ML20009C415
| ML20009C415 | |
| Person / Time | |
|---|---|
| Issue date: | 11/06/2019 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| Burkhart, L, ACRS | |
| References | |
| NRC-0679 | |
| Download: ML20009C415 (156) | |
Text
Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION
Title:
Advisory Committee on Reactor Safeguards Open Session Docket Number:
(n/a)
Location:
Rockville, Maryland Date:
Wednesday, November 6, 2019 Work Order No.:
NRC-0679 Pages 1-89 NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers 1323 Rhode Island Avenue, N.W.
Washington, D.C. 20005 (202) 234-4433
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com 1
1 2
3 DISCLAIMER 4
5 6
UNITED STATES NUCLEAR REGULATORY COMMISSIONS 7
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 8
9 10 The contents of this transcript of the 11 proceeding of the United States Nuclear Regulatory 12 Commission Advisory Committee on Reactor Safeguards, 13 as reported herein, is a record of the discussions 14 recorded at the meeting.
15 16 This transcript has not been reviewed, 17 corrected, and edited, and it may contain 18 inaccuracies.
19 20 21 22 23
1 UNITED STATES OF AMERICA 1
NUCLEAR REGULATORY COMMISSION 2
+ + + + +
3 668TH MEETING 4
ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 5
(ACRS) 6
+ + + + +
7 OPEN SESSION 8
+ + + + +
9 WEDNESDAY 10 NOVEMBER 6, 2019 11
+ + + + +
12 ROCKVILLE, MARYLAND 13
+ + + + +
14 The Advisory Committee met at the Nuclear 15 Regulatory Commission, Two White Flint North, Room 16 T2B10, 11545 Rockville Pike, at 1:00 p.m., Peter 17 Riccardella, Chairman, presiding.
18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
2 COMMITTEE MEMBERS:
1 PETER RICCARDELLA, Chairman 2
MATTHEW W. SUNSERI, Vice Chairman 3
JOY L. REMPE, Member-at-Large 4
RONALD G. BALLINGER, Member 5
DENNIS BLEY, Member 6
CHARLES H. BROWN, JR. Member 7
VESNA B. DIMITRIJEVIC, Member 8
WALTER L. KIRCHNER, Member 9
JOSE MARCH-LEUBA, Member 10 DAVID PETTI, Member 11 12 DESIGNATED FEDERAL OFFICIAL:
13 CHRISTOPHER BROWN 14 15 ALSO PRESENT:
16 MIKE BLOOM, Duke Energy 17 JOSH BORROMEO, NRR 18 MARK DeWIRE, Duke Energy 19 STEVE EVANS, Duke Energy 20 TIM HARDIN, EPRI 21 ALLEN HISER, NRR 22 ANDY HON, NRR 23 RAJ IYENGAR, RES 24 MEENA KHANNA, NRR 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
3 ALSO PRESENT (CONTINUED):
1 2
JOHN LEHNING, NRR 3
LOUISE LUND, RES 4
SCOTT MOORE, Executive Director, ACRS 5
JEFF POEHLER, RES 6
DOUG PRUITT, Framatome 7
MARY JANE ROSS-LEE, NRR 8
DAVID RUDLAND, NRR 9
ASHLEY SMITH, NRR 10 DAN TINKLER, Framatome 11 ROBERT TREGONING, RES 12 DAN WIDREVITZ, NRR 13 AARON WYSOCKI, ORNL 14 PETER YARSKY, RES*
15 STEPHEN YODERSMITH, Duke Energy 16 17 18 19 20 21 22 23 24
- Present via telephone 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
4 CONTENTS 1
Call to Order and Opening Remarks........
5 2
White Paper Addressing Adequacy of 7
3 RG 1.99, "Radiation Embrittlement of 4
Reactor Vessel Materials," Revision 2, 5
and Working Group Efforts 6
Opportunity for Public Comment
......... 68 7
Committee Discussion
.............. 69 8
Brunswick Atrium 11 Fuel Transition and..... 77 9
Application/Framatome 10 Opportunity for Public Comment
......... 89 11 Adjourn..................... 89 12 13 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
5 P R O C E E D I N G S 1
1:03 p.m.
2 CHAIRMAN RICCARDELLA: (presiding) 3 The meeting will come to order.
4 This is the first day of the 668th 5
meeting of the Advisory Committee on Reactor 6
Safeguards. I'm Pete Riccardella, Chairman of 7
the Committee.
8 ACRS was established by the Atomic 9
Energy Act and is governed by the Federal 10 Advisory Committee Act, or FACA. The ACRS 11 section of the U.S. NRC public website provides 12 information about the history of the ACRS and 13 provides FACA-related documents, such as our 14 Charter, Bylaws, Federal Register notices for 15 meetings, letter reports, and transcripts of full 16 and subcommittee meetings, including slides 17 presented at the meetings. The Committee 18 provides advice on safety matters to the 19 Commission through its publicly-available letter 20 reports.
21 The Federal Register notice announcing the 22 meeting was published on October 16th, 2019, and 23 provides an agenda and instructions for interested 24 parties to provide written documents or request 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
6 opportunities to address the Committee, as required by 1
FACA.
2 In accordance with FACA, there is a 3
Designated Federal Official, or DFO, for today's 4
meeting. The DFO for this meeting is Mr. Christopher 5
Brown.
6 During today's meeting, the Committee will 7
consider the following: the White Paper addressing 8
the adequacy of Regulatory Guide 1.99, "Radiation 9
Embrittlement of Reactor Vessel Materials," Rev 2, 10 and, two, Brunswick Atrium 11 fuel transition and 11 application. And three, preparation for our December 12 meeting with the Commission. And then, also, if we 13 have time today, preparation of reports, which I doubt 14 we'll get to.
15 There's a phone bridge line. To preclude 16 interruption of the meeting, the phone will be kept in 17 listen-only mode during the presentations and 18 Committee discussions.
19 We have received no written comments or 20 requests to make statements from members of the public 21 regarding today's meeting. There will be an 22 opportunity for public comments, as we have set aside 23 10 minutes in the agenda for comments from members of 24 the public attending or listening to our meetings.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
7 Written comments may be forwarded to Mr. Christopher 1
Brown, the DFO.
2 A transcript of the open portions of the 3
meeting is being kept, and it is requested that the 4
speakers use one of the microphones and identify 5
themselves and speak with sufficient clarity and 6
volume so that they can be readily heard.
7 And I would also request that everybody 8
turn off or silence their cell phones to avoid 9
interruption of the meeting.
10 With that, I will turn the meeting over to 11 Louise Lund, who will introduce the first topic.
12 MS. LUND: Yes. Thank you. And thanks to 13 the Committee for having us here today.
14 My name is Louise Lund. I'm the Division 15 Director for the Division of Engineering in the Office 16 of Research.
17 And today, we're here to talk about 18 Regulatory Guide 1.99, Revision 2,
"Radiation 19 Embrittlement of Reactor Vessel Materials," which is 20 used by licensees in the operating fleet to determine 21 the change in reactor vessel properties due to 22 radiation.
23 So, the guidance in the Reg Guide is used 24 to determine the adjusted reference temperature, which 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
8 is used to determine the fracture toughness, a key 1
input to pressure-temperature limits. The Reg Guide 2
also provides guidance for assessing the change in 3
upper shelf energy, which is related to ductile 4
fracture resistance. And both of these properties are 5
needed to show compliance with 10 CFR 50, Appendix G.
6 And I think that it's been noted that the 7
current Reg Guide dates from 1988. Since that time, 8
there has been a large amount of test data on 9
irradiated reactor vessel materials which have been 10 obtained from test or surveillance materials. So, 11 with licensees planning to operate plants out to 60 or 12 80 years, it is important that guidance for assessing 13 embrittlement be adequate for the neutron fluence 14 levels that are projected for those plants. New 15 plants will use reactor vessel materials with 16 different chemistries than the operating fleet.
17 Therefore, the staff performed an 18 assessment of the adequacy of Reg Guide 1.99, Rev 2, 19 based on these considerations. And the presentation 20 today summarizes the multiyear effort to assess Reg 21 Guide 1.99 through a modern data-driven approach.
22 And based on a lot of the discussions we 23 had at the Subcommittee, we also have discussed in the 24 presentation the next steps that we have been 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
9 formulating for the staff based on this assessment, 1
including the formulation and activities of a working 2
group for this purpose and an oversight group to 3
develop revision of Reg Guide 1.99.
4 So, without further ado, I'm going to go 5
ahead and turn it over to the staff. You have Allen 6
Hiser and Dan Widrevitz and Jeff Poehler in front of 7
you today.
8 MR. POEHLER: Good afternoon.
9 I'm going to introduce the presentation 10 today. So, this is Regulatory Guide 1.99, Revision 2, 11 assessment results and follow-on activities. This 12 presentation summarizes a
multiyear effort to 13 thoroughly assess Reg Guide 1.99 through a modern 14 data-driven approach; plus, the activities underway to 15 address the assessment results.
16 And here's an outline of today's 17 presentation. First, we're going to go through some 18 background on the current Regulatory Guide 1.99.
19 Then, we're going to talk about the assessment 20 results, which cover RTndt; upper shelf energy; 21 credibility criteria; plant-specific data results; 22 attenuation results; common additions to the Reg Guide 23 that are not actually in the Reg Guide, but these are 24 basically common practices; conclusions. We will, 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
10 then, discuss the Reg Guide 1.99 working group charter 1
and activities. And then, we'll discuss the Reg Guide 2
1.99 oversight group charter and activities.
3 MEMBER KIRCHNER: Jeff, so you don't get 4
a crook in your neck, why don't you borrow Dan's 5
microphone? That one seems to work better.
6 MR. POEHLER: How's that? Is that better?
7 MEMBER KIRCHNER: Much better.
8 MEMBER BLEY: Much better.
9 MR. POEHLER: Okay. Great.
10 So now, I'm going to turn it over to Dan 11 to cover the results of the assessment.
12 (Laughter.)
13 MR. WIDREVITZ: Good afternoon, everyone.
14 My name is Dan Widrevitz, and I have a working 15 microphone.
16 (Laughter.)
17 And I'm also the lead author of this 18 report. Obviously, it's the product of a number of 19 employees, current and former, all of whom who are 20 currently in this room.
21 And so, just to give a little bit of 22 background -- I know many of you were here for the 23 Subcommittee meeting, but I know several of you 24 weren't -- the Regulatory Guide 1.99 provides guidance 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
11 with regards to radiation embrittlement of reactor 1
vessel materials. As we will repeat ad nauseam, it 2
was issued in May of 1988.
3 MEMBER BALLINGER: Which is, by the way, 4
probably before you were born.
5 MR. WIDREVITZ: Not quite. I remember 6
Ghostbusters.
7 (Laughter.)
8 The curve itself, it contains a number of 9
aspects, one of which is an embrittlement trend curve 10 purporting to predict the shift in ductile-to-brittle 11 transition temperature of reactor vessel steels, which 12 we call RTndt. This curve was fit to what was done 13 177 data points, datums. There was a relative paucity 14 of data at higher fluences because the fleet was 15 relatively young. And there was a relative paucity of 16 data concerning low copper materials because the fleet 17 designs were relatively old.
18 The Regulatory Guide also includes trend 19 curve for upper shelf energy, which is a measure of 20 toughness of a material once you've entered the 21 ductile regime, which is used in concordance with an 22 acceptance criteria in 10 CFR 50, Appendix G. I'll 23 discuss what we do with RTndt in the next slide.
24 It contains what we call the credibility 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
12 criteria, which are a set of criteria for plant-1 specific surveillance data that allow a certain 2
manipulation of the embrittlement trends curves if you 3
have good data, like the credibility criteria 4
determined to be good data.
5 And finally, it has an attenuation formula 6
based on dpa studies which allows you to predict the 7
change in embrittlement as you go through the steel of 8
the vessel and, essentially, as your neutron flux 9
attenuates, you'll receive less damage as you go 10 through. And this is a way of estimating where you 11 are at different depths within the vessel.
12 So, what do we do with RTndt? We 13 calculate something called adjusted reference 14 temperature, which is you take your estimate of 15 unirradiated RTndt, you add a delta RTndt, which is 16 essentially from the curve in the Reg Guide, and you 17 get a margin where the margin's trying to capture your 18 uncertainties in measurement in the database, et 19 cetera.
20 The ART is used to determine fracture 21 toughness when developing pressure-temperature limits 22 for operating the plant. And the entire Reg Guide's 23 approach is contained with 10 CFR 50.61, which uses 24 the same model as Reg Guide 1.99.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
13 So, in terms of results, RTndt results, 1
our analysis was conducted using what we call the 2
baseline dataset, which was generated by ASTM E10.02, 3
which is available on the ASTM website. The dataset 4
includes domestic and international power reactor 5
data. Something like half of it is domestic. It's a 6
total of 1900 data points. It's a much, much larger 7
database than was available in the '80s.
8 And we performed a number of in-depth 9
statistical analyses. I won't get into these. There 10 is a lot of information in the assessment report 11 itself.
12 In terms of interesting results, what we 13 have plotted here are two plots of the residuals.
14 That's essentially your prediction minus measured 15 values. The top plot is for what we call base 16 materials, plates and forgings, et cetera. The red 17 dots are U.S. data. The gray dots are international 18 data.
19 And if you look way at the high end -- the 20 X-axis is fluence -- you see that the measured values 21 begin to significantly outweigh the predicted values, 22 so larger. So, that's a non-conservative prediction 23 of embrittlement results that we're seeing for 24 measured data comparing to prediction. So, that's a 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
14 bit concerning since PWRs are all heading in that 1
direction in terms of fluence.
2 CHAIRMAN RICCARDELLA: Just for the 3
record, negative values on that are non-conservative.
4 MR. WIDREVITZ: Yes, negative values mean 5
that the measure is larger than the predicted.
6 CHAIRMAN RICCARDELLA: Right.
7 MR. WIDREVITZ: Which is non-conservative 8
for our use.
9 Now the similar plot for weld materials 10 does not clearly show this trend. And whether that's 11 an artifact of the paucity of data at high fluence or 12 whether it's an artifact of how the curves were 13 originally put together -- if we want, there's a paper 14 by Randall that discusses how the Reg Guide 1.99 15 curves are put together, if you want more detail.
16 There doesn't appear to be the same effect 17 for welds. And traditionally, welds have been some of 18 our more concerning materials in terms of how we 19 generate our safety approach.
20 So, a lot of interesting questions. This 21 is really the top result. So, to sum it up, our --
22 CHAIRMAN RICCARDELLA: Also just for 23 clarity, the horizontal dashed lines on that, they're 24 two sigma lines, but not of this data. They're two 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
15 sigma lines that are in the Reg Guide that we use to 1
determine the margin, right?
2 MR. WIDREVITZ: Yes, those are the Reg 3
Guide standard deviations, which I will say in a few 4
minutes are wrong.
5 CHAIRMAN RICCARDELLA: Well --
6 MR. WIDREVITZ: Well, they're off.
7 CHAIRMAN RICCARDELLA: They're wrong, 8
yeah --
9 MR. WIDREVITZ: Compared to the modern 10 dataset.
11 CHAIRMAN RICCARDELLA: Yes.
12 MR. WIDREVITZ: I mean, you're always 13 limited by your dataset.
14 CHAIRMAN RICCARDELLA: But, again, if you 15 look at the weld, the lower horizontal dashed line on 16 the weld plot, you only have a couple of data points 17 that actually are outside of that main --
18 MR. WIDREVITZ: And if you look at the 19 print version, which is less saturated than these TVs, 20 what you'll see in welds is there's quite a bit of 21 number of points above the dotted line for welds --
22 CHAIRMAN RICCARDELLA: Yes.
23 MR. WIDREVITZ: -- which is an overly-24 conservative estimate.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
16 CHAIRMAN RICCARDELLA: Yes.
1 MR. WIDREVITZ: Which, from a safety 2
perspective --
3 CHAIRMAN RICCARDELLA: Yes.
4 MR. WIDREVITZ: -- is not an issue.
5 CHAIRMAN RICCARDELLA: Uh-hum. Okay.
6 Thank you.
7 MR. WIDREVITZ: So, in terms of results, 8
our primary conclusions: non-conservative high 9
fluence results for base metals become prominent at 10 fluences above, we generally say 6 times 10 to the 11 19th neutrons per centimeter squared. And that's 12 where your mean of your results goes below 2 sigma, 13 the Reg Guide's 2 sigma. So, you would expect your 14 results, on average, to be too low there.
15 And also, you get inaccurate low copper 16 results because the Reg Guide simply wasn't fed with 17 low copper data. The interpolation just didn't 18 capture those behaviors.
19 In terms of secondary conclusions, the 20 standard deviation of delta RTndt in the Reg Guide 21 seems too low when compared to a modern database.
22 There's a conservative bias in low to mid fluences, 23 which some mind consider a burdensome estimate. And 24 there's a lack of temperature adjustment, which is an 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
17 accuracy issue. For the current fleet, how big that 1
accuracy issue is is something you can discuss. But 2
when you think about applicants like NuScale who are 3
off the chart in terms of temperature, that becomes 4
more of a deficiency if you wanted to apply the Reg 5
Guide to that.
6 In terms of delta used results, the 7
assessment for delta used was based on the REAP 8
dataset, which is a collection of surveillance, plant 9
surveillance results merged with the actual material 10 properties in baseline. The dataset, again, included 11 domestic and international power reactor data, 1200 12 data points, considerably more than we had in the 13
'80s.
14 What you see here is a plot attempting to 15 show the significance of the results. So, the 16 question in upper shelf energy is, Appendix G has an 17 energy criteria, 50 foot-pounds. And when you go 18 below that criteria, you have to conduct what's called 19 an equivalent margins analysis, which is a more 20 sophisticated analysis in terms of your survivability 21 with transients. No plant has failed to succeed one 22 of these. They've gone considerably above 50 foot-23 pounds.
24 However, the concern is, if you have a 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
18 material where there should be an equivalent margins 1
analysis, and yet you have a prediction that does not 2
indicate that you need to have done one, are you in a 3
difficult spot? So, when we conducted the analysis, 4
on this plot, plants where an equivalent margin 5
analysis should have been triggered, but is not 6
currently predicted, is the lower righthand quadrant.
7 And as you can see, it's just barely anything is in 8
there. It's a statistically insignificant number of 9
our materials, to the best of our knowledge, are in 10 that state. And so, our concern there simply isn't to 11 the same order of magnitude as for RTndt.
12 So, sort of a conclusion: what we see 13 with the modern datasets is something like 19 percent 14 of the materials measured delta used is not bounded by 15 the Reg Guide model. The Reg Guide was originally 16 drawn as an upper-bound model. So, this is not good 17 behavior for its original intent. There's a very 18 limited number of materials that are currently 19 misprojected or remain above 50 foot-pounds and do not 20 trigger the PNAs.
21 However, there's a minimal impact. The 22 safety criteria that's supported by upper shelf 23 estimation is known to be extremely conservative. And 24 so, consequently, that's not really an area of further 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
19 focus.
1 In terms of credibility criteria, the Reg 2
Guide has five credibility criteria. You apply these 3
credibility criteria to measured data, which you, 4
then, quote, "refit" in the Reg Guide, where you take 5
what they call a chemistry factor, which is a linear 6
multiplier on your fluence function, and you least 7
squares fit it to your actual data. So, if you've got 8
credible data, you allow a shift of the curve into 9
your data, and then, it allows you to reduce your 10 margin term by a certain amount. If your data is not 11 deemed credible through these five criteria, you do 12 not consider it. You're indicated not to consider it.
13 And this criteria typically has failed because of 14 excessive scatter in your data. A lot of this is fed 15 by Sharpie testing, and Sharpies have a fair amount of 16 scatter. The assessment goes into a great deal of 17 detail on this.
18 In terms of issues, there's a lot of 19 issues of construction of the credibility criteria.
20 One of them is, the more surveillance data points you 21 have, the more likely it is for data to be deemed non-22 credible due to scatter. It's not well insulated to 23 have an outlier test that scales with the number of 24 samples you're taking. It has no apparent basis for 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
20 the reduction in margin for credible data.
1 A real critical consequence is that the 2
benefit of fitting predictions to surveillance data is 3
nullified in many cases by the credibility criteria 4
rejecting data that do not perform to the fluence 5
shape function of Reg Guide 1.99. So, when you try to 6
fit a curve to something that just doesn't have the 7
same slope characteristics, you're going to have more 8
rejections. And that's what it does.
9 Why we're concerned about that here is 10 high fluence and low copper data are not expected to 11 conform to the fluence shape function of Reg Guide 12 1.99.
13 MEMBER PETTI: So, in essence, you're 14 potentially throwing out the data?
15 MR. WIDREVITZ: You are going to throw out 16 data that doesn't look as you expect it. So, that's 17 exactly right.
18 MEMBER PETTI: And data specific for your 19 plant material.
20 MR. WIDREVITZ: If your actual material 21 doesn't match the Reg Guide, you're told not to use 22 it. So, it's not --
23 MR. HISER: Except in one scenario. If 24 your data indicates a higher trend, you don't get to 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
21 reduce the margin, but you do get to use the higher 1
chemistry factor.
2 MR. WIDREVITZ: No.
3 (Laughter.)
4 No, you don't enter the procedure if your 5
data is not credible. Common practice has superseded 6
the Reg Guide text. And that gets to something I'll 7
discuss after attenuation.
8 Sorry. Sorry, Jeff. Go back.
9 The attenuation formula given in the Reg 10 Guide was based on dpa for studies. All modern data 11 confirmed that it's a fairly good representation of 12 essentially the reduction damage as you go through the 13 steel. The formula only works for areas horizontally 14 adjacent to the active fuel area.
15 So, what we see now that we have these 16 plants that have been operating for a considerable 17 long time is that our area of interest grows. We 18 essentially say you must look at the area that has 19 fluence 10 to the 17th or greater neutrons per 20 centimeter squared. As your plant ages, that creeps 21 up, and it creeps down on both ends of the active 22 fuel. In areas that are not exactly adjacent to the 23 fuel, you get geometry effects, get a lot of things 24 going on where this formula just doesn't apply.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
22 And so, we actually have a research 1
project on determining fluence outside the beltline in 2
operation. This is to be incorporated in revision of 3
Reg Guide 1.190. Certainly the materials folks here, 4
we're certainly aware, and we keep in touch, but we 5
don't review that directly. So, I'm not going to say 6
anything more about that.
7 So, common additions. There are several 8
practices not addressed in the Reg Guide that have 9
been commonly accepted by the NRC. In short, a great 10 number of users had difficulties, had questions. The 11 credibility criteria are not particularly tightly-12 written. And as a result, the NRC had a presentation 13 in 1998 which we tend to call the Wichman presentation 14 where we discussed how to deal with a number of issues 15 that our applicants were encountering.
16 One of them was the use of sister plant 17 data to supplement plant-specific surveillance data.
18 In this case, this is material that's a heat-for-heat 19 match which has been irradiated in another plant.
20 There, you adjust it by temperature. There's also 21 quite a bit of discussion of what to do with outliers, 22 particularly low fluence
- outliers, in that 23 presentation. And there are a lot of sort of art-of-24 practice tips that are critical to successfully using 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
23 the Reg Guide that are not entirely consistent with 1
how the Reg Guide is written, depending on how you 2
read it. So, again, the Reg Guide is not very 3
tightly-drawn. A lot of it text. And that's where I 4
think Allen and I, as you just saw, there are some 5
textual ambiguities that certainly this presentation 6
aimed to address.
7 So, in conclusion, correcting the non-8 conservatism of the embrittlement trend from higher 9
fluences is the most significant recommendation of the 10 assessment. The credibility criteria should also be 11 revised to make more effective use of plant-specific 12 surveillance data, particularly in those cases where 13 you're failing the credibility criteria because your 14 data is showing genuine trends that do not match the 15 shape function of the Reg Guide. The Reg Guide really 16 only has one shape function, which is sort of on the 17 higher kappa range.
18 And there are several common practices not 19 addressed in the Reg Guide that should be addressed in 20 a revision, such as the use of sister plant data, 21 implementation of credibility criteria to agree with 22 things that are in the presentation, or otherwise 23 necessary to discuss clarification. This is more 24 deficiency of omission.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
24 I should note that, depending on what one 1
would like to do with the Reg Guide, some of these 2
things would be simply dealt with inherently in a 3
solution. For example, if you replace the trend curve 4
with something that has a temperature term, you no 5
longer need to even consider this degree of the 6
correction in the Wichman presentation.
7 MR. POEHLER: Okay. Now I'm going to 8
discuss what the staff is doing in response to the 9
issues that were identified in the Technical Letter 10 Report. So, in response to the Technical Letter 11 Report, the NRC has initiated an effort to revise Reg 12 Guide 1.99, and we have chartered two groups for this 13 effort. The first one is a working group, and the 14 second one is an oversight group. And I'm going to 15 talk on the next slide about the working group and 16 what it's doing.
17 So, the working group's primary task is to 18 develop a revision to Reg Guide 1.99, Rev 2, that will 19 address the recommendations of the TLR. And we're 20 going to give priority to those recommendations with 21 the most significant potential reduction in safety 22 margin, particularly in the non-conservatism in high 23 fluence being the highest priority.
24 The upper shelf energy model in the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
25 Regulatory Guide will not be addressed because, 1
potentially, non-conservative upper shelf energy 2
predictions have no credible impact on safety, as has 3
been demonstrated by NRC accepted equivalent margin 4
analyses.
5 The working group is going to concentrate 6
on the technical content of the Regulatory Guide 7
rather than regulatory or implementation issues. And 8
the working group consists of both NRR and Office of 9
Research staff. And Office of Research staff serves 10 as the lead for the working group.
11 So now, I'm going to talk about some of 12 the tasks or the major tasks that the working group 13 has underway. The first one of those was to recommend 14 an alternate embrittlement trend curve, and that one 15 we have completed. We have selected the ASTM E900-15 16 embrittlement trend curve as the basis for the 17 revision of Reg Guide 1.99, Revision 2.
18 Yes?
19 CHAIRMAN RICCARDELLA: Excuse me, Jeff.
20 MR. POEHLER: Sure.
21 CHAIRMAN RICCARDELLA: Does that curve 22 have any sort of temperature adjustment in it?
23 MR. POEHLER: It has. Temperature is a 24 variable.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
26 CHAIRMAN RICCARDELLA: It's a parameter in 1
it?
2 MR. POEHLER: Yes.
3 CHAIRMAN RICCARDELLA: Great. Okay.
4 MR. POEHLER: The next major task is to 5
determine the limitations -- and these aren't in any 6
particular order, by the way; the next four of these 7
are things that we're working in parallel --
8 determining the limitations of the embrittlement trend 9
curve implementation. So, these would be limitations 10 on the ranges of the variables. Like all the chemical 11 composition variables, of which there are four, 12 temperature, fluence -- and these would be defined, 13 basically, from the database from which E900 was 14 developed areas of the database. Areas of the ranges 15 of the variables where the data gets thin is going to 16 help us determine what the limitations are.
17 The next major task is determining how to 18 apply surveillance data. You know, it's just going to 19 be a bit different, how we would do it with the E900 20 as opposed to how we do it now, just by the nature of 21 the differences in the two different trend curves.
22 The next --
23 CHAIRMAN RICCARDELLA: Is this referring 24 to the credibility criteria, this one?
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
27 MR. POEHLER: Yes. Yes, in part, yes.
1 And whether you could refit the curve to actual data 2
or not and things like that. Yes.
3 Reg Guide 1.99 was relatively simply to 4
refit because of the smaller number of variables.
5 MEMBER KIRCHNER: Jeff, is there any 6
requirement there? If you start refitting the E900 7
curve, it almost begs the question of why don't you 8
just use the surveillance data that's prototypical for 9
the plant in question. Or at least it leads me to 10 think that, rather than fall back on a curve that's 11 populated by lots of plants with lots of potential 12 variations in material and forging or manufacturing, 13 et cetera.
14 MR. POEHLER: Yes, I mean, that's a good 15 point, but I think the philosophy is we would rather 16 use the trend curve based on a large amount of data 17 if --
18 MEMBER KIRCHNER: No, I understand that, 19 right.
20 MR. POEHLER: -- if your actual plant data 21 is fitting that with respect --
22 MEMBER KIRCHNER: Yes. Well, it would beg 23 the question of how much data do you have, 24 surveillance data that you could construct your own 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
28 curve without having a large uncertainty band, yes.
1 MR. HISER: Yes, I think one of the things 2
that we would be after is looking at the trend from 3
E900 as a function of fluence because the data can 4
have quite a bit of scatter.
5 MEMBER KIRCHNER: Yes, yes.
6 MR. HISER: So, it may indicate that it's 7
well above or well below the E900 prediction.
8 MEMBER KIRCHNER: Right.
9 MR. HISER: But we want to be able to 10 maintain some coherency, so that you can calculate a 11 projected embrittlement for the vessel.
12 MR. WIDREVITZ: Just to clarify, I mean, 13 this is exactly the question that we're working very 14 hard to consider. The question is, when do you gain 15 more from your plant-specific data in terms of 16 handling the uncertainties --
17 MEMBER KIRCHNER: Exactly. And how many 18 data points do you need?
19 MR. WIDREVITZ: Right. Because of you're 20 in the middle of your data cloud and you've got a very 21 dense amount of data in your trend curve, then you 22 have a statistical quality there that your plant-23 specific data can never achieve. There just aren't 24 enough samples. And that's exactly what we're 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
29 thinking about.
1 CHAIRMAN RICCARDELLA: Yes, I mean, how 2
many data points do you typically have after, say, 3
you've pulled three or four surveillance capsules at 4
a plant? What are we talking about?
5 MR. WIDREVITZ: For material, three or 6
four.
7 CHAIRMAN RICCARDELLA: Three or four 8
points per material?
9 MR. WIDREVITZ: Yes.
10 CHAIRMAN RICCARDELLA: So, that's why.
11 MR. WIDREVITZ: You're spending a lot of 12 your samples to get that, to find your transition, but 13 now you've only got one measurement of transition out 14 of so many samples.
15 CHAIRMAN RICCARDELLA: Right.
16 MR. WIDREVITZ: And that's sort of the 17 data science challenge.
18 CHAIRMAN RICCARDELLA: Aspect of it.
19 Okay.
20 MR. POEHLER: Another task would be to 21 determine the margins on the embrittlement trend 22 curve. So, the E900 standard provides an equation for 23 the standard deviation of the prediction. But we 24 would want to structure it more like the current Reg 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
30 Guide most likely, which has, you know, it's based on 1
2 standard deviations, but there's also a standard 2
deviation associated with the initial unirradiated 3
RTndt. So, we would probably keep the same structure, 4
but use the standard deviation of the shift, which is 5
obviously different, for that.
6 CHAIRMAN RICCARDELLA: Is there a way to 7
risk-inform that as to what the margin should be?
8 MR. HISER: I think that's probably one of 9
the things we're going to look at. I mean, we don't 10 have at this point preconceived notions of here's what 11 we're going to do.
12 CHAIRMAN RICCARDELLA: Yes, yes.
13 MR. HISER: It's more open-ended.
14 MR. WIDREVITZ: Yes, absolutely. I mean, 15 that's one of the questions we're asking ourselves.
16 CHAIRMAN RICCARDELLA: It seems to me, I 17 mean, heat-up/cool-down curves really aren't a high-18 risk part of plant operation, I don't think, are they 19 in terms of creating a --
20 MR. HISER: It all depends. They're one 21 part of the operation that you use every cycle. So, 22 from that perspective, it's a known transient, if you 23 will.
24 CHAIRMAN RICCARDELLA: I know, but you 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
31 assume a Goretti flow. You put safety factors on the 1
loads. I mean, just to pick 2 sigma, just to say, 2
okay, we'll use 2 sigma -- I just wonder if we could 3
risk-inform that somehow.
4 MR. HISER: In the next couple of slides, 5
I'm going to talk about some of the things we're going 6
to do to look at the safety impacts of this.
7 CHAIRMAN RICCARDELLA: Okay. Thank you.
8 MR. POEHLER: Another task is to determine 9
the default values for inputs that are not available.
10 Typically, you have the information on the chemistry, 11 but the E900 trend curve uses manganese and 12 phosphorus, which is not needed for the current Reg 13 Guide. Most plants should have this data on their 14 materials. But, in case they don't, you know, what 15 defaults do you use? And that could even apply to 16 copper and nickel for some occasions.
17 And
- also, things like irradiation 18 temperature, which some plants have a really good idea 19 what it is, and I'm not sure they all do. They didn't 20 have to have that before for the current Reg Guide.
21 So, that's another major task.
22 And the final piece of this is to actually 23 write the Draft Regulatory Guide for internal review.
24 But the second four bullets would be probably the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
32 major elements, technical elements, that go into that 1
Reg Guide.
2 Okay. Now I'm going to turn it over to 3
Dr. Allen Hiser to talk about the oversight group 4
activities.
5 MR. HISER: Okay. Thanks, Jeff.
6 Now the oversight group looks to provide 7
guidance and direction for the overall effort.
8 Numbers of the group are from both Research and the 9
Office of Nuclear Reactor Regulation to, you know, a 10 Branch Chief in each office who has cognizance over 11 the vessel integrity issues. And also, we have a 12 Branch Chief who has a deep history. So, he's 13 included as well. And we have three SLs in the two 14 offices, the materials SL.
15 Responsibilities of the oversight group 16 are listed here. One is to recommend implementation 17 options. So, we take the results from the working 18 group, products, determine or recommend ways that this 19 could or should be implemented for plants. And I'll 20 talk a little bit more about that in a couple of 21 slides.
22 The main thing is we're looking to assess 23 adequate protection aspects of changing the 24 embrittlement predictions. And that really is one of 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
33 the key things that we need to sort through.
1 And finally, we will identify/recommend if 2
there are any needs for rulemaking. As Jeff 3
mentioned, the PTS rule, 50.61, has Reg Guide 1.99, 4
Rev 2, methodology embedded within the rule. If the 5
Reg Guide were to be changed, how does that impact the 6
need to change the 50.61?
7 Next slide.
8 We lost some of our colors. This was 9
supposed to be a timeline. On the printout, there's 10 actually an arrow. Yes, here you go. I noticed that 11 yesterday; the screen drops some of the colors.
12 This is clearly not a linear trace, and it 13 just hits on some of the important past milestones and 14 future dates of interest.
15 The Reg
- Guide, as was mentioned 16 previously, was issued in 1988. So, the next gap 17 there takes us to 2019 with the research report. The 18 entire rest of the timeline is an equivalent 31 years.
19 The items that are on the top represent plants where 20 we think the fluence make exceed 6 times 10 to the 21 19th neutrons per centimeter squared. We've used that 22 sort of as a trigger point where, from the charts that 23 Dan showed, at least for plate materials we may be 24 getting a deviation that's too much. I think that's 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
34 on the order of, I don't know, 40-50 degrees 1
Centigrade.
2 So, in this case for Turkey Point Plant, 3
which you've heard the SLR, I guess the whole 4
Committee heard about the SLR application and our 5
review. We expect that their surface fluence for the 6
vessel will exceed 6 times 10 to the 19th somewhere in 7
2025. Turkey Point, the 1/4T fluence will exceed that 8
level in 2044.
9 CHAIRMAN RICCARDELLA: I thought long and 10 hard to try to figure out what "PTN" stood for. Thank 11 you.
12 (Laughter.)
13 MR. HISER: That is their name for their 14 plant.
15 CHAIRMAN RICCARDELLA: I know.
16 MR. HISER: I think the alternative is not 17 maybe as friendly.
18 Then, in 2048, Surry Unit 2, the surface 19 of the vessel would exceed 6 times 10 to the 19th.
20 So, in general, P-T limits are one of the main areas 21 of concern that we have. Twenty-five years is really 22 the point at which we would need to have something in 23 place to satisfy Turkey Point's needs.
24 Surface fluence is important for PTS. At 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
35 least from the information we have today on Turkey 1
Point, they have a significant margin for their plate 2
and base materials for PTS.
3 CHAIRMAN RICCARDELLA:
- But, as I
4 understand it, Reg Guide 1.99 doesn't really apply to 5
PTS, does it?
6 MR. HISER: The methodology, the chemistry 7
factor times fluence factor, the credibility criteria, 8
all of that was built into 50.61.
9 CHAIRMAN RICCARDELLA: I thought it was 10 embedded. It thought the PTS rule had a different 11 irradiation.
12 MEMBER BALLINGER: They used a different 13 correlation.
14 MR. HISER: 50.61(a), the alternate PTS, 15 has a different correlation.
16 CHAIRMAN RICCARDELLA: Ah, okay. Okay.
17 MR. HISER: Yes, but the original, which 18 changed actually after the Reg Guide was changed in I 19 believe 1991-92, so that it would have coherency with 20 the Reg Guide.
21 MEMBER BALLINGER: If the criteria was 3 22 times 10 to the 19th, how far would this shift to the 23 left?
24 MR. HISER: Turkey Point would shift a 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
36 bit.
1 MEMBER BALLINGER: A bit?
2 MR. HISER: Yes. Maybe to earlier than --
3 MEMBER BALLINGER: Maybe to now?
4 MR. HISER: Yes.
5 (Laughter.)
6 MEMBER BALLINGER: Yes.
7 MR. WIDREVITZ: We would have a lot more 8
plants right now and soon.
9 MR. HISER: I think Turkey Point may have 10 about the highest fluences of what we expect to see.
11 Surry also had --
12 CHAIRMAN RICCARDELLA: Implying they don't 13 have real bad chemistry, is that --
14 MR. HISER: Referring back to Dan's 15 charts, he shows for welds that there doesn't appear 16 to be an adverse trend at high fluence. The materials 17 that are limiting materials for Turkey Point are 18 welds.
19 CHAIRMAN RICCARDELLA: Okay.
20 MR. HISER: The base materials, there is 21 a significant margin.
22 Okay. So, that's on the top where we need 23 to worry about the plants and we need to make certain 24 that we have the best methodology in place that we can 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
37 come up with to estimate embrittlement trends. The 1
bottom bullets, starting in 2025 through 2032, 2
represent where we will get additional data. 2025, 3
the PWR Owners Group has put together a Supplemental 4
Surveillance Program, the PSSP, and those capsules 5
will be coming out in, I guess somewhat after 2025, 6
but that data will become available in the late 2020s.
7 Turkey Point is scheduled to withdraw a 8
capsule in 2026, and it would be tested shortly 9
thereafter and the data would be available for Turkey 10 Point, again, the highest fluence PWR. That data will 11 cover their subsequent license renewal fluence regime.
12 So, that will give them data, plant-specific data, to 13 satisfy their needs.
14 Surry Unit 1 will pull a capsule in 2027.
15 Surrey Unit 2 in 2032. Those are the capsule 16 withdrawal dates. The data availability, obviously, 17 will be somewhat after that. So, there will be data 18 available for the initial SLR plants.
19 Okay. Next slide.
20 Now, in terms of the activities of the 21 oversight group, I talked about assessing the safety 22 significance. We're looking at both the safety 23 significance of deviations in prediction. So, if you 24 underpredict the embrittlement, what is the impact for 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
38 PTS and for P-T limits? We're looking at doing FAVOR 1
runs to try to assess through-wall crack frequency and 2
leveraging some of the analyses from the PTS 3
reevaluation. That was what was used to develop 4
50.61(a). As I mentioned, if you have a shift in the 5
RTndt, maybe you underestimate the shift. What's the 6
impact on through-wall crack frequency?
7 In addition, we're looking at a plant 8
impact study on a subset of plants. So, looking at 9
what we call a smart sample with features such as high 10 fluence, high copper, the plants that would tend to 11 have the highest embrittlement. We just want to 12 understand what the impact of using the E900 trend 13 curve would be on the embrittlement estimates for 14 those plants.
15 And we will use the working group's update 16 recommendations and criteria that had been developed 17 previously. Using all of these results, we'll look at 18 what implementation options we should have for the Reg 19 Guide. And the gamut could be from not making any 20 change to the Reg Guide, choosing to modify the Reg 21 Guide, and then, implement it on all plants, no 22 plants, some subset of plants, to be determined. We 23 have a lot of work to do to get to that point.
24 Next slide, Jeff.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
39 And looking at the milestones, I guess 1
probably the ones that I would zero-in on the most 2
would be a public meeting with stakeholders in the 3
middle of February that we're looking at, where we 4
would discuss the results that we have to date and 5
talk about our options going forward.
6 Two weeks after that, we would have a 7
decision point for whether we continue the work to 8
revise the Reg Guide and implement the new 9
embrittlement trend curves or whether we would 10 terminate this effort. Assuming that we continue with 11 the Reg Guide revision, then we would expect to have 12 a Draft Reg Guide for internal staff review the end of 13 April. So, that's five and a half months away.
14 And next slide.
15 MEMBER REMPE: I think you really mean 16 2020 on that day, right?
17 (Laughter.)
18 MR. HISER: Thank you.
19 MEMBER REMPE: Yes.
20 MR. POEHLER: My bad. I put that line 21 item in there.
22 MR. HISER: Revisionist history.
23 MEMBER BALLINGER: I'm trying to get a 24 handle on the longer term. What happens after April 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
40 30th, 2020? The internal review, how long does that 1
take? In other words, when can we get surety, if you 2
want to use that word, on what's going to happen?
3 MR. HISER: That's a good question. We 4
have an internal review process overall. I mean, at 5
this point, this would be the working group and 6
oversight group recommendation going forward. I would 7
expect we'd have a lot of internal discussions.
8 MR. IYENGAR: So, once we have the draft 9
and it's reviewed by our colleagues in NRR and us, 10 we'd have a Draft Reg Guide and it goes to the Reg 11 Guide Branch. And it goes through a process of 12 releasing it for public comment. And it receives 13 public comment to give them 60 days. And then, 14 finalize it. And then, issue it.
15 It could typically range from four to six 16 months, and depending on the public comments, if the 17 public comments are extensive and need resolution 18 time, it might take longer. But that's the timeframe 19 we are shooting for.
20 And one of the reasons for this engagement 21 with you all is to gain some efficiency upfront. We 22 released the TLR publicly in August. So that there's 23 enough public awareness of this topic and we have this 24 engagement. So, as we move along, we will probably 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
41 gain some efficiency in the Reg Guide process.
1 MEMBER BALLINGER: So, my high school math 2
tells me that the end of 2020 is probably a good date 3
to think about.
4 MR. IYENGAR: Well --
5 MEMBER BALLINGER: Because you said six 6
months and I just added six to four, got 10, assumed 7
it was 12.
8 MR. IYENGAR: You are a professor. So, 9
you certainly want to pin down the students on dates.
10 (Laughter.)
11 I certainly don't want to make the 12 commitment, but I just typically give an average time.
13 MS. LUND: Yes, and I would be remiss if 14 I also didn't mention that, you know, the area of 15 regulatory guidance is subject to some Executive 16 Orders that have come out recently as well. So, we 17 and other agencies are also trying to see if that 18 impacts any of our processes as well. We hope not.
19 I mean, because we already put out our guidance for 20 public comment and things like that. But that's 21 another thing that we're also engaged in as well, is 22 responding to that.
23 MEMBER BALLINGER: So, sometime before Dan 24 has white hair?
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
42 MR. WIDREVITZ: If it goes that way, for 1
certain.
2 CHAIRMAN RICCARDELLA: I guess speaking 3
for myself -- and this isn't for the Committee, just 4
my personal opinion -- and I understand we have to be 5
cautious not to put an undue burden. But to be in a 6
situation where we have a trend curve based on 177 7
points that was published in 1988, and now we have 8
1900 data points that show there is some inadequacy, 9
I mean, to not revise the Reg Guide just doesn't make 10 sense.
11 MS. LUND: And we heard that, you know, 12 and we agree with that. Okay? And I think that we've 13 got the working group together to address that.
14 CHAIRMAN RICCARDELLA: Yes, yes.
15 MS. LUND: And I think that what you see 16 here is I think a fairly assertive schedule to address 17 that with the working group. Trying to put an actual 18 end date to this obviously takes it out of the working 19 group's hands because they don't actually process the 20 Reg Guides. But I think that we find it important, 21 just as I think the Committee finds it important as 22 well. So, I can tell you that.
23 MR. HISER: Yes, I guess, you know, the 24 point I'd make from the timeline, 31 years ago we 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
43 revised the Reg Guide. This is not something we do 1
frivolously or every five years or 10 years. So, this 2
one, we really want to make sure that we get it right 3
because this may be the last opportunity that we have.
4 CHAIRMAN RICCARDELLA: There will be more 5
data another 10 years from now.
6 MEMBER KIRCHNER: There will be more data 7
from your capsules, yes.
8 MR. HISER: Compared to 1900 data points, 9
there won't be too many. But the one advantage is a 10 lot of that data will be high fluence. So, it will 11 help us to understand the trends that Dan showed.
12 MEMBER KIRCHNER: And plant-specific with 13 material, understanding the materials at high fluence.
14 MR. HISER: Yes, that's correct.
15 And I think that was the end of our 16 presentation.
17 MEMBER KIRCHNER: Well, Allen, may I ask 18 a question? How are you going to determine change in 19 through-wall crack frequency as a function of shift in 20 RTndt? My question is, are you going to take samples 21 that are irradiated, and then, test them? How does 22 one go about developing a database to change the 23 estimate of frequency of through-wall cracks?
24 MR. HISER: Now this will be using 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
44 probabilistic fracture mechanics, the FAVOR code.
1 CHAIRMAN RICCARDELLA: Let me just say, 2
through-wall crack frequency under the PTS analysis is 3
sort of a surrogate for vessel failure.
4 MEMBER KIRCHNER: I know, yes.
5 CHAIRMAN RICCARDELLA: So, they're doing 6
it. They're not saying "observing through-wall 7
cracks". They're saying, if we have a thermal-shock 8
event --
9 MEMBER KIRCHNER: Right.
10 CHAIRMAN RICCARDELLA: -- what's the 11 probability of a vessel fracture?
12 MEMBER KIRCHNER: Right.
13 CHAIRMAN RICCARDELLA: Really is what it 14 says. That's what they mean. It's not an NDE kind of 15 thing. Now it makes some assumptions about the 16 existence of flaw distribution, of a preceding flaw 17 distribution --
18 MEMBER KIRCHNER:
- Right, flaw 19 distribution. Right.
20 CHAIRMAN RICCARDELLA: -- in the vessel, 21 but they're very small compared to through-wall.
22 MEMBER KIRCHNER: It's not my field. But 23 it's a question that's motivated in part, just 24 technical curiosity. I'm just trying to think, so on 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
45 what basis now are you going to change this frequency 1
curve, or however you capture this?
2 MR. HISER: What the calculations will do, 3
we'll look at defining a pressure-temperature limit 4
curve that's allowable in accordance with ASME 5
requirements.
6 CHAIRMAN RICCARDELLA: Right, with the 7
code, yes.
8 MR. HISER: Right, with the FAVOR code.
9 Then, we'll assume that is actually an error by 50 10 degrees. The real material is -- or some number of 11 degrees -- so, the real material is actually more 12 brittle than what is assumed in the P-T limits.
13 CHAIRMAN RICCARDELLA: Right.
14 MR. HISER: And therefore, we will run the 15 FAVOR code to see what the through-wall crack 16 frequency is for that -- I hate to use the word 17 "error" -- but that miscalculation of delta T. And as 18 you do that over a range of delta T incorrect 19 calculations, then you will understand what the impact 20 is.
21 MEMBER KIRCHNER: But how is the FAVOR 22 code benchmarked and validated to use for this 23 purpose?
24 MR. WIDREVITZ: So, the FAVOR code, as I 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
46 think some of us have experienced it, is a wildly 1
complicated code. So, to answer in detail, I think we 2
would have to think about it. But the short answer is 3
a lot of similar calculational work was done with 4
FAVOR as the basis for the 50.61(a) rulemaking. And 5
so, we would likely be building on that schema to 6
produce these results. So, this isn't like page 7
engineering we're doing here.
8 MEMBER KIRCHNER: Maybe we have some 9
additional input.
10 MR. TREGONING: This is Rob Tregoning from 11 Research.
12 We've done extensive validation efforts 13 over the years on FAVOR. I would argue that it's not 14 wildly complicated. We have codes that we use in the 15 agency that are much more complicated than the FAVOR 16 code that we base our decisions on.
17 (Laughter.)
18 And we've validated it in a variety of 19 ways. We've had small-scale testing and experiments, 20 large-scale testing and experiments, experiments 21 specifically designed to simulate thermal-shock 22 events, experiments with lower transients. And then, 23 there's been extensive round robin benchmarking 24 internationally. Other countries have similar codes.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
47 We've had a number of benchmarking and international 1
round robin exercises that have been used to benchmark 2
the FAVOR code. And currently, now we are doing an 3
official, more rigorous sort of Part 50 V&V effort, so 4
that we can bring our verification documentation and 5
quality assurance up to typical regulatory standards.
6 CHAIRMAN RICCARDELLA: Okay. Thank you.
7 Just another question. As I understand 8
it, the FAVOR code or what was used for the PTS rule 9
used a different correlation than Reg Guide 1.99, Rev 10
- 2. How does that compare to the E900? I mean, is it 11 about the same? Is it more conservative, less 12 conservative than --
13 MR. WIDREVITZ: To handle this one, over 14 the span of which the base data that both interpolate, 15 the two agree fairly well. However, E900 has a lot 16 more data. It has international data. It has some 17 higher fluence. But over the span where they're both 18 interpolating over the same data, they compare quite 19 closely. Statistically speaking, if you saw the 20 statistics in that range, you wouldn't be able to 21 guess which one was which.
22 MR. IYENGAR: May I just add to what Rob 23 had mentioned? This code FAVOR is not only our 24 flagship code, it's one of the best codes that's been 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
48 used extensively by not only us, but also industry.
1 And there's been a lot of good comments on the use of 2
the code.
3 I do want to tell you that I just fully 4
agree with what Rob mentioned, and these recent 5
efforts that we are undertaking actually had enabled 6
our staff to run these codes across the agency better.
7 And we have even a developer training for this code 8
where some of our staff can actually intensely develop 9
and change stuff, if the need arises.
10 And I do want to tell you, also, DOE has 11 asked us to use this code to benchmark the Grizzly 12 code that they are using for similar type of analysis.
13 And the Japanese have, with Mark Kirk, the lead 14 earlier on, benchmarked this code against the Pascal.
15 And so, if there is any code that has been on the 16 component integrity side using probabilistic fraction 17 mechanics that has been used extensively and been 18 fruitful, it's FAVOR, I think No. 1 choice. And next 19 comes xLPR. Just so that I don't want to make Dave 20 Rudland happy here.
21 (Laughter.)
22 MEMBER MARCH-LEUBA: This is thinking 23 outside the box, somebody that doesn't really know 24 what's going on in this area. But I understand the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
49 industry is really concerned that you are reopening 1
the Reg Guide and doesn't know how it is going to hit 2
them. I mean, if I was a plant operator, I would be 3
scared. Okay?
4 CHAIRMAN RICCARDELLA: Industry is going 5
to be talking.
6 MEMBER MARCH-LEUBA: I know. I know. But it 7
seems to me, by looking at your timeline and 8
everything, that is only going to apply to one or two 9
PWRs well into the thing. And what I'm thinking is, 10 could this be resolved by a penalty for those two 11 plants?
12 MR. HISER: Well, that's partly what we 13 want to assess. The couple of plants that were 14 identified on the timeline are those that have high 15 fluence.
16 MEMBER MARCH-LEUBA: Yes.
17 MR. HISER: So, we know that they're in an 18 area that on Dan's charts would indicate that there's 19 a potential if you use non-conservative predictions --
20 MEMBER MARCH-LEUBA: But I thought you 21 said --
22 MR. HISER: What we want to do is 23 understand, if we were to implement this, what is the 24 impact on plants.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
50 MEMBER MARCH-LEUBA: I thought you said 1
that one of the plants is not really there because 2
it's limited by the welding, which really is not 3
effective.
4 MR. HISER: Correct.
5 MEMBER MARCH-LEUBA: I don't know. I'm 6
typically not an industry applicant, but if I was in 7
this area, I would feel more confident if, instead of 8
reopening everything, we just give a penalty to the 9
two or three cases where it really matters.
10 CHAIRMAN RICCARDELLA: But do we want to 11 perpetuate a 30-year-old technology in our Reg Guides?
12 MEMBER MARCH-LEUBA: I'm just putting it 13 out there.
14 MR. HISER: I don't think this is -- I 15 mean, we have the same concerns as the industry. I 16 mean, we don't want to penalize plants if their 17 predictions are reasonable currently. From what Dan 18 showed up to maybe 3 or so times 10 to the 19th, we 19 seem to get good predictions from Reg Guide 1.99. One 20 possibility would be to say, if you're above that 21 fluence, you need to use this other model. I mean, 22 there are many ways that this can be implemented, and 23 this is not something that staff will just edict. We 24 have a public meeting the middle of May that we're 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
51 looking at. We will have lots of public input and 1
industry input on this.
2 MEMBER MARCH-LEUBA: My concern is that we 3
might elevate this to too much of a problem, when it 4
really may not apply to everybody.
5 MR. WIDREVITZ: And I just wanted to just 6
provide a little bit of the new reactors perspective, 7
because, until very recently, that's the only thing I 8
was doing. And we are here in the presentation -- you 9
know, our primary concern is the operating fleet. It 10 exists. It's real. They're moving these very high 11 fluence zones. As a new reactor person, Reg Guide 12 1.99 gives me very, very inaccurate predictions, to 13 the point that I really don't know from the Reg Guide 14 what's happening because the Reg Guide was concerned 15 with a different body of chemistries.
16 And so, it really isn't just these high 17 fluences where the Reg Guide is going to produce bad 18 results; it's also things like the low copper modern 19 materials that all new plants, including the AP1000, 20 are going to use. And so, whether that's significant 21
-- you know, I think Allen put it very well -- like 22 how many plants? How significant is that going to be?
23 But we have to think about the full spectrum of 24 inaccuracies.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
52 CHAIRMAN RICCARDELLA: And in most cases, 1
it's overburdensome. It can be overburdensome.
2 MR. WIDREVITZ: Well, for new reactors, 3
it's a large underprediction, but their ultimate shift 4
might not be that serious anyway because their 5
materials are so good.
6 CHAIRMAN RICCARDELLA: Right.
7 MR. WIDREVITZ: So, it doesn't matter if 8
your prediction is that far off. And it's not a 9
single modal problem.
10 CHAIRMAN RICCARDELLA: Yes. Well, I think 11 we're about 10 minutes behind schedule and we have --
12 MR. RUDLAND: Just a quick question, Pete, 13 or just a quick comment. And I think that what you 14 were pointing out is that one of the reasons why we 15 want to do this assessment is because there could be 16 benefit to going to an updated trend curve, also. It 17 could actually help industry save money. So, we want 18 to make sure that we assess the whole range of whether 19 it's going to be a safety problem, but also it may be 20 a benefit. So, we may want to make it a voluntary 21 effort for them to do that if they can get a benefit 22 out of it.
23 CHAIRMAN RICCARDELLA: Okay.
24 MS. KHANNA: So, Meena Khanna.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
53 Go ahead. Do you want to say your name?
1 MR. RUDLAND: Oh, sure. I'm Dave Rudland.
2 MS. KHANNA: So, just to add onto what 3
Dave Rudland said -- and I think this has been a 4
really good discussion -- I think that's going to be 5
the challenge for the staff, right? They're going to 6
be coming up with options for management to consider.
7 We also have to be aware of the fact that we have 8
forefitting and backfitting guidance out there now 9
that the Commission has relayed to us. So, we'll have 10 to factor that. And as Dave mentioned, there may be 11 opportunities where the Commission -- I'm sorry --
12 where the licensees will be able to get some benefit, 13 but, then, we'll have to look at the other venue, too.
14 We'll have to look at, if we're adding 15 additional burden to the licensees, we'll have to go 16 through the Commission's policy now on MD 8.4. So, 17 that's all going to be part of the assessment. So, 18 great questions. Please know that that will be 19 considered thoroughly by management before any kind of 20 recommendation goes forth.
21 CHAIRMAN RICCARDELLA: Okay. So, do we 22 want to change speakers and listen to what EPRI has?
23 MR. HARDIN: Good afternoon, everyone.
24 My name is Tim Hardin. I'm with the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
54 Materials Reliability Program at the Electric Power 1
Research Institute, and appreciate the opportunity to 2
come and share our perspectives on this very important 3
issue.
4 I wanted to talk about a number of things.
5 Hearing the staff's immediately prior presentation, it 6
was very interesting that we were not aware of the 7
positions that the staff was going to take. And so, 8
I think there's a lot of overlap, and I don't need to 9
dwell on some of these issues.
10 I do want to talk about the upper shelf 11 energy prediction model, the shift model, or the delta 12 T 41J model and some of the scoping study that we've 13 done on the impact that it will have on materials when 14 you go from Reg Guide 1.99, Rev 2, shift prediction 15 model, to the E900 shift prediction model.
16 I do want to stress that this presentation 17 is from the EPRI staff. It should not be regarded as 18 representing an industry position of the U.S.
19 utilities, although they have been briefed on it.
20 So, in July, with the release of the TLR, 21 EPRI reviewed the TLR and provided some comments to 22 this Subcommittee on August 22nd. And since that, as 23 I mentioned, we have studied the potential impact, and 24 I want to summarize that in the next slides.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
55 Just briefly, we found the TLR to be 1
succinct and thoroughly documented and the 2
recommendations to be supported by the analyses that 3
were presented in the TLR. We did have some comments, 4
specific areas that we pointed out, some disagreement, 5
but, by and large, the conclusions we think are 6
accurate.
7 But we also thought that the TLR and the 8
data in the TLR support a phased implementation of any 9
change to the Reg Guide. And the conversation that we 10 just had in here was very much, I think the thinking 11 is along those lines.
12 So, on the upper shelf energy prediction 13 model, what we were going to say is that, you know, in 14 the August 22nd meeting, we mentioned that EPRI has 15 done some research in this area. And we developed a 16 revised upper shelf energy decrease prediction model.
17 It was not because we thought the Reg Guide was not 18 conservative so much as we were concerned with the 19 lack of technical basis for the existing model in the 20 Reg Guide.
21 And so, we developed this new prediction 22 model in MRP-414 that bounds 2 sigma of the data.
23 But, then, when we evaluated the impact of this new 24 model on the fleet, we found that at 80 years the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
56 number of vessels that have materials that are 1
predicted to fall below 50 foot-pounds is about the 2
same, if not exactly the same, as under the Reg Guide.
3 So, it wasn't clear that adopting this new model would 4
really achieve anything. And so, the only point I 5
wanted to make was we do not believe there is any 6
benefit, a significant safety benefit, to revising the 7
upper shelf model, and we don't think it should be.
8 And it sounds like there's agreement there from the 9
staff, since this is not an area of focus in their 10 current effort.
11 Now, on the shift prediction model, we 12 note that the staff, as we heard -- and I have quoted 13 some lines out of the TLR here -- that the non-14 conservatism of the Reg Guide starts in the area from 15 3 to 6 times 10 to the 19th. It seems that the 16 emphasis today has been on 6 times 10 to the 19th, 17 which we would understand and agree with.
18 And because the Reg Guide is used for 19 embrittlement predictions to show compliance with 20 10 CFR 50, Appendix G, then the appropriate metric to 21 look at in terms of deciding who is affected, and who 22 needs to implement a potential change, would be the 23 1/4T fluence for the vessel, because that's the 24 fluence that is used in generating pressure-25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
57 temperature limit curves, per ASME Section 11, 1
Appendix G.
2 So, based on that, TLR's assessment of 3
where the Reg Guide starts becoming non-conservative, 4
if you look at the BWR fleet, we have this report that 5
was recently generated, and I believe the staff have 6
it for review, BWRVIP-321. It's the plan for 7
extension of the integrated surveillance program into 8
the second license renewal period.
9 Most of the plants, the BWRs, have 1/4T 10 fluence, and the range is about 1 to 5 times 10 to the 11 18th. And some of the higher fluence BWRs have 12 fluences as high as 9 times 10 to the 18th. But 13 that's substantially below the fluence at which the 14 Reg Guide starts becoming non-conservative at 3 to 6 15 times 10 to the 19th. So, we do think that the BWR 16 fleet could be safety exempted from any change, and 17 they could continue to use the current Reg Guide shift 18 prediction model through the 80-year life.
19 PWRs, of course, fluence is considerably 20 higher and there are going to be some that are 21 impacted in that area of where the Reg Guide is 22 potentially non-conservative. Now, in order to make 23 a prediction of what the materials would, you know, 24 the shifts would be, you need to know what 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
58 embrittlement prediction model is going to be used.
1 And coming into this meeting, we did not know what the 2
staff was going to recommend. So, we thought the two 3
logical choices would be either E900-15 or EONY, which 4
is the trend correlation that's adopted in the 5
alternate PTS rule.
6 So, we made a chart here comparing the 7
two. E900, developed in 2015, an international 8
dataset of 1878 data. The result of a consensus 9
process; whereas, EONY was developed 10 years before 10 with only 855 U.S.-only data, as a result of an NRC 11 contractor effort.
And so, our preferred 12 recommendation would be for ASTM E900-15, and I see 13 that the staff agrees with that.
14 So, what we've done is we've assumed that 15 it's E900-15 and taken a look at what happens to 16 specific materials if you change from a prediction 17 using the Reg Guide to E900. So, on the following 18 plots we have plotted the difference, where we take 19 the prediction using E900 and subtract the prediction 20 of the shift using the Reg Guide. And so, a positive 21 difference above zero would mean that there would be 22 an increased shift due to the use of E900; a negative 23 difference means that the shift would decrease and 24 that E900 would provide relief in comparison to the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
59 Reg Guide.
1 We looked at 80 years of operation.
2 That's the fluence that we assumed. And we looked at 3
two different groups, high copper and low copper, 4
since they behave very differently.
5 So, this is the first. This is the low 6
copper comparison, where, again, you have anything 7
above the zero line indicates that, if you went from 8
the Reg Guide to E900, the shift would increase, and 9
below the line it means that the prediction would be 10 reduced.
11 So, the first takeaway is the base 12 materials, the plates and foragings, would pretty much 13 see an increase across the board, but the welds would 14 see a mixed story. So, there would be some 15 improvement and there would be some penalties for 16 those materials.
17 And again, this is the mean shift plus 2 18 sigma. And from 18 to 19 is the BWRs, and higher than 19 19 is the PWRs. This is the low copper. For the high 20 copper --
21 CHAIRMAN RICCARDELLA: Each of these are 22 actual plants at 80 years or?
23 MR. HARDIN: This is projected --
24 CHAIRMAN RICCARDELLA: Projected.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
60 MR. HARDIN: -- fluences at 80 years.
1 CHAIRMAN RICCARDELLA: But each data point 2
represents a plant, basically?
3 MR. HARDIN: Yes.
4 CHAIRMAN RICCARDELLA: Okay.
5 MR. HARDIN: From the surveillance 6
database.
We took that material and just 7
extrapolated.
8 And then, for the high copper steels, a 9
little bit of the same story. Generally, the plates 10 increasing across the board. There are a few plates 11 that would benefits. But the welds are very much 12 mixed. A lot of the PWR welds would see a significant 13 benefit if they went from the Reg Guide to E900. So, 14 I offer that information for consideration.
15 Now, considering how this would impact a 16 specific vessel and its P-T curves is a little bit 17 more complicated, because, as we saw in those plots, 18 some materials are going to increase in shift; some 19 are going to decrease, and often that will happen 20 within the same vessel. So, you're not going to know 21 what the impact on the vessel is until you sit down 22 for each material and do the calculation, and then, 23 you take the highest adjusted reference temperature 24 for that vessel and, then, compare it to the adjusted 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
61 reference temperature that is currently used as the 1
basis for the P-T curves, the current P-T curves.
2 So, the vessel-specific evaluations have 3
not been done. And we can't do them really until, 4
well, I said here until we know what prediction model 5
we're going to use, but now I think we know that, 6
E900. And also, as the staff alluded in their 7
presentation, how is surveillance data going to be 8
accounted for? And so, until we know the proper 9
procedures to use for that in the new Reg Guide, we 10 won't be able to sit down probably and do a fully-11 informed vessel-specific evaluation.
12 But what's clear is that performing that 13 evaluation, a plant going out and having to do that 14 evaluation would be a significant burden. And if 15 their 1/4T fluence is below the level at which the Reg 16 Guide is non-conservative, then there's really no 17 benefit to that. So, we're offering that for 18 consideration, some sort of metric at the 1/4T for 19 screening on who has to take action and who does not 20 take action.
21 So, the conclusion and recommendations.
22 No change is recommended to the upper shelf energy 23 mode. E900-15 is the preferred alternative shift 24 model. And BWRs appear to be able to use the existing 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
62 Reg Guide shift model through 80 years of operation; 1
PWRs would, of course, have to be screened according 2
to their 1/4T fluence. And it will be helpful to the 3
staff, to the plants, as they assess -- you know, many 4
of them are considering second license renewal. Well, 5
they need to look ahead and see what their operating 6
envelope is going to be. And so, a timely 7
understanding of what these new requirements are going 8
to be will help their efforts in deciding if it's 9
feasible to go for a second license renewal.
10 I think that's all I had. Are there any 11 questions?
12 MEMBER MARCH-LEUBA: Yes, I have one.
13 MR. HARDIN: Yes, sir?
14 MEMBER MARCH-LEUBA: What type of burden 15 is there for a plant if the Reg Guide is revised? Is 16 it a burden on the paperwork to calculate the new 17 numbers and send it to the NRC? Or the burden is 18 because, if they calculate a number that is bad, they 19 have to change their core and modify their peaking 20 factors and --
21 MR. HARDIN: No, it's not a matter of 22 changing the core, but it's a very expensive 23 calculation.
24 MEMBER MARCH-LEUBA: So, the burden is on 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
63 the paperwork and --
1 MR. HARDIN: And then, what happens is, 2
you know, you not only have your P-T curve, so you 3
give it some administrative paperwork of your P-T 4
curves, but, then, you take those P-T curves that you 5
get from your vendor that does the calculation, and 6
then, you have to change plant procedures and you have 7
to conduct training of the plant personnel to 8
implement those new --
9 MEMBER MARCH-LEUBA: Say that I'm a BWR.
10 I'm in the 5 times 10 to the 18.
11 MR. HARDIN: Yes.
12 MEMBER MARCH-LEUBA:
I'm going to 13 calculate numbers with a new correlation. They're 14 going to be the same as the old one, hopefully. Then, 15 what is the burden?
16 MEMBER PETTI: I don't think that's -- no, 17 that's not the point. He said the BWRs aren't 18 affected. The changes will only affect PWRs, a subset 19 of which -- and we really won't know because of the 20 different materials until you actually have to -- you 21 have to try before you buy.
22 MEMBER BALLINGER: And how often are the 23 P-T limit curves calculated on a kind of sort of 24 routine basis for a plant? In other words, at some 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
64 point during the going forward, would there be a case 1
where a new P-T limit curve would be required in any 2
case? And so, this would all get factored into that?
3 MR. HARDIN: Well, certainly the P-T 4
curves are recalculated whenever a plant withdraws a 5
surveillance capsule, and then, has to consider that 6
data.
7 MEMBER BALLINGER: So, is that the same 8
calculation that would be done?
9 CHAIRMAN RICCARDELLA: Yes. I mean, 10 they're going to -- I'm trying to understand the 11 difference. If they're going to do a calculation when 12 they pull a surveillance capsule, whether they use the 13 old trend curve or the new trend curve, there's no 14 more of a paperwork burden, right?
15 MR. HARDIN: That's right. That's right.
16 CHAIRMAN RICCARDELLA: The paperwork 17 burden is about the same.
18 MR. HARDIN: It would appear to make sense 19 to use --
20 CHAIRMAN RICCARDELLA: Yes.
21 MR. HARDIN: -- the new one, if you're --
22 MEMBER BALLINGER: So, is it safe to say 23 that all the plants would have withdrawn, have had 24 occasion to withdraw a capsule between now and going 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
65 forward?
1 MR. HARDIN: I believe that is not 2
generally true.
3 MEMBER BALLINGER: Not true?
4 MR. HARDIN: Well, to withdraw a capsule?
5 CHAIRMAN RICCARDELLA: Yes, sir.
6 MR. HARDIN: Some plants have P-T curves, 7
I am informed, I am told, that are good through the 8
end of license.
9 MEMBER BALLINGER: Okay.
10 CHAIRMAN RICCARDELLA: Yes, yes, yes.
11 MEMBER BALLINGER: So, that would be a 12 burden?
13 CHAIRMAN RICCARDELLA: But, for license 14 renewal or subsequent license renewal, the ones we've 15 reviewed, they have a schedule to pull a capsule, 16 right?
17 MR. RUDLAND: This is Dave Rudland.
18 It really depends on the plant and whether 19 or not they have pulled their 80-year capsule or not, 20 whether they have pulled a capsule that represents 21 their fluence at 80 years. If they have, then there's 22 no reason for them to pull another.
23 MEMBER MARCH-LEUBA: Right, but if you 24 issue the new reg, the new revision of the Guide, will 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
66 everybody have to do the new P-T curve?
1 MR. RUDLAND: And that's to be determined.
2 I mean, that's what we're doing as part of our safety 3
impact, is determine how we want to implement this 4
thing --
5 MEMBER MARCH-LEUBA: Yes.
6 MR. RUDLAND: -- whether it's for one 7
plant or 94 plants. What does it need to be for? We 8
don't know that yet.
9 MEMBER MARCH-LEUBA: That's something that 10 the staff needs to -- I mean, for example, for BWRs 11 and they are 5 times 10 to the 18, they don't need to 12 do anything.
13 MR. RUDLAND: That's right. So, we have 14 to do a safety assessment. If there's no safety, then 15 there's probably no reason to implement.
16 MEMBER MARCH-LEUBA: Yes.
17 MR. RUDLAND: Or I should say there's no 18 reason to mandate that it's being used.
19 CHAIRMAN RICCARDELLA: Reg Guides are 20 mandatory? Are they? They're guidance.
21 MEMBER BALLINGER: 10 CFR 50, Appendix 22 G --
23 CHAIRMAN RICCARDELLA: Huh 24 MEMBER BALLINGER: Appendix G is.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
67 CHAIRMAN RICCARDELLA: Yes, yes. But does 1
Appendix G direct you to 1.99, Rev 2?
2 MEMBER BALLINGER: Yes.
3 CHAIRMAN RICCARDELLA: Oh, okay.
4 MEMBER KIRCHNER: Sometimes things in Reg 5
Guides are incorporated in the -- I'm not the expert 6
on this, but they will be taken almost verbatim, and 7
then, put into like 10 CFR 50 regulations. And then, 8
they have -- yes, right.
9 MR. HISER: Yes, this is Allen Hiser.
10 And that's what was done in the early '90s 11 with the PTS rule, was that the procedure in Reg Guide 12 1.99, Rev 2, was put into the rule. Now the question 13 about -- Reg Guides are guidance documents. They're 14 not regulation. Reg Guide 1.99, Rev 2, was actually 15 implemented using a Generic Letter that asked plants, 16 how will you predict embrittlement? Will you use Reg 17 Guide 1.99, Rev 2, or will you use some other 18 approach? Virtually I think everybody at that point 19 used Reg Guide 1.99, Rev 2. There have, subsequently, 20 been plants that have used, let's see, the master 21 curve approach, and they've used -- no, not.61(a).
22 I think some of the plants have developed their own 23 curves, like the BMW (sic) Owners Group developed 24 curves that were used.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
68 So, the Reg Guide 1.99, Rev 2, is not 1
required, but plants chose to implement it through the 2
GL process.
3 MR. HARDIN: I'm not aware of any plant 4
that does not use the Reg Guide 1.99, Rev 2, for 5
prediction of embrittlement. I understand the point 6
about Reg Guides being guidance, but this particular 7
Reg Guide is as close to being a rule as it can get.
8 I just wanted to also correct an earlier 9
statement, with all due respect to Dr. Hiser, that the 10 PWR Supplemental Surveillance Program is an EPRI 11 Materials Reliability Program, a program, not a BWR 12 Owners Group program.
13 CHAIRMAN RICCARDELLA: Okay. Well, thank 14 you, Tim, for your presentation. You effectively got 15 us pretty close to back on schedule.
16 So, with that, I'm going to open the 17 meeting for public comments. So, if someone will 18 check to make sure the phone line is open? And while 19 you're doing that, I'll check to see, is there anybody 20 in the room from the public who would like to make a 21 comment?
22 (No response.)
23 Hearing none, we'll go to the phone line.
24 Is there anybody on the public phone line? If so, 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
69 would you speak up, please, and just let us know 1
you're there?
2 (No response.)
3 It sounds like nobody is there. So, I 4
assume nobody wants to make a comment.
5 You can close the line.
6 So, we have some time for Committee 7
discussion. Ron, do you want to lead that?
8 MEMBER BALLINGER: I think we pretty much 9
covered everything that we thought we would be putting 10 in a letter of some kind. We have a draft letter and 11 will be producing another draft tonight --
12 CHAIRMAN RICCARDELLA: Yes.
13 MEMBER BALLINGER: -- for consideration of 14 the Committee. But that's about all I would have.
15 Everything has been covered pretty accurately.
16 MEMBER BLEY: Do you want to summarize 17 here what the main points you would be making in the 18 letter are?
19 MEMBER BALLINGER: The main points -- and 20 actually, the staff, unbeknownst to myself, the staff 21 has actually come around to our conclusions. And that 22 is that, (a) we should do this. We should reconsider, 23 we should revise the Reg Guide. We should revise it 24 on a schedule which is defined. We should also -- we 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
70 can't say much towards implementation, but we suggest 1
that the implementation would be in a very phased 2
approach, so that plants could pick and choose where 3
it was essential that they needed it.
4 But the fact that we have truth nowadays, 5
as opposed to 187 data points, we have 1800 data 6
points, is a sufficient reason for revising the Reg 7
Guide, especially when you consider subsequent license 8
renewal going forward. Now is the time to do it.
9 MEMBER BLEY: From what I heard, I think 10
-- I think from what I heard from both the staff and 11 EPRI, it is that no existing plant in a spot at this 12 point in time where the current approach would be non-13 conservative, such that nobody would have to do 14 anything right now.
15 MEMBER BALLINGER: If you choose 6 times 16 10 to the 19th.
17 MEMBER BLEY: Yes.
18 MEMBER BALLINGER: If you choose 3, then 19 there are plants that are on the edge. Now, if you 20 switch to the 1/4T --
21 MEMBER KIRCHNER: Yes, I was going to ask, 22 just for clarification, if you switch to 1/4T --
23 MEMBER BALLINGER: Yes, I'd have to think 24
-- probably not. If we go to 1/4T, then there's 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
71 probably 8 or 10 years before -- I'm just guessing now 1
whether Turkey Point would be. Maybe the staff would 2
know or the EPRI folks would know.
3 MEMBER KIRCHNER: And the current RG 1.99 4
is inner-surface.
5 MEMBER BALLINGER: Yes, that's inner-6 surface.
7 MEMBER KIRCHNER: Right.
8 MEMBER BALLINGER: Yes.
9 CHAIRMAN RICCARDELLA: There was that 10 staff flowchart where the arrow got -- the timeline 11 where the arrow got lost. I thought that had 1/4T 12 and --
13 MEMBER BALLINGER: Yes, but that was 6 14 times 10 to the 19th.
15 CHAIRMAN RICCARDELLA: Yes.
16 MEMBER BALLINGER: Yes, the fact of the 17 matter is 1.99, Revision 2, that correlation goes 18 awry.
19 CHAIRMAN RICCARDELLA: At the inside 20 surface Turkey Point hits 6 times 10 to the 19th at 21 2025, which is essentially five years from now, and at 22 1/4T, it hits it at 2044.
23 MEMBER BALLINGER: Yes, the 10th value 24 layer for neutrons is usually 2 inches, roughly 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
72 speaking. So, if you would want to go to 1/4T, you've 1
got to drop --
2 MEMBER KIRCHNER: And EPRI points out that 3
the code, the ASME code, to which, if I read their 4
viewgraphs correctly, Section 11 is 1/4T.
5 MEMBER BALLINGER: That's for PTS.
6 CHAIRMAN RICCARDELLA: No. No, no, no.
7 MEMBER BALLINGER: No?
8 CHAIRMAN RICCARDELLA: No. The 1/4T is 9
the size flaw you assume when you're developing your 10 heatup/cool-down curves.
11 MEMBER BALLINGER: Okay.
12 CHAIRMAN RICCARDELLA: For PTS, I'm not 13 sure what the original rule assumed. That just picked 14 a temperature. The alternate PTS rule uses a flaw 15 distribution and does a probabilistic fraction 16 mechanics analysis.
17 MEMBER BALLINGER: We're about to get 18 clarification here, I think.
19 MR. HISER: This is Allen Hiser.
20 Actually, PTS is the fluence at the clad-21 base-metal interface.
22 CHAIRMAN RICCARDELLA: So, that's a 23 quarter inch.
24 MEMBER BALLINGER: That's not 1/4T.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
73 CHAIRMAN RICCARDELLA: It's quarter inch.
1 (Laughter.)
2 MEMBER BALLINGER: It's quarter inch, yes.
3 MR. HISER: And.61(a) I believe uses the 4
same methodology. It's just within the probabilistic 5
fraction mechanics analysis it uses a
flaw 6
distribution. So, you still benchmark it to the clad-7 base-metal interface fluence.
8 MEMBER KIRCHNER: And for clarification, 9
Allen, that's how you apply 1.99, at the stainless 10 steel clad inside the ion emitter or where it hits the 11 base metal?
12 MS. KHANNA: It's the ID.
13 MR. HISER: Well, at the --
14 MEMBER KIRCHNER: ID, not the base-metal 15 clad? Okay.
16 MR. HISER: Yes, Reg Guide 1.99 has an 17 attenuation function. You can plug in a surface 18 fluence, and then, you can get the spectrum of 19 fluences and embrittlements through the wall. The PTS 20 rule uses the stainless steel-to-base-metal interface.
21 MEMBER KIRCHNER: And when individual 22 plants do their analysis, when you talked about that 23 fourth factor, do they typically at least take credit 24 for the stainless and go in a quarter inch and do an 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
74 attenuation?
1 MR. HISER: The P-T limits use a 1/4T 2
flaw. And for PTS, it's just the fluence at that 3
interface.
4 MEMBER KIRCHNER: Okay.
5 MR. HISER: There's no attenuation of it 6
through the vessel wall. It's just where it's 7
benchmarked to is all.
8 CHAIRMAN RICCARDELLA: Yes. Yes, because 9
you're concerned about thermal shock which affects 10 really small, you know, it affects flaws closer to the 11 inside surface.
12 MEMBER KIRCHNER: Yes.
13 CHAIRMAN RICCARDELLA:
For the 14 heatup/cool-down curves, you're mainly worried about 15 more gradual through-wall transients, and the 1/4T is 16 the size that you use. But I think what you're saying 17 is that the 1.99, that both the points on your time 18 chart were relevant, the surface one and the 1/4T.
19 MR. HISER: That is correct. The surface 20 one is more applicable to PTS calculations, in 21 accordance with 50.61. The 1/4T is for Appendix G, 22 10 CFR Part 50, for pressure-temperature limits.
23 CHAIRMAN RICCARDELLA: And I have to plead 24 guilty to being one of the people that introduced the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
75 1/4T flaw in Section 3.
1 (Laughter.)
2 MR. RUDLAND: This is Dave Rudland. I 3
just want to make another comment.
4 Dr. Ballinger keeps talking about 3E to 5
the 19. I think the staff spent a lot of time looking 6
statistically at the data, and, you know, well, the 7
mean trend may begin to deviate at 3; it really does 8
not become statistically significant until you get to 9
about 6. And so, we've based a lot of our assessments 10 on the 6 value, and at this point we strongly feel 11 that there's no immediate issue in terms of the safety 12 significance because of that assessment.
13 Now, again, you can argue about whether 14 the mean or the deviation from mean is significant or 15 not, but in the scope of the scatter of the data I 16 think we really think that, by the time you get to 6, 17 that you have a noticeable shift in the predictions.
18 CHAIRMAN RICCARDELLA: Because you're 19 using that 2 sigma. As I understand it, the 6 times 20 10 to the 19th is where the mean curve --
21 MR. RUDLAND: Goes beyond 2 sigma.
22 CHAIRMAN RICCARDELLA: -- it goes beyond 23 the 2 sigma.
24 MR. RUDLAND: And the 3 curve is just 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
76 where it begins to deviate from basically the zero 1
line, right.
2 CHAIRMAN RICCARDELLA: But I'm just 3
wondering if we ought in the process take a relook at 4
that 2 sigma.
5 MR. RUDLAND: We have. We have. We have 6
so far, at least in the data that we've analyzed, 7
whatever the baseline data had.
8 CHAIRMAN RICCARDELLA: Oh, I know there's 9
new data that will generate a new 2 sigma, but is 2 10 sigma the appropriate margin?
11 MR. RUDLAND: And I was just talking to my 12 old colleague, Dr. Kirk, here about risk-informing 13 those margins. And some of the analyses that we're 14 doing with FAVOR might be able to help us do that.
15 CHAIRMAN RICCARDELLA: Yes. Good. Okay.
16 Well, with that, we have -- look me look at the 17 overall schedule.
18 MEMBER MARCH-LEUBA: All the people for 19 the next presentation are next door. So, we just need 20 to tell them to come in.
21 CHAIRMAN RICCARDELLA: Should we take a --
22 we don't have a break scheduled.
23 MEMBER MARCH-LEUBA: No?
24 CHAIRMAN RICCARDELLA: No. All right.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
77 So, we're right on schedule, 2:30.
1 MEMBER MARCH-LEUBA: Well, it's barely 2
behind.
3 CHAIRMAN RICCARDELLA: So, we don't even 4
have to recess the meeting or do I --
5 MEMBER MARCH-LEUBA: Well, yes, recess the 6
meeting.
7 CHAIRMAN RICCARDELLA: Recess while we 8
change speakers.
9 (Whereupon, the foregoing matter went off 10 the record at 2:31 p.m. and went back on the record at 11 2:40 p.m.)
12 CHAIRMAN RICCARDELLA: Okay. So, we'll 13 start with, I'll turn the meeting over to Jose for 14 opening comments.
15 MEMBER MARCH-LEUBA: Okay. So, we are 16 going to be talking about Brunswick's transition to 17 Atrium 11 fuel using the Framatome methods. For the 18 members that were not in the Subcommittee, this is a 19 complicated review, a complex review. We can get 20 confused very easily because at the time Brunswick 21 submitted a license amendment request many of the 22 Framatome methods were not approved. While they have 23 been in the review, most of the methods from Framatome 24 have been approved generically, and we have seen them, 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
78 like the AURORA-B AOO or LOCA. So, we've seen all 1
those.
2 So, you'll be hearing about the plant-3 specific methods and the generic methods during the 4
whole presentation. That's what the issue is. Okay?
5 Go for it, Andy.
6 MS. ROSS-LEE: So, thank you.
7 Just to introduce myself, my name is M.J.
8 Ross-Lee. I am the new Deputy Director for the 9
Division of Safety Systems.
10 So, I want to thank you for the 11 opportunity today. As mentioned, we're here to 12 present the staff's evaluation of the Brunswick 13 license amendment request to allow the use of the 14 Framatome methodologies necessary to support 15 transition to Atrium 11 fuel design.
16 I'll be turning the rest of the 17 introduction over to Andy Hon, who is from the 18 Division of Licensing.
19 MR. HON: Thank you, M.J.
20 Good afternoon, Mr.
Chairman and 21 Distinguished Members of the ACRS.
22 I'm Andy Hon. I'm the Licensing Project 23 Manager for Brunswick within the NRR. Today I'm going 24 to give like a three-minute, quick overview in the 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
79 public session to check on the project and the status 1
of it.
2 For Brunswick Units 1 and 2, currently, 3
they are using Framatome Atrium XM10 fuel, which was 4
also approved to operate in the MELLLA+ domain. That 5
integrated both the analysis with GE and Framatome 6
last year, and the Committee endorsed that review.
7 The licensee submitted a license amendment 8
request about a year ago for adopting the new, 9
advanced Framatome methodology called Atrium 11. They 10 also requested us to a quick turnaround to support 11 their fuel-loading schedule in February next year.
12 This is the first application of seven new 13 Framatome analyses for this new fuel design. And most 14 of the new methodology actually has been approved just 15 in time for generic use in the Topical Reports. And 16 as far as the specific change request, it is to change 17 the references in the coder table in Chapter 5 of the 18 Tech Spec.
19 As far as what the staff did in the last 20 year, it's really, for those approved Topical Reports 21 they adopted, we reviewed mostly the limitations and 22 conditions to ensure that they are addressed properly 23 by the licensee. And for one plant-specific 24 stability analysis, we spent most of our energy on 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
80 that review. We identified the necessary license 1
conditions for implementation
- and, later on, 2
enforcement, if necessary.
3 Our team conducted two audits of the 4
analyses, one in February, one in March of this year.
5 As a result of the audit, we generated 32 requests for 6
additional information, all of which were addressed by 7
the licensee to our satisfaction.
8 And the license condition for the 9
methodology not in the Topical Report, we identified 10 the necessary license conditions. The licensee 11 responded and adopted those license conditions for 12 implementation, and we'll give you more details later 13 on.
14 In addition, our Office of Research 15 performed the
- phased, independent confirmatory 16 analysis of the FSI stability to ensure the analysis 17 by the licensee is consistent. We will give you a 18 summary of that result.
19 So, the last bullet is really the reason 20 we're here. At this point, we completed our safety 21 evaluations and generated a Draft Report submitted to 22 the Committee to review. We opted here to make a 23 decision, pending any significant comments from the 24 Committee that staff needs to address.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
81 So, that's the status of the project at 1
this point.
2 And the last slide, I'd like to recognize 3
the members of our team who contributed to this very 4
challenging, like M.J.
- said, very challenging 5
schedule, and we worked really hard to complete the 6
review by the one-year timeframe. Typically, it takes 7
two years. So, I would like to use the last slide to 8
recognize members.
9 So, with that, I'm going to save some time 10 and turn it over the time to our licensee to give a 11 non-proprietary summary.
12 MEMBER MARCH-LEUBA: I apologize. We are 13 going to be even more rushed than just today. Because 14 we started late and we need to finish a little earlier 15 because I would like the opportunity to read the 16 letter in the closed session to verify there is no 17 proprietary information.
18 So, if we can find the slides? Perfect.
19 So, we are still in open session. Please 20 go ahead.
21 MR. DeWIRE: All right. Good afternoon.
22 My name is Mark DeWire. I'm the Assistant 23 Office Manager, Shift, or the Senior License Holder 24 for the Brunswick Nuclear Station.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
82 I'm here to give you just a brief overview 1
of the Brunswick station. We are a General Electric 2
BWR-4 with Mark I containment. Began commercial ops 3
in 1975-1976 for Unit 2 and Unit 1 with an original 4
license of 2436-megawatt thermal. Completed extended 5
power uprate which brought us up to 2923 megawatts in 6
the mid-2000s. And we're currently on a 24-month 7
operating cycle. And as mentioned, we are on Atrium 8
10XM fuel with eight lead assemblies of the Atrium 11.
9 We did transition over to Framatome fuel in 2008-2009 10 for Unit 1 and Unit 2.
11 We are licensed for the increased core 12 flow. And as mentioned, we did complete the MELLLA+
13 last year, and we completed them both. And both units 14 are currently operating in the MELLLA+ region now.
15 The purpose of this is to do a Tech Spec 16 update to Section 565(b), which has our analytical 17 methods. We're going to be revising that section to 18 allowing Framatome methodologies to be used for Atrium 19 11 fuel. And we're pursuing the Atrium 11 fuel due to 20 improved fuel cycles. The 11x11 array reduces the 21 fuel duty by approximately 19 percent. It has 22 improved channel performance, improved fuel cycle 23 economics, and near and dear to me is the improved 24 debris protection features with BWR fuel.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
83 The following methodologies will be 1
removed from the section in the Tech Specs, which 2
would be the X-COBRA-T and the Cotransit 2 and the GE 3
detect-and-suppress confirmation density. And we will 4
be replacing those with the Framatome methodologies, 5
as mentioned, with the AURORA-B for the transient 6
analysis, LOCA, and control rod drop, and the best 7
estimate enhanced option 3 for the OPRMs.
8 And that's it for my section. I'll turn 9
it over to Mr. Mike Bloom.
10 MR. BLOOM: Good afternoon, everybody.
11 My name is Mike Bloom. I'm the Director 12 of Fuel Management and Design for Duke Energy. So, I 13 take care of the 11 reactors' core design and 14 licensing analysis. For this group, the interest will 15 be the Brunswick Unit 1 and Unit 2, which we work in 16 collaboration with Framatome on for our reloads.
17 On slides 8 and 9, you'll see a listing of 18 a number of Framatome reports and one Duke report that 19 are included in the license amendment request. And 20 these reports, these are the application of the 21 methodologies we talked about are going into 565(b) 22 and the MELLLA+ and power uprate domain.
23 MEMBER MARCH-LEUBA: Other than adding 24 these reports to the technical specifications, have 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
84 you made any modifications in the LAR? Does the LAR 1
include other modifications other than adding these 2
references?
3 MR. BLOOM: Sorry, I didn't catch that.
4 MEMBER MARCH-LEUBA: The LAR incorporates 5
these references in your Tech Specs?
6 MR. BLOOM: Yes.
7 MEMBER MARCH-LEUBA: Are there any other 8
modifications of any relevance?
9 MR. YODERSMITH: Yes, the name is Stephen 10 Yodersmith. I work for Duke Energy.
11 There as a Note F in one of the Tech Spec 12 tables that was added for DSS-CD. That note is being 13 removed. It was for first-time application of DSS-CD.
14 It allowed you to basically not arm the DSS-CD trip 15 function when you applied it for the first time. So, 16 since, one, we're operating with DSS-CD and, two, 17 we're removing it from the Tech Specs, that note is 18 coming out. It's a very minor change there.
19 And the only additional change, as Andy 20 mentioned, is we did add a license condition to our 21 operating license, Section B, to incorporate the best 22 estimate, and hence, Option 3 license conditions.
23 Those are the only changes.
24 MEMBER MARCH-LEUBA: Thank you.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
85 MR. BLOOM: Okay. I've move on to slide 1
- 10. So, on slide 10 is a list of some Brunswick Unit 2
1, Cycle 23, cycle-specific documents for just in the 3
reload design. These were provided. They are not 4
part of the LAR, but were provided as supplemental 5
information for the NRC to review, so they could see 6
how the methods were applied to an actual reload and 7
be able to do some comparisons between the results 8
from the cycle that was in the methodologies to the 9
actual cycle application.
10 Slide 11 talks a little bit about our 11 AURORA-B LOCA methodology that we're applying. As 12 Jose was talking about earlier, this particular 13 methodology was not approved generically at the time 14 we submitted our license application. However, it had 15 gone through quite a bit of review, as we already had 16 a Draft SE. So, in collaboration with the NRC, it was 17 agreed that we could reference that Draft SE and Draft 18 Generic Licensing Topical in our initial LAR.
19 And since then, of course, the final LAR 20 has been approved. And we circled back with any 21 additional changes from the Draft SE and made a 22 supplement in July to our LAR to capture those. And 23 again, we do appreciate the NRC's flexibility for 24 allowing us to do that, so we can maintain our 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
86 schedule.
1 On slide 12 is a discussion on our 2
stability method for both units. Currently, we use 3
the detect-and-suppress solution confirmation density, 4
otherwise known as DSS-CD. As we transition to Atrium 5
11 fuel for this upcoming outage, we will be 6
transitioning to the Framatome best estimate option 3, 7
with a confirmation density algorithm.
8 In doing so, we will maintain our current 9
confirmation density algorithm trip. And so, the OPRM 10 subpoints are going to be the same. So, there's 11 really no change for how the plant's going to respond.
12 Within the LAR, we did provide a sample 13 problem for the Atrium 11 fuel. And we followed that 14 up with, for one of the supplementary documents, the 15 Reload Safety Analysis Report, as a cycle-specific 16 application of that methodology to that unit.
17 So, 13, we'll talk just a minute about 18 ATWS-I. So, this is a fuel-type-dependent analysis.
19 So, right now, for the resident fuel, which is Atrium 20 10XM, it's a TRACG-based analysis. As we introduce 21 Atrium 11, we'll be moving to the RAMONA5 Framatome-22 based methodology for that.
23 Similar to the AURORA-B, that generic 24 methodology had not been approved at the time. And 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
87 so, in this case, it had made it quite as far down the 1
review process. We elected to do a Brunswick-specific 2
methodology application for that. Since then, the 3
ANP 10.346 has been approved, and the methodology that 4
we have is identical to the generic methodology.
5 On slide 14, it's basically the pathway 6
that got us here today. We submitted our LAR in 7
October 2018. We did have the two NRC audits that 8
were previously discussed in February on best estimate 9
option 3. And then, we followed up in March with an 10 audit over all the methodologies.
11 Since then, we've been working on the 12 reload in terms of one Cycle 23 reload documents and 13 providing those as they become available.
14 We responded to the set of RAIs in June, 15 and then, we provided that AURORA-B supplement in 16 July.
17 Just last month, with the best estimate 18 option 3 methodology, there were some additional 19 licensing conditions that were applied through an RAI.
20 We responded to that in October and agreed to put 21 those into our license.
22 And now we're here in November with the 23 ACRS full Committee.
24 MEMBER MARCH-LEUBA: I must say that, for 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
88 an LAR of this complication, the first application of 1
methods and plant-specific methods, having it reviewed 2
in this short time, I know for you it doesn't look 3
very short, but it is pretty short.
4 MR. BLOOM: Oh, I can certainly appreciate 5
that.
6 MEMBER MARCH-LEUBA: Both the staff and 7
you should be congratulated for having done this job.
8 MR. BLOOM: Thank you.
9 And so, like you said, we are looking for 10 the LAR approval here in February, and that's going to 11 support our startup of Brunswick I, Cycle 23, in March 12 of 2020.
13 That was all I had to present, pending any 14 questions.
15 MEMBER MARCH-LEUBA: Do the members have 16 any questions in the open session?
17 (No response.)
18 Any members of the public want to present 19 a question?
20 (No response.)
21 Anybody on the open phone line, would they 22 like to do a question?
23 Is the open line open right now?
24 Questions and comments, yes, sorry.
25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
89 Comments?
1 (No response.)
2 Yes, okay. So, hearing none, there are no 3
comments from the public.
4 Any members want to say any final words in 5
the open session?
6 (No response.)
7 So, can you hit the gavel, Pete?
8 (Whereupon, at 2:55 p.m., the Committee 9
went from open session to closed session.)
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
Regulatory Guide 1.99, Rev. 2 Assessment Results and Follow-on Activities November 6, 2019
Introduction
- This presentation summarizes a multi-year effort to thoroughly assess RG 1.99 through a modern data-driven approach, plus the activities underway to address the assessment results.
2
Todays Presentation Regulatory Guide 1.99 Background Assessment Results
- RTNDT results
- USE results
- Credibility Criteria/Plant-Specific Data results
- Attenuation results
- Common additions
- Conclusions RG 1.99 Working Group Charter and Activities RG 1.99 Oversight Group Charter and Activities 3
Regulatory Guide 1.99 Background Radiation Embrittlement of Reactor Vessel Materials Rev. 2 issued May, 1988.
- RTNDT curves fit to 177 datum. Few data at higher fluences or with lower Cu.
- USE developed as upper-bound curves. Acceptance criteria for USE is in 10 CFR 50 Appendix G.
- The basis for the credibility criteria and use of surveillance data is not well understood.
- Attenuation formula based on dpa studies.
RTNDT - A measure of brittle-ductile transition temperature (considered equivalent to T41J);
USE - Upper Shelf Energy 4
Regulatory Guide 1.99 Background
- ART = unirradiated RTNDT + RTNDT + M (Where RTNDT = shift in RTNDT due to irradiation and M = margin term)
- ART is used to determine fracture toughness when developing pressure-temperature limits.
- The Pressurized Thermal Shock Rule, 10 CFR 50.61, uses the same RTNDT model as RG 1.99.
5
RTNDT Results 6
Limited weld data at high fluence precludes assessment of whether weld trends with base metals at high fluences.
Assessment based on BASELINE dataset generated by ASTM E10.02.
Dataset includes domestic and international power reactor data (~55%
domestic).
1901 data points In-depth statistical analysis performed.
RTNDT Results
- Primary conclusions:
- Nonconservative high fluence results (base metals)*, becomes prominent at fluences 6x1019 n/cm2.
- Inaccurate low Cu results
- Secondary conclusions:
- Standard deviation of RTNDT () in RG is too low
- Conservative bias in low-to-mid fluences (burden)
- Lack of temperature adjustment (inaccuracy) 7
- Limited weld data available at fluences near or above 1x1020 n/cm2 (E > 1MeV)
USE Results 8
EMA - Equivalent margin analysis.
Limited number of materials are likely to be measured below 68J (50 ft-lbs) but not have been subject of an EMA.
50 ft-lb = 68J Assessment based on REAP dataset merged with properties from BASELINE.
Dataset includes domestic and international power reactor data.
1223 data points.
USE Results
- 19% of materials measured USE not bounded by RG model.
- Limited number of materials are misprojected to remain above 50 ft-lb and not trigger equivalent margins analyses (EMAs).*
- Minimal impact - the safety criteria supported by USE estimation (50 ft-lb) is known to be extremely conservative.
9
- - To date no material has failed to be justified by EMA; including many well below 68J.
Credibility Criteria
- RG 1.99, Rev. 2 has five credibility criteria.
- Criteria compare measured data to refit (chemistry factor)
RG 1.99 prediction results with a requirement of shape-function of RG 1.99
- If surveillance data is deemed credible, RG 1.99 allows reduction in margin term.
- If data is not deemed credible, data is not considered.
- The criteria typically failed is excessive scatter.*
10
- - One or more surveillance data point outside 2 where is the standard deviation of the ETC (17 °F for base materials and 28 °F for welds)
Credibility Criteria - Issues
- The more surveillance data points, the more likely it is for data to be non-credible due to scatter.
- No apparent basis for reduction in margin for credible data.
- Critical consequence: benefit of fitting predictions to surveillance data is nullified in many cases by credibility criteria that reject data not conforming to the fluence shape-function of RG1.99.
- High fluence and low Cu data not expected to conform to fluence shape function of RG1.99.
11
Attenuation Formula Attenuation formula matches well to modern results.
Formula only works for areas horizontally adjacent to the active fuel.
NRC has a research project on determining fluence outside of the beltline region. This will be incorporated in a revision of RG 1.190.
12 Fractional Thickness Geometric Beltline Jones, E. N., "Comparison of Regulatory Guide 1.99 Fluence Attenuation Methods," Journal of ASTM International, Vol. 9, No. 4, 2012, pp. 1-7.
Common Additions Several practices not addressed in RG have been commonly accepted by NRC. Several of these were explained in 1998 presentation by Wichman et al.*
- Use of sister plant data to supplement plant-specific surveillance data.
- This is material that is a heat-for-heat match irradiated in another plant.
- Data is adjusted for difference in irradiation temperature.
- Exclusion of low-fluence outliers from credibility assessment.
13
Conclusions
- Correcting the nonconservatism in the embrittlement trend curve at higher fluences is the most significant recommendation of the RG 1.99 assessment.
- The credibility criteria should also be revised to make more effective use of plant-specific surveillance data.
- Several common practices not addressed in the RG should be addressed in a revision, such as use of sister plant data, implementation of credibility criteria, degree-for-degree, etc.
14
Staff Response to Issues Identified in TLR
- In response to TLR, NRC initiated effort to revise RG 1.99.
- Two groups chartered for this effort:
- RG 1.99 Working Group
- RG 1.99 Oversight Group 15
RG 1.99 Working Group Charter The working groups primary task is to develop a revision to RG 1.99, Rev. 2 that will address the recommendations of the TLR.
Prioritization will be given to those recommendations with the most significant potential reduction on safety margin, e.g., the non-conservatism at high fluences is the highest priority.
USE model in the RG will not be addressed because potentially non-conservative USE predictions have no credible impact on safety as demonstrated by NRC accepted equivalent margins analyses.
The working group will concentrate on the technical content of the RG rather than regulatory or implementation issues.
The working group consists of NRR and RES staff. RES staff serves as the lead for the WG.
16
RG 1.99 Working Group Tasks
- Recommend an alternate ETC
- The working group selected the ASTM E900-15 ETC as the basis for the revision of RG 1.99, Rev. 2.
- Determine limitations of ETC implementation.
- Determine how to apply surveillance data.
- Determine margins on ETC.
- Determine default values for inputs that are not available.
- Write draft RG for internal review.
17
RG 1.99 Oversight Group
- Provides guidance and direction for this effort
- Members from RES and NRR:
- Branch chiefs - RES (1) and NRR (2)
- Senior level staff - RES (1) and NRR (2)
- Responsibilities:
- Recommend implementation options
- Assess adequate protection aspects
- Identify/recommend need for rulemaking 18
Timeline 1988 Issue RG 1.99 Rev.
2 2019 RES Report 2025 PTN Surface 2025 PWR PSSP Capsule 2027 Surry 1 Capsule 2026 PTN Capsule Test 2032 Surry 2 Capsule 2044 PTN 1/4T 2048 Surry 2 Surface 19 2050 Fluence Exceeds 6x1019 n/cm2
RG 1.99 Oversight Group Current Activities Determine criteria to assess safety significance - PTS and PT-limits
- Using FAVOR and leveraging analyses from PTS re-evaluation
- Determine change in through-wall crack frequency as a function of shift in RTNDT.
Conduct a plant impact study of a select number of operating reactors
- Smart sample of plants with applicable features, e.g., high fluence, high copper, etc.
- Use working groups RG update recommendations and criteria developed above.
Using the results of these analyses, determine Regulatory Guide implementation options.
20
Project Milestones 21 Milestone WG t Target OG Target Develop Criteria for PT-limits 12/1/2019 Develop Criteria for PTS 12/1/2019 Conduct initial plant impact 1/14/2020 Develop and assess initial framework for revised regulatory guide 1/14/2020 Public meeting with stakeholders 2/14/2020 Decision point for continuation of RG development 3/1/2020 3/1/2020 Draft RG for internal review 4/30/2019
List of Acronyms ART - adjusted reference temperature EMA - equivalent margins analysis ETC - embrittlement trend curve FAVOR - Fracture Analysis of Vessels, Oak Ridge (computer code)
RG - regulatory guide RTNDT - reference temperature, nil-ductility transition RTNDT - change in RTNDT due to irradiation PSSP - PWR Supplemental Surveillance Program P-T Limits - pressure-temperature limits PTS - pressurized thermal shock TLR - technical letter report USE - upper shelf energy 22
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m Tim Hardin Technical Executive Meeting of the ACRS November 6, 2019 Rockville, MD Potential Revision of Regulatory Guide 1.99, Revision 2
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 2
Contents
Background
Potential Revision of the Regulatory Guide 1.99, Revision 2 (RG1.99R2) Upper Shelf Energy (USE) Prediction Model Potential Revision of the RG1.99R2 T41J (T30 ft-lb) Prediction Model
- Considerations for BWR and PWR Fleets
- Recommended Alternative T30 Prediction Model
- Preliminary Assessment of Fleet Impact Conclusions and Recommendations This presentation offers the professional opinions of EPRI staff only and does not represent an Industry position of the U.S. utilities
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 3
Background (1/2)
In July 2019, NRC released Assessment of the Continued Adequacy of Revision 2 of Regulatory Guide 1.99, Technical Letter Report TLR-RES/DE/CIB-2019-2 (hereafter, the TLR)
EPRI provided review comments on the TLR at the August 22, 2019 meeting of the ACRS Subcommittee on Metallurgy & Reactor Fuels (ADAMS ML19260E007)
Since that meeting, EPRI has studied the potential impact that a revision of Regulatory Guide 1.99 may have on the fleet This presentation summarizes the results of that study and provides recommendations regarding implementation of a new RG Shift refers to the change in Ductile-Brittle Transition Temperature (DBTT), as measured by Charpy T30 (T41J)
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 4
Background (2/2)
Principal conclusions from EPRI review of the NRC TLR
- The conclusions of the TLR are succinct and their technical bases are thoroughly documented
- The recommendations of the TLR are reasonably supported by the analyses presented
- The data and analyses presented in the TLR support a phased implementation of any revision to RG1.99R2 to minimize unnecessary burden on the operating fleet
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 5
Potential Revision of the RG1.99R2 Upper Shelf Energy (USE) Prediction Model
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 6
USE Decrease Prediction Model In the August 22, 2019 Subcommittee Meeting, EPRI discussed a new USE prediction model developed by the Materials Reliability Program (MRP) in 2017 and published as MRP-414
- Materials Reliability Program: Prediction Model for the Decrease in Upper Shelf Energy of Reactor Vessel Steel Due to Neutron Embrittlement (MRP-414). EPRI, Palo Alto, CA: 2017.
This model achieves a 2 bound of the USE decrease data An evaluation of the impact of the MRP-414 model on the PWR fleet found:
- The total number of plants which have materials predicted to fall below 68 J at 80 years is the same, when either the RG1.99R2 or the MRP-414 model is used Also, it is noted that no material which has been predicted to fall below 68 J has been unacceptable when evaluated using an Equivalent Margins Analysis (EMA) per RG1.161 and ASME XI Appendix K
- The 68 J minimum USE criterion in 10CFR50, Appendix G is conservative Changing the USE prediction model in RG1.99R2 would result in negligible safety benefit but would cause a significant reanalysis burden for the fleet
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 7
Potential Revision of the RG1.99R2 T41J (T30 ft-lb) Prediction Model
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 8
T41J Prediction Model The NRC TLR concluded
- the estimates of embrittlement provided by RG1.99 appear to become non-conservative at fluence levels approaching 3 to 6 x 1019 n/cm2 (E > 1 MeV) and
- it is evident that a fluence limit should be established to indicate when RG1.99 ceases to adequately predict T41J for regulatory purposes.
The NRC TLR does not specify a fluence limit, but the implication is somewhere 3 - 6 x 1019 n/cm2 (E > 1 MeV)
Because RG1.99R2 is used for embrittlement predictions to show compliance with 10CFR50, Appendix G, the appropriate metric for assessing the need for a change to the RG1.99R2 T41J prediction model is the RPV 1/4T fluence, since that is the fluence upon which operating limits are based per ASME XI App. G
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 9
Considerations for the BWR Fleet Based on the TLRs assessment that the predictions of RG1.99R2 T41J prediction model are non-conservative at fluences ~ 3 - 6 x 1019 n/cm2 (E > 1 MeV), we note:
- From BWRVIP-321, Plan for Extension of the BWR Integrated Surveillance Program (ISP) Through the Second License Renewal (SLR):
most plants have projected target 1/4T SLR fluences that fall within a similar range, between 1 x 1018 and 5 x 1018 n/cm2while several plants are projected to have much higher fluence, on the order of 6 to 9 x 1018 n/cm2
- The 80 year fluences for the BWR fleet are well below the range at which RG1.99R2 becomes non-conservative The BWR fleet can safely use the existing RG1.99R2 T41J prediction model through an 80 year life
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 10 Considerations for the PWR Fleet For PWRs, 1/4T fluence is higher than for BWRs and many vessels will exceed the ~ 3 - 6 x 1019 n/cm2 (E > 1 MeV) threshold For a particular vessel, a new embrittlement prediction model may increase shift for some materials, but decrease shift for other materials To assess the potential impact on the Adjusted Reference Temperatures (ART) of each vessel material, it is necessary to identify what new prediction model will be used
- Probable candidates are assessed in the next slide
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 11 Comparison of Potential Alternative T30 Trend Curve Prediction Formulae ASTM E900-15 is the Preferred Alternative T30 Trend Curve Prediction Formula ID ASTM E900-15 EONY Year 2015 2006 Fit to LWR surveillance data from 13 countries (USA, Japan, France, Germany, S. Korea, Belgium, )
LWR surveillance data from the USA
- of DT30 Data 1,878 855 Development Result of ASTM consensus process, 2010-2015 Result of NRC contractor effort, 1998-2006 Status Adopted as an ASTM Standard Guide Adopted in 10 CFR 50.61a Applied to one plant (Palisades) to date
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 12 Assessment of Fleet Impact of E900-15 T30 Trend Curve Methodology Differences between alternative (E900-15) and RG1.99R2 DT30 predictions quantify the impact
- Impact Positive difference: increases shift / possible impact on operations Negative difference: decreases shift / provides relief
- Values compared Mean + 2s predictions at Inside diameter (ID)
- Evaluated Low copper (Cu) steels and High Cu steels separately Data used / assumptions
- Evaluated projected fluence values at 80 years
- Capacity factor: 90%
- The evaluation was conducted using available US surveillance data, not all RPV materials Difference = T30 90015 T30 RG1.99R2
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 13 Comparison of ASTM E900-15 to RG1.99R2 Trend Curve Low-Cu Steels ASTM E900-15 Time Fleet @ 80 years Copper Low (<0.1 wt%)
Location Inner Diameter Prediction Mean DT30 +2s Mean DT30+2s: Alternative - RG1.99R2 [oF]
Log10(Fluence) [n/cm2]
New prediction reduces DT30 New prediction increases DT30 PWR Base Weld BWR Base Weld
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 14 Comparison of ASTM E900-15 to RG1.99R2 Trend Curve High-Cu Steels ASTM E900-15 Time Fleet @ 80 years Copper High (>0.1 wt%)
Location Inner Diameter Prediction Mean DT30 +2s Mean DT30+2s: Alternative - RG1.99R2 [oF]
Log10(Fluence) [n/cm2]
BWR PWR Base Weld New prediction reduces DT30 New prediction increases DT30
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 15 Impact of Adopting ASTM E900-15 Shift Prediction Model (1/2)
Changing from the RG1.99R2 T41J model to ASTM E900-15 will increase predicted embrittlement shift for some materials, and decrease it for others
- In general, base metals (plates and forgings) will see increases; welds will see both increases and decreases.
Estimating the net impact of these trends on the licensing basis (e.g.,
the P-T curves) for any particular vessel requires a vessel-specific evaluation
- Depending on the direction and magnitude of the shift change for each vessel material, there may or may not be an impact on the plant operating limits (P-T curves).
- Whether or not the shift changes would require revision of the plant P-T curves is unknown until the specific ART values are calculated for all materials in that vessel and compared to the ART values used as the basis for the plants existing P-T limits
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 16 Impact of Adopting ASTM E900-15 Shift Prediction Model (2/2)
Vessel-specific evaluations until are not possible until NRC provides guidance on:
- What prediction model should be used
- How surveillance data is to be used /
credited Plant-specific data, sister plant data, etc.
Performing the vessel specific evaluation would involve significant burden that has no safety benefit if the plants 1/4T fluence is < 3 - 6 x 1019 (e.g.,
the fluence level below which the TLR identifies no safety concern for the current RG1.99R2 prediction model)
A 1/4T fluence metric is recommended for screening plants that need (or do not need) to use a new RG1.99 T41J prediction model Figure from TLR-RES/DE/CIB-2019-2 illustrating consideration of surveillance data
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 17 Conclusions and Recommendations Changing the USE prediction model in RG1.99 would result in negligible safety benefit but would cause a significant reanalysis burden on the fleet ASTM E900-15 is the preferred alternative T41J prediction formula Because RG1.99 is used for embrittlement predictions performed to show compliance with 10CFR50, Appendix G, the appropriate metric for assessing the need for plants to adopt a new T41J prediction model is the RPV 1/4T fluence, since that is the fluence upon which operating limits are based per ASME XI Appendix G
- BWRs do not reach the threshold of concern through 80 years of operation and can be exempted from the burden of adopting a new T41J shift prediction model
- PWR adoption of a new shift prediction model would appropriately be based on a 1/4T fluence metric Because embrittlement prediction models have significant impact on the RPV operating envelop, it will be helpful to plants considering SLR for the regulator to identify the shift model that will be adopted in RG1.99R3, and guidance for consideration of surveillance data
© 2019 Electric Power Research Institute, Inc. All rights reserved.
w w w. e p r i. c o m 18 TogetherShaping the Future of Electricity
Brunswick Steam Electric Plant Units 1 and 2 License Amendment Request Application of Advanced Framatome Methodologies
BSEP Station Overview and Advanced Framatome Methods LAR Overview Mark DeWire - BSEP Assistant OPS Manager - Shift 2
BSEP Station Overview General Electric BWR-4, Mark I Containment Began commercial operation in 1975 (Unit 2) and 1976 (Unit 1), OLTP 2436 MWt EPU (120% OLTP) 2923 MWt fully implemented in 2004 (Unit 1) and 2005 (Unit 2) 24 month operating cycle Transitioned to Framatome Fuel in 2008 (U1) and 2009 (U2)
Currently Full Core Framatome ATRIUM 10XM Fuel 8 ATRIUM 11 Lead Test Assemblies installed on Unit 2 in 2015 Licensed for Increased Core Flow (ICF)
(i.e., 110% Core Flow at reduced power, 104.5% Core Flow at CLTP 2923 MWt)
Licensed for Maximum Extended Load Line Limit Analysis Plus (MELLLA+)
(i.e., 85% Core Flow at CLTP 2923 MWt) 3
LAR Overview Technical Specification (TS) 5.6.5.b lists the analytical methods used to determine core operating limits.
This license amendment request (LAR) revises TS 5.6.5.b to allow application of Advanced Framatome Methodologies for determining core operating limits in support of loading Framatome fuel type ATRIUM 11.
Duke Energy is pursuing the ATRIUM 11 fuel type due to the improved fuel cycle economics and safety margins.
11x11 array reduces fuel duty ~19%: LHGR margin Improved debris protection features (fuel failure risk reduction)
Improved channel performance Improved fuel cycle economics 4
LAR Overview (continued)
The following Methodologies will be removed from TS 5.6.5.b as they will no longer be applicable with the addition of the Advanced Framatome Methods.
XN-NF-84-105(P)(A) Volume 1, XCOBRA-T: A Computer Code for BWR Transient Thermal-Hydraulic Core Analysis ANF-913(P)(A) Volume 1, COTRANSA2: A Computer Program for Boiling Water Reactor Transient Analyses NEDC-33075P-A, GE Hitachi Boiling Water Reactor, Detect and Suppress Solution - Confirmation Density, Revision 8, November 2013 5
LAR Overview (continued)
The Advanced Methodologies that will be added to TS 5.6.5.b are listed below.
ANP-10300P-A, AURORA-B: An Evaluation Model for Boiling Water Reactors; Application to Transient and Accident Scenarios, Rev. 1, January 2018 ANP-3703P, BEO-III Analysis Methodology for Brunswick Using RAMONA5-FA, Rev. 0, August 2018 DPC-NE-1009-P, Brunswick Nuclear Plant Implementation of Best-estimate Enhanced Option-III, Rev. 0, September 2018 BAW-10247P-A, Supplement 2P-A, Realistic Thermal-Mechanical Fuel Rod Methodology for Boiling Water Reactors Supplement 2: Mechanical Methods, Rev. 0, June 2018 ANP-10340P-A, Incorporation of Chromia-Doped Fuel Properties in AREVA Approved Methods, Rev. 0, May 2018 ANP-10335P-A, ACE/ATRIUM 11 Critical Power Correlation, Rev. 0, May 2018 ANP-10333P-A, AURORA-B: An Evaluation Model for Boiling Water Reactors; Application to Control Rod Drop Accident (CRDA), Rev. 0, March 2018 ANP-10332P-A, AURORA-B: An Evaluation Model for Boiling Water Reactors; Application to Loss of Coolant Accident Scenarios, Rev. 0, March 2019 6
Advanced Framatome Methods LAR Supporting Information and Schedule Milestones Mike Blom - Director - Fuel Management and Design 7
LAR Supporting Information The following documents, included in the LAR, demonstrate acceptability of the LAR supporting operation of the new ATRIUM 11 fuel type in the currently approved operating domain (i.e.,
including EPU and MELLLA+).
ANP-3705P, Applicability of Framatome BWR Methods to Brunswick with ATRIUM 11 Fuel ANP-3686P, Mechanical Design Report for Brunswick ATRIUM 11 Fuel Assemblies ANP-3643P, Brunswick Unit 1 Thermal-Hydraulic Design Report for ATRIUM 11 Fuel Assemblies ANP-3644P, Brunswick Unit 2 Thermal-Hydraulic Design Report for ATRIUM 11 Fuel Assemblies ANP-3668P, ATRIUM 11 Fuel Rod Thermal-Mechanical Evaluation for Brunswick LAR ANP-3661P, Brunswick ATRIUM 11 Equilibrium Cycle Fuel Cycle Design ANP-3667P, Brunswick Unit 1 ATRIUM 11 Equilibrium Cycle Nuclear Fuel Design Report ANP-3702P, Brunswick ATRIUM 11 Transient Demonstration 8
LAR Supporting Information (continued)
ANP-3674P, Brunswick Units 1 and 2 LOCA Analysis for ATRIUM 11 Fuel ANP-3694P, ATWS-I Analysis Methodology for Brunswick Using RAMONA5-FA ANP-3703P, BEO-III Analysis Methodology for Brunswick Using RAMONA5-FA DPC-NE-1009-P, Brunswick Nuclear Plant Implementation of Best-estimate Enhanced Option-III ANP-3714P, Brunswick ATRIUM 11 Control Rod Drop Accident Analyses with the AURORA-B CRDA Methodology 9
Cycle Specific Reports Duke Energy provided the B1C23 reload reports outlined in the table below to the NRC for information during this review.
10 Report B1C23 Fuel Cycle Design Report B1C23 Nuclear Fuel Bundle Design Report B1C23 Safety Limit MCPR Report B1C23 Fuel Rod Design Report B1C23 Reload Safety Analysis Report
AURORA-B LOCA The Final Safety Evaluation (SE) for Topical Report (TR) ANP-10332P AURORA-B LOCA had not been issued at the time of LAR submittal.
The Draft SE along with ANP-10332P were referenced for the initial LAR submittal.
All the limitations and conditions presented in the Draft SE were addressed in the initial LAR submittal.
Following issuance of the Final SE for ANP-10332P, the LAR was supplemented with reference to the approved AURORA-B LOCA TR (i.e., ANP-10332P-A) and changes in limitations and conditions were addressed.
11
BEO-III w/ CDA With the addition of these methodologies, BSEP is transitioning from the Detect and Suppress Solution - Confirmation Density (DSS-CD) stability methodology to the Best Estimate Enhanced Option-III with Confirmation Density Algorithm (BEO-III w/CDA) stability methodology.
As with DSS-CD, the CDA will remain the licensing basis trip, and identical Oscillation Power Range Monitor (OPRM) licensing basis setpoints will be used for BEO-III w/CDA thereby minimizing the impact to BSEP.
The LAR provided a demonstration of the implementation of BEO-III with CDA with sample Brunswick ATRIUM 11 results from the currently approved operating domain.
12
RAMONA5-FA ATWS-I In addition to the methodology changes, with the transition from ATRIUM 10XM fuel to ATRIUM 11 fuel, BSEP will transition to RAMONA5-FA for the licensing basis Anticipated Transient Without Scram with Instability (ATWS-I) analysis.
The current ATWS-I licensing basis analysis is the TRACG ATWS-I evaluation performed with ATRIUM 10XM fuel for MELLLA+ which was approved for BSEP in 2018.
The LAR documents ATWS-I licensing basis analysis for Brunswick with ATRIUM 11 fuel and the currently approved operating domain.
The Brunswick specific methodology is identical to ANP-10346P, ATWS-I Analysis Methodology for BWRs Using RAMONA5-FA, which is pending final approval by the NRC for generic application.
13
LAR Schedule Milestones 14 October 2018 LAR submittal December 2018 Duke / NRC schedule review to support February 2020 need date February 2019 Audit - BEO-III Plant Specific Methodology March 2019 Audit - All Methodology March to October 2019 Implementation cycle reports submitted for information June 2019 RAI Responses July 2019 LOCA Supplement October 2019 Additional License Condition for BEO-III w/CDA November 2019 ACRS Sub-Committee and Full Committee February 2020 LAR approval March 2020 Implementation cycle startup
Questions 15
Additional Info 16
ATRIUM 11 Fuel Design and Methodology Information In accordance with the process described in ANF-89-98(P)(A) Revision 1 and Supplement 1, a summary of the evaluation of the ATRIUM 11 design against the NRC-approved generic design criteria was provided to the NRC during this review.
ANP-3653P, Fuel Design Evaluation for ATRIUM 11 BWR Reload Fuel A compendium of Framatome methodologies and design criteria, which are described in Topical Reports that the NRC has found acceptable for referencing in BWR licensing applications was provided to the NRC during this review.
ANP-2637P, Boiling Water Reactor Licensing Methodology Compendium 17
Methodology Application Methodology Application ANP-10333P-A, AURORA-B: An Evaluation Model for Boiling Water Reactors; Application to Control Rod Drop Accident (CRDA), Revision 0, March 2018 7 ANP-10300P-A, AURORA-B: An Evaluation Model for Boiling Water Reactors; Application to Transient and Accident Scenarios, Revision 1, January 2018 2 ANP-3703P, BEO-III Analysis Methodology for Brunswick Using RAMONA5-FA, Revision 0, August 2018 Attachment 15 DPC-NE-1009-P, Brunswick Nuclear Plant Implementation of Best-estimate Enhanced Option-III, Revision 0, September 2018 6 BAW-10247P-A, Supplement 2P-A, Realistic Thermal-Mechanical Fuel Rod Methodology for Boiling Water Reactors Supplement 2: Mechanical Methods, Revision 0, June 2018 ANP-10340P-A, Incorporation of Chromia-Doped Fuel Properties in AREVA Approved Methods, Revision 0, May 2018 ANP-10335P-A, ACE/ATRIUM 11 Critical Power Correlation, Revision 0, May 2018,
8, 10, and 12 ANP-10332P-A, AURORA-B: An Evaluation Model for Boiling Water Reactors; Application to Loss of Coolant Accident Scenarios, Revision 0, [DATE] 3 18
LAR Attachment Numbers Attachment Number Attachment Name 5
ANP-3705P, Applicability of Framatome BWR Methods to Brunswick with ATRIUM 11 Fuel 6
ANP-3686P, Mechanical Design Report for Brunswick ATRIUM 11 Fuel Assemblies 7
ANP-3643P, Brunswick Unit 1 Thermal-Hydraulic Design Report for ATRIUM 11 Fuel Assemblies 8
ANP-3644P, Brunswick Unit 2 Thermal-Hydraulic Design Report for ATRIUM 11 Fuel Assemblies 9
ANP-3668P, ATRIUM 11 Fuel Rod Thermal-Mechanical Evaluation for Brunswick LAR 10 ANP-3661P, Brunswick ATRIUM 11 Equilibrium Cycle Fuel Cycle Design 11 ANP-3667P, Brunswick Unit 1 ATRIUM 11 Equilibrium Cycle Nuclear Fuel Design Report 12 ANP-3702P, Brunswick ATRIUM 11 Transient Demonstration 13 ANP-3674P, Brunswick Units 1 and 2 LOCA Analysis for ATRIUM 11 Fuel 14 ANP-3694P, ATWS-I Analysis Methodology for Brunswick Using RAMONA5-FA 15 ANP-3703P, BEO-III Analysis Methodology for Brunswick Using RAMONA5-FA 16 DPC-NE-1009-P, Brunswick Nuclear Plant Implementation of Best-estimate Enhanced Option-III 17 ANP-3714P, Brunswick ATRIUM 11 Control Rod Drop Accident Analyses with the AURORA-B CRDA Methodology 19
BSEP Unit 1 and Unit 2 Differences BSEP Unit 1 and Unit 2 are essential identical. The key differences in system configuration are in the core inlet region and the turbine bypass system.
Fuel Support Casting Central Orifice Diameter
Unit 1: 2.43
Unit 2: 2.09 Turbine Bypass System
Unit 1: 4 Valves
Unit 2: 10 Valves As a result of these minor differences, Unit 1 is more limiting with respect to long term stability and ATWS-I. Therefore, Unit 1 was the modeled plant for these methodology demonstrations.
20
Power Density Comparison for MELLLA+ Submittals Plant GGNS PB BFN BSEP NMP2 MNGP EPU Thermal Power (MWth) 4408 3951 3952 2923 3988 2004 Licensed Core Flow (Mlb/hr) 112.5 102.5 102.5 77 108.5 57.6 MELLLA+ knee % power 80.6 78.8 77.6 77.6 77.6 82.5 MELLLA+ knee % flow 55 55 55 55 55 57.4 power/flow ratio at MELLLA+ knee 57.42 55.23 54.40 53.56 51.86 50.01 power density (kW/ft) 5.5 4.8 4.8 4.9 4.9 4
power density (kW/L) 62.3 58.4 58.4 59 59 48.3 21
Introduction Brunswick Fuel Transition to Framatome ATRIUM-11 License Amendment Request Presented to the ACRS November 6, 2019 Andy Hon, PE Project Manager Division of Operation Reactor Licensing Office of Nuclear Reactor Regulation 1
Brunswick ATRIUM 11 Fuel License Amendment Background
- Current Framatome ATRIUM-XM10 fuel was approved to operate in MELLLA+ domain - Integrated analyses from GEH and Framatome in September 2018.
- LAR was submitted in October 2018 for the adopting advanced Framatome methods for ATRIUM 11 - Decision needed by February 2020 to support fuel loading in scheduled outage.
- First application of seven new Framatome analyses and one Duke Energy method for the new fuel design.
- Most of the new methodologies have been already approved generically as topical reports.
- Requested changes are mainly COLR references to the new Framatome methodologies in Chapter 5 of the Tech Specs.
2
NRC Staff Safety Evaluation Reviewed the L&Cs of approved generic topical reports are addressed by BSEP.
Reviewed plant specific stability technical reports and identified the necessary license conditions for implementation and enforcement.
Conducted two audits of safety analyses - February and March 2019.
32 RAIs - all accepted and satisfactorily responded by Duke Energy. License conditions were adopted for implementation.
RES performed an ATWSi confirmatory study using TRACE code to model the new fuel performance at Brunswick.
Prepared to make a decision on the request after finalizing SE to incorporate the ACRS comments to support the requested completion date.
3
NRC Staff Review Team 4
Office of Nuclear Reaction Regulation A. Smith (Lead)
J. Borromeo C. Cheung E. Dickson A. Hon R. Grover J. Lehning S. Krepel D. Woodyatt A. Wysocki (ORNL)
Office of Nuclear Regulatory Research P. Yarsky (Lead)
A. Bielen C. Gingrich N. Hudson S. Marshall T. Zaki