.
17 MEMBER HALNON: I understand the fluence 18 prediction. Does that also take into consideration 19 potential past trainings and excessive cool-down rates 20 or over-cooling transients that may have occurred?
21 Technically when that occurs an analysis 22 is done and you pay your respects and you move on.
23 Would you consider these operating events as well?
24 MR. RUDLAND: Those numbers come directly 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
54 from the plant's license renewal application so I'm 1
assuming they probably do.
2 MEMBER HALNON: I was just curious how 3
that was all factored in.
4 CHAIR BALLINGER: Is it my understanding 5
that prior to startup after a refueling outage a new 6
PT curve has to be established, right?
7 MR. RUDLAND: PT curves have a limited 8
life depending on what they did, it could be for 9
whatever EFP why they determined to calculate them 10 for.
11 So, they're good to that particular time 12 and then if they pull a surveillance capsule, they may 13 need to make a change to the PT curve depending on 14 what their data tells them.
15 CHAIR BALLINGER: But do they update it 16 for history between the last outages and things like 17 that? For example, if a cool-down occurs that's not 18 part of the original number that people have used, 19 they have to readjust?
20 MR. RUDLAND: I don't think they do 21 typically.
22 CHAIR BALLINGER: I thought prior to each 23 startup after a refueling outage, they update their PT 24 curve.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
55 MR. RUDLAND: I don't believe that's the 1
case.
2 MR. HISER: Do you want me to take a crack 3
at it?
4 MR. RUDLAND: If you know the answer, 5
sure.
6 MR. HISER: A couple things because there 7
was discussion of cool-down transients and things like 8
that. Those do not impact the PT limits, they are 9
evaluated against the PT limits to make sure the plant 10 didn't violate the PT limits.
11 But then also, normally the PT limits 12 themselves are calculated out ahead. Maybe if a plant 13 is operating about 20 years, they may have curves that 14 are good for 30 years so they don't have to 15 continually change their curves.
16 They have some stability in operations.
17 MR. RUDLAND: Again, remember, they use 18 the reg guide as a prediction of future embrittlement 19 so they can set their PT curves at a higher fluence 20 basically, at a higher EFPY.
21 And they will make a change, again, if 22 they pull a surveillance capsule or something that 23 tells them they need to change that particular 24 prediction.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
56 MR. HISER: Dave, if I can correct or 1
maybe clarify one other thing? Where we believe there 2
could be an issue is in about 10 years, in about 2032 3
as indicated on this slide.
4 The 2026 date on there is when the next 5
capsule will be tested by this plant so there's 6
nothing between now and between 2032 that we believe 7
is critical for this issue.
8 I saw 2025 a couple of times and just 9
wanted to clarify that part of things.
10 CHAIR BALLINGER: Other questions?
11 MR. RUDLAND: Somebody has their hand up.
12 CHAIR BALLINGER: I've got keep my eyes on 13 this. Steve Schultz?
14 MR. SCHULTZ: This is Steve Schultz.
15 The calculations that you're talking about 16 that the licensees do, Ron, you're probably referring 17 to a licensee's evaluation of their vessel fluence and 18 that's being tracked as a function of how the plant 19 has operated year by year over its lifetime.
20 And it's also corrected based upon new 21 information about technology associated with the 22 calculation of fluence, which has changed over the 23 years but the most up to date fluence calculation 24 methodologies are periodically used to adjust the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
57 plant's vessel fluence.
1 MR. SCHULTZ: I thought there was some 2
ongoing adjustments that could be made. Okay, thanks.
3 CHAIR BALLINGER: I don't see anymore 4
hands.
5 It's 3:10 p.m. and we're way ahead, we're 6
not scheduled for a break until 4:00 p.m. so I would 7
propose that if the EPRI folks are ready to go, we 8
just push ahead and get the EPRI presentation going.
9 I'm not sure, I think that's the way we 10 should probably go. So, let's see...
11 MR. BROWN: Elliot, are you okay with 12 sharing a screen?
13 MR. RUDLAND: I can do it if you'd like, 14 I just need to pull it up.
15 CHAIR BALLINGER: One thing, we have a 16 public comment that's been requested. Chris, us being 17 way ahead of schedule, does that impact this at all?
18 MR. BROWN: No, the person contacting me, 19 they're not going to make the comment after all.
20 CHAIR BALLINGER: Thanks a lot.
21 MR. LONG: Can everyone hear me clearly?
22 This is Elliot Long with the Electric Power Research 23 Institute. I am giving again a set of slides that I 24 gave to the NRC Staff at the October 18th public 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
58 meeting on RPD integrity.
1 And I will touch base on a few industry-2 wide actions that we have ongoing to generate 3
additional high fluence data. Next slide, please, 4
Dave.
5 There are two current ongoing industry 6
initiatives to generate high fluence capsule data 7
within the PWR fleet. The first is the Coronated 8
Reactor Vessel Surveillance Program, documented in MRP 9
326, now Revision 1.
10 The second is the PWR Supplemental 11 Surveillance Program, PSSP, and that's documented in 12 MRP 412. We'll note, the red asterisk there, this is 13 mainly PWR initiative. They have the higher fluence 14 of the plant designs in the U.S.
15 The PWRs have their own NRC-approved 16 integrated surveillance program through 60 years of 17 operation and that is documented in BWR VIP 86 18 Provision 1-A.
19 There is a plan for subsequent license 20 renewal, SLR, that has been accepted by the NRC and 21 this plan that's documented in BWR VIP 3.1A shows the 22 highest BWR unit will not exceed the threshold for 23 fluence of 6 times 10 to the 19th neutrons per 24 centimeter squared.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
59 I will elaborate on that at the end of my 1
presentation. Additionally, I want to reiterate and 2
discuss the prior EPRI conclusions from the November 3
2019 ACRS meeting. My colleague, Tim Harden, made a 4
presentation at that meeting and I wanted to revisit 5
those.
6 I will also touch base briefly on the 7
potential impact to PT limit curves, something that 8
Dave began his presentation with as well. So, next 9
slide. We'll first discuss the CRVSP.
10 So, the original intent of this program 11 was to optimize the existing and remaining U.S. PWR 12 surveillance capsule withdrawal schedules so that we 13 can increase the amount of high fluence data.
14 This high fluence data can then be used to 15 inform embrittlement trend curves for prediction of 16 RPD operation beyond 60 years and higher fluences.
17 This project originally completed in 2011, 18 where EPRI reviewed the reactor vessel surveillance 19 programs for the entire U.S. PWR fleet, made some 20 recommended changes to certain plants to delay 21 capsules such that they would be withdrawn with higher 22 fluence.
23 Our target date was by 2025 to ensure it 24 was slightly timely but also that we got the high 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
60 fluence data we sought. 10 years have since elapsed 1
since the Revision O version.
2 In 2021 we revised the report, basically 3
reviewing how we did, what's changed, what's left to 4
do, and what's going to happen over the next few 5
years. These updates included all the capsules that 6
have been withdrawn and evaluated since 2011.
7 Future pull schedules, updated capsule 8
fluence values, and then we did look at how any plants 9
that were closed or are going to be closed would 10 impact the overall goal of the CRVSP.
11 The next slide summarizes some of the data 12 and where we are. Next slide, please. Okay, so the 13 cover page of the report is on the right. Currently, 14 we have withdrawn and tested or plan to test 16 out of 15 the 30 capsules in the program.
16 Of the remaining 14, 7 will not be tested 17 or not plan to be tested, either due to a shutdown for 18 instance or they'll be delayed beyond 2025. In 19 summary, going back to Dave's plots, the red dots, 48 20 U.S. capsules have been tested at a fluence greater 21 than 3 times 10 to the 19.
22 And then 4 of those are greater 8 times 10 23 to the 19. By 2025 we'll have 7 more, the final 7 24 capsules will be tested at fluences greater than 3e19 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
61 and two more of these will be a fluence greater than 1
8.
2 This report also reviewed and put a 3
withdrawal schedule together for when the PSSP 4
capsules will be withdrawn. The first in Unit 1 will 5
be withdrawn in the spring of 2027 and then Sharon 6
Harris's PSSP capsule will be available for testing in 7
the fall of 2028.
8 Which leads me into the PSSP program 9
itself. I'll go ahead to the next slide, Dave.
10 So, the PSSP program was again designed to 11 generate high fluence surveillance data, much the same 12 purpose as the CRVSP, to inform developed of ETCs, of 13 embrittlement trend curve for higher and higher 14 fluence and longer-term operation.
15 This specific project aimed to fill in 16 gaps in the tested surveillance capsule database and 17 to utilize RPD material that was irradiated and 18 commercial reactors, not based on test reactor data.
19 The end game of this program was to 20 irradiate two supplemental surveillance capsules for 21 approximately ten years, which would have withdrawn, 22 tested, evaluated and then published.
23 Within these two capsules, there are 288 24 sharpie specimens from 27 unique plates, forgings, and 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
62 welds. I should stress here that all of these 1
materials in these two capsules were previously 2
irradiated specimens, so they were taken from capsules 3
that were already withdrawn, already tested for the 4
plants in the program.
5 And then the materials were reinserted 6
into these capsules for additional irradiation. Once 7
everything has been withdrawn and tested, we will have 8
24 new transition temperature shift results and then 9
3 additional upper shelf energy-only results.
10 The fluence ranges for these specimens 11 will be from about 4019 up to 1.2e20 or 12e19 for the 12 various materials in these two capsule. Go to the 13 next slide, please?
14 So, the history of when we manufactured 15 and designed two capsules containing previously 16 irradiated and reconstituted PWR materials. The MRP 17 sponsored this fabrication and two U.S. utilities 18 agreed to host the materials.
19 The first one was inserted in the Fall 20 Unit 1 in October of 2016 and the second one was 21 inserted into Sharon Harris in April of 2018. The 22 fabrication report was then published in 2016, MRP 23 412.
24 Next slide, please, Dave. As I noted from 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
63 the updated CVRSP report, Fall Unit 1 will be 1
available to be withdrawn and tested in the spring of 2
2027, Sharon Harris in the fall of 2028.
3 Shown to the figure on the right, you can 4
see how these specimens were made. The top image is 5
that of a previously broken and irradiated sharpie 6
specimen so that specimen was tested.
7 One half of the specimen was machine, to 8
the second row the middle piece specimen insert is the 9
top piece that was machined into the middle. The end 10 tabs were then welded on to make a new sharpie 11 specimen.
12 The middle image was then machine cleaned 13 and cut the size and then that was inserted into the 14 capsule, the bottom image there. So, that's how we 15 refabricated and remade sharpie specimens out of 16 previously broken, previously irradiated material.
17 After the withdrawal and testing of these 18 capsules, they'll be evaluated in 2028 through 2030.
19 The capsule reports will be delivered to the NRC Staff 20 for review approximately 18 months after each 21 capsule's withdrawal dates.
22 There will be two separate reports, one 23 for each capsule, and thereafter we will analyze the 24 data and its impact on future ETCs. That bottom right 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
64 image is just a picture of a PSSV capsule itself in a 1
specimen holder in one of the two plants.
2 Before I continue forward, are there any 3
questions in regard to either of these industry 4
programs, again designed to generate high fluence 5
capsule data for the U.S. fleet?
6 MEMBER HALNON: This is Greg. The welding 7
and this machining and whatnot, what's the delta 8
between that and a virgin, I guess, material?
9 MR. LONG: Great care was taken and a 10 qualified process was developed to ensure that the 11 welding process was done at a temperature such that 12 the irradiation temperature was not exceeded on the 13 specimens.
14 So, they were designed and qualified and 15 then there was a benchmarking assessment in MRP412 16 that showed the sharpie curves between original 17 material and material made in the fashion of the top 18 right image were essentially the same.
19 MEMBER HALNON:
So, you're pretty 20 confident that any differences are not going to show 21 up in the data so it's going to be just like an 22 original specimen?
23 MR. LONG: That's correct, we are 24 confident that it will be within the typical scatter 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
65 of a sharpie specimen curve. They look fairly 1
identical.
2 MEMBER HALNON: Good, thanks.
3 MEMBER BROWN: This is Charlie Brown, I've 4
got a question on this, I'm not sure it's an 5
intelligible question.
6 It's a vague memory that sharpie venotch 7
specimen sizes -- I'm springboarding out of Greg's 8
comment -- there's some minimum sizes.
9 And once you chop one in half like this 10 and try to reconstitute it, how do you know you 11 haven't compromised it with results from the welding 12 and the other testing that the other ends that have 13 been involved in?
14 MR. LONG: There are standards within ASTM 15 for reconstitution of sharpie specimens and those 16 standards were adhered to in the development of these 17 reconstituted specimens.
18 So, the sizing limits of the ASTM 19 standards were achieved for the specimens included in 20 the program.
21 MEMBER BROWN: So, there's no effect of 22 the welding then? When you put them back together 23 that's far enough away from the well zone?
24 MR. LONG: That's correct, they're far 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
66 enough away from the material itself and we again used 1
a very specific welding process that did not exceed 2
the temperature in which the original specimens were 3
irradiated.
4 And then this was again benchmarked and 5
qualified by the fabrication vendor and developers of 6
the castles themselves.
7 MEMBER BROWN: Thank you very much.
8 MR. LONG: Dave, go onto the next slide, 9
we'll change gears here a little bit and review some 10 of the conclusions from the prior ACRS meeting on this 11 topic. Go onto the next slide.
12 Back in the November 2019 timeframe, my 13 former colleague and predecessor Tim Harden, who 14 retired earlier this year, made a presentation to the 15 ACRS in regard to potential changes to Reg Guide 199 16 Revision 2.
17 The conclusions from that meeting on the 18 right are shown here and they have not changed. If in 19 the future a revision to the reg guide is implemented, 20 the ASTM model, E900-15 remains the preferred model.
21 It is understood and Dave spent a lot of 22 time discussing those performance values. Below 6019 23 the reg guide at this point remains adequate for 24 predicting RPV and embrittlement.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
67 And noting that PT limit codes from 10 CFR 1
50 Appendix G utilized the quarter T fluence. I took 2
some time to put together when certain plant designs 3
will see a fluence of 6019 at the quarter T location.
4 So, go onto the next slide. So, this 5
chart at the right, we're focusing on what surface 6
fluence is needed to generate a quarter T and a three-7 quarter T fluence of 6 times 10 to the 19.
8 So, the top right chart was developed 9
using the current fluence attenuation formula from the 10 Reg Guide 199 Revision 2, and as you can see for the 11 various plant designs, Westinghouse 2-loop, 3-loop, 12 BMW, various 4 loops with different RPD manufacturers, 13 some of the CE designs.
14 Everyone has a slightly different vessel 15 thickness and the fluence needed at the surface to 16 reach a quarter T value at 6019 is shown at the right.
17 So, for example, a Washington 2 Loop plan 18 would need a surface fluence of 8.86 times 10 to the 19 19 to have a quarter T fluence of 6.
20 And similarly for the 2 loop plans, at 21 three-quarter T you would need a surface fluence of 22 19.3 times 10 to the 19 to have a three-quarter T 23 fluence of 6019.
24 So, that's obviously very high for the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
68 three-quarter T location and as the vessels get 1
thicker from the various designs you have a higher and 2
higher surface fluence needed to generate a quarter T 3
fluence of 6019.
4 Generally
- speaking, the 3-Loop 5
Westinghouse design have the highest surface fluence 6
at end of life. I then went in and summarized the 7
various plants that have thus far applied for a 8
subsequent license renewal, and that is the bottom 9
right chart.
10 The bottom right chart shows the surface 11 fluence at SLR for the six PWR plants that have thus 12 far applied for a second license extension or SLR.
13 Only 1 of the 6 will exceed the surface fluence 14 necessary to have a quarter T fluence of 6019 before 15 end of life.
16 That's Plant A and this is predicted to 17 occur at about 65 BFPY for Plant A. The remainder of 18 the plants have a surface fluence value that will 19 never then subsequently lead to a quarter T fluence of 20 6 times 10 to the 19.
21 And then finally, I wanted to reiterate 22 the BWR plans are not shown anywhere on this chart as 23 they will never reach these fluence levels in any 24 reasonable operational timeframe. 1 BWR SOR plant has 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
69 a surface fluence value of 5 times 10 to the 18.
1 So, very much orders of magnitude below 2
the 6019 where the reg guide predictions begin to 3
break down.
4 CHAIR BALLINGER: This is Ron again, 5
probably a question for Dave regarding 61A analysis.
6 Did that analysis consider the fluence gradient 7
through the vessel each time?
8 MR. RUDLAND: No, because 61A only looks 9
at the surface fluence.
10 CHAIR BALLINGER: This is a big factor 11 because now one of the compensating things, I suppose, 12 for a lot of the uncertainty is the fact that any 13 crack would be growing into a tougher material by 14 quite a bit.
15 MR. RUDLAND: That's correct. In the PFM 16 stuff that we did, of course we took the attenuation 17 into account.
18 CHAIR BALLINGER: You did? Okay.
19 MR. RUDLAND: For the PT curve stuff we 20 did but for the PTS you only have to look at the 21 surface because the screening criteria is based only 22 on the surface fluence.
23 CHAIR BALLINGER: I got it, thanks.
24 MR. LONG: Yes, this chart was just 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
70 developed, these calculations here, just to show where 1
the 6019 would be at the quarter T and is focused on 2
pressure temperature limits in this case.
3 CHAIR BALLINGER: That's why in the 4
presentation I said the staff would eventually be 5
concerned.
6 I had different time periods between the 7
PTS and the PT limits because of the difference 8
between the fluence levels at the ID versus the 9
quarter T.
10 MR. LONG: Right, and this is my last 11 slide so if there's any other questions, please let me 12 know.
13 Hearing none, I will turn this back over 14 to you all. Thank you for the opportunity to present 15 today.
16 CHAIR BALLINGER: Thank you for your 17 presentation. Questions from the Members and 18 consultant?
19 Hearing none, I think now the public line 20 is available I suppose, are there any public -- oh, 21 there's a hand raised. Where's the hand?
22 MR. DENIS: This is Den Denis from 23 Structural Integrity Associates. I wanted to comment 24 on previously if there are any questions regarding PT 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
71 limits curve generations, I noticed also Westinghouse 1
is on the line.
2 But I
know Westinghouse, GE, and 3
Structural Integrity Associates, we provide the plan-4 specific PP limits curves as a vendor. If there are 5
any questions on the generation or the processing of 6
those analyses, we can take a conversation offline if 7
desired.
8 I just wanted to offer that out.
9 CHAIR BALLINGER: Thank you.
10 MEMBER BROWN: I wasn't able to formulate 11 my question, I'm trying to summarize it. After 12 listening to all this, I'm trying to condense it down.
13 What I heard, and if I'm wrong please tell 14 me, is that based on all the development data in the 15 testing they've got coming, getting it for 2WRs, it 16 looks like embrittlement is not going to be a limiting 17 factor for 80-year SLRs?
18 Is that right or am I totally off base?
19 CHAIR BALLINGER: I think that's more or 20 less correct. Dave can correct me. I think the issue 21 is the uncertainty.
22 MEMBER BROWN: I understand that.
23 MR. RUDLAND: I think there's a couple of 24 issues. I think one is the uncertainty, I think Allen 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
72 made the point that there isn't many plants that are 1
going to reach the quarter T fluence of interest by 80 2
years.
3 But we still have to understand the 4
behavior in PTS space to make sure we're bounding and 5
what we've done to make sure there's not a concern 6
there.
7 Plus, as Ron pointed out, there are a lot 8
of uncertainties in the analyses at least that we did 9
in terms of whether or not we've covered all the 10 plant-specific issues.
11 I think in general what you said is not 12 incorrect.
13 MEMBER BROWN: Are there plans to even be 14 able to eventually agree, or you all agree, not we, 15 propose going to an 80-year life that there are ways 16 to at least withdraw a capsule to see if they have 17 exceeded something during that last 20 years?
18 Is there some process in place for that?
19 MR. RUDLAND: Plants have capsules in 20 there to be able to do it, they just need to pull them 21 to test them.
22 MEMBER BROWN: But you could do that to 23 make sure your uncertainties didn't kill you? That's 24 what I'm trying to say.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
73 MR. RUDLAND: That's correct. They've got 1
that data and this data that Elliot talked about 2
that's coming, which is separate from the actual 3
plant-specific data.
4 And then again, between that and if we are 5
able to fix the embrittlement trend curves, I think 6
that would go a long way to really reducing the 7
uncertainty.
8 MEMBER BROWN: Understood, thank you.
9 MR. HISER: Just to add a little bit, each 10 of the PWRs that has an SLR at this point has good 11 means to test a capsule to get a higher fluence. So, 12 there are commitments to do the testing, as Dave said.
13 If they follow through on those 14 commitments, we'll have the data, if they continue to 15 delay the withdraw on testing, then we won't have the 16 data.
17 MR. RUDLAND: Just to be clear on that 18 point, Allen, the regulations allow them to delay or 19 not test that capsule.
20 So, even though they said they'll do it as 21 part of their renewed license, there's no commitment 22 from them to do that and they can take the path that 23 others have to delay or not test the last capsule.
24 MEMBER BROWN: How does this then allow 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
74 you as the NRC to say everything is okay if they 1
choose to delay and not take that capsule out once 2
they're into that SLR period?
3 MR. RUDLAND: I think that's why we 4
believe a change is needed, to guarantee that they 5
are.
6 MEMBER BROWN: That they will take them 7
out?
8 MR. RUDLAND: Yes, that's something we're 9
trying to determine the best path forward.
10 MEMBER BROWN: Does that require something 11 from the Commission to go tell them that?
12 MR. RUDLAND: If they were to change the 13 word, yes.
14 MEMBER BROWN: I'm just trying to 15 understand that, I don't know as much about it.
16 (Simultaneous Speaking.)
17 MR. RUDLAND: That's why we wanted to 18 specifically change Appendix H to say you need to test 19 the last capsule. It was required in the rulemaking, 20 we have to go to the Commission.
21 First, we have to go to the Commission to 22 ask permission to change the rule and then we'd have 23 to go through the rulemaking process.
24 CHAIR BALLINGER: That's a good segue, by 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
75 the way. Slide 30, options being considered, can you 1
give us some kind of feeling for the timeline of when 2
these decisions will be made, and then path forward?
3 Because there was a member and that's my 4
question as well, sooner or later, we would probably 5
get to a point where we would see something where we 6
could get a letter.
7 This has been going on for quite a while.
8 MR. RUDLAND: Yes, it has and the Staff is 9
actively trying to move quickly to a decision on what 10 to do, I can say that.
11 We want to make sure that we do it but 12 again, the issue is twofold, we want to make sure that 13 we don't do something that is burdensome to those that 14 won't have the issue for a very long time.
15 We want to make sure that the change is 16 focused and if we do we need to talk to the Commission 17 about it before and they have to decide whether or not 18 it's appropriate to change the regulations or not.
19 So, that helps us take a little bit of 20 time and it's difficult for me to give you a schedule 21 on that right now.
22 CHAIR BALLINGER: So, the NuScale people 23 will have to apply for an exemption?
24 MR. RUDLAND: I don't know, I can't 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
76 comment on the PRM right now. I don't know if 1
somebody else can? The NuScale PRM is still under 2
consideration, I do know that and so the decisions on 3
that have not been finalized.
4 CHAIR BALLINGER: Because they want to 5
directly use E900, right?
6 MR. RUDLAND: I believe that was what the 7
request was. Bob Caldwell may have a comment.
8 MR. CALDWELL: This is Bob Caldwell, I'm 9
the Director of the Division of Renewed Licenses. I 10 just wanted to go back to a little bit about making 11 the industry do stuff.
12 Right now they don't have to and to make 13 them do stuff we have to go through and do the backfit 14 analysis. That's not trivial so it's hard for us to 15 promise one thing or another at the moment with 16 regards to what our actions forward are.
17 The effort that Dave and the team went 18 through was to help establish what the safety basis 19 is, the safety case for what we want to do and go 20 forward. And that will be a primary consideration as 21 we move forward and make our decision on what needs to 22 be done.
23 CHAIR BALLINGER: Thank you.
24 MEMBER HALNON: This is Greg, I've just 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
77 got one more question and then I'm not even sure if I 1
know how I'm asking it makes any sense.
2 But getting a subsequent license renewal 3
to 80 years, I see there's some prediction on the 4
amount of fluence that will occur during that 20-year 5
period such that you can assure there's a safety 6
margin in the PT curves.
7 The first 20 years of operating these 8
plants, the capacity factors are in the 60 to 75 9
percent. Now we're running these plants up to 95-plus 10 percent capacity factor.
11 How does that factor in? Earlier, the 12 first 20-year fluence is not going to equal the last 13 20-year fluence while operating it. But I guess also 14 it would have low leakage cores and other things now 15 too.
16 How does all that factor into the 17 analysis? Are those all just inputs to it?
18 MR. POEHLER: This is Jeff Poehler from 19 the Office of Research, I can maybe comment on that.
20 Plants specifically when they're doing the 21 60 to 80-year estimate will use a conservative 22 estimate for capacity factor for fluence going 23 forward.
24 Typically, 90 or 95 percent.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
78 MEMBER HALNON: So, they take actuals and 1
then they predict out at a very high capacity factor?
2 MR. POEHLER: Generally speaking, yes.
3 MEMBER HALNON: 40 years of fluence is not 4
going to be equal to the second 40 years, it's going 5
to be a lot less.
6 MR. HISER: Really, what they want to do 7
is to make sure they're not non-conservative.
8 There are time-limited aging analyses on 9
reactor vessel embrittlement so plants project those 10 forward and I think in at least one case they were 11 actually using the capacity factor that exceeded 100 12 percent just to make sure they would have a 13 conservative prediction at 80 years.
14 MEMBER HALNON: It's front and center in 15 all these analyses then.
16 MR. HISER: Absolutely, and what they do 17 is they calculate cycle by cycle the accumulated 18 fluence and then up to maybe the last cycle before 19 their application submitted, and then they project 20 forward from there.
21 MEMBER HALNON: Good, thanks.
22 CHAIR BALLINGER: Okay, I think I need to 23 apologize. I think I've confused public comments from 24 general questions. Are there any members of the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
79 public that would now like to make a comment?
1 Okay, any last final comments from 2
Members? Well, in that case I would like to thank you 3
and I'm sure the rest of the Committee would thank you 4
for doing this.
5 It's an important, at least in my mind, a 6
very important topic and it behooves us to keep track 7
of this. We appreciate the extensive effort, 8
actually, that's been made by the Staff since the last 9
time we wrote a letter.
10 So, that was another reason for asking for 11 this update. So, if there are no other comments from 12 the members then I think this meeting is adjourned.
13 (Whereupon, the above-entitled matter 14 went off the record at 3:45 p.m.)
15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1716 14th STREET, N.W., SUITE 200 (202) 234-4433 WASHINGTON, D.C. 20009-4309 www.nealrgross.com
Reactor Pressure Vessel Embrittlement Monitoring and Prediction in Long-Term Operation ACRS Metallurgy and Reactor Fuels Subcommittee November 15, 2021
Meeting Purpose
- Discussion of Issues
- Regulatory Guide 1.99 Rev 2 (RG 1.99) and 10 CFR 50.61 embrittlement trend curve
- Appendix H surveillance testing
- Discussion of holistic risk-informed analysis of these issues and its potential impact on reactor pressure vessel (RPV) integrity
- Path forward options 2
=
Background===
Monitoring and Prediction of Embrittlement Embrittlement Trend Curve (ETC) provides estimates of change in fracture toughness (T or RTNDT) as a function of fluence Surveillance capsule testing provides monitoring to ensure ETC predicts plant specific behavior properly Together they are used to determine pressure-temperature (PT) limits for normal operation 3
RTNDT(u)
Embrittlement (T)
Operating Time (years)/Fluence ART ETC Max Pressure per 10 CFR 50 App. G Coolant Temperature Heatup & Cooldown Data 40yr 60yr 80yr 40yr 60yr 80yr ART = Adjusted Reference Temperature
Ideal Scenario
- ETC provides conservative predictions of embrittlement
- Surveillance data covers all operating periods 4
RTNDT(u)
Embrittlement (T)
Operating Time (years)/Fluence ART ETC Future Data?
40yr 60yr 80yr RTNDT(u)
Embrittlement (T)
Operating Time (years)/Fluence ART ETC
?
40yr 60yr 80yr IF ETC under-predicts measurements IF Limited Surveillance Data is Available Current Data
?
?
?
Potential Uncertainty Sources
Embrittlement Uncertainty 5
Embrittlement Operating Time (years)/Fluence Embrittlement Ideal withdrawal schedule &
accurate ETC Delayed withdrawal schedule &
inaccurate ETC Expected uncertainty Large uncertainty Operating Time (years)/Fluence Fluence Capsule test date Future data Embrittlement (T)
Operating Time (years)/Fluence Embrittlement (T)
Operating Time (years)/Fluence Fluence Capsule test date Holistic study needed to understand impact of uncertainty Surveillance data ETC Large Uncertainty
?
Current Perspective of Potential Issue
- High confidence that currently operating plants remain safe
- Recent licensing actions remain valid
- Insufficient embrittlement monitoring and under predictions of reactor vessel embrittlement will eventually (after about 10 years for PTS; after about 23 years for P-T limits and upper shelf energy) impact the staffs confidence in the integrity of the reactor pressure vessel in long-term operation, i.e.,
both safety margins and performance monitoring may be impacted
- Further work is needed to determine which plants are impacted by this potential issue 6
Embrittlement Trend Curve
- May 1988, NRC published RG 1.99, which contained an improved embrittlement trend curve (ETC)
- Fit based on 177 datapoints
- June 1991, NRC updated 10 CFR 50.61 to include the ETC from RG 1.99
- Addressed lower than measured predictions (up to 60°F) of embrittlement in some vessels
- This ETC was re-evaluated for continued adequacy in 2014 (ML13346A003) and in more detail in 2019 (ML19203A089) 7
Issue - ETC 8
-180°F Deviates from mean Statistically significant
+180°F DT41J = T41J is a measurement of embrittlement representing the shift in transition temperature from brittle to ductile fracture at an impact toughness of 41J
Issue - ETC 9
Scatter greater than RG 1.99 standard deviation Limited data at high fluence
Issue - ETC Fluence Function 10 Fluence function begins to flatten at the same fluence level underprediction occurs in Slide 8
Surveillance Capsule Delays Appendix H to 10 CFR Part 50 requires periodic monitoring of changes in fracture toughness caused by neutron embrittlement
- ASTM standard (E185-82) allows final capsule fluence to be 2X RPV design fluence - plants change (intended 40-year) design fluence to current license length (e.g., 60 or 80 years)
- ASTM standard (for 40 years) permits holding last capsule without testing Commission finding (Perry decision NRC Administrative Letter 97-04) that staff review of requests to change capsule withdrawal schedules is limited to verification of conformance with the ASTM standard (i.e., not based on technical or safety considerations)
- Capsule withdraw and testing repeatedly delayed in some cases to achieve higher fluence 11
- Regulations are unchanged; surveillance program addressed in guidance
- Guidance provides flexibility for licensees to demonstrate adequate management of RPV embrittlement due to varying plant-specific circumstances
- Aging Management Program XI.M31, Reactor Vessel Material Surveillance
- Continues reliance on Appendix H program using ASTM E185-82
- GALL Report (NUREG-1801, Rev. 1) for license renewal (40 to 60 years)
- shall have at least one capsule with a projected neutron fluence equal to or exceeding the 60-year peak reactor vessel wall neutron fluence prior to the end of the period of extended operation
- Describes use of reconstituted specimens and use of operating restrictions (neutron flux, spectrum, irradiation temperature, etc.)
- GALL-SLR Report (NUREG-2191) for subsequent license renewal (60 to 80 years)
- withdrawal and testing of at least one capsule... with a neutron fluence of the capsule between one and two times the peak neutron fluence of interest at the end of the subsequent period of extended operation - or data from a prior tested capsule
- Specifies - it is not acceptable to redirect or postpone the withdrawal and testing of that capsule to achieve a higher neutron fluence that meets the neutron fluence criterion for the subsequent period of extended operation 12
License Renewal in Practice
- Licensees have changed capsule withdrawal schedules prior to application for license renewal or subsequent license renewal
- Change is evaluated under current approach of conformance verification
- (Updated) current licensing basis surveillance program for license renewal/subsequent license renewal is then consistent with the program in GALL/GALL-SLR 13
Issue - Appendix H Performance Monitoring 14 Plant Capsule
- of times delayed Turkey Point 5
4 Robinson 5
2 Surry U1 5
2 Surry U2 5
2 North Anna U1 4
2 North Anna U2 4
2 St. Lucie U2 4
1 Point Beach 5
1 Many licensees have delayed capsules (time and/or fluence),
some recent examples:
Not all plants have delayed withdrawal of capsules Capsule withdrawal schedule changes include delays in both time and/or fluence
Potential Impact of Issue 15
-180°F 1
(1974) 3 (1985) 4 (2001) 60 Years (2032) 80 Years (2052) 5 (2026) 2 (1978)
+180°F
0 50 100 150 200 250 300 350 400 1.0E+17 2.0E+19 4.0E+19 6.0E+19 8.0E+19 1.0E+20 1.2E+20 1.4E+20 DRTNDT,F Fluence, n/cm^2 Data RG1.99 Fit through Plant data Updated fit with additional data Potential additional data Potential Impact of Issue 16 150 F No correction Fit to original data Embrittlement Fit with potential additional data 75 F
Risk-informed Analysis Integrated Decision Making Defense in depth Increase in risk is small Performance Monitoring Change meets current regulations Safety Margins
- Considered combined effects of surveillance and embrittlement predictions
- Leveraged 5 principles of risk-informed decision making
- Targeted sample of plant data used, but much plant specific information not available 17 17
Analysis Assumptions
- Comparisons based on ASTM E900-15 ETC
- The NRC staff found that the ASTM E900-15 ETC provided the most accurate characterization of this database*
18 Plates and Forgings Welds
- Basis for a Potential Alternative to Revision 2 of Regulatory Guide 1.99, TLR-RES/DE/CIB-2020-11, ML20345A003
Analysis Assumptions - Fleet Impact Study
- A targeted sample of 21 plants
- Emphasis on high fluence plants, with a few low Cu plants and BWRs to round out
- Determined changes in adjusted reference temperature resulting from switching ETCs -
embrittlement shift delta (ESD)
- Results used to benchmark ESD range of risk analysis 19
Results - Fleet Impact Study
- There is a tendency for material reference temperatures to increase when switching from RG 1.99 to ASTM E900-15.
- Base materials are more likely to see increases in reference temperatures than weld materials.
- Only a handful of plant limiting materials will have ESDs > 50 °F, and these tend to be at fluences
~6x1019 n/cm2.
- Range of ESDs assumed in risk study bounds fleet impact findings.
20
Risk of Failure Large Uncertainties:
Unknown frequency of transient Actual plant fluence variations Are these analyses bounding?
Unknown plant-specific considerations How much protection do administrative and other operational limits provide against violating the PT limit?
21 ESD represents the underprediction of RTNDT RG 1.99 Revision 2 Update FAVOR Scoping Study, May 6, 2021, TLR RES/DE/CIB-2020-09, Rev. 1, ML21126A326
Through-Wall Crack Frequency Results 22 Transient Type Shallow Flaw 1/4T Flaw Comment BWR P-T Limit Cooldowns CPF 1x10-6 for all ESDs CPF 1x10-6 for ESD > 40 °F BWRs must cooldown on saturation curve, so cooldown on licensed limits not plausible.
BWR Saturation Cooldown CPF 1x10-6 for all ESDs CPF 1x10-6 for all ESDs BWR Leak Test, Cooldown rate 50 °F/hour CPF 1x10-6 for all ESDs CPF 1x10-6 for ESD > 100 °F Additional information is desired to determine if high cooldown rates are possible, or ASME Code action will be pursued to prohibit.
BWR Leak Test, Cooldown rate
> 50 °F/hour CPF 1x10-6 for all ESDs CPF 1x10-6 for ESD > 100 °F PWR P-T Limit Cooldowns CPF >1x10-6 for ESDs 50 °F CPF > 1x10-6 for ESD 20 °F Additional information on event frequencies is desired to confirm TWCF< 1x10-6
/year.
PWR Cooldown, Actual Transients CPF < 1x10-6 for most transients n/a
Pressurized Thermal Shock Considerations
- 10 CFR 50.61 uses ETC from RG 1.99
- RTPTS from 10 CFR 50.61 might be impacted
- Limits of 270 °F for plates, forgings, and axial weld materials, and 300 °F for circumferential weld materials
- However, through-wall crack frequency calculated with corrected embrittlement less than 1x10-6 for all cases investigated 23
Safety Margins
- Uncertainties in risk calculations are high and increasing with time
- Even though the risk appears low, resolving these issues will help maintain the fundamental safety principles that are the basis of plant design and operation
- Safety margins, as provided by regulations and current license bases, provide reasonable assurance against brittle fracture 24
Pressure, psi Temp, F Safety Margins Illustration Uncertainties increasing due to lack of surveillance, but margin is less due to embrittlement underprediction 25 Margin Uncertainty PT-curve using RG 1.99 Reduced Margin Operating Margin Structural limit Accurate PT-curve Adequate Margin
Performance Monitoring
- Performance monitoring ensures
- Analysis results remain valid with time
- No unexpected (or unmodelled) adverse safety issue occurs
- Delaying capsule withdrawal for an extended period with the possibility of no future data represents a lack of performance monitoring 26
Analysis Summary
- With the current state of knowledge, a generalized analysis suggests the overall risk of brittle fracture is low
- The uncertainty in these results is high and increases with time
- Plant specific details not considered
- Under certain conditions, safety margins are impacted and are decreasing as uncertainty increases
- Delaying capsules at high fluence represents a lack of sufficient performance monitoring
- Issues are plants with fluences > 6x1019 n/cm2 27
Who is Impacted?
- Embrittlement Underprediction
- Plant specific details (e.g., limiting material, etc.) may contribute to which plants are impacted
- More work is needed to determine which plants are impacted
- Lack of Surveillance Data
- Any plant renewing license that chooses to delay last capsule 28 Percentage of Fleet Surpassing Fluence Levels Percentage of PWRs Surpassing Fluence Levels Year\\Fluence 6 x 1019 n/cm2 8 x 1019 n/cm2 6 x 1019 n/cm2 8 x 1019 n/cm2 60 years 6%
0%
9%
0%
80 years 22%
10%
34%
15%
Staff Goals
- Currently, regulations are sufficient for reasonable assurance of adequate protection against brittle fracture of vessel
- Staff wants to ensure continued reasonable assurance in long-term operation
- Provide remedies for the identified issues with RPV surveillance requirements and embrittlement predictions, on a risk-informed, performance basis
- Do not impact those plants that are not adversely affected by the issues
- Plant-specific surveillance data that covers end of license fluence level
- Projected fluence at end of license < ~3 x 1019 n/cm2 29
Options Being Considered to Meet Goal
- Plant-specific action
- Focused regulatory action
- Generic communication
- No action 30
Summary
- High confidence that currently operating plants remain safe, and recent licensing actions remain valid
- Issue will eventually (after about 10 years for PTS and 23 years for P-T limits) impact the staff confidence in the integrity of the reactor pressure vessel in long-term operation, i.e., both safety margins and performance monitoring may be impacted
- Further work is needed to determine which plants are impacted by this issue
- Proactively ensure continued reasonable assurance though a risk-informed, performance-based solution
- Staff is considering options - desires focused solution to only those conditions adversely impacted by this issue 31
Thank You 32
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m Elliot J. Long Principal Technical Leader NRC Public meeting on RPV Integrity October 18, 2021 RPV Embrittlement Monitoring and Prediction in Long-Term Operation October 2021 EPRI MRP Technical Brief
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 2
Presentation Outline*
Future Sources of High Fluence Capsule Data
- PWR Coordinated Reactor Vessel Surveillance Program (CRVSP)
MRP-326, Revision 1
- PWR Supplemental Surveillance Program (PSSP)
MRP-412 Review of Prior EPRI MRP Conclusions from the November 2019 ACRS Meeting
- Potential to impact plant Pressure-Temperature (P-T) limit curves
- Note that BWR Units have an NRC approved Integrated Surveillance Program (ISP) for up to 60 years for the U.S. BWR Fleet (BWRVIP-86, Rev. 1-A). The implementation plan for Subsequent License Renewal (SLR) has also been accepted by NRC, with the highest BWR Unit projected not to exceed the threshold for fluence of 6 x 1019 n/cm2 (E>1.0 MeV) as described in this report (BWRVIP-321-A). See Slide 12 for more details.
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 3
Update to the CRVSP, MRP-326, Revision 1
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 4
Coordinated PWR Reactor Vessel Surveillance Program (CRVSP)
Materials Issue Being Addressed:
- Optimize the U.S. PWR surveillance capsule withdrawal schedules to increase the amount of high-fluence (f > 3.0 x 1019 n/cm2) surveillance data which can be used to inform development of embrittlement trend correlations (ETCs) applicable for RPV operation to high fluence (60+
years).
Objectives of the Project
- Revision 0 (2011): Review the reactor vessel surveillance programs (RVSPs) of the operating U.S. PWR fleet and recommend changes to selected RVSP withdrawal schedules in order to increase the amount of high fluence surveillance data by 2025.
- Revision 1 (2021): Review of how we did, what has occurred, whats left to do, and when it is most likely to happen across the US fleet Updates to the evaluation include
- Evaluated capsules withdrawn since 2011
- Future capsule pull schedules
- Capsule fluence values
- Analysis of closed (or to be closed) plants
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 5
Update to the CRVSP, MRP-326, Revision 1 Current high fluence capsule withdrawal results
- 16 out of 30 CRVSP Capsules are tested or planned to be tested
- There are 14 remaining CRVSP Capsules Half of these are not planned to be tested (i.e., due to plant shutdown) or will be delayed beyond 2025 Summary of available high fluence data
- 48 U.S. capsules have been tested at f > 3.0 x 1019 n/cm2 4 of these are f > 8.0 x 1019 n/cm2
- By 2025, the remaining 7 planned CRVSP capsules will be tested at f > 3.0 x 1019 n/cm2 2 of these are predicted to be f > 8.0 x 1019 n/cm2 PSSP withdrawal schedule update
- Farley 1 Capsule P to be withdrawn in Spring 2027
- Shearon Harris Capsule P in Fall 2028 There are no NEI 03-08 requirements of any kind associated with this report revision
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 6
PWR Supplemental Surveillance Program (PSSP)
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 7
PWR Supplemental Surveillance Program (PSSP)
Materials Issue Being Addressed:
- Additional high-fluence (f > 5.0 x 1019 n/cm2) surveillance data is needed to inform development of embrittlement trend correlations (ETCs) applicable for RPV operation to high fluence (60+ years).
Objectives of the Project:
- Fill projected gaps in the tested surveillance capsule database
- Inform future ETCs using actual RPV surveillance materials from commercial PWRs (not test reactor data)
End game: Irradiate two supplemental surveillance capsules for ~10 total years before withdrawal, testing, evaluation and publication of capsule test reports
- These two surveillance capsules have 288 Charpy Specimens from 27 unique plates, forgings and welds
- The data generated from these capsules will ultimately yield 24 new transition temperature shift results and 3 additional upper shelf energy results Fluences levels of the to be evaluated specimens will range from ~4.5E+19 up to ~1.2E+20 (n/cm2)
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 8
PWR Supplemental Surveillance Program Project History
- Program designed and fabricated 2 supplemental surveillance capsules containing previously-irradiated, reconstituted PWR materials
- EPRI MRP sponsored the fabrication of these 2 surveillance capsules:
ALA-P; 14 materials (Host: Farley 1), inserted October 2016 CQL-P; 13 materials (Host: Shearon Harris),
inserted April 2018
- MRP-412 (PSSP Capsule Fabrication report) was published in 2016
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 9
PWR Supplemental Surveillance Program Current Project Status and Timeline
- Farley 1 Capsule P to be withdrawn in Spring 2027; Shearon Harris in Fall 2028 per MRP-326,R1
- Testing of surveillance capsules and data evaluation in 2028-2030
- Anticipated Project Deliverable Date:
Capsule report within ~18 months of each capsules withdrawal date (2 reports total)
- Data evaluation and impact on future ETCs in 2030-2032 PSSP Capsule seated in its holder
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 10 Prior EPRI Conclusions on the Potential Revision of RG1.99R2
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 11 Conclusions from the EPRI Presentation to the ACRS in November 2019 EPRI MRP previously presented on the potential revision of RG1.99R2 to the ACRS in Nov. 2019 The conclusions from that meeting (shown at right) have not changed If a future revision to RG.199R2 is implemented, ASTM E900-15 remains the preferred ETC model (today)
It is understood that for fluence values below 6E+19 n/cm2, RG1.99R2 remains adequate for predicting RPV embrittlement The next slide details when certain plant designs will see that fluence level at 1/4T
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 12 What surface fluence = 6E+19 @ 1/4T and 3/4T?
The chart at right was developed using the current fluence attenuation formula from RG1.99R2*
It is understood that the 3-Loop Westinghouse design has the highest surface fluence at end-of-life SLR plants most limiting surface fluence values are also summarized herein Only Plant A will hit a 1/4T fluence of 6E+19 n/cm2 for 80-years
- This is currently predicted to occur well into the SLR operating period
- BWR plants are not shown as they will never reach these fluence levels in any reasonable operational time-frame [BWR SLR Plant 70 EFPY surface fluence < 5E+18 n/cm2]
1/4T (E+19) 3/4T (E+19)
WEC 2-Loop B&W/CE 6.5 8.86 19.3 WEC 3-Loop/Smaller CE CE 7.875 9.62 24.8 B&W NSSS B&W 8.44 9.96 27.4 WEC 4-Loop RDM 8.45 9.96 27.5 WEC 4-Loop B&W 8.5 9.99 27.7 WEC 4-Loop/Larger CE CE 8.625 10.1 28.3 CE - Special CE 8.79 10.2 29.2 CE - Sys80 CE 11.2 11.7 45 Design RPV Maker Vessel T (in) Surface Fluence Needed to Reach 6E+19 @
Plant Design EFPY Surface Fluence (E+19) Potentially Impacted?
A WEC 3-Loop 72 10.8 at ~65 EFPY B
WEC 3-Loop 68 7.26 No C
Smaller CE 72 6.56 No D
WEC 3-Loop 72 7.34 No E
WEC 2-Loop 72 7.80 No F
B&W 72 2.02 No
© 2021 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 13 TogetherShaping the Future of Electricity