ML19046A392

Meeting Transcript

ML19046A392
Person / Time
Issue date:	01/23/2019
From:	Office of Nuclear Regulatory Research
To:
Kenneth Hamburger 415-2022
Shared Package
ML19046A388	List: ML19046A389 ML19046A390 ML19046A391 ML19046A392 ML19046A393
References
20181317, NRC-0048
Download: ML19046A392 (170)

Text

Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION

Title:

Informational Update on High Energy Arc Faults (HEAF) Involving Aluminum Docket Number: (n/a)

Location: Rockville, Maryland Date: Wednesday, January 23, 2019 Work Order No.: NRC-0048 Pages 1-169 NEAL R. GROSS AND CO., INC.

Court Reporters and Transcribers 1323 Rhode Island Avenue, N.W.

Washington, D.C. 20005 (202) 234-4433

1 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3 + + + + +

4 INFORMATIONAL UPDATE ON HIGH ENERGY ARC FAULTS 5 (HEAF) INVOLVING ALUMINUM 6 + + + + +

7 WEDNESDAY, 8 JANUARY 23, 2019 9 + + + + +

10 ROCKVILLE, MARYLAND 11 The public meeting convened in the 12 Commissioners' Hearing Room at the Nuclear Regulatory 13 Commission, One White Flint North, 11555 Rockville 14 Pike, at 9:00 a.m., Ken Hamburger, Facilitator, 15 presiding.

16 PRESENT:

17 KEN HAMBURGER, RES/DRA/FXHAB, Facilitator 18 MICHAEL CHEOK, RES/DRA 19 TOM BOYCE, RES/DE/RGGIB 20 MICHAEL FRANOVICH, NRR/DRA 21 STANLEY GARDOCKI, RES/DE/RGGIB 22 NICK MELLY, RES/DRA/FXHAB 23 KENN MILLER, RES/DE/ICEEB/EET 24 MARK HENRY SALLEY, RES/DRA/FXHAB 25 GABE TAYLOR, RES/DRA/FXHAB NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

2 1 C O N T E N T S 2 Page 3 Opening Remarks . . . . . . . . . . . . . . . . . 3 4 Introduction & Meeting Objectives . . . . . . . . 6 5 Generic Issue Program Status for PRE-GI-018 . . . 6 6 Preliminary Observations of Sep 2018 Full-Scale 7 Confirmatory Tests on Aluminum HEAFs . . . 44 8 Future Test Planning and Equipment Selection . . 87 9 NRC/EPRI Working Group Charter . . . . . . . . . 99 10 Public Question and Comment . . . . . . . . . . 116 11 Concluding Remarks . . . . . . . . . . . . . . 169 12 Adjourn . . . . . . . . . . . . . . . . . . . . 169 13 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

3 1 P R O C E E D I N G S 2 9:01 a.m.

3 MR. HAMBURGER: Good morning, everybody.

4 It's nine o'clock, so we're going to get started.

5 My name is Ken Hamburger. I work in the 6 Office of Nuclear Regulatory Research. I'm an NRC 7 Facilitator, and I'm here today in that capacity.

8 Mostly just to keep us on time and on 9 schedule. And make sure everybody has a chance to say 10 what they came here to say.

11 A couple of logistical notes. For 12 visitors, while you're here, please keep your visitor 13 badges visible. Keep them on you. You shouldn't need 14 an escort wherever you are today.

15 This room is fine. The hallway out here 16 is fine. The cafeterias are fine. You don't need an 17 escort for any of that.

18 If you need to access any of the other 19 floors, please find an NRC staff member to escort you.

20 The bathrooms are through this door, make a left.

21 Women's bathrooms are on the left. Men's bathrooms 22 are on the right. Again, don't need an escort.

23 The cafeteria is also open to you if you'd 24 like to grab coffee or a snack. It's out that door to 25 the left. Follow the hallway down, and the cafeteria NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

4 1 is on the left.

2 If there is any sort of fire or emergency, 3 you have emergency exists in the rear of the room.

4 And we have the two exits in the front of the room.

5 We are transcribing today's meetings. We 6 have our Court Reporter in the back there. So, a 7 couple of things for those in the room.

8 Please identify yourself when speaking.

9 Use your name so he can capture that in the 10 transcript.

11 Please speak into the microphones. He has 12 an audio feed, so he can only hear what's said over 13 the microphones.

14 And we have webinar participants. So they 15 can only hear what's said into the microphones. So 16 please don't forget.

17 I'll try to remind you. Even if you're 18 right up here in the front and you're speaking to 19 somebody over here, the webinar hears silence if you 20 don't have one of the mics.

21 So, we have an aisle mic over here and 22 over here. I have two wireless mics. If you have any 23 trouble getting to the aisles, you can just raise your 24 hand, I'll bring over a wireless mic to you.

25 And for those on the webinar, best way to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

5 1 ask a question or make a comment is to use the raise 2 hands' feature on the webinar. Or use the chat 3 feature and you can ask Tom. Tom is manning our 4 webinar.

5 And he will either relay the question to 6 us. Or open up the phone lines. We do have a phone 7 line in the room so you can ask the question yourself.

8 And we have you muted at the moment, but 9 periodically we'll open up the phone lines. Anybody 10 has a comment or question.

11 So I suggest that you keep yourself muted 12 until you're ready to speak in case we open up the 13 phone lines.

14 We have a public comment period scheduled 15 for the last two hours of today. That's open for 16 anybody to present or say anything they would like to 17 say.

18 I do think we're pretty lightly loaded in 19 terms of the agenda. I expect to have two hours plus 20 for that public comment period.

21 But just so that we can get through all 22 the items on the agenda and make sure that everyone 23 has a chance to speak, I may ask you to hold certain 24 comments or questions until that public comment 25 period.

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

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

6 1 Just to make sure we get through 2 everything we have to get through. Again, I expect 3 more than enough time for that public comment period.

4 If anybody has any questions or issues 5 with today's meeting or with the webinar, you can 6 reach out to me, Ken Hamburger. I'll put my contact 7 information up on the board during the breaks.

8 And with that, I would like to introduce 9 our Division Director, Mr. Michael Cheok.

10 MR. CHEOK: So thank you Ken. And good 11 morning. Welcome. And thank you for participating in 12 this public meeting on the Pre-Generic Issue on High 13 Energy Arching Faults involving aluminum.

14 Again, my name is Mike Cheok. I am the 15 Director for the Division of Risk Assessment in the 16 Office of Nuclear Regulatory Research.

17 Today we will provide a brief status of 18 where we are in the generic issue. We will also talk 19 about our proposed next steps.

20 But most importantly, we would like to 21 hear your comments, questions, or feedback on the 22 process.

23 As Ken said, I would like to welcome those 24 on the Webinar. And I would like to say that we would 25 also love to hear from you. For those of -- for those NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

7 1 people who are not available today, this meeting is 2 being transcribed.

3 So, today you will hear two main teams 4 throughout the meeting. First, you know, we will 5 follow the GI Process, the Generic Issue Process, 6 which has several well-defined steps.

7 The first step that you'll hear about is 8 the screening process. And you will hear about the 9 results of that screening process.

10 We are now in the assessment phase of the 11 process. And you know you will again hear about what 12 we have done so far and where we plan to go.

13 We will have a methodical process. We 14 will be methodical in getting to the most realistic 15 and the most applicable offering experienced data and 16 information.

17 We will also have to be methodical in 18 terms of defining our tests going forward. We will 19 have to have realistic and applicable tests to get the 20 best information on how we can characterize the HEAF 21 phenomenon.

22 We will -- we will have to do this so 23 that, you know, when they come up with our 24 recommendation, the recommendation will be robust. It 25 will be supportable. And it will be defendable.

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

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

8 1 Also, you know, I would like to say that 2 not all HEAFs are the same. We have to characterize 3 the different classifications or categories of HEAFs 4 correctly.

5 We have to determine the corresponding 6 frequencies of these classes of HEAFs correctly. We 7 have to model the effects of these HEAFs on plants 8 correctly.

9 Also, not all plants are the same. Not 10 all plants have aluminum. Plant configurations and 11 plant equipment are different.

12 Plant procedures and plant practices are 13 different. So, you know, we have to get this right in 14 terms of the generic issue.

15 A single solution for all plants may not 16 be applicable. We may have to have solutions that 17 maybe different for different plants.

18 So the second thing you will hear today is 19 that we will need to continue to improve on our 20 communications. We will need to involve all the 21 public stakeholders throughout the process, early and 22 often.

23 We plan to hold as many public meetings as 24 needed. And we hope that you will show up and provide 25 us with the feedback.

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

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

9 1 So with that, I would like to go ahead and 2 turn the meeting over to Mark Salley for his --

3 bringing us through the Agenda.

4 MR. SALLEY: Thanks Mike. To pick up 5 where Mike left off there, this is about 6 communications today.

7 And as a part of the Generic Issue, the 8 Generic Safety Issue process, there is a requirement 9 that they have periodic updates.

10 And this fulfills the periodic update.

11 But more importantly, when we talk to Tom Boyce and 12 Stan, there was additional information that we wanted 13 to share today. So, we wanted to share as much 14 information as possible.

15 You know, also with communications, good 16 communications are very important. We've done a lot 17 of internal briefings inside the NRC through the 18 Commissioner assistants, up through the DEDOs.

19 And it's surprising some of the questions 20 and information we hear that's floating around in the 21 industry.

22 Without getting into that, we'd like to 23 take the high road here and to answer any questions 24 you have. And to communicate anything we can with 25 you.

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

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

10 1 Also, with good communications is a two-2 way street. We need dialog. So, at the request of 3 the stakeholders and industry, we've allowed two hours 4 this afternoon as Kenny said, for any open discussion.

5 So again, we want to hear from you. So 6 today is about communication.

7 I want to give you a quick overview of 8 what we're going to try to accomplish today. And then 9 we'll get right into it.

10 Again, this is part of the Generic Issue 11 Program. We're doing this update. We've also 12 performed some testing in September.

13 And we'll refer to this as confirmatory 14 testing. If you remember the first phase of testing 15 we did with the Internationals, the OECD NEA Program, 16 we had two tests, Test 23 and Test 26 that kind of 17 stuck out a little bit.

18 And we had some ideas and some thoughts 19 that aluminum could be the issue. But we wanted to be 20 -- we don't want to cycle industry.

21 We want to be sure. We want to be 22 thoughtful in how we move forward. And we wanted to 23 do some confirmatory testing.

24 We did some in September. We did the 25 medium voltage. It's basically broken up into three NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

11 1 pieces, medium voltage, low voltage, and the bus 2 ducts.

3 We did four tests with the medium voltage.

4 Frank Cielo is here from KEMA. If anyone has any 5 questions about the testing lab, you can grab Frank at 6 a break and talk with him.

7 But we did the confirmatory testing, and 8 Nick's going to give a presentation. And we're going 9 to share some of the insights.

10 There's a lot of data. And it's going to 11 take time for us to analyze the data and go through 12 it.

13 But I think at this point we can share 14 some observations and let you know what we see.

15 Again, it's about giving you the information and 16 sharing what we have.

17 We'll also be moving into the next series 18 of tests. We hope to run the low voltage and the bus 19 ducts this summer.

20 So again, we'll talk about that. If you 21 remember the meeting back in April, we had two and a 22 half days, very good discussion with everyone. We're 23 moving along with that matrix and some of the testing 24 there.

25 And finally, the last thing that we'll NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

12 1 talk about after lunch is, NRC has an MOU with EPRI to 2 work on technical matters in a number of areas. One 3 of them is fire risk.

4 Under the fire risk we've put together a 5 working group and a charter with EPRI that we'll look 6 at the data and do some of the technical analysis.

7 We'll discuss that this afternoon.

8 So that's the basic overview of what we're 9 going to have today. Again, this afternoon I want to 10 emphasize that we've got a two-hour slot for any of 11 your questions and open discussion.

12 I see that Victoria, you've got a 13 presentation from NEI, which we'll put up first. And 14 Kelly, I believe you have one too that we'll have.

15 So, with that, I'll turn it back to you 16 Kenny and get started.

17 MR. GARDOCKI: Well good morning. My name 18 is Stan Gardocki. I work with the Generic Issues 19 Program in the Department of Research here at the NRC.

20 If you have any questions dealing with the 21 generic issue process, you can always give me a call 22 anytime. Or send me an email.

23 I want to first step back a little bit for 24 those people that weren't here at the last public 25 meeting, and introduce the Generic Issues Program and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

13 1 how it -- the generic issue processes through the 2 program.

3 It basically has three stages, as you can 4 see on the slide there. Now the first time we see the 5 GI, it comes into what they call the screening stage.

6 And we take a first cut at it to see if it 7 meets the seven criteria. If it does not meet all 8 seven criteria, it doesn't proceed forward.

9 We spend as much time on this screening 10 stage as necessary to make a good evaluation. Because 11 we don't want to waste, you know, resources, time and 12 personnel doing something that's not necessary.

13 So, we've revamped the program several 14 years ago and came up with these seven criteria. And 15 we think it's a pretty good screen to go to the next 16 stage.

17 After it goes through the screening stage, 18 we go to what they call the assessment stage. And we 19 spend about one to two, maybe possibly, you know, two 20 years on the issue to determine the risk and the 21 safety significance of the issue.

22 This is important for us here to determine 23 whether it goes to the next stage, call the Regulatory 24 Office of Mutation Stage. Which, you know, we have 25 the applicable NRC office that's working on the issue.

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

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

14 1 This particular issue would go like to 2 NRR. So we don't want to waste NRR's time. So, we do 3 what they call an assessment.

4 We do the testing up front. We do the 5 modeling up front. And we do what they call an 6 assessment report at the very final part of it.

7 And we put it out to the public. We have 8 meetings with ACRS and we discuss what they call the 9 regulatory analysis. We'll get into that a little bit 10 later on.

11 And then if it passes the screen -- the 12 assessment stage, it goes into a regulatory office 13 stage. And this is where the regulatory office here 14 would be like NRR.

15 It develops generic communications. They 16 might send out generic letters, bulletins, INs. They 17 meet with the industry. Develop an action plan if 18 there's any necessary actions that industry has to 19 take.

20 That process could take several years.

21 Once you get into a regulatory office, depending on 22 how quickly the industry reacts.

23 And we stay in the regulatory office stage 24 until all actions have been implemented for all plants 25 affected. And that the NRC has verified that all NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

15 1 actions have been taken.

2 The next slide here is a generic process 3 overview. It's a very encompassing slide. But it 4 captures everything from the screening to the 5 regulatory office.

6 And it shows who's responsible. Who's 7 working on it. And at the bottom you can see all the 8 notices and the public involvement and ACRS 9 involvement.

10 So it's a pretty good slide. It gives you 11 an overview of the entire process.

12 Where we are now, is we went through the 13 screening process. And we're in the assessment stage.

14 You can see there the drop downs that we put out the 15 necessary reports.

16 And during the assessment stage, you can 17 see the research still has the issue. So we're still 18 in charge of the research office.

19 And we have what they call an assessment 20 team working on it. These personnel are just, you 21 know, working on it.

22 And they go back and report to the generic 23 issues, you know, group. And put out an assessment 24 report at the end. Okay. Next side.

25 Specifically where we are on this generic NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

16 1 issue is, when we first get the generic issue into the 2 NRC, we screen it out to see if it's an immediate 3 safety concern. We ask the regulatory office to 4 document that.

5 And we don't want to procrastinate through 6 a process if there's an immediate safety concern in 7 the plants.

8 NRR has the LIC-504. And I don't know if 9 anyone's been introduced to that process. But it's 10 pretty quick.

11 You know, they can say, you now, we have 12 a problem with a valve, or a valve stem or whatever it 13 is. And the plants need to go out there and replace 14 immediately.

15 But, if the NRR comes back with immediate 16 safety concerns and says it's not an immediate safety 17 concern, we don't know how much of a safety concern it 18 is, so put it in the generic issues process.

19 Do the entire process. The in depth 20 development and the assessments. And see how much 21 risk and safety importance it has.

22 So they've done that. You can see that 23 back in March 2016. In May 2016 we put it in the 24 program.

25 You can see the mark -- the ML numbers at NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

17 1 the end there. If you ever want to go into the ADAMS, 2 and you can see all these documents.

3 We also have a generic issues' dashboard.

4 It's on the public website. You can follow the issue.

5 It can show you where we are.

6 All these information documents are tabbed 7 at the bottom there under documents. And then once 8 you get into regulatory office space, you can see 9 which plants are affected.

10 You know, how they're progressing. What 11 actions they need to take. What steps need to be 12 taken to take the generic issue to completion. It's 13 a pretty good tool.

14 All right. Once we get through this 15 screening process, you can see May 2016 we did a 16 generic issues program staff screening. Which is just 17 myself and my department and my supervisor and 18 division director signing off to see if there's any 19 specific screening criteria not being met. I can 20 screen it out.

21 So if this only applies to one plant in 22 the United States, it doesn't meet the criteria of two 23 or more plants. It's not a generic issue.

24 So, I or my supervisor at that point 25 cannot waste putting together a committee. Review NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

18 1 panel, they call it G-I-R-P, GIRP to waste time and 2 effort.

3 So it's very economical this program. We 4 take various steps to say push it forward or step 5 back, depending on the criteria.

6 So, we took a look at it. And we said, we 7 think it meets all seven criteria. We formed a review 8 panel.

9 We did form a review panel. And you can 10 see the screen report produced by the GIRP was on July 11 2017.

12 So you can see we've gone through about a 13 year to get the screening report done. And now we're 14 progressing into the assessment stage.

15 We issued an assessment plan. And you can 16 see up there. The -- if you want to take a look at 17 that.

18 It pretty much highlights what the 19 activities are going to be done during the assessment 20 stage. And it gives it kind of a time line so we can 21 kind of hold people, you know, responsible to get 22 something done at a certain time.

23 If we think it's going to be protracted 24 and go beyond two, three, four or five years, then we 25 might pull back and say we might take it out of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

19 1 program. Put it in a study. And then come back in.

2 So that's what the assessment stage does.

3 We try to put a time. And we put milestones. And we 4 try to meet those and keep people to those milestones.

5 Okay, next.

6 What you can look forward with this 7 generic issue coming up. We're going to put out -- or 8 hope to put out an assessment report near the end of 9 the year.

10 We should get all the tasks and milestones 11 done. And if it meets the risk and safety criteria to 12 go on into the regulatory office, we think it's a 13 safety or risk important item, then we formally call 14 a transition team.

15 We don't want to just hand it over to 16 another department without a smooth transition. So we 17 do a transition team.

18 We put members on the team that are on the 19 GIRP review panel. And then we put people from the 20 office it's going to on the team.

21 And we hand over all the information. And 22 the Office Director will send a letter to the other 23 Office Directors saying, you have it. And you have 24 sufficient resources, you know, knowledge about the 25 issue.

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

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

20 1 So we just don't want to hand it over to 2 -- across the fence and say it's now -- it's yours.

3 So, it's a very smooth transition.

4 Once it's transitioned to the regulatory 5 office, the regulatory offices, like I said before, 6 will determine what regulatory actions need to be 7 taken.

8 It could be, like I said, INs, GLs, 9 generic letters. It could be even orders issued by 10 the Commission.

11 They're looking forward to see if there's 12 anything that the industry needs to do. To look and 13 see if there's any, not loopholes, but missing 14 regulatory items and regulations that need to be 15 adjusted.

16 And once they've determined that, they put 17 out that information to the industry. They could put 18 it out in a 54F letters to say industry, here it is.

19 And you know, respond back as appropriate.

20 And one industry has taken all the 21 actions, we review them. We verify them. And then we 22 close out the generic issue.

23 I think that's the last slide, right? Oh, 24 one more. Okay.

25 All right. The assessment plan. This was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

21 1 a -- you can find this inside the assessment plan.

2 And this is a very good overview of what milestones 3 and actions are being taken during the assessment.

4 You can see which parts have actually been 5 completed. And you can see what actions we plan on 6 going forward.

7 I will ask Nick to chime in a little bit 8 on what activities he foresees and adjustments to the 9 schedule.

10 MR. MELLY: This is Nick Melly in the 11 Office of Research. I just wanted to kind of key in 12 on some of the important things that you're seeing up 13 here in terms of the time line.

14 And in terms of some of the things that 15 Stan was talking about in terms of the risk numbers 16 that are used in the assessment plan. And what causes 17 things to move forward. And what causes things to 18 fall out of the generic issue plan.

19 So, currently right now, we're looking up 20 there, and we're right around the first of the year 21 we're looking at the Phase Two, large scale test plan.

22 Right now we've finished four tests since September on 23 the medium voltage.

24 And we're looking to also get low voltage 25 testing done and bus duct testing done so that we can NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

22 1 have the information we need in order to create a zone 2 of influence model for aluminum moving forward for 3 these risk studies.

4 That's very important that we have all the 5 data we need. So that we're not cycling the industry.

6 Or we're not creating overly conservative models for 7 the damage states.

8 To accomplish that, the NRC is working 9 with EPRI to create the zone of influence. And we're 10 going to have discussions here about a charter that's 11 in the process so that the NRC can look at all the 12 data with EPRI, and create a zone of influence that's 13 appropriate for aluminum taking into account many 14 different factors.

15 We want to look at the frequency of the 16 event occurrence. Couple that with plant 17 configurations such as a unit connected design and 18 things like that, in order to create a good picture of 19 what the zone of influence is.

20 We do not want to move forward in the 21 generic issue with a conservative model giving us 22 conservative risk estimates, because it's just not 23 realistic.

24 That's where the pilot plant selection, 25 the pilot plant studies come into play. We really NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

23 1 need help from industry because if we use our in-house 2 modeling techniques such as the SPAR to move things 3 forward as we have done for quick assessments, it may 4 be conservative because there are a lot of large 5 assumptions that have to be made.

6 If we can leverage the plant design, the 7 plant PRA models when we move forward using pilot 8 plants, we can get a better picture of what the risk 9 is at the plants.

10 And we want to link that with moving 11 forward through the generic issue, so that we can 12 adequately assess the risk at plants.

13 It's an aggressive schedule right now.

14 And we believe that we can accomplish these. We have 15 a lot of things moving in parallel in terms of the 16 testing, the data analysis with the working group.

17 We believe we are close to being able to 18 select pilots, or hope to get pilot plants in this 19 process, so we can start figuring out what needs to be 20 done in order to adequately assess the risk. Which is 21 necessary for the assessment plan right now.

22 MS. ANDERSON: Are we -- are we taking 23 questions?

24 MR. GARDOCKI: One last comment. Like I 25 said, the dashboard has all the information as -- in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

24 1 regards to this generic issue.

2 You can always refer to that on the NRC 3 public website. And in the future, we try to hear 4 periodic meetings, you can probably expect another 5 meeting right before we start the assessment report.

6 And we're going to have a public meeting 7 with ACRS if they want to have a meeting to introduce 8 this. And they kind of get involved with the 9 regulatory analysis part of the assessment report.

10 So he can expect those as future 11 activities as far as generic issues program public 12 involvement. That concludes my presentation.

13 MR. HAMBURGER: Yes. So we will welcome 14 questions from those in the room first.

15 MS. ANDERSON: Victoria Anderson from NEI.

16 Nick, you mentioned that one of the things you'll be 17 doing is working on making sure you have all the data 18 you need to develop an aluminum zone of influence.

19 Are you looking at developing a single 20 zone of influence? Or a spectrum of ZOIs based on 21 different physical conditions?

22 MR. MELLY: A spectrum.

23 MS. ANDERSON: A spectrum?

24 MR. MELLY: We're not trying to just do a 25 one worst case zone of influence approach for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

25 1 aluminum. We're trying to link it in more a 2 probabilistic manner so that we can look at the 3 potential at the plant in order to get a more dynamic 4 zone of influence based on important parameters such 5 as the current, the voltage, plant design.

6 We don't want one conservative zone of 7 influence that would be used for the study. We're 8 really trying to get a -- to move the method forward.

9 MS. ANDERSON: Okay.

10 MR. TAYLOR: And if I could -- this is 11 Gabe Taylor from the NRC. If I could just add to 12 that.

13 At the workshop last April we talked about 14 different approaches that we could do. And I think 15 the idealized approach would be the dynamic ZOI that 16 Nick alluded to.

17 It's similar to what the IEEE does for 18 their arc flash calculations. So if we could get 19 there, that would be great.

20 If we can't, there's other fall back 21 options where we basically bin different categories 22 based on the hazard and the frequencies. And have a 23 ZOI for each individual bin.

24 And then, you know, the lease detail would 25 be something like we have today. Which is, you know, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

26 1 a one size fits all, which is kind of not where we 2 want to go.

3 But what path we go hasn't been determined 4 yet.

5 MR. MELLY: And it's actually very 6 important to understand that when even looking at the 7 test program that we have developed. Because we are 8 testing at two seconds and four seconds.

9 And for the low voltage cases, we're going 10 to be doing minimal tests at eight seconds. But we're 11 not doing -- we're doing that for various reasons just 12 so we can look at the spectrum of possible fall cases 13 in the plant.

14 We're not trying to say, the four second 15 test had this damage state. So now we're going to 16 link all zones of influence to that four second test, 17 which was the worst case.

18 That is not what we're trying to do with 19 the test program.

20 MS. ANDERSON: Okay. Are we going to be 21 talking more later about the considerations that go --

22 that will go into that development of the ZOI 23 spectrum?

24 MR. MELLY: Right now I believe we're 25 trying to work that through the EPRI working group --

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

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

27 1 MS. ANDERSON: Okay.

2 MR. MELLY: Charter. In that that is 3 essentially what the whole conversation of the charter 4 is going to be.

5 MS. ANDERSON: Right.

6 MR. MELLY: Is that -- we're going to be 7 looking at the data that we have in-house, and coming 8 up with a method to develop that.

9 MS. ANDERSON: All right. If we don't get 10 into it in detail today, I think it may be useful to 11 have another public interaction on that topic. Even 12 if it's outside of the generic issue process. I think 13 there may be some substantial technical input that 14 stakeholders may have.

15 MR. MELLY: I agree. And we've had 16 initial discussions already that we're -- we're 17 looking at the data now. And we've just developed 18 this charter.

19 We do plan on having several more 20 interactions with the public when we move a little 21 more forward in the process of developing an initial 22 method. Looking at the data.

23 We plan on sharing the working group 24 conclusions or the working group assessment with the 25 public before any finalized method is developed.

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

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

28 1 MS. ANDERSON: Okay. Thank you.

2 MR. HAMBURGER: Thanks Victoria.

3 MS. UHLE: Hi. This is Jennifer Uhle from 4 NEI. So, my question here is talking about this 5 approach, which is great to hear.

6 However, the amount of test data you would 7 need to really develop that doesn't seem to jive well 8 with the, you know, having the, you know, four tests.

9 So, -- or then I would ask in reverse, if 10 you plan on developing these zones of influence, then 11 shouldn't that inform your method in how you develop 12 them?

13 Shouldn't that inform the tests that you 14 do run? So it -- you know, where are we in likelihood 15 of getting to this state?

16 MR. MELLY: So to respond to that, we did 17 hold a PIRT panel to discuss the test results of the 18 Phase One testing.

19 And that was really to inform the Phase 20 Two of testing of what we would need to do in order to 21 create this spectrum of zone of influence, and what 22 data we would need.

23 That's really what went into the 24 development of the test matrix that is in the current 25 test plan.

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

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

29 1 And while there have only been four tests 2 performed right now on the medium voltage at two and 3 four seconds and various currents, we're really trying 4 to isolate parameters in this test series so that we 5 can look parameter to parameter differences as we move 6 forward.

7 Also, we are looking at the next series of 8 testing this summer to have four additional tests at 9 low voltage, various parameters. As well as four 10 tests looking at the bus ducts.

11 Additionally, we're trying to look at when 12 we can perform decrement curve testing, which I'll get 13 into a little bit later, in order that we can have 14 enough information to supplement the operating 15 experience data that we do currently have in-house, as 16 well as a wealth of information from previous testing 17 at shorter durations using the IEEE 1584 data and 18 things like that.

19 So we're trying to bring all of that 20 together to create this. But as you say, it is a 21 limited number of tests, but it is a resource issue.

22 Testing is fairly expensive, fairly slow, 23 and there's a lot that goes into the testing itself.

24 MR. SALLEY: Yeah, Jennifer, and just to 25 put the final piece on that, as a researcher you know NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

30 1 you can never have too much data. The more data the 2 better.

3 I think we've -- from what we've worked in 4 the April workshop, we have a pretty good matrix.

5 Which we'll talk about in a little bit, as to how we 6 parse this out.

7 And at the end of this afternoon's session 8 on the working group, you'll see how that it fits into 9 the NRC's scheme with the OECD NEA and how we bring 10 that to bear too, with it.

11 So, we have parts long presentation. And 12 I think it will provide you more information.

13 MR. MELLY: Yeah. And this was also one 14 of the main issues that was brought up by the GIRP 15 panel for the generic issue.

16 Is that we didn't want to come out with a 17 conservative zone of influence for aluminum before we 18 did this assessment plan. So, the real discussion by 19 the members was, what test do we need to accomplish in 20 order to create a realistic zone of influence?

21 And it was decided that we needed medium 22 voltage tests, both low voltage and bus duct testing 23 if we wanted a whole picture of what was going to be 24 necessary for the working group to establish 25 something.

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

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

31 1 MS. UHLE: Okay. And then you're, I would 2 expect working with EPRI to plan these tests and 3 ensure that they are prototypic in their set up?

4 MR. SALLEY: Yes. Like I said, we -- in 5 April we've done that. And as a matter of fact for 6 the September tests, we had members from EPRI, witness 7 the testing with us.

8 And that and we plan to do the same in the 9 spring and fall. That we have the working group that 10 will be there to see the tests and get the information 11 first hand.

12 MS. UHLE: Yeah. It's not so much the 13 witnessing the test as is, you know, taking the 14 comments that are proposed and factoring them in then, 15 rather than dispositioning.

16 You know, going forward I think that would 17 have to be done to really get to the point that we 18 would, I think, all like, with the multiple zones of 19 influence.

20 MR. SALLEY: Yeah. And I agree. And it's 21 one of the reasons we made all the public comments and 22 how we changed the test plan as a public document.

23 So you can -- the ML number is out there.

24 You can see the effects it had from the meeting. And 25 the changes it made.

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

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

32 1 You know also, one of the things is, we 2 never really were looking at the decrement curve. And 3 that was something that Ken brought up in the -- in 4 the April meeting.

5 And we've expanded our test program to 6 include that for the value we hope to bring to the 7 testing. So we are listening.

8 MS. UHLE: Thanks.

9 MR. HAMBURGER: Thanks Jennifer. I know 10 we have a question on the webinar. Do we have any 11 more questions from those in the room before we jump 12 over to the webinar?

13 (No response.)

14 MR. HAMBURGER: Okay. And just before we 15 go to the webinar, both Nick and Stan will be here all 16 day.

17 So if any questions occur to you as we 18 move forward, please you can jump in and ask them or 19 find them at one of the breaks. Yes. You may.

20 MR. AIRD: We have one question from the 21 webinar from Mohammad Mustafa. He asks, did the GIRP 22 review include existing codes and standards?

23 For example, National Electric Code, 24 IEEE's et cetera, related to aluminum? Are they 25 inadequate?

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

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

33 1 MR. SALLEY: I'll start the answer and 2 I'll let me guys finish. You know, one of the things 3 that we did when we were doing internal briefings, 4 when Fred Brown was the DEDO and we first started 5 going through the internal briefings with this, you 6 know, Fred made the comment to us that the high energy 7 arc faults are not unique to nuclear power plants.

8 You know, this is an electrical 9 phenomenon, you know, it's not unique. Now granted in 10 a nuclear power plant, you're all familiar we do move 11 a lot of electricity around. We've got a lot of 12 things happening.

13 So, it may amplify a little bit, or we've 14 picked up on it. But, one of the things again with 15 Fred, was to reach out and to see that very question.

16 If you remember the April workshop, we had 17 the chief electrical engineer from the NFPA come down.

18 And of course in the common vernacular in the industry 19 is arc flash.

20 You know, you'll hear this referred to a 21 lot as arc flash and arc flash events. So, we are 22 working with the NFPA.

23 And again, I think too, with the test lab, 24 the different scenes that we're taking it a little 25 further then the arc flash events that commonly occur.

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

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

34 1 And you see them here.

2 We see them in the Metro in D.C. We've 3 lost it. The Atlanta airport has had the one a couple 4 of years ago.

5 Around Christmas you just saw it in New 6 York City with the blue in the sky again. So, these 7 events are happening out there.

8 We're going to talk a little bit about the 9 differences in how we bin them in '16 with that. So 10 that will be part of the discussion with looking 11 between an arc flash, an arc blast, and a HEAF.

12 And again, working with the NFPA, one of 13 the things Ken Miller and us have worked on, is 14 definitions. To clearly define which of the different 15 phenomenon.

16 So, that's kind of where we're at.

17 Anything you guys want to add?

18 MR. TAYLOR: This is Gabe Taylor, Office 19 of Research. So, I was on the GIRP panel.

20 And the information that we looked at was 21 the operating experience that we communicated in the 22 Information Notice 2017-04. As well as other 23 international operating experience that was documented 24 in the OECD reports.

25 In addition to that, Nick mentioned, we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

35 1 looked at the SPAR models, SPAR all hazards model to 2 see what the potential risk increase could be.

3 Understanding that they are somewhat course, and not 4 as detailed as the plant's PRA models.

5 We also looked into the NFP 805 transition 6 plants, into their SEs and information available there 7 to see what the potential HEAF contributions were.

8 And that's not -- that's using the 9 existing models, so it wouldn't include the 10 increasing, if there is an increase for the aluminum 11 type HEAF.

12 So that was kind of the core information 13 that we looked at to evaluate the HEAF and the 14 screening, the aluminum HEAF and the screening phase.

15 As far as the NEC, that's more of a 16 utility specific, whether they follow that code or 17 not. And I really can't make any comment to whether 18 it's adequate.

19 MR. MELLY: This is Nick Melly in the 20 Office of Research. Yeah, as far as the NEC codes and 21 the IEEE, they did a lot of the wealth of their data 22 collection on copper.

23 They haven't specifically looked at 24 aluminum. They do have some theoretical comparisons 25 between copper and aluminum.

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

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

36 1 But their main focus is on personnel 2 protection. Whereas we are looking a little beyond 3 that for extended duration faults.

4 They are following closely what we're 5 doing in terms of definition space. And they're --

6 we're going to be looking at their data in a little 7 bit more detail as we move forward.

8 But in terms of saying whether their 9 regulations are inadequate based on what we've seen, 10 we can't really make that determination.

11 MR. SALLEY: And I guess a final thing on 12 the aluminum, Frank Cielo is here. Frank, if I could 13 ask on you, you do testing every day.

14 Is there any insights or anything you can 15 provide to that?

16 MR. CIELO: Yeah, this is Frank Cielo from 17 KEMA labs. I was just talking to my colleague Alex 18 Feldman.

19 We do arc flash testing on a regular basis 20 for lots of manufacturers. And it is absolutely for 21 personnel protection.

22 And we've done your testing of course.

23 And we've done high energy arch fault testing for some 24 other customers.

25 It's just completely -- the consequences NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

37 1 are just -- they're completely different. In orders 2 of magnitude different.

3 So, and one of the things you may want to 4 think about is showing somebody a video of a regular 5 arc fault test for, you know, to comply with the IEEE 6 standard versus the HEAF. It's just like, pfft. I 7 mean, it's so obvious, the amount of destruction. The 8 amount -- the impact on our laboratory, the physical 9 impact on our laboratory and what actually happens, 10 it's just -- it's just absolutely incredible.

11 And you know, we test for every -- we're 12 an independent test lab, all right. We're ISO/IEC 13 17025 certified.

14 We have no affiliation with any 15 manufacturer, anybody. And we test for every 16 manufacturer basically on the planet between our three 17 laboratories.

18 And most of the arc fault testing, we do 19 in the U.S. against the IEEE standard. And you really 20 have to see it to kind of appreciate what this is.

21 Because the first time we did this for the 22 NRC, we weren't sure we were going to continue, 23 because of just the amount of destruction that we 24 experienced at our lab. And the damage that we 25 experienced at our lab.

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

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

38 1 But it's important testing. And it needs 2 to be done. And we, you know, we just continue to 3 work through it. Hope that helps.

4 MR. GARDOCKI: And this is Stan Gardocki.

5 Back to your original question about the generic 6 issues program involvement with industry standards.

7 We do look at the NRC endorses the 8 industry standards to the regulations. So we'll find 9 out if there's any adjustments that need to be made in 10 the industry standard.

11 And then they have to go back and re-12 endorse it for the regulations.

13 MR. HAMBURGER: Before we get to you, we 14 have a follow up question from the webinar.

15 MR. AIRD: Yeah. We have one follow up 16 question. It goes, is this an electrical circuit 17 protection issue? Or an aluminum conductor issue?

18 There are fast acting devices to preclude 19 arc flash and protection schemes.

20 MR. TAYLOR: So at this stage we're trying 21 to understand the hazard. We are aware that there are 22 some protection schemes out there that are available 23 commercially that could be use to fastly open up a 24 protective device and clear the fault and the cycle, 25 a few cycles.

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

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

39 1 So, we're aware of that. And you know, 2 once we understand what the hazard is, and the 3 frequence is, and understand what -- where the risk 4 falls out in the delta risk, then we'll be able to 5 have a clearer picture of where it goes moving 6 forward.

7 And if it does make it into the regulatory 8 office limitation stage. At that point there would 9 likely be a lot of involvement between the NRC and 10 industry in understanding the ways to protect the 11 plant from the type of hazard, whether that be 12 prevention, protection, or mitigation features.

13 So, we are aware of those technologies.

14 But as far as this program, we're not trying to 15 evaluate the adequacy of those feature devices.

16 And then regarding the comment on 17 aluminum, you know, we did see the two tests in the 18 first phase that showed some much larger damage areas.

19 There were a few aluminum tests that didn't create 20 any, you know, additional hazard, then what we're 21 seeing in the copper.

22 They actually didn't even hold the fault 23 in. So, not all aluminum type events may lead to some 24 type of larger damage footprint then what we're seeing 25 form the copper.

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

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

40 1 But from the September tests, when we ran 2 those, the four tests that we ran, and Nick will talk 3 about those in the next presentation, all four -- they 4 were medium voltage. But all four did last for the 5 entire duration.

6 And we did see, you know, a much larger 7 damage footprint. And more so the energy release, 8 because the energy is related to the duration of the 9 arc.

10 And when you're running them for two to 11 four seconds, there is a difference between that.

12 MR. MILLER: This is Ken Miller, 13 Electrical Research. As Gabe said, we've about 14 recognized that protection schemes in fact are 15 supposed to clear these kinds of faults within very 16 short periods of time.

17 We understand that. And but of course 18 operating experiences has shown that sometimes that 19 doesn't occur. There's failures that occur. And 20 these kinds of HEAFs, you know, occur as a result of 21 that.

22 And again, the important point here is to 23 study and understand how destructive those can be as 24 far as effecting the zone of influence.

25 So, again, we understand that protection NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

41 1 normally would take case of something like this. But, 2 we're seeing cases where it doesn't.

3 MR. MELLY: Yeah. And to the second part 4 of your question as well, this is Nick Melly from the 5 Office of Research.

6 In terms of aluminum, we're also not 7 saying at this moment in time that the -- if you have 8 aluminum as part of your conductive pathway that it's 9 more prevalent to have a higher G arching fault.

10 In terms of frequency, we're not seeing 11 that from the data yet. It is an issue that we're 12 going to be looking at when we take a deeper dive into 13 the data.

14 But the frequency of occurrence from 15 aluminum versus copper, right now we're not seeing as 16 a change.

17 MR. HAMBURGER: Okay. Let's take the next 18 question from the room, please.

19 MR. FLEISCHER: This is Ken Fleischer from 20 EPRI. Not so much a question, but I did want to share 21 some OE of one major U.S. manufacturer of switch gear 22 in regards to type testing of copper bus bar switch 23 gear versus aluminum bus bar switch gear back in the 24 '70s.

25 Now again, to reemphasize that these were NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

42 1 two second tests, because the IEEE standard requires 2 switch gear to be rated to two seconds. They noticed 3 no discernable difference between the two tests.

4 The only thing that the test did to 5 demonstrate was that the aluminum bus bar would have 6 to be a little larger than the copper bus bar with a 7 little bit of different spacing. But they did not see 8 any of the extra aluminum byproduct from those two 9 second tests.

10 We are also working to contact a few more 11 additional manufacturers. But we did get one that 12 actually told us they did the type tests with 13 aluminum.

14 MR. HAMBURGER: Thank you Ken.

15 MS. VOELSING: Hi, this is Kelli Voelsing, 16 also from EPRI. So, I think we need to clarify so 17 that we're not mixing an apples and oranges 18 discussion.

19 The reason that Ken shared that 20 information that we received from a manufacturer 21 regarding the difference between aluminum and copper 22 type testing, is because from the comments from the 23 gentleman from KEMA, we assumed when he said the 24 testing is completely different, that he was referring 25 to testing between copper and aluminum.

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

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

43 1 In offline discussions, he clarified that 2 he was talking about the difference between flash 3 testing and arc fault testing. So the duration of 4 testing.

5 So I did want to clarify that apparently 6 he was intending to communicate a difference in the 7 duration of tests. Not a difference between copper 8 and aluminum, which is the generic issue that we're 9 supposed to be talking about here today.

10 And I'd also like to clarify that I think 11 we have to be careful about referring to the damage 12 footprint. Because we haven't necessarily validated 13 any damage.

14 What we've validated is, you know, 15 byproducts being deposited at a certain radius. Not 16 damage.

17 MR. TAYLOR: Thank you Kelli. And that's 18 a -- that's a good point. I was searching for the 19 right words.

20 And it's really the observations from what 21 we saw between the copper and the aluminum, where at 22 the two aluminum tests, well, six aluminum tests now 23 showed the larger area.

24 MR. CIELO: Yeah. This is Frank Cielo.

25 Thank you for clarifying that. I didn't intend to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

44 1 cause any confusion there.

2 It was the difference between the regular 3 arc fault testing and this testing. Thank you for 4 clarifying that.

5 MR. HAMBURGER: Okay. Any more comments 6 from anyone in the room?

7 (No response.)

8 MR. HAMBURGER: Tom, do we have anything 9 else on the webinar?

10 (No response.)

11 MR. HAMBURGER: Okay. We are ahead of 12 schedule. So Nick, if you're ready to go, we'll talk 13 about those preliminary observations.

14 MR. MELLY: All right, next slide. Again, 15 this is Nick Melly from the Office of Research. I'm 16 going to be discussing the testing that was performed 17 in September.

18 Kind of a big picture look at some of the 19 observations, some of the preliminary insights and the 20 project goals.

21 I'm also going to be talking about some of 22 the challenges, the measurement devices that we're 23 using, and the general observations.

24 I also want to discuss the future testing 25 parameters and schedules and some important milestones NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

45 1 as we move forward. Next slide.

2 So to start this all off, we did share the 3 draft test plan to our OECD NEA members, as this 4 initially started as an international project on June 5 30 of 2017.

6 We also put this out for public comment in 7 the Federal Register in August for a month long public 8 comment period. That closed on September 1, 2017.

9 Next slide.

10 As part of that initial test plan, we did 11 receive five comments from NEI. And we had 32 12 comments from our OECD members.

13 Since then we have received numerous 14 comments both from NEI, EPRI, as well as some of the 15 industry itself. Which we have discussed at the 16 public workshop that we held in April 2018.

17 And we've addressed all of these comments 18 and dispositioned them. Formally put that into our 19 public document system, ADAMS, essentially addressing 20 all comments and how the NRC planned to address them.

21 Whether they were going to be put into and 22 incorporated into the test plan. And if any physical 23 changes were made to the test plan, both parameter and 24 testing method. We have the ML link that is up there.

25 So right here on the screen you're seeing NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

46 1 the Phase Two HEAF test program. This is the matrix 2 that we discussed earlier.

3 This came about after the PIRT panel 4 phenomenon ranking table -- parameter identification 5 and ranking table exercise that we held in January 6 2017.

7 This was our best guess of, these are the 8 amounts of tests that need to be conducted in order to 9 answer the question and get the data that we need.

10 What you're seeing on the screen here is 11 four box tests. Test 2-19, 2-21, 2-22, and 2-24.

12 These are the tests that were performed in September 13 that I'm going to be discussing today.

14 These were all medium voltage, 6.9 kV 15 tests at 25 kA and 35 kA for two seconds and four 16 seconds respectively.

17 Another aspect that you're going to see on 18 the slide here, you see the two tests that are on the 19 far right-hand side with Xs. These were identified 20 that we may need to do additional testing.

21 And we wanted to be budget ourselves to be 22 able to perform tests as things came up.

23 MR. TAYLOR: Nick, just to clarify. This 24 is the -- showing 35 kA for the 6.9 kV test. We 25 actually tested at 32 kA.

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

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

47 1 So this was from the original test plan.

2 MR. MELLY: And that is because the test 3 facility itself, there are some limitations on the 4 capabilities of the generator. So we actually had to 5 reduce the current in order to achieve the duration of 6 testing that we wanted to achieve.

7 The two tests that I was discussing 8 earlier, the SPAR tests on the right-hand side, we 9 have allocated those for the decrement curve testing 10 that was a direct result of the comments from EPRI.

11 These are going to be tests where we're 12 actually going to be simulating actual operating 13 experience event at a unit connect design plant where 14 the generator essentially fed the fault.

15 This is not a trivial exercise. We have 16 a separate contracting action with the laboratory in 17 order to develop the capabilities needed to do this 18 test.

19 KEMA has never done a test like this 20 before. So, getting their generators ready to 21 simulate this, is a separate effort. Which is why we 22 did not do them in September.

23 And we're currently pushing that forward 24 right now, working with EPRI to establish what needs 25 to happen. And the KEMA test engineers.

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

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

48 1 Another thing that you'll notice on this 2 test matrix is that there are some tests in blue and 3 some tests in orange. We're doing this as part of an 4 international program.

5 The tests in blue represent tests that are 6 going to be funded through the OECD NEA test program.

7 And largely these are copper tests versus the ones in 8 orange, which are the aluminum tests.

9 We're performing the orange tests first, 10 because they're associated with the need of the 11 generic issue program itself. So we are prioritizing 12 the generic issue and performing the test that you see 13 in orange before we do any OECD NEA tests.

14 Next. I'll take questions during the 15 presentation. So, sure.

16 MR. HAMBURGER: Sure. This is a long 17 presentation. So if you --

18 MS. UHLE: Okay. Thanks. Jennifer Uhle 19 from NEI. So, I want to go back to a question that I 20 previously asked.

21 And I think in your discussions about this 22 test -- particular test series -- excuse me, I have a 23 cold. You indicated that the industry comments, the 24 EPRI comments, you dispositioned.

25 And this is the point about that we really NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

49 1 want to make or get across so that it's clear to 2 everybody. The industry comments that were given to 3 NRC, they were, I agree, dispositioned.

4 But that does not mean addressed in a 5 substantial manner. There were numerous comments that 6 the industry felt essentially affected the proto-7 typicality of the tests that were -- we got the 8 answer, we'll deal with this in a next test series.

9 And so, we've been saying that industry 10 has -- comments have been addressed. But, that's not 11 our opinion with regard to the test series.

12 And that's why when we want -- we want to 13 highlight the importance of truly listening to the 14 comments. And not dispositioning, meaning, I heard 15 you. I don't agree with you. And I'm not changing 16 the test set up.

17 MR. MELLY: Do you have any specifics in 18 terms of which comments were not addressed?

19 MS. UHLE: Yeah. And I'm going to --

20 MR. HAMBURGER: Well, let's -- if we're 21 going to discuss those, I suggest we wait for the 22 public comment period. Because that might be a 23 detailed discussion.

24 And so it's --

25 MS. UHLE: Well, I just -- if you don't NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

50 1 mind, this can be very quick if we just -- and then 2 maybe we discuss it later.

3 MR. HAMBURGER: Okay. Sure.

4 MR. FLEISCHER: Ken Fleischer from EPRI.

5 One particular comment that actually we also 6 verbalized during the April workshop, is we still 7 didn't really see the benefit or -- really the benefit 8 for the low voltage eight second test.

9 Unless that test could be justified by 10 some means. And I know that it was originally 11 mentioned at the April workshop that it was probably 12 going to be discontinued and go back to the four 13 seconds.

14 But we see that it's still considered as 15 eight seconds. I don't know if that ever really got 16 finally dispositioned as one example.

17 MR. MELLY: So this has been 18 dispositioned. And the final test plan has not been 19 released, because it is part of the OECD agreement.

20 However, you still see the eight second 21 test as part of the low voltage. We did reduce the 22 number of eight second tests.

23 However, we've added to the test plan that 24 the eight second tests are not going to be used to 25 inform the eight second or the low voltage zone of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

51 1 influence spectrum.

2 However, we are trying to still do an 3 eight second test only in term -- so we can draw 4 comparisons between the low voltage and medium voltage 5 tests where we might see a medium voltage case hold 6 for longer than four seconds.

7 So the eight voltage second test is kind 8 of just a place holder, because we are at the limits 9 of the capabilities of the KEMA facility. So we 10 cannot do an eight second test at KEMA for a medium 11 voltage case.

12 MR. FLEISCHER: That's interesting.

13 MR. MELLY: However, we're trying to draw 14 some comparison of what we see from the low voltage to 15 what we can maybe expect to see in a medium voltage 16 case where it does last for potentially longer then 17 four seconds.

18 So, we have made that distinction in the 19 test plan which we plan on releasing very shortly.

20 That is the only reason that you see that eight 21 seconds on the low voltage case currently.

22 MS. VOELSING: Kelli Voelsing from EPRI.

23 And Ken can probably speak to this a lot more 24 eloquently.

25 But, I'm not -- we see in the OE very NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

52 1 different situations that lead to potentially high 2 energy arc faults in low voltage versus medium voltage 3 equipment. So very different set of conditions that 4 lead to those.

5 And so I'm not sure that there's a solid 6 technical basis for extrapolating the results of an 7 eight second low voltage test and saying that that's 8 somehow applicable or representative to what you might 9 see for medium voltage if the facility had the 10 capability to do that.

11 And I would say that the generator voltage 12 to k curve is another example of a comment. And I 13 realize, you know, you're planning to address that in 14 the future.

15 But, it's not going to be particularly 16 relevant to the low voltage testing that's up coming.

17 And so, is there a time line when those two reserve 18 testing places are planned?

19 Because it would seem that the results of 20 those tests might be critical to the resolution of 21 this issue. And I'm wondering how those are going to 22 happen before the planned time line for the first 23 presentation of resolving that by December.

24 MR. TAYLOR: Okay, Gabe Taylor from 25 Research. So, I'll touch on the last part of your NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

53 1 comment first.

2 The two orange boxes on the far right, the 3 optional tests, as Nick mentioned, we're going to save 4 those for the decrement curve evaluation. And later 5 in the presentation we're going to talk about the 6 decrement curve.

7 One thing that -- that's the type of 8 experiment that hasn't been done at KEMA, trying to 9 match that type of decay response from the generators.

10 Typically they use super excitation to maintain a 11 constant power source.

12 So because of that, there's a lot involved 13 on their end analytically. And also validating that 14 they'll be able to perform that type of testing.

15 So because it's not in our current 16 contract, and there's the work involved that it 17 requires us to go into our contracting process to make 18 changes that are new. A new contract to get the work 19 done.

20 So, through that process, it's not a --

21 it's going to take some time to get that in place.

22 MS. VOELSING: I think the complexity is 23 understood, I'm just asking about the time line, with 24 respect to the resolution of the GI that you propose 25 by the end of December.

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

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

54 1 MR. TAYLOR: Yes, I apologize for the long-2 winded explanation, I was trying to provide as much 3 detail as possible. So, when we get that process 4 completed, we'll be able to perform the testing.

5 So, because where we're at in the 6 contracting process phase, I don't have an exact time 7 frame for you. But if I were to guess, it would be 8 sometime later this year.

9 MS. VOELSING: Do we want to talk about the 10 extrapolation of eight-second?

11 MR. TAYLOR: Yes, that was the next place 12 that I was going to pick up. So, you remember from 13 our April workshop, we talked about some small-scale 14 testing and modeling. And over time, that has -- we 15 have changed directions and the process that we're 16 going through, because of some subject matter expert 17 changes.

18 So, currently, the plan is to work with 19 Sandia, using their area model, and basically, be able 20 to develop that to run simulations to estimate 21 different physical parameters, such as heat fluxes, 22 pressures, particle ejecta, that sort of thing.

23 And by being able to perform these small-24 scale, as well as medium-scale, tests, we could have 25 models and simulations that could -- basically, the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

55 1 data would validate the models and then, you could use 2 the simulations to develop ZOIs or that sort of type 3 of information that would be useful in informing the 4 eventual model that's used in the PRA risk 5 assessments.

6 MS. VOELSING: So, obviously, modeling and 7 simulation is always a valuable part of the process, 8 but developing models and validating them based on an 9 extrapolation, I'm not sure that that provides solid 10 technical basis for the models.

11 MR. TAYLOR: They actually already have the 12 model in use out there, and I believe EPRI is actually 13 using it for some of their DC work.

14 So, what we really need to do is provide 15 some changes to the sub-model, to accurately represent 16 what we're trying to characterize. So, it's a multi-17 pronged approach, to try to understand this hazard.

18 We see some value in performing the small-19 scale work, but also, regardless of that pronged 20 approach, there's also some value in having more than 21 two points and you're trying to extrapolate for low 22 and medium voltage.

23 So, while we did reduce the number of 24 eight-second tests that were originally planned in the 25 small-scale test, or in the full-scale test plan, we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

56 1 still see value in performing that type of duration.

2 And in addition, we're seeing some 3 operating experience that shows longer duration low 4 voltage OpE. However, I think we are in agreement 5 that the frequency of low voltage versus medium 6 voltage and the hazard that it contributes are 7 different between the two.

8 MR. HAMBURGER: If we have more questions 9 about the small-scale testing or about the resolution 10 of specific technical comments, let's hold those and 11 let Nick --

12 MR. MELLY: I'd like to answer one thing.

13 The extrapolation is important, we think, because we 14 just simply don't have the ability to do an eight-15 second test at the KEMA facility.

16 We are trying to match, we would like a 17 data point, so that we can develop this spectrum type 18 approach to the zone of influence models.

19 We did see some OpE, such as Robinson, 20 where it was between an eight-second to 11-second arc, 21 in medium voltage. We do not have the capability at 22 the KEMA facility to replicate this event.

23 So, we are looking at how we can make 24 comparisons between the low voltage energy release 25 versus the medium voltage energy release, in terms of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

57 1 duration. We do see that as a key important factor.

2 The PIRT documented duration as one of the 3 most important factors for the high energy arcing 4 fault and we're, again, seeing that from testing. So, 5 we do see limited number of low voltage tests as an 6 extrapolation point, an important test parameter.

7 Okay, moving forward.

8 I want to discuss some of the challenges 9 that we saw with the Phase 1 testing, and some the 10 measurement limitations that we're running into. We 11 had been testing with plate thermocouples, essentially 12 using an Inconel plate to measure the heat flux and 13 the temperature.

14 And you can see the picture on the left, 15 the before image, versus the picture on the right, 16 after it was subjected to Test 23, where we completely 17 melted the Inconel plate and were not able to record 18 any data at the three-foot location.

19 This was a limitation that we had not 20 anticipated and we were not able to get any data from 21 that test, because we damaged our instrumentation.

22 This was a main issue that we were trying to deal with 23 as we went towards Phase 2, is how are we going to 24 collect the data that we need to inform the model?

25 Next slide.

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

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

58 1 So, what we came up with with NIST was a 2 tungsten slug calorimeter, essentially, a more robust 3 measurement device using one-inch diameter tungsten 4 slug.

5 It was very durable and the hope was that 6 it would survive a direct impact from the high energy 7 arcing fault at the three-foot location and beyond, or 8 wherever we put these devices.

9 We were using a calcium silicate 10 insulating board, a Marinite type board, around the 11 tungsten slug, to essentially protect the device and 12 get the data that we needed.

13 You see a picture on the left there, that 14 is a picture of it under the heat release rate hood at 15 NIST itself, cone calorimetry, testing the principles 16 and developing that tool, so that we can use it in 17 Phase 2 testing. Next slide.

18 Another challenge that we faced with the 19 Phase 1 testing was, how to get the pressure results 20 from the events themselves.

21 We had a lot of interference with 22 electromagnetic interference for our pressure devices 23 that we were using in the Phase 1 testing, so we 24 didn't fully believe the results that we were getting, 25 because of this large interference.

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

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

59 1 Pressure is a very important aspect of 2 these high energy arcing faults. As we said, we can 3 have blast effects with these, where we can create 4 pressure waves within a compartments and potentially 5 damage boundaries between fire barriers, or between 6 fire areas, or have pressure effects on other 7 equipment.

8 A good example of this is the event that 9 occurred at Turkey Point in 2017, where we did breach 10 a fire barrier and damage a fire door between two 11 adjacent switch gear rooms.

12 This is one area that we're looking at and 13 we really needed to dial in on how we were performing 14 these pressure measurements, so that we can get a 15 better picture of what is going on. Next slide.

16 So, as we move forward to Phase 2, we 17 decided to move to a fiber optic measurement device.

18 KEMA has upgraded, essentially, their facility and 19 we're getting much cleaner results.

20 All the pressure measurements that we have 21 received in the Phase 2, the first four confirmatory 22 tests, do appear to have worked adequately and just as 23 we expected and things are looking good in that area.

24 Next slide.

25 We are also looking at video as data, as NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

60 1 part of these test series. It's a very fast event.

2 We're using a lot of different videography tools, such 3 as our infrared cameras, essentially to look at the 4 temperature of the event as a whole.

5 We are -- there are some challenges with 6 this, because it is such a quick event, and we're 7 seeing temperature ranges going from ambient to 35,000 8 degrees at the initiation of the arc itself.

9 So, there are some compromises that we 10 needed to look at, in terms of our infrared cameras, 11 such as the field of view, the speed, and the 12 temperature ranges that we can achieve. Next slide.

13 We have improved the infrared camera 14 technologies that we're using, both through NIST and 15 Sandia, trying to look at a dynamic picture of what is 16 going on during the test itself.

17 For our Phase 2 camera capabilities, we 18 have a greater temperature range and higher 19 resolution.

20 And all of these infrared techniques that 21 we're using will be shared, potentially, in a separate 22 NUREG report, that our colleague Dave Stroup is 23 working on currently, and we're going to be pushing 24 that forward to our Phase 2 test program. Next slide.

25 Another important tool that we're using NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

61 1 and that we have recently developed for this test 2 program is our data acquisition system.

3 Right now, we're using an isolated data 4 acquisition system with an independent power supply.

5 We've seen in other test programs that the arc itself 6 can short out the data acquisition.

7 And like we said, these are fairly slow 8 tests, we can only perform a few a week, and if we 9 lose a data acquisition system and all the data, it's 10 a complete waste of a test.

11 So, we have spent a lot of effort in how 12 we can create redundant systems in order to collect 13 all the data that we need. Our current system has 72 14 channels, we're only using about half of those 15 channels currently.

16 And I'll discuss different ways that we're 17 adding capability to this data acquisition system to 18 potentially collect more information, as we move 19 forward. Next slide.

20 So, what you're currently seeing on the 21 screen is our test setup for the Phase 2. We have 22 different instrumentation racks, located at three-foot 23 on all sides of the cabinet and above.

24 And additionally, we have a six-foot rack 25 located in the direction where we believe the arc is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

62 1 going to travel. We've had fair success in the past 2 with predicting the arc path, based on the electrical 3 setup of the cabinet itself and where we're initiating 4 the arc.

5 However, it's not an exact approximation 6 for every test case. Sometimes, the arc will tend to 7 migrate, as we've seen done in previous tests, so we 8 really need to have a way to collect information on 9 all sides of the cabinet itself, so that we do not 10 miss the arc ejecta and all the energy release, to 11 have a useful test in all cases.

12 Some of the challenges with this is that 13 it's a lot of instrumentation, it's a lot of data, and 14 it's a lot of time to set up the tests. Next slide.

15 What you're looking at on the screen here 16 is what one of those typical racks looks like and the 17 types of instrumentation that is on each rack. We 18 have ASTM E thermocouples that we're using, that link 19 to the IEEE standards.

20 We have plate thermocouples. We have the 21 tungsten slugs that we discussed. We also have a lot 22 of passive instrumentation on these test racks, such 23 as cable coupons, aerogels, and carbon tape, that 24 we're using to measure the ejecta from the event 25 itself, looking at the byproducts that are created at NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

63 1 different distances away from the arc.

2 We're hoping that all of this will feed 3 into the model that we select, as we move forward with 4 the working group.

5 You can see here, this is a lot of data.

6 So, we're trying to spend our resources wisely, we 7 want to work with our modeler and figure out what data 8 exactly they need before we go and post-process all 9 data and all collection points that we took during the 10 test. Next slide.

11 Additionally, we're working with Sandia, 12 trying to leverage a lot of their DOE advancements and 13 their camera capabilities, high speed capabilities, 14 leveraging that, so that we can use it for this test 15 program. We have various cameras located away from in 16 front of the test cell itself.

17 And additionally, we've developed a system 18 to put a camera inside the test cell, using mirrors 19 and, essentially, to look at the arc location without 20 damaging their expensive cameras.

21 Worked fairly well in the first series of 22 tests and we hope to continue that as we move forward, 23 because, again, the video data is very important as we 24 try and assess what happened during the event, both 25 the infrared, as well as the high speed. We're able NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

64 1 to understand things now that we were not able to see 2 in the first series of testing. Next slide.

3 Another thing that we're doing in this 4 test series is, we're doing the full weights and 5 measurement of all of the test equipment before we 6 test and after we test.

7 We want to get a picture of how much 8 material was vaporized, both in the conductive 9 pathways, the aluminum in the different phases of 10 power, as well as how much was vaporized of the 11 cabinet itself, if we had cabinet breach.

12 This is something that was not done in the 13 first series of testing and does take considerable 14 resources post-test arrangement. Next slide.

15 This is the piece of equipment that we 16 used in the first four tests that were conducted in 17 September. It was a single compartment GE-Magne-Blast 18 Metal-Clad Switchgear, where aluminum busbar was 19 located in the rear of the compartment that you're 20 seeing there.

21 This was the equipment that was discussed 22 in the April workshop and was selected with some 23 interaction with the industry. It was deemed to be a 24 typical cabinet that we would see in the U.S.

25 operating plants. Next slide.

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

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

65 1 This is the actual in-KEMA design setup 2 before one of the tests. You're looking at the 3 cabinet. The front door of the cabinet is located on 4 your left-hand side. You see pressure transducers 5 located on the sides of the cabinet, under that 219 6 symbol.

7 As well as, you're seeing our protected 8 instrumentation racks located, again, at the three-9 foot and six-foot locations, in order to collect the 10 information that we needed from the ejected arc.

11 This was a -- for each one of these test 12 series, you see the racks there, themselves, we could 13 not reuse a single instrumentation rack, because of 14 the damage from the event itself.

15 There is a considerable amount, about a 16 week of prep time that went into building these 17 instrumentation devices at the KEMA facility before 18 they are used.

19 So, it is a considerable amount of work 20 that goes into a single test and they're not reusable, 21 so that kind of explains why these test take as long 22 as they do to prepare for, as well as to run a single 23 test. Next slide.

24 What you're seeing here, and I'm not sure 25 if you can make it out clearly, is the arcing wire NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

66 1 that we're using to initiate the arc itself across 2 three phases of power. It's a thin -- Ken, can you go 3 over that with the mouse by any chance?

4 MR. TAYLOR: Yes, it's a 24-gauge single-5 strand tin-copper wire, per the C37.27 standard, IEEE. 6 MR. MELLY: And this was a change that we 7 made based on some of the comments we received during 8 the public workshop. We initially had been initiating 9 these arcs using the low voltage standard, because 10 they had a rationale for the ionization of the wire 11 that was created itself.

12 However, from one of the comments, we 13 decided to change that to the medium voltage case that 14 was used, and we went with a thinner gauge wire. That 15 wire was able to maintain and hold the arc in all 16 cases during the test week. Next slide.

17 So, here are the four tests that we did 18 perform. These are pictures of the rear of the 19 cabinet itself, post-test. The first upper left-hand 20 picture you see there is our Test 219. This was a 21 two-second arc at our 25kA.

22 And moving across, you have our two-second 23 arc at 35kA. On the bottom left, you have our four-24 second arc at the 25kA. And the bottom right is our 25 four-second arc at the 35kA, 32kA, I'm sorry.

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

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

67 1 You can see the progression in damage 2 states, however, we did breach the cabinet in all 3 cases. We do see this linking up with operating 4 experience, where an arc occurs in the rear of the 5 cabinet itself, it does create a cabinet breach.

6 From the video itself, we can actually 7 witness the metal in the back of the cabinet 8 essentially vaporizing and ejecting the energy towards 9 our instrumentation devices.

10 One thing to note here is that we do see 11 fairly repeatable results. That was a challenge when 12 we had equipment donated from a ton of other 13 countries, where we essentially didn't have the same 14 piece of equipment.

15 Which is why, in the second phase of 16 testing, we're trying to use the same make, model, and 17 initiate the arc in a similar location, so that we can 18 have comparable data as we move forward.

19 We were successful in locating our 20 instrumentation and we're using the arc initiation 21 point as our main parameter to decide where we 22 locating things.

23 In all four tests, we did see the door in 24 the front of the compartment open, due to pressure 25 effects. It was a single-latching door. We were able NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

68 1 to re-close the cabinets after the experiment itself.

2 We saw minimal bending, but we didn't see many thermal 3 effects at the front facing of the cabinet itself.

4 Next slide.

5 Another aspect of this I want to discuss 6 is some of the observations we're seeing from our 7 measurement devices themselves. We were again using 8 that calcium silicate Marinite board for our ASTM E 9 slugs, as well as our tungsten slugs. And we saw some 10 unusual effects occurred during the test series. Next 11 slide.

12 What you're seeing there is a before and 13 after picture of one of our tungsten slugs that was 14 located on the three-foot rack. On the right-hand 15 side, you see this green silicate glass-like material, 16 that has bubbled up from the Marinite itself.

17 And we're looking into trying to match 18 material property characteristics for the Marinite to 19 the temperatures of the exposed surface, with the data 20 that we've collected from that tungsten slug.

21 We have -- those devices did survive and 22 we were able to actually collect valuable temperature, 23 as well as heat flux information from the majority of 24 our tungsten slug devices.

25 So, we do see that as a success moving NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

69 1 forward in the ability to capture these high energy 2 cases at a close proximity to the cabinet. Next 3 slide.

4 Again, our plate thermocouples did not 5 survive the highest heat fluxes that we saw there, 6 which did expose us to a new problem that we had not 7 anticipated, which was once the plate thermocouple 8 vaporizes, the wires and the wire pathway that is 9 behind that device becomes exposed, which damages all 10 the other wires behind it.

11 So, there were some cases where we still 12 lost data because of the high energy that's released 13 from the cabinet, and we're currently working at ways 14 to enhance the protections, so that we do not lose any 15 of the valuable data. Next slide.

16 Again, another thing that we're looking at 17 to improve our cost-efficiency of the test program is, 18 we have been using the ASTM E slugs.

19 That is largely so that we can have a 20 direct comparison between testing that was done as 21 part of the IEEE work, as well as ASTM work. However, 22 we may see this as a redundant testing device, which 23 we do not need to use moving forward.

24 It's one area that we hope to bring up 25 with the working group, after we can look at all of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

70 1 the data for these devices, to see if we even need to 2 use these moving forward and maybe make each test a 3 little bit more cost-effective. They're fairly 4 robust, however, they did not survive the high 5 intensity arc itself. Next slide.

6 Another big ticket item that we're looking 7 at for this test program is the deposition of ejected 8 byproducts, as we move away from the cabinet itself 9 for these aluminum tests.

10 What you're seeing here in these two 11 pictures is some of that ejected material, which is 12 the white powder coating that is on the wall of the 13 test cell itself.

14 We did take black carbon tape samples of 15 this material and we're going to be performing post-16 test analysis at Sandia to look at exactly what that 17 byproduct was, both at the wall, on the three-foot 18 test stand, and the six-foot test stand.

19 One important note is that we didn't 20 really see this type of ejecta during the two-second 21 arcs that we had in this test program. The majority 22 or the bulk of the material was during the four-second 23 test.

24 So, that's another issue that we're 25 looking at, that these aluminum tests might not have NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

71 1 the amount of byproduct material ejected, depending on 2 the duration of the arc itself.

3 So, this was a question that came up just 4 recently in this series of testing that we're going to 5 be evaluating. We have a question?

6 MS. VOELSING: Yes, sorry, this is Kelli 7 Voelsing from EPRI, again. I'm sorry, I may have been 8 confused, but I thought the latest information I 9 received was that, although you took samples with the 10 carbon tape, you did not plan to have Sandia analyze 11 them, because of resources, and you were counting on 12 the small-scale test results to do the material 13 characterization. Did I misunderstand what I was 14 previously told?

15 MR. TAYLOR: Yes, there must have been some 16 misunderstanding. So, we put samples in a number of 17 locations. On every rack, we had a number of 18 different types of samples, aerogel, black carbon 19 tape, PVC samples, and we have hundreds of samples to 20 analyze.

21 There's probably five or six different 22 ways that can analyze each sample. And really, 23 putting the samples in the testing is not the 24 expensive part, the expensive part is the post-test 25 analysis.

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

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

72 1 So, what we're trying to do right now is 2 understand exactly what the model needs and what 3 samples need to be analyzed and how they need to be 4 analyzed, before we move forward and expend the 5 resources to analyze the samples.

6 So, they've done some preliminary, picking 7 off one on each rack and analyzing it for different 8 parameters.

9 But as far as analyzing the samples that 10 are needed to inform the working group's work or the 11 modeling work, we haven't initiated that type of 12 resource expenditure. So, we're trying to be as 13 efficient and as effective as we can to get the 14 information that's needed.

15 MS. VOELSING: So, to date, either from the 16 small-scale or from this testing, there's no material 17 characterization results available?

18 MR. TAYLOR: Not to the public, no.

19 MS. VOELSING: Or to the working group?

20 MR. TAYLOR: Or to the working group, yes.

21 They have done some preliminary analysis, but even on 22 the small-scale, the small-scale is really focusing on 23 the modeling.

24 And until they have -- because the changes 25 in direction on the modeling and the approaches, until NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

73 1 we have the modeling approach nailed down, we didn't 2 want to analyze the samples yet.

3 So, all the samples are still out at 4 Sandia, there's some preliminary information that was 5 provided to the NRC, as just a quick look, but as far 6 as a comprehensive analysis of the samples, we don't 7 even have that yet.

8 MS. VOELSING: It might be helpful to the 9 -- even to have the quick look, to know what the 10 material was characterized at. And will there be any 11 potential for stakeholder engagement on these models 12 that you're working on with Sandia? Because that 13 doesn't seem to be a part of any of the discussions.

14 MR. TAYLOR: I'll let -- do you want to 15 answer that or -- okay. So, yes, when we get far 16 enough down the line, I think that's, as Victoria made 17 mention, it would be very valuable to have feedback on 18 the modeling and what we're getting out of that.

19 MS. VOELSING: It might be helpful to have 20 feedback on the front-end, to make sure that what 21 comes out the back-end is appropriate.

22 MR. TAYLOR: Agreed.

23 MR. MELLY: And we do see this as a very 24 important piece, as we move forward. Again, the 25 material composition, as well as the conductivity that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

74 1 we're talking about affecting equipment is a big part 2 of the aluminum issue.

3 And we recognize that and we have been 4 working towards ways of improving the data that we've 5 collected, so that we can draw this conclusion. We do 6 see this as an important aspect. Next slide.

7 And again, this is a picture of some of 8 the metal ejecta that we did take samples of at 9 various distances from the cabinet itself, that's what 10 you're seeing on the ground here. Next slide.

11 And to even push that forward, we have 12 looked into new ways to actually address, actively, 13 conductivity measurements within the room itself, 14 during the test that we do conduct.

15 We're going to be using a developed tool, 16 instrument, in order to look at the conductivity 17 during a test cell, as we do these four low voltage 18 and four medium or bus duct tests in the summer.

19 These are essentially resistant devices, 20 where their initial intent was to measure the 21 corrosion effects in smoke environments, to look at 22 how they affect conductivity changes in real-time.

23 And we're hoping that, if we do use these 24 as we move forward, as we coat these in byproducts, we 25 can get a real-time estimate of conductivity changes 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 within the test cell itself at various distances.

2 This is a new device that we've not tested 3 before, we're currently working with NIST, not right 4 now, because they're furloughed, but we're working 5 with NIST to add these into the test program itself 6 and put these in the future tests that we're running.

7 We're doing some shake-out tests right now 8 to see if these will be useful and the type of data 9 that we can collect using these.

10 As I said earlier, the next four tests 11 that we have planned, or the next eight tests that we 12 have planned, include four low voltage tests, at 15kA 13 and 25kA, the tests that are identified here. A two-14 second test and eight-second test is slated for the 15 next series of testing. Next slide.

16 And additionally, we have our bus duct 17 tests. We're going to be doing a copper bus duct with 18 an aluminum enclosure, which is typical that we see 19 and that we saw from a survey of how much aluminum is 20 out there, that was performed by NEI.

21 As well as, we're going to be looking at 22 aluminum bus in an aluminum enclosure. So, these are 23 the tests that we have slated for the spring/summer.

24 Next slide.

25 Some of the important OECD, the next phase NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

76 1 actions and some of the important milestones that we 2 have upcoming. We've completed most to this date. We 3 also are going to be getting some international 4 equipment for the OECD NEA tests. We have just 5 received that in-house.

6 And the next big milestone is the second 7 series of tests that we're going to be performing in 8 the spring, May, summer time frame. We have not 9 established a firm date yet, we're going to be working 10 with KEMA, as well as NIST, when they're off furlough, 11 to establish the dates for the upcoming test series.

12 Any questions on anything that I've went 13 over in the last portion of this presentation?

14 MR. HAMBURGER: We'll take any questions 15 from the room first.

16 MR. SCHAIRER: Yes, this is Mark Schairer 17 with EPM. Can you elaborate a little bit more on how 18 the data you're collecting will support the zone of 19 influence spectrum?

20 In particular, are you going to 21 differentiate between ignition and damage of cables 22 and different cable types, like thermoplastic, 23 thermoset, Kerite? So, is the data you're collecting, 24 will that be sufficient to support that spectrum?

25 MR. TAYLOR: So, we're -- Gabe Taylor, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

77 1 Office of Research. So, that's one thing that we've 2 been trying to get a grasp around is, we're doing a 3 lot of work to understand the hazard, but on the back-4 end, the fragility that you're speaking to, does 5 equipment or cables become damaged or do they ignite?

6 That's an area where -- the applicability 7 of the current data, that's more thermal fires and 8 damage via these high energy short duration exposures.

9 We're trying to get a grasp of how we would go about 10 doing that.

11 So, again, we're looking at possible heat 12 conduction type modeling, similar to the THIEF model 13 that we have in FDTs, to try to characterize what the 14 thermal fragilities, from a function standpoint and 15 from an ignition standpoint, are.

16 But if there's -- it's an area where we're 17 definitely seeking feedback from anybody that has 18 expertise in this area, to help inform the process 19 going forward, because that's a very important concept 20 of determining what the ZOI is, is what those 21 thresholds or fragilities are.

22 MR. SHUDAK: This is Tom Shudak with 23 Nebraska Public Power. As far as your other testing 24 that you're going to do on copper buses, were you 25 anticipating doing those before you released your NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

78 1 report or are you going to do some comparison between 2 the copper and aluminum from just a zone of influence 3 from these control tests?

4 MR. TAYLOR: So, we have the original 5 testing program for the Phase 1 OECD test. So, they 6 potentially provide us, for a comparison. But as far 7 the GI Program, that test data from Phase 1 and 8 anything else that we can come up with from either 9 NFPA or IEEE to compare it from, we will.

10 As far as the other tests in this matrix, 11 we don't envision those testing being performed before 12 the assessment phase would need to be completed. So, 13 the comparison of those wouldn't be available for the 14 GI Program.

15 However, later on, when we complete all 16 the testing and we work with EPRI through the working 17 group and come up with kind of the updated guidance 18 for 6850 Appendix M replacement, that obviously would 19 be a very valuable comparison.

20 MR. PRAGMAN: Chris Pragman, Exelon 21 Nuclear. I just want to follow up on Tom's question, 22 because my understanding of the GI process is that the 23 decision is based on a change in CDF, or delta CDF.

24 So, if you don't have the base model involving copper, 25 how do you quantify the delta CDF for aluminum?

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

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

79 1 MR. MELLY: So, the CDF or the delta that 2 we're going to be using right now is compared to the 3 current 6850 Appendix M zone of influence model. So, 4 we're not changing that.

5 However, if we need to look at the zone of 6 influence as potentially greater, through the working 7 group or through the charter, that is how we would 8 establish a delta.

9 From the current modeling techniques that 10 are used, that three-foot, five-foot, with the 11 caveats, versus what we establish for the aluminum, 12 that would be the current delta that we're talking 13 about for the Generic Issue assessment phase.

14 MR. PRAGMAN: But it seems like to make an 15 apples-to-apples comparison, you'd have to compare 16 four-seconds copper to four-seconds aluminum. 6850 17 isn't four-seconds anything, it's more a empirical 18 number based on past experience.

19 MR. MELLY: And that's why we're not trying 20 to simply say, it's going to be a four-second model.

21 We're going to be developing a model specifically for 22 aluminum, using this data, and then, applying that to 23 the plant.

24 However, the current model that is in 25 regulatory use is that 6850 model, so that would be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

80 1 the baseline risk, what plants have currently modeled 2 to, versus what we come up with as a working group for 3 the generic assessment phase.

4 The eventual models, as we move forward 5 and complete the entire test program, will be looking 6 to change both the 6850 and the aluminum, so we may be 7 altering the copper zones of influence as we move 8 forward.

9 However, for the Generic Issue assessment 10 phase, the baseline is the current modeling techniques 11 that are part of the regulatory basis versus what the 12 working group comes up with for the zone of influence 13 of aluminum.

14 Again, not a single four-second 15 conservative model, but a spectrum that uses the 16 frequency of occurrence, plant design, as well as all 17 the data available, both the spectrum of two-second, 18 four-second tests.

19 MR. PRAGMAN: So, it sounds like we're 20 saying we're going to -- the baseline is 6850, which 21 is based on experience. The GI process, we'll use a 22 model -- I mean, when we're using models as licensees, 23 we're expected to benchmark them somehow.

24 So, I don't understand how you would 25 benchmark your model versus experience. I don't feel NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

81 1 like you've explained that well enough yet.

2 MR. MELLY: And hopefully, that will be the 3 next meeting that we have, where the working group 4 establishes the basis for this new model that we've 5 created, using both data, operating experience from 6 the plants and the frequency to explain what we will 7 be using moving forward as a zone of influence for 8 these aluminum cases. Again, we're not changing the 9 baseline, which is the Appendix M of 6850, though.

10 MR. PRAGMAN: Thanks.

11 MR. TAYLOR: I just had one thing to add to 12 the previous comment from Mark. One thing I forgot to 13 mention, in our -- and Nick probably brought it up.

14 But on our instrument stands, we also place cable 15 samples, so we can look at those afterwards and see 16 the effects from the event or the test.

17 MR. MELLY: Yes.

18 MR. TAYLOR: Cable coupons, yes.

19 MR. MELLY: We use various materials of 20 cables as well on those. Tom, do we have questions on 21 the webinar?

22 MR. AIRD: We have one question on the 23 webinar, from Mark Hewlett, APS. He asks, are the 24 connectivity sensors mounted in a manner 25 representative of location inside a nearby cabinet or NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

82 1 direct exposure?

2 MR. MELLY: Currently, the thought is that 3 we're going to put them direct exposure at various 4 distances away from the cabinet, such as on our 5 instrument racks at the three-foot, six-foot, and 6 maybe beyond, where we anticipate deposition of 7 materials.

8 However, the current test plan does have 9 adjacent cabinets as a potential. And if those tests 10 are performed, then we can put the conductivity 11 measurement inside of that cabinet, to see if we're 12 getting any influence inside the cabinet itself from 13 deposition of materials.

14 MR. TAYLOR: And from those sensors being 15 used in the smoke type testing, orientation plays an 16 impact on the measurement.

17 So, if there's feedback on -- we have a 18 solid state protection cabinet that's got circuit 19 cards in this orientation or we have this other 20 cabinet that has circuit cards in this orientation, 21 that would be useful information for us to incorporate 22 into the testing, to ensure that we're being 23 representative of the configurations.

24 MR. HAMBURGER: Any other questions from 25 those in the room? Anything on the webinar. Okay.

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

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

83 1 And again, I think all of the staff here is going to 2 be here for the rest of the day, so if any questions 3 occur to you, don't hesitate to ask. And I think Mike 4 Cheok has a --

5 MR. CHEOK: So, Ken, before we go to the 6 next presentation, I would just want to revisit what 7 I said at the beginning and then, we can go from 8 there.

9 So, I think what I said at the beginning 10 was we are going to have to be very methodical in 11 doing our tests and evaluating this issue. And the 12 second thing was, improving our communications. So, 13 I will, I guess, say it again.

14 So, in being methodical, we have to do 15 things correctly. We have to do tests that are 16 realistic, that are in conformance to some standards.

17 We have to do tests that are practical.

18 So, what we heard this morning was some 19 discussion of the tests, but on the -- and the second 20 part of the Generic Issue is not just the testing, but 21 to look at operating experience. And we haven't got 22 to that part of the test yet. And that becomes a very 23 important part of resolving this Generic Issue.

24 So, we also need to be very realistic, in 25 terms of looking at OpE, operating experience, so that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

84 1 we can be realistic and so that we can be using 2 parameters that are plant-specific, that we see in our 3 plants.

4 So, what's going to happen -- what we 5 continue to have to do is to marry the results from 6 the testing with a frequency and a look at what we 7 have out in the plants.

8 So, what we are hearing this morning is 9 only a smaller part of the whole Generic Issue. We 10 will continue to have to involve all our stakeholders, 11 in terms of trying to be looking at operating 12 experience.

13 We have to be looking at plant-specific 14 information. Like I said this morning, we have to 15 look at the design, plant-specific designs, 16 configuration-specific designs. We have to look at 17 plant practices.

18 And so, on the other hand, on testing, we 19 also have to understand what the zone of influence is.

20 And we will continue to work with our testing partners 21 and with EPRI on, once we get the tests back, we will 22 have to look at what the results show us, how we look 23 at the raw data, and how we can interpret the raw data 24 into something that could be useful.

25 And so, yes, Sandia modeling on the data NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

85 1 will be available to us, hopefully soon, and when we 2 get it, we'll be glad to share it with EPRI and the 3 rest of the industry: this is what the data is showing 4 us, this is what we will -- this is how we will use 5 it, in terms of modeling.

6 And at that point, we -- it's going to be 7 a public meeting, everyone's going to be asked for 8 feedback on how they think the data and the modeling 9 is going to match.

10 Again, so, we are a little bit ahead of 11 the schedule. I mean, off-the-cuff right now, we are 12 hoping to get some data soon. NIST being not in a 13 work status now kind of delayed some of our data 14 collection and analysis.

15 But once we get that also, obviously, we 16 will walk that through the work group, which you will 17 hear this afternoon, and then, with the rest of the 18 public in a forum where we will present the data for 19 everyone to analyze.

20 So, the second part of what I said this 21 morning was communications. And I would like to have 22 everyone's input, and I'm not sure I got that feeling, 23 and I'm not sure I want to be talking past some of the 24 questions and comments this morning.

25 So, I think what I heard was, we said we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

86 1 dispositioned a lot of the industry comments, but that 2 in dispositioning the comments, we may not have come 3 to a -- we did not address them, for lack of a better 4 term.

5 So, I think what we need to do, maybe this 6 afternoon or maybe after this meeting, we will sit 7 down with our staff and we'll -- if you all will 8 provide us with very, more specific comments on, this 9 is what we thought we gave you, we did not see the 10 resolutions to this, we'd be glad to discuss this 11 again in the next public meeting.

12 We'll be very specific as to, this is your 13 comments, this is how we will address this in the next 14 phase of testing or in the next task, in terms of 15 frequencies or in the characterization of the zone of 16 influence.

17 So, I think we can be more specific, in 18 terms of addressing comments in a public forum, so 19 that we understand your comments better and you can 20 understand our responses better.

21 So, let's see what else I've got in my 22 notes from the last two hours. I think I got it all.

23 I mean, so, we will talk about the decrement curve.

24 We will talk about using the need for an eight-second 25 duration test.

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

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

87 1 I understand that there's some 2 sensitivities to test extrapolation. So, things like 3 that, I think we can address and we will address it 4 specifically in the public forum.

5 MR. HAMBURGER: Thank you, Mike. If we 6 don't have any more questions from the room, I think 7 we're overdue for a break. So, let's come back at 8 11:00. Come back at 11:00.

9 Again, bathrooms are out the door on this 10 side of the room, to the left. Cafeteria is open, 11 again, that's to the left, down the hallway, and that 12 will be on your left as well. My contact information 13 is up on the screen. If you have any issues, give me 14 a call, it's probably the best way to get a hold of me 15 at the moment.

16 (Whereupon, the above-entitled matter went 17 off the record at 10:42 a.m. and resumed at 11:07 18 a.m.)

19 MR. HAMBURGER: All right. So, we're going 20 to resume with Gabe Taylor.

21 MR. TAYLOR: Okay. Welcome back. So, 22 myself and Kenn Miller are going to go through future 23 test planning for the aluminum HEAF testing and the 24 equipment selection. Next slide.

25 So, during this presentation, we're really NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

88 1 trying to communicate what we're planning on doing, 2 the upcoming tests, and also, receive feedback on 3 anything that we can do to improve how we go about 4 testing the equipment as we move forward.

5 Just as a refresh, I know Nick brought 6 this up in this past presentation, but if we look at 7 the overall test program, including the OECD test 8 report, you can see that we've completed the four 9 medium voltage 6.9kV tests, in the fall, September.

10 One clarification is that it was 32kA, not 35, as the 11 slide shows.

12 So, for the aluminum HEAF, we're really, 13 on this enclosure testing matrix, we're really looking 14 at the right portion of the matrix, and the individual 15 tests that are highlighted in the peach or orange 16 color are the ones that are NRC-sponsored.

17 So, that leaves us with four low voltage 18 tests, at 480 volts, two current levels, 15kA and 19 25kA, and durations of two and eight seconds. Note 20 that we also will be planning on running four-second 21 tests as part of the OECD program.

22 On the far right, you can also see two 23 tests that aren't connected to any part of the test 24 tree matrix.

25 And as we communicated previously, we plan NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

89 1 on using those two spare or optional tests to evaluate 2 the decrement curve. The decrement curve will be 3 evaluated at the 6.9kV, and it's yet to be determined 4 what current level we'll be testing that at.

5 But those are information that we're 6 trying to work with EPRI and understand the best way 7 to represent realistic plant decrement type 8 conditions. So, just -- and I have a later slide on 9 the decrement curve. So, that's low voltage.

10 If we look at the equipment that we're --

11 we haven't procured any equipment yet. We've put out 12 bids.

13 And right now, what we're look at for the 14 aluminum bus low voltage enclosures, either a 15 Westinghouse DS type switchgear or a General Electric 16 AK series switchgear. So, these are the low voltage 17 switchgear load centers that will be just downstream 18 of your step-down transformer.

19 As far as frame size, we're looking at 20 1600 amp or 200 amp continuous frame size rating. And 21 then, also, for interrupting capacity, we're looking 22 at anywhere between 42kA and 65kA.

23 And those numbers really came out of the 24 evaluation we did to support the workshop last April, 25 looking at what was common for plants out there. And NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

90 1 although our sample size wasn't the entire fleet, it 2 was about a quarter of the fleet that we sampled from 3 the information that we had.

4 And as far as -- these were the common 5 ranges of equipment that was out there. So, if you 6 looked at an average, interrupting capacity was right 7 around 50kA for what we looked at. Next slide.

8 Okay, moving on to the medium voltage bus 9 ducts. For this type of testing, we're looking at a 10 combination of enclosure and busbar material types.

11 And also, there's a combination of NRC solely-12 sponsored tests and OECD-sponsored tests. The voltage 13 level we chose for this was 4160 volts and a single 14 current at 25kA.

15 The combinations, as you can see, are 16 copper bus/steel enclosure, which is OECD only.

17 Copper bus/aluminum enclosure, which is similar to the 18 test we ran in Phase 1, the Zion bus duct experiment, 19 so we have two tests there at two and four seconds.

20 Then, we have an aluminum bus/steel 21 enclosure. In that case, we'll run one of the two-22 second tests, there's an OECD test for four seconds.

23 And then, aluminum bus/aluminum enclosure, we have two 24 tests there for two and four seconds.

25 We do have two spares test, and at this NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

91 1 time, we haven't determined a need for those yet. But 2 there could be a potential there to evaluate decrement 3 curve or some other type of configuration, as we see 4 fit.

5 So, if there's feedback to support those 6 optional tests, we'd be interested in learning that.

7 So, and one thing, just to be clear, these aren't the 8 iso phase bus ducts, these are typically non-9 segregated phase bus ducts. Next slide.

10 As far as the general characteristics of 11 the medium voltage bus ducts, again, non-segregated 12 phase bus ducts. I gave you the voltage already.

13 As far as the rating, we're looking at 14 1200-2000 amps rating on the busbars and then, also 15 from a withstand rating anywhere from 31.5 to 63kA, 16 which that's typical out of the standard that's 17 identified there for medium voltage bus ducts.

18 One area that's still uncertain at this 19 time is, for our measurements, we need to stabilize 20 the arc in one location.

21 And if you're familiar, from the testing 22 that we did in the first phase, the OECD program 23 testing, because of electromagnetic effects, it 24 basically propelled the arc out the end of the bus 25 duct configuration.

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

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

92 1 And if you look at the OpE, there's 2 typically some feature that stabilizes the arc, 3 whether it's a bolted connection or a grounding bar or 4 some other cause.

5 Recent OpE, there was an event that 6 happened exterior to the plant, they had an exterior 7 bus duct, and because of the coatings that they had on 8 the busbar, that acted to stabilize the arc in one 9 location.

10 So, right now, what our thought process is 11 is what's shown in the graphic there, is that if you 12 see the power supply from the laboratory, coming down 13 the left side of the image, the copper colored 14 conductors, it then runs horizontally into the bus 15 duct.

16 And in the bus duct, we make a physical 17 separation of the bus duct and, on the left side, we 18 place a shoring wire at that location to generate the 19 arc there. The idea there is that would stabilize it 20 in that general location, where we'd then have the 21 instruments located above and below.

22 Another option could be to use the epoxy 23 insulation that was used in the industry testing.

24 What we found in some communications with external 25 parties is that that type of insulation isn't common NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

93 1 inside the plant. So, that's one non-conformity that 2 we would potentially introduce into the testing, if we 3 performed that.

4 Another option that we also had some 5 feedback from the Japanese experiments is that they 6 put a grounding bus perpendicular to the orientation 7 of the phases, the bus phases, and that helped to 8 stabilize the arc as well. But then again, that's 9 something that's not common to the types of 10 configurations that are in the plant.

11 So, just to kind of summarize, right now, 12 we're looking to put a gap in the bus ducts to 13 stabilize the arc, but if there's any feedback or 14 logic to support some other configuration that would 15 provide a more reliable means to maintain the arc in 16 one location, so that our instruments could capture 17 reasonable and usable data, we'd be very interested in 18 that type of feedback.

19 And then, from informal surveys of some of 20 the equipment vendors, common enclosure housing 21 thickness, looking at 11-gauge aluminum housing. The 22 conductor for aluminum, for the aluminum experiments, 23 to follow the ASTM standard that's specified there.

24 So, if there's any other types of design 25 considerations that are out there we should be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

94 1 considering, we're very interested in learning those.

2 If not, we'll work with our electrical contractor to 3 try to find and procure bus ducts that are 4 representative of the types of installations in the 5 field.

6 Okay, now it's moving on to the decrement 7 curve. And we talked about this a little bit earlier.

8 Last January of 2018, EPRI published some white 9 papers.

10 In the one white paper, they went through 11 a number of different scenarios. And one of the 12 scenarios they to be most common to exhibit a long-13 duration arc was cases where it's a generator-fed 14 fault.

15 So, basically, you'd have a fault between 16 a generator and a protected device, so you wouldn't 17 have a main generator breaker to clear the fault. And 18 then, the generator would basically feed the fault 19 until the exciter voltage dropped out and stopped 20 providing energy to the fault.

21 So, we found that information very useful 22 and, as I mentioned earlier, we're trying to work with 23 our testing facility to be able to replicate some of 24 the field experience that was provided to us via EPRI.

25 It does require new contracting, which takes time.

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

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

95 1 And also, it requires some analysis and 2 verification on the laboratory's part, to make sure 3 that they can run the experiment that meets our needs, 4 to replicate that type of exposure.

5 So, it's not a rudimentary process of 6 going out and doing this, this isn't very common or 7 we're not aware that it's been done in the past, but 8 we do have confidence that it can be done moving 9 forward.

10 And that's feedback we received from EPRI 11 and during the workshop and during the public 12 comments, and we're definitely taking that into 13 consideration and making changes to improve our test 14 program.

15 So, that's one example where we made 16 changes, but we're also looking for any other feedback 17 to improve our testing, because in the end, we have to 18 have something that's usable to make realistic methods 19 and models to inform our risk analysis.

20 So, just to summarize, our initial focus 21 here is to support the Generic Issue Program. We're 22 focusing on the NRC-sponsored aluminum type tests, to 23 support the GI Program in its assessment phase, to 24 provide useful information to determine what the ZOI 25 or the ZOI model is, to support the risk assessment.

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

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

96 1 And although the decrement curve, the 2 process for getting all that is slow, we're trying to 3 expedite it as much as possible.

4 There could be possibilities, if they're 5 interested, for industry to perform that type of 6 testing. We could support it on the measurement side 7 as well, if that is a faster means of getting it done.

8 I think that's all I had. So, I really 9 appreciate any feedback on our planned approach this 10 coming year.

11 MR. HAMBURGER: Gabe, if people have 12 feedback that occurs to them after this meeting, can 13 they get in touch with you by email?

14 MR. TAYLOR: Right. You can email 15 gabriel.taylor@nrc.gov. My information's on the 16 meeting notice. As well as Nick Melly, 17 nicholas.melly@nrc.gov.

18 MR. HAMBURGER: Okay. Any questions from 19 those in the room? About any of the equipment Gabe's 20 presented?

21 MR. TAYLOR: I'll just make one additional 22 thing, while we're waiting. We'll run it no 23 differently than we've done previously.

24 In the first phase, we identified what 25 equipment was available and then, we worked with EPRI NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

97 1 to make sure that was consistent with the type of 2 equipment that's found in the field.

3 So, we do have some initial leads on 4 equipment last week, but there's a few bids still out 5 there to see. And once we get that information, we'll 6 work with the working group to understand which 7 equipment is most applicable for the series of 8 aluminum tests.

9 MR. HAMBURGER: Okay. Any questions or 10 comments? Tom, anything on the webinar? Okay.

11 Well, I think the next presentation is 12 slated for an hour and I think there's going to be 13 some comments and questions on that, so I don't want 14 to get into that before lunch, because I have a 15 feeling that may delay lunch.

16 So, unless anyone objects strenuously, I 17 think we can go to lunch now and take a little 18 extended lunch. Come back at 12:45? 1:00? You okay 19 with 1:00? Mark, unless you think you can get it done 20 in -- okay, we'll come back at 1:00.

21 (Whereupon, the above-entitled matter went 22 off the record at 11:21 a.m. and resumed at 1:00 p.m.)

23 MR. HAMBURGER: All right. Thank you all 24 for returning promptly.

25 Just a reminder, for questions and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

98 1 comments, please state your name before your comments, 2 so the court reporter can transcribe it properly, and 3 try to use the microphones for any comments or 4 questions, so other webinar participants can hear it.

5 Mark?

6 MR. SALLEY: All right. I take it 7 everybody had a good lunch. We've got a lot to cover 8 this afternoon. So, we'll pick it up.

9 Back on May 11th, 2017, in this room, 10 there was a briefing for the Commission. And one of 11 the industry execs was at the table and briefly 12 described the HEAF and the work that was going on.

13 And he made sure to make the point that the industry 14 really wants, speaking for the industry world, the 15 industry really wants to work with NRC research on 16 resolving this. The vehicle that we typically use in 17 research is the MOUs I described earlier with EPRI.

18 And we try to get the scientific arms together to do 19 that work.

20 For those of you that have been around 21 fire protection and have been here a while, if you 22 think back 20 years ago or so, again, in this room we 23 had a whole different topic in fire protection that we 24 were working, and that was multiple spurious --

25 everybody remember that? -- with the hot shorting of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

99 1 circuits and all that. It was a big issue in the late 2 nineties and at the turn of the millennium.

3 Today we don't talk too much about it. It 4 seems to be solved. Everybody's comfortable doing it.

5 The utilities know what they need to do. The 6 inspectors know how to inspect it. And we have done 7 a lot of work.

8 And as a part of the success of that, it 9 was that MOU between research and EPRI. If you 10 remember, we did a whole series of tests. The 11 CAROLFIRE program was the first one, and we worked 12 through it, again working together to get a good 13 scientific solution.

14 So, we've seen success with this before, 15 and it's worked quite well. I think we really 16 advanced the science, and it's good science. So, 17 again, with that in mind, we want to reach out and 18 work with EPRI on the technical solution to the HEAF.

19 Next slide.

20 So, what we did in the past few weeks, 21 working with Kelli and her folks and mine on this 22 side, is we had a number of calls and discussions, and 23 we thought we'd put together a Charter for what this 24 Working Group is going to attempt to do. We held it 25 out as a draft until today, again, for any feedback NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

100 1 from stakeholders or industry on where we're looking 2 to move forward.

3 Again, the basic mission statement here, 4 I think it's quite obvious: to improve the 5 understanding of risk from electrical arcing fault 6 hazards in nuclear power plants. And you'll notice we 7 chose the words "arcing fault hazards" and not 8 directly to HEAFs. I think that's part of it and the 9 fact that we want to get the correct binning and get 10 the correct selection of the events. So, that's the 11 basic Charter for the group.

12 Next slide.

13 So, the goals of this Working Group:

14 Obviously, a better understanding of the 15 key factors contributing to the occurrence and 16 severity.

17 You can read them here.

18 Advancing the HEAF fire PRA modeling.

19 Again, we have the single model out of 6850 for the 20 zone of influence in Appendix M. We also have the one 21 FAQ for the bus ducts. We think we can do a lot 22 better with that.

23 And then, analyzing the plant impacts and 24 the risk implications. If you look at that, that 25 should look very familiar to everyone here. That's NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

101 1 basically HEAF in the form of a risk triplet. So, 2 again, following the sound principles is how we want 3 this group to move forward.

4 Next slide.

5 So, this is where it kind of gets a little 6 interesting. There's a lot of famous quotes and a lot 7 of people have said things about, you know, getting 8 the right group of people together. I think one pops 9 into my mind, Lee Iacocca. If you remember Lee 10 Iacocca from the automotive world, this is the man, 11 basically, that came up with the Mustang in 1964, and 12 then, he saved Chrysler later. So, he was a pretty 13 heavy hitter in there.

14 He had a saying that kind of went 15 something like this: I hire people smarter than me, 16 and then, I get out of their way. So, that's kind of 17 the philosophy that I want to look at with this 18 Working Group, and this is the Working Group that 19 we've assembled between Kelli and myself working for 20 some of our best experts out there. You may of the 21 people. Some or a lot of them are in this room here 22 today. And again, they're very good people.

23 Also what's here, we put two levels of 24 management. The Project Managers, again, Kelli and 25 myself. If you have questions or want to talk to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

102 1 someone, we're there. And then, our sponsors that are 2 above it for the whole program, which is Mike Cheok, 3 my supervisor, and then, Tina Taylor.

4 A lot being said with putting the right 5 people together and getting out of their way. We've 6 talked a lot about zones of influence and things of 7 that nature, and how we want to do this. And I really 8 think that, as we're putting this group together --

9 and they're just starting out, by the way. So, we 10 want data, we want this, we want a lot of things. But 11 this group is just getting together. Like I said, the 12 Charter is a draft. Any input from anyone today was 13 welcome, and then, we'll get started with this.

14 But looking at HEAF -- and I'll set the 15 bar real high for this group -- don't just lock me 16 into a ZOI or something like that. I mean, the 17 sheet's clear. We're starting out.

18 If we look at HEAF and we talk about the 19 tests and why we're doing some of the tests and 20 different things, we basically have one equation with 21 six unknowns. Okay? And if you think back to college 22 and your early days in engineering, most of us went 23 through you could do one equation, two unknowns, and 24 work it through and substitute, and bring it back, and 25 all that.

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

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

103 1 But this is a complex phenomena. I mean, 2 what effect does the voltage play? What effect does 3 the current play? And what effect does duration play?

4 What effect does the material play? What effect does 5 the enclosure do? We know there's a lot of variables 6 here.

7 And I think, as a part of this Working 8 Group, or what I challenge them to do is tell me this.

9 Does voltage matter? Does current matter? We've 10 fought a lot about duration. We're seeing the 11 material properties seem to matter between copper and 12 aluminum.

13 So, this group has a lot of work ahead of 14 them to do it. And I don't want to tie their hands, 15 but I want to let them do their thing.

16 In a perfect world, if this group was, 17 I'll say it as a manager, very successful, we would 18 have an equation where you could plug in the voltage, 19 you could plug in the current, the enclosure, and come 20 up with the zone of influence.

21 And maybe it's not a box like we've been 22 doing 3 foot x 5 foot. If you look at some of the 23 data we've gotten from video, it's more of a sphere.

24 Is that the correct thing?

25 Or the FAQ group, when they worked the bus NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

104 1 duct out, I think, Industry, what do you call that, 2 "the cone of death," or something, for the cone that 3 comes down from the bus duct? Again, I don't want to 4 limit the group. I want to let the physics and the 5 science do what it needs to do.

6 So, these are the people that we've 7 selected. I think we've got a good group here, very 8 capable. And we want to let them get on with it.

9 Let me see what I've got. Next slide.

10 Again, deliverables. To improve the risk 11 models. The PRA stuff is driving it. That's what's 12 got us here. We want to look at the frequency and the 13 binning and the zone of influence. And again, I'll 14 zone of influence loosely. I think the group can come 15 up with what they need to for that. It may be a 16 sphere. It may be an equation. It may be different 17 things they can come up with.

18 Again, our goal was to look at the risk to 19 the plants. We talked a little bit about pilots, and 20 that's something, again, where we'll need some help 21 from industry. If you go back and look at 6850 and 22 the history of that, again, for the folks that have 23 been around since the 2000 timeframe -- we published 24 that in 2005 -- one thing that we constantly at the 25 end of AFP-805 we're hit with is, you know, we never NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

105 1 really piloted it. We piloted it in pieces. We never 2 did a full pilot. If hindsight is 20/20, everyone 3 agrees back in 2000 we would have piloted 6850, and it 4 would have been a better tool for all to use.

5 So, here's an area, too, where we're 6 looking at pilots. Again, we're still in the 7 discussion phase as a part of the Working Group.

8 Remember, this is just starting up. But that the 9 Working Group would develop the models and the tools 10 and the data and how to do that, and then, let the 11 licensees or the consultants implement it in the plant 12 to do the pilot. So, that's kind of where we're at 13 with that right now.

14 Again, we really hope to advance the 15 science, to reduce uncertainty, to improve realism, 16 all the things we strive for, and to update the 17 current guidance. It is funny, when we started out in 18 the testing way back with the internationals, our goal 19 was to go and to provide the confirmatory test that 20 Appendix M, 6850, was a good model.

21 And for the most part, we were doing 22 really great and we were thinking, hey, this is a good 23 -- man, if you had to pick a one-size-fits-all HEAF 24 model, up until about that 22nd test, we're saying 25 this was right on. And we were, you know, Nick and I, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

106 1 patting each other on the back that this is really 2 going to be successful until those later tests came 3 and showed us that there's more to it than that.

4 So, again, we want to get that accuracy 5 and we want to get that fidelity in it; and again, 6 understand the variables and the parts they play, 7 which, again, will be the update to the guidance for 8 6850, Appendix M, and the supplemental, the FAQ -- I 9 think it's 37 or 47 -- for the bus duct.

10 Communications, we talked this morning is 11 important. We will strive to have open communications 12 with you. And I guess, as the Working Group, like I 13 said, the key is we've got to let them do a little 14 work. It's just getting together. And I think as 15 they get to key points in that, we can agree upon 16 things that we want to share with the larger group as 17 to how we're moving forward.

18 Next slide. So, I've got one more slide 19 here.

20 Priorities. And these are the NRC's 21 priorities as to how this Working Group moves forward.

22 I kind of want to give you the big picture of this.

23 This is not a short-term thing that we do in six 24 months or a year, but it's a little bigger.

25 Obviously, the first thing we want to look NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

107 1 at is the aluminum HEAF. That's the driver for us in 2 the GSI program, and we want to solve that portion 3 first.

4 You saw the matrix that Gabe and Nick used 5 earlier this morning, how we laid all the testing out.

6 And we want to pick those aluminum pieces up first and 7 get this Working Group going.

8 Again, the decision from the GIRP was that 9 we want to go out with the most accurate model. We 10 don't want to go out with something that's very 11 conservative, do some more research, have it combine, 12 and have people do the work two and three times, but 13 to come out with the most accurate, realistic model as 14 possible. That's the goal that we're striving for.

15 It's a timely goal. If we get our testing 16 done, we get this group going, we believe we can 17 achieve it, and we have the schedule. So, that's our 18 first priority.

19 You also know that the way we originally 20 started this program was with an international 21 program. HEAF is not unique to the United States 22 nuclear plants. It's worldwide. We see these, 23 approximately 10 percent of the most significant 24 events, we see with our international partners are 25 HEAF. There's a lot to learn, and everybody pretty NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

108 1 much understands that.

2 So, the OECD/NEA group, we're in the 3 process of signing out the agreement now, where we 4 want to go forward with the second form of testing.

5 And this will do a lot more of the copper and it will 6 be a lot more confirmatory, and again, bring a lot 7 more of the realism into the total understanding of 8 HEAF.

9 As a matter of fact, the German equipment 10 has already been shipped, and we have already received 11 equipment from Germany. So, there's a lot of interest 12 from our international partners to work on this.

13 Our goal with our international partners 14 was to keep this Working Group going because they 15 would have momentum, they would have understanding, 16 they would have experience, and to look at refining 17 the other models, again, the Appendix M, for the 18 copper and the different conditions.

19 We could, then, take that and put it out 20 for public comment and get our OECD partners, our 21 experts in the foreign countries, and let them do a 22 peer review of it, so that we could have a good, high-23 quality product.

24 So, that's kind of the big picture looking 25 forward, that we go for the aluminum piece first, and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

109 1 then, we work into the second phase of testing with 2 the OECD and the NEA.

3 Again, the whole goal at the end of the 4 day is to improve the realism and fidelity of our 5 models.

6 Kelli worked very closely with us. Kelli, 7 that's kind of the end of my presentation. If you 8 have anything you'd like to say on the side of EPRI or 9 anything, this would be a good time.

10 MS. VOELSING: Are you saying you want me 11 to give my presentation now?

12 MR. SALLEY: If you're ready. Is it on 13 the Working Group?

14 MR. HAMBURGER: Sure. That's fine.

15 Before you give your presentation, does 16 anybody have any questions for Mark?

17 MS. ANDERSON: I have a question.

18 MR. HAMBURGER: Oh, okay.

19 MR. SALLEY: Okay.

20 MR. HAMBURGER: And, Mark, would you mind 21 just telling people where they can find the current 22 version of the Charter?

23 MR. SALLEY: Yes. If you look at the 24 meeting notice today, the one -- we updated it last 25 week. It's a two-page Charter. It's a link at the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

110 1 very bottom. You can just load it. It's an EPRI 2 Dropbox, is it?

3 MR. HAMBURGER: I believe it's a link to 4 the EPRI box.

5 MR. SALLEY: Box, and it's in there, a 6 copy. Again, it's a draft. So, we'll take any 7 comments/suggestions.

8 MR. HAMBURGER: Victoria?

9 MS. ANDERSON: So, this is Victoria 10 Anderson with NEI.

11 That was actually my first question. You 12 said that stakeholder input on the Charter was 13 welcome. And I checked out the meeting notice this 14 morning, and I don't think I noticed that link. So, 15 even if it is available, maybe there's a way we can 16 better publicize it. Maybe make it available in 17 ADAMS, if we really want to get as much public 18 feedback as possible. I mean, I was hunting for it 19 and I couldn't find it.

20 MR. HAMBURGER: I believe it's a link at 21 the bottom of the meeting description.

22 MS. ANDERSON: Right. But, I mean, again, 23 I was motivated --

24 MR. HAMBURGER: I'll put that down as --'

25 MS. ANDERSON: -- to try to find it, and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

111 1 I still struggled.

2 MR. HAMBURGER: Okay. So, I'll put that 3 down as an action item for NRC.

4 MS. ANDERSON: If we're really looking for 5 broad stakeholder input, that's not really going to 6 help it along.

7 MR. SALLEY: The millennials got me on 8 that one, Victoria.

9 MS. ANDERSON: What's got you?

10 MR. SALLEY: The millennials got me on 11 that.

12 MS. ANDERSON: I'm not a millennial. So, 13 I can't help you there.

14 MR. SALLEY: My millennials.

15 MS. ANDERSON: Okay. They'll help you.

16 MR. HAMBURGER: And of the documents, 17 including the Charter and the presentations from today 18 will be packaged up and placed into ADAMS and made 19 publicly available.

20 MS. ANDERSON: Okay. So, I think the 21 sooner that can be done with the Charter, the better 22 to help with soliciting feedback.

23 MR. GARDOCKI: And we'll try to put stuff 24 like that on the GI Dashboard.

25 MS. ANDERSON: Excellent. That's an NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

112 1 outstanding idea.

2 MR. GARDOCKI: So, that's a great place 3 for all this information. Okay?

4 MS. ANDERSON: Outstanding idea.

5 Definitely fully support that.

6 How will other stakeholders and the public 7 be able to interact with this Working Group? Like 8 will information come out of this group? Will there 9 be draft reports put out? Will there be public 10 meetings on where they're going?

11 MR. SALLEY: Yes, we can decide that 12 however it's best for everyone. And again, I'll work 13 with EPRI and Kelli. I think that once we maybe get 14 a product that we think is far enough along would be 15 a good time to put a draft report out, give folks 30 16 days to comment on it, whatever. If we see 17 significant comments, we could hold a public meeting 18 and discuss it.

19 Or, again, how we talked with the pilots, 20 and I think that the Working Group itself could do 21 some presentations as to what their logic was, how 22 they came up with it, and what the intent was. So, we 23 can do that in a public forum, if you like.

24 MS. ANDERSON: Great. You mentioned 25 Kelli, and she showed up behind me.

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

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

113 1 (Laughter.)

2 MS. VOELSING: Yes, I was going to say, 3 the names that are on the list for the Working Group 4 predate me or my involvement. And to your point, 5 Mark, if there's other industry stakeholders and 6 experts that have technical expertise to contribute to 7 the Work Group, I would think we would want to 8 consider their participation. So, there should be 9 room for that.

10 MS. ANDERSON: Okay. And just one other 11 question or I suppose a question/comment hybrid is:

12 you underscored the need to have the pilots to support 13 this Working Group and the pre-GI resolution. Right 14 now, I think plants would be hesitant to volunteer for 15 that because we heard this morning is we're not 100 16 percent sure what we're going to be evaluating or what 17 we're looking at, or how to deal with the fragilities, 18 or how to deal with this or how to deal with that.

19 The licensees in the audience can correct 20 me if I'm wrong, but, until some of that is better 21 defined, I don't think you're really going to get a 22 volunteer. So, I think the sooner we can better 23 define some of that, and we can get some realistic 24 inputs, the sooner we'll be able to get some pilots to 25 volunteer.

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

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

114 1 MR. MELLY: Yes, we absolutely understand 2 that. And the Working Group is currently also working 3 towards a pilot plant resource need or resource 4 allocation, what would be expected of the pilot 5 plants. We drafted that late at the end of last year, 6 and we're working that through the Working Group to 7 say, would this be the type of resources that any 8 pilot plant would need to allocate to doing this?

9 What type of information are we looking for? How it 10 would move through the process. So, we're still 11 fairly early in that step, and that draft is being 12 shared within this Working Group to see, is this the 13 type of information we need to collect and how would 14 it move forward?

15 MS. ANDERSON: So, I think you correctly 16 identified the resources as one specific concern the 17 pilot plants would have. I think another concern we 18 need to pay attention to is whether or not they're 19 going to be asked to consider information that they 20 believe, based on OE, to be unduly conservative.

21 I don't think a plant is going to 22 volunteer to put their information out there to be 23 evaluated with information or analysis based on 24 extremely conservative testing or results, to possibly 25 come up with erroneous conclusions. I just can't see NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

115 1 a plant being willing to do that.

2 MR. TAYLOR: Yes, I agree. I think one 3 thing from other projects we worked on, such as the 4 spurious operation one, where they basically use data 5 from all the tests, we use operating configurations, 6 and then, the experts' judgment to modify or adjust 7 the data. So, I agree, we want to get a realistic 8 model to characterize the risk. And I think letting 9 the Working Group work through that process should be 10 able to get us there. But we're too early to --

11 MS. ANDERSON: And I think that 12 underscores the need to have continuing public 13 interactions, so that potential pilots have more 14 confidence in that. And then, there's potential for 15 them to volunteer.

16 MR. HAMBURGER: Thank you. Any more 17 questions for NRC staff about the NRC/EPRI MOU or the 18 Working Group?

19 (No response.)

20 MR. HAMBURGER: Any questions on the 21 webinar?

22 (No response.)

23 MR. HAMBURGER: Okay. And just a note for 24 those on the webinar. Before the end of the day, we 25 are going to open up the phone lines to make sure NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

116 1 nobody has any questions that they haven't put down in 2 the chat.

3 All right. We are now officially in the 4 public question and comment session. We have two 5 prepared presentations.

6 Kelli, would you like to go first?

7 Okay. The first presentation was provided 8 to me by Kelli from EPRI.

9 And you can use the podium, and I can step 10 through. Or, if you want to take the buttons?

11 MS. VOELSING: First of all, thanks for 12 the opportunity to comment. My name is Kelli 13 Voelsing. I'm the RSM Program Manager at EPRI.

14 So, we have been participating and 15 following the development of what's going on with high 16 energy arcing faults and in preliminary discussions of 17 the Working Group Charter with Mark and his team, and 18 with Mike Cheok and Mark Thaggard as well.

19 I will say that at a high level I think we 20 are aligned on the tasks in the Working Group Charter, 21 the issues that need to be worked through in order to 22 have a realistic understanding of the risk and the 23 potential contribution to risk from aluminum. But I 24 think there's significant work and technical 25 challenges that we're going to have to work through in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

117 1 order to get to that answer.

2 So, maybe perhaps for those of you that 3 don't know, my technical lead on fire PRA is out right 4 now on a medical leave, having just had a baby. And 5 so, I have been given the opportunity to learn a whole 6 lot about high energy arcing fault in a very short 7 period of time. And the analogy that keeps coming in 8 my head, as I think about these, I thought it might 9 resonate with some of you all, which is, when we see 10 the videos, when we see the pictures, when we talk 11 about these enormous volted three-phase long-duration 12 faults, I absolutely understand how the first reaction 13 to seeing those is obviously one of concern. That 14 could be a significant issue with significant damage 15 at a nuclear plant.

16 But I think back to loved ones -- and 17 hopefully, none of you -- but if you've been given a 18 diagnosis and you hear the words "You have cancer" 19 from a doctor, the first thing that's going to go 20 through your head, oh, you know, invasive surgery.

21 We're going to cut out large portions of tissue and 22 poisonous chemotherapy and radiation, and obviously 23 the worst-case scenario. It's very easy to envision 24 that.

25 But I think very quickly you get into the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

118 1 details, and you start hearing from the doctor about 2 locations and size and phases and genetic markers of 3 the tumor, and all of the discriminating factors that 4 go into the appropriate treatment of a cancer. And I 5 think what I've learned about high energy arc fault is 6 that there's a lot of those discriminating factors in 7 not only what we see in the testing, but what we see 8 in the OE and in the experience of our plants as well.

9 And just like proceeding with a worst-case 10 cancer treatment, sometimes proceeding with a worst-11 case treatment or a one-size-fits-all model or 12 approach to understanding HEAF would ultimately do a 13 disservice to the patient; and that for the best 14 outcomes, we have to understand those factors and we 15 have to take them into account, as we understand both 16 the risk and the mitigation plans for how we deal with 17 those.

18 So, I guess I was just saying, it keeps 19 coming into my head that I understand why the concern.

20 And obviously, these worst-case HEAFs would be a 21 significant or could be a significant safety impact, 22 depending on the configuration, the location of the 23 plant, the equipment that could in a zone of 24 influence. But there's a lot of other factors that we 25 have to consider before we get to that conclusion.

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

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

119 1 So, if we could go to the next slide?

2 I'll start here, and I think we've already 3 heard this today. But I just wanted to put in context 4 that, currently, the Bin 16 treatment in NUREG-6840 or 5 EPRI 1011989, the Fire PRA Methodology in Appendix M, 6 that we've been referring to is a one-size-fits-all 7 for everything in Bin 16. There's one zone of 8 influence. There's one initiating event frequency.

9 And it's certainly easy to imagine that, if you double 10 or triple or quadruple the zone of influence, you're 11 talking about taking out entire rooms or large zones 12 of potentially impacted equipment. So, we have to 13 consider what all goes into Bin 16 and is it really a 14 some-size-fits-all approach that we should be 15 pursuing.

16 So, if we'll go to the next slide?

17 Some differentiating factors that may come 18 into play here. We do see distinct differences in the 19 types of events, the damage profiles of events between 20 medium voltage and low voltage equipment. I'll talk 21 some more about that.

22 The location. It's probably true that 23 there is aluminum out there in bus ducts, whether non-24 segregated or iso phase bus ducts. But that equipment 25 is outside or in the turbine deck, and doesn't impact NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

120 1 safety-significant equipment, it may not still be a 2 significant contributor to risk.

3 We see differences within Bin 16 between 4 switchgear and motor control centers, bus ducts, that 5 all of those have different probabilities in terms of 6 is this likely to be a long-duration or short-duration 7 event; is it likely to be a flash or a blast versus a 8 HEAF? And I think we can start to draw some insights 9 there.

10 Source of the fault may be another 11 important consideration that we haven't really talked 12 about yet. We saw a picture of the wire that was 13 being used to generate the fault on the aluminum 14 busbars, which would be, obviously, on the switchgear 15 on the load side of the breaker. Whereas, a good 16 portion, more than half of the industry experience is 17 that the fault is generally generated on either the 18 breaker side or at the stabs, which is generally not 19 where the aluminum is present. So, that may be an 20 important consideration in terms of testing, in terms 21 of where do we put the fault.

22 Electric configuration. At the time of 23 the white papers last year, we had reviewed a good 24 portion of the industry. We've now reviewed all 103-25 104 units to determine which units have safety-related NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

121 1 buses or buses that are subject to unit-connected 2 generator design without a generator breaker, such 3 that you would have a long-duration fault that has to 4 spin down. And that applies to some plants. It also 5 does not apply to a lot of plants. And so, I think we 6 need to consider that as well.

7 Material. Obviously, the reason we're 8 here today is the presence and location of aluminum.

9 And then, again, operational issues. We 10 would be remiss to not acknowledge that the best way 11 to prevent the risk of a HEAF is to not have a HEAF.

12 And although there are not one smoking gun, we do see 13 in the OE that a large majority of HEAFs have, 14 although not the same preventable cause, they do have 15 preventable causes that relate to cleanliness, 16 maintenance practices, human performance.

17 So, next slide.

18 Again, we are in alignment on the major 19 tasks of the Working Group Charter. We think those 20 are the right tasks that need to be worked through and 21 appreciate the commitment from Mike Cheok and Mark 22 Salley and their teams to work through that in a 23 methodical fashion that considers the spectrum of risk 24 and the probabilistic aspects of that spectrum of 25 events.

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

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

122 1 So, the first task is to improve the risk 2 models, supported by all available information, be 3 that testing, data, whether type testing, 4 manufacturing data or manufacturers' data, or the 5 current test programs, or other testing.

6 OE from the industry as well as expert 7 engineering judgment because the bottom line is I 8 don't think we're ever going to have enough resources 9 to test all the configurations, all the possible 10 configurations that we might like to test.

11 And then, there are basically two tasks 12 that we've aligned on that support that. The first is 13 a realistic representation of the frequency of events 14 represented in Bin 16 and an appropriate modeling of 15 the various zones of influence for the events included 16 in Bin 16.

17 And then, following that, once we have 18 developed those methods and agreed and aligned on 19 those approaches, and tested them, and make sure that 20 we feel like they're practical and implementable, 21 then, obviously, that translates into updated 22 guidance. So, I think, in general, we are aligned on 23 the major tasks of the Working Group. Perhaps the 24 timeline and the ability to get through those tasks I 25 think will be an evolving issue, as we really start to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

123 1 dig into the technical details.

2 So, I have a slide with a lot of words for 3 kind of what -- and again, I don't have answers for 4 all of these or necessarily any of these, and 5 certainly some technical experts in the crowd who can 6 support, as necessary, but we wanted to identify some 7 of what we think, in working through these tasks in 8 the Working Group, what are some of the key insights 9 that we think we need to take into account as the 10 Working Group, as we work through these?

11 First, with respect to the frequency work, 12 you've heard about binning or sub-binning or this idea 13 that Bin 16 includes multiple different kinds of 14 events represented. And so, I just wanted to give you 15 a flavor of what we're talking about there.

16 So, right now, the total population in Bin 17 16 which is the frequency bin and the fire events 18 database that supports the arcing fault inputs into 19 the fire PRA. So, from 1964 to 2017, the total 20 population of that bin is 28 events across the U.S.

21 nuclear industry.

22 It does include a wide variety of 23 equipment. There's not a lot of events, but there's 24 a lot of potential sources that come into that bin.

25 So, it includes medium voltage, low voltage. It NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

124 1 includes switchgear. It includes motor control 2 centers. It includes bus ducts, and it includes --

3 basically, in the Charter we refer to it as the 4 arching fault hazard because it includes things that 5 would be characterized as flashes or blasts that 6 really didn't do damage outside the component and did 7 not, essentially, have a zone of influence or create 8 a post-event fire, as well as events that are 9 definitely high energy arcing faults and do have that 10 larger zone of influence and immediate damage or fire 11 or immediate damage to components.

12 The data includes events with and without 13 post-event fires and with and without post-event 14 damage. And perhaps it's obvious to say, but with 15 only 28 events, it may be that we have to use some 16 engineering judgment and for different types of 17 components or different aspects, what we actually know 18 about the probability of having a flash event or a 19 blast event versus a HEAF event. Because with 28 20 events, you can only break that out down into so many 21 sub-bins. And you've got 30 bins and three-quarters 22 of them have one event or zero events in it, I'm not 23 really sure how much that tells you from a statistical 24 perspective. So, we're going to have to be willing to 25 bring in other expertise and data and judgment into NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

125 1 that consideration.

2 But I do think, nevertheless, the OE in 3 the ISAs database and the OE in Bin 16 does give us 4 some pointers about things that are critically 5 important. So, I'll highlight a few of those here.

6 First of all, medium voltage equipment, 84 7 percent of the events in Bin 16 that were damaging 8 occurred in medium voltage equipment. So, we're not 9 seeing large numbers of these events in low voltage or 10 high voltage equipment.

11 I'll also add that, for low voltage 12 equipment, there were two damaging events in low 13 voltage equipment in this population of 28 events. In 14 both of these cases, they were allowed to exist for a 15 prolonged period of time because there was not enough 16 energy to trip the protection features in the design.

17 So, they simply didn't have enough current to cause a 18 trip, and that's why they persisted for a long period 19 of time.

20 And so, although they're in the bin and 21 they do indicate post-event damage in some cases, we 22 need to do some more digging into what that damage 23 actually was. And it appears that the damage was due 24 to a post-event fire and not related to the HEAF 25 itself.

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

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

126 1 And then, of the events with external 2 damage, the graph at the lower right of the slide 3 shows a breakdown of these events. In Bin 16, there 4 were 11 of them that caused external damage. Seven of 5 11 of those were in bus ducts, either iso phase or 6 non-segmented bus ducts.

7 And so, I would point out that our 8 experience in fire PRAs tells us that, generally, due 9 to the location of the bus ducts -- you know, they 10 tend to be outside or in the turbine building, and not 11 impacting safety-significant equipment -- so, they 12 tend to be lower contributors in the fire PRA. So, 13 more than half of those causing external damage would 14 have been in that type of application.

15 So, if I can go to the next slide?

16 Similarly, some early thoughts or insights 17 related to the zone of influence work that we've been 18 referencing. And I think we've pointed to this as 19 well. Currently, the zone of influence that is used 20 is based on one event that at the time it was created 21 was kind of the worst-case known event. It was a 22 SONGS event, and it's the basis for the 3x3x5 zone of 23 influence that's currently used.

24 So, just consideration of the test 25 protocol that we're involved in now and trying to come NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

127 1 up with a zone of influence versus what has been done, 2 it's important to recognize the differences in those 3 approaches if you're trying to come up with a 4 reasonable delta CDF.

5 And then, secondly, to date, small-scale, 6 no large-scale, test results on HEAF or the byproduct 7 characterization have been available for analysis.

8 And so, to fully have a robust set of comments about 9 future test plans and the data that needs to be 10 collected to make sure that we can make good 11 decisions, we don't really feel like we can do that 12 until we've been able to analyze the existing data.

13 So, useful insights which might play into 14 that feedback on future testing related to zone of 15 influence work is, first of all, aluminum oxide is an 16 excellent insulator material with very low 17 conductivity. And aluminum in most situations will 18 oxidize almost immediately to aluminum oxide.

19 And so, one of the reasons we're 20 particularly interested in the small-scale or material 21 characterization properties is because, if the white 22 dust or very fine particles that we're observing at 23 distant locations are primarily aluminum oxide, and 24 they're not conductive, it may be that they don't 25 really -- you know, they may be very fine and spread NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

128 1 easily like volcanic dust, but if they don't 2 contribute to damage, then that may not constitute a 3 zone of influence necessarily in a PRA.

4 Finally, we know from testing and OE that 5 the arc will travel away from the source. And so, in 6 the testing that's been done, as we said, we induced 7 the arc at the switchgear, at the busbars, and it's 8 going to immediately blow out the back of the cabinet.

9 And if you put cable coupons directly in front of 10 that, is that -- we've talked about the location of 11 the arc and the potential impact that that could have 12 on the testing. But also, when you're trying to 13 understand the fragility of components, it's important 14 to represent cables and cabinets and other SSEs in a 15 manner that's consistent with how they might be 16 configured in the plant.

17 Again, just the fourth bullet there is 18 kind of a similar concept. I know some of the OECD 19 testing on bus ducts, they, I believe, had the bus 20 duct on the ground. And so, if you're talking about 21 a conical zone of influence, which is what we 22 currently use in the fire PRAs for bus ducts, well, if 23 you did the testing on the ground, then you would, 24 obviously, not be able to get that downward zone of 25 influence that you would expect to see in the plant.

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

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

129 1 And that wasn't part of the NRC's current 2 test plan, but I just point that out as indicative of 3 the kind of things we need to make sure that, when 4 we're configuring this equipment and collecting test 5 data, that the data we're collecting is representative 6 of the way the equipment would be installed and 7 operated in the field.

8 And then, we've alluded to this as well.

9 Bin 16 does include significant frequencies that 10 include arc blasts and arc flash events. Of those 28 11 events, we have looked at them, and at least 10 of 12 them -- there's a few that we want to understand more 13 about -- but at least 10 of them would be classified 14 as flashes or blasts, which, per the definitions 15 proposed by the NRC, would cause no damage external to 16 the component, and therefore, essentially, a 17 negligible zone of influence versus the one-size-fits-18 all zone of influence.

19 Early insights on plant risk. I think 20 this is going to be a challenge for the Working Group.

21 Although EPRI is not involved in the generic issue 22 process, it's certainly something we're going to have 23 to be able to address as part of the generic issue.

24 This idea of an average plant risk is an 25 elusive concept, especially in this situation, because NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

130 1 there are some drivers, some designs, some risk 2 factors that would certainly make you more susceptible 3 to this type of risk, the most obvious being, do you 4 have aluminum installed in these components or not?

5 A good fraction of the industry may not have aluminum 6 or may have aluminum only in bus ducts where it's not 7 particularly significant to risk.

8 A review of all of the U.S. units shows 9 that 70 percent of plants are not susceptible to a 10 long-duration generator-fed fault on a safety-related 11 bus. So, we've identified this as one of the primary 12 drivers. You know, where can you get a long-duration 13 fault? When there's only one breaker and that's where 14 the fault occurs, then you have to wait for the 15 generator to spin down while you're still feeding that 16 fault. But 70 percent of the U.S. industry would not 17 have a safety-related bus that is even susceptible to 18 that kind of fault.

19 Even when the long-duration generator-fed 20 fault can occur, the following factors must be 21 considered. So, we've talked about the voltage decay 22 as the generator spins down. And I'm glad to hear 23 that we're looking at how do we make that in future 24 tests reasonably represented.

25 The presence or absence of aluminum and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

131 1 the proximity of aluminum to the fault location, and 2 then, other equipment in the zone of influence, and 3 the actual fragility of that equipment.

4 With respect to the bus ducts, I will say 5 our not 100-percent survey of the U.S. industry, but 6 our reasonable survey of the industry suggests that 7 it's not rare to have aluminum in bus ducts, mostly 8 due to weight and cost. So, we probably are going to 9 find aluminum there, but the risk is dependent upon 10 the SSEs within the zone of influence. And aside from 11 some spatial considerations at very specific plants, 12 it's generally a low contributor to risk in most fire 13 PRAs. So, having a larger zone of influence in those 14 situations is probably not going to make much of a 15 difference.

16 Testing considerations. We talked at the 17 beginning about differentiating factors, the breaker-18 initiated versus switchgear-initiated fault. And I 19 have a side views of a breaker, and in the NRC's 20 presentation they showed the shortening wire across 21 the switchgear busbars that they used to initiate the 22 fault.

23 I will say, we found that more than 50 24 percent of the experience in the industry is that it 25 is the moving parts where you rack breakers in and out NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

132 1 where the fault is likely to initiate, either due to 2 maintenance or human performance, or other causes.

3 But it generally will tend to be on the breaker side 4 or at the breaker stabs or where those moving parts 5 come together.

6 And those parts of the breaker are made 7 out of copper, and they travel through some porcelain 8 bottles and through a torturous path, through some 9 more porcelain components before they move into the 10 segment of the cabinet where the switchgear busbars 11 would be.

12 And so, we don't know, but it could be 13 important to consider that because, you know, if the 14 fault initiates there, it has to sustain itself and 15 move far enough along, and continue to sustain energy 16 in order to make it to the aluminum busbars, and then, 17 still have enough energy to cause this larger-than-18 expected oxidation. That may be important to 19 consider.

20 I think we've touched several times on the 21 data for evaluating damage. Just the mere observance 22 of damage to instruments at 3 feet or presence of 23 substance at a distance from the test does not, in and 24 of itself, dictate a zone of influence. To agree with 25 the NRC, that's not an easy question to answer. You NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

133 1 know, how are we going to take these -- you know, what 2 instrumentation, what equipment? How are we going to 3 take measurements? How are we going to know what part 4 of that is due to aluminum versus everything else?

5 That's part of the switchgear. How are we going to 6 have a baseline to compare to in order to come up with 7 a zone of influence? I think we have to consider that 8 as part of the design of the future test program.

9 And I am personally not familiar with the 10 Sandia model. So, I don't have any comments on that 11 model or the fidelity of the model, or the ability of 12 the model to predict damage, and therefore, a 13 representative zone of influence.

14 So, target selection is critical to assure 15 the results are representative of plant equipment.

16 Usually, in a switchgear room you're going to have 17 multiple cabinets. And therefore, does the byproduct 18 penetrate through the cabinet? Does it cause damage 19 inside another cabinet? Does it move through the 20 louvers? That's a little different than just having 21 a rack directly in front of the blast.

22 Actual location and data collection to 23 assess impact on operation. To date, the OE does not 24 suggest large-scale deposition of byproducts or damage 25 outside the cubicle. So, when there has been NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

134 1 byproducts, you know, in general, we have been able to 2 wipe them off and close the cabinet and everything 3 still works. So, we need to make sure that we're 4 using representative equipment to assess fragilities 5 and zones of influence.

6 And then, clarity on what conductivity and 7 what material is being measured will be a challenge 8 for the test program. And along with representative 9 conditions that lead to the longer-duration faults, 10 this is why I asked earlier about the extrapolation of 11 an 8-second fault in low voltage equipment to be 12 representative of high voltage equipment. And so, 13 this really gets to that point.

14 In medium voltage equipment, we see very 15 clearly in the OE that the dominant contributor there 16 is this unit-connected design where the generator has 17 to wind down as the voltage collapses and continues to 18 supply energy into the fault for a long-duration 19 period of time.

20 For low voltage equipment, what we see as 21 the contributor most frequently there is it's a long-22 duration event simply because there's not enough 23 energy to trip the protection scheme. And therefore, 24 the current is just below what you need to cause a 25 trip in the circuit. And so, it just continues to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

135 1 propagate. So, there may be long-duration arcing 2 fault events, but they may not have the high energy, 3 because, by definition, if they did, the equipment or 4 the protection system would be expected to trip.

5 And so, I think you can see how what leads 6 to a long-duration fault in low voltage is a different 7 condition than leads to a long-duration fault in 8 medium voltage. And so, the extrapolation of one to 9 the other I think would have to be reviewed by the 10 Working Group to make sure that we're comparing apples 11 to apples in our extrapolation.

12 So, to kind of circle back to where we 13 started, at least our hope, like what the NRC said, is 14 that we come out of this with a refined Bin 16 15 treatment and supplemental guidance to update 16 NUREG-6850, and an acknowledgment that all arcing 17 fault events are not the same, and that risk should 18 consider the probability of various types of events 19 and associated zones of influence, neither of which 20 are probably a one-size-fits-all.

21 In the diagram there, maybe the cabinet in 22 the middle is a low voltage cabinet and probably is 23 most susceptible to high energy -- or not high energy 24 -- but an arcing fault with a relatively or non-25 existent zone of influence. And maybe there's some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

136 1 probability of a larger zone of influence. And then, 2 maybe it's the bus duct with aluminum that has a 3 larger zone of influence because it does have a low 4 minimum. And we do see in the OE damaging types of 5 HEAFs occurring in that type of equipment.

6 So, this is by no means intended to 7 illustrate the answer. Just that we think that 8 there's a nuance in the answer, and that we hope what 9 we end up with at the end of this is a reasonable 10 representation based on all the input and the expert 11 judgment of the probability of the various types of 12 events represented in Bin 16. That's kind of 13 component- or application-specific. And then, also, 14 a representation that zones of influence for each of 15 those probabilities is also probably variable.

16 So, with that being said, we look forward 17 to continuing to work with the Working Group. As I 18 said, it's been a good collaboration on the Charter, 19 making sure that we had the right tasks identified in 20 order to get to a technically-solid basis for 21 improving our state of knowledge and our modeling of 22 the risk of high energy events or arcing fault events.

23 MR. HAMBURGER: Thank you, Kelli. Would 24 you be willing to answer any questions from the 25 audience, if they have any?

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

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

137 1 MS. VOELSING: I'll try or --

2 MR. HAMBURGER: Okay.

3 MS. VOELSING: -- I'll defer to the 4 technical experts. But sure.

5 MR. HAMBURGER: Okay. Any questions for 6 Kelly from the room?

7 (No response.)

8 MR. HAMBURGER: From the webinar?

9 (No response.)

10 MR. HAMBURGER: Okay. Thank you.

11 And I have one more presentation that was 12 provided to me in advance by Victoria Anderson from 13 NEI.

14 And after this presentation, the floor is 15 open for any questions that anybody might like to ask.

16 MS. ANDERSON: All right. So, we wanted 17 to talk a little bit about the context of the 18 technical information that this effort is gathering 19 and the operating experience and the technical 20 information we have so far, and put that in the 21 context of the risk evaluations that are done as part 22 of the pre-GI process.

23 Because any decisions associated with the 24 generic issues process are designed to be risk-25 informed, and really the specific thing we're looking NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

138 1 at here, as I'll get onto in the next slide, is the 2 change in core damage frequency due to potential 3 aluminum HEAFs needs to be treated as the core damage 4 frequency due to aluminum HEAFs minus the core damage 5 frequency baseline. And I think one thing we talked 6 about a little bit this morning was that we aren't 7 really sure yet how we're defining that baseline, and 8 it's very important that we define it appropriately.

9 Next slide, please.

10 So, most past generic issues have been 11 dispositioned according to delta CDF instead of CDFs.

12 If you look at this chart here that talks about which 13 issues you exclude from further consideration and 14 which ones you continue, the X-axis is core damage 15 frequency and the Y-axis is change in CDF.

16 Based on our knowledge of the fire PRA 17 CDFs out there in the industry, which would really 18 tell us about our HEAF baseline, we believe the 19 results for this pre-GI evaluation would be in this 20 green region. We don't think anything would be over 21 1E minus 4 on core damage frequency. So, therefore, 22 the determination is going to be based on delta CDF, 23 which really underscores the importance of properly 24 defining that CDF baseline.

25 In either factor in that determination of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

139 1 the delta CDF, any conservative assumptions could 2 result in an inappropriate decision. It could wind up 3 putting you in the incorrect region. And if we try to 4 make a decision based on a single assessment of the 5 baseline risk or we just use one specific event, 6 initiating event, frequency, or consequences, we'll 7 wind up with an inaccurate and unrepresentative 8 answer, which is not what you want in a generic issue 9 program.

10 Next slide, please.

11 So, this chart gives sort of a pictorial 12 of where we think we might be and what kind of 13 decisions we might be making. So, I'll just give an 14 explanation of what the different dots are for those 15 in the room.

16 The orange dots are perhaps guesses of 17 where we think we might be, based on some of the 18 initial bounding testing that's been done. The blue 19 dots are where we think perhaps realistically some of 20 the plants may lie in that delta CDF versus CDF region 21 if we applied some of the insights from the tests done 22 to date. And then, that green dot is perhaps what we 23 would decide our assessment would be, based on just 24 those orange dots, if we aren't really adequately 25 considering all the input from various plants.

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

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

140 1 So, there's really two things to consider 2 from this chart. The first thing is, should we be 3 making determinations based on just a couple of small 4 samples interpreting data from bounding results so 5 far, when that may actually not be representative of 6 what we see in the industry? And the other 7 consideration is, are those orange dots themselves 8 based on conservative assumptions? And perhaps we 9 don't have anything outside of the excluded region.

10 I'm not saying that these are the answers, but this is 11 a stylized representation of where we may be finding 12 ourselves and underscores what it's very important to 13 make sure that we are accurately assessing both the 14 baseline risk and the potential delta risk from 15 including aluminum.

16 So, moving on to the next slide, we have 17 been talking about this in terms of, well, what have 18 we done with past GIs? It wound up being a matter of 19 delta CDF.

20 So, if you look at what we did with the 21 seismic GI, when that came out, we didn't even have 22 the amount of information relative to seismic PRAs 23 that we have relative to fire PRAs right now. But, 24 even with that limited risk-informed approach, there 25 was a pretty high variability in the decisionmaking NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

141 1 that we came up with.

2 So, again, if you go back to that previous 3 chart, even without detailed information, we've found 4 a lot of variability. So, we anticipate that you 5 would find, with additional plant detail, you would 6 continue to find more variability.

7 So, we believe that it's extremely 8 important that we include realism in the assessment 9 and in resolution of the pre-GI. It's important to 10 make sure that we're really isolating the specific 11 impacts associated with the inclusion of aluminum 12 because that's a scope of this GI, and appropriately 13 shore up what that CDF for that baseline is, so that 14 we can accurately calculate a delta CDF and accurately 15 assess this issue.

16 We believe that a lot of plants may not 17 require action. We think a lot of plants will fall 18 into that region to exclude from further 19 consideration. If we have a properly-conducted pilot 20 with proper insights and proper consideration of OE 21 and properly-constructed tests, I don't think that 22 we'll find ourselves in that delta CDF region that has 23 us moving forward.

24 So, if we move on to our last slide, I 25 think what our priorities for the evaluations are NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

142 1 -- and these are not just priorities in terms of what 2 we think is most important, although they are that, 3 too -- these steps really need to be taken 4 sequentially for us to get to a proper answer. So, we 5 need to make sure that we fully understand and 6 evaluate data from previous tests. I think we heard 7 a few times today that some of the previous test 8 results weren't fully understood; we didn't have full 9 material characterizations complete yet. And we think 10 that that's critical to get that done, so we can get 11 a better picture of what phenomena we're looking at.

12 We need to achieve consensus on the 13 technical approach for frequency and zone of influence 14 work. It sounds like the Working Group will be taking 15 that on. So, we fully support continuing with that.

16 Once that's complete, I think that would 17 be the appropriate time to move on to any additional 18 testing to resolve any identified gaps, so that we 19 make sure that our tests are carefully constructed to 20 close those gaps.

21 And finally, complete a realistic 22 evaluation of the plants using all that test data in 23 OE.

24 So, I think just one more thing I want to 25 underscore in my presentation is that it was mentioned NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

143 1 earlier that the industry comments were dispositioned.

2 We turned in comments following the workshop last 3 April. I know that there still have been some changes 4 to the test plan made. But, based on the comment 5 resolution we saw, it doesn't look like the test plan 6 would really support a fully realistic evaluation of 7 this pre-GI. And we do think it's important to 8 perhaps look at those comments again before we proceed 9 with further testing. Because, right now, we don't 10 think that the testing plan fully aligns with what is 11 needed for a realistic evaluation.

12 MR. HAMBURGER: Would you be willing to 13 answer any questions?

14 MS. ANDERSON: I'll answer any questions.

15 MR. HAMBURGER: Okay. Anybody in the room 16 who has questions or comments for Victoria?

17 MR. HYSLOP: Hi, Victoria. J.S. Hyslop, 18 NRR.

19 You talked about the baseline and not 20 really having a baseline. I know we had some comments 21 earlier about 2 seconds versus a different model. Is 22 that what you're getting at when you say that?

23 Because there is a baseline, at least one that's been 24 used in regulatory analysis, and the NRC has accepted 25 it, and licensees have acted on it. So, can you NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

144 1 clarify that for me?

2 MS. ANDERSON: What I mean is, if we're 3 trying to determine what the delta CDF is for the 4 specific impact of aluminum involvement in a high 5 energy arching fault event, our baseline needs to be 6 isolating out -- needs to be defined such that we're 7 isolating out the unique impacts of having aluminum 8 involved. And right now, I don't think that we have 9 the detailed information on the phenomena associated 10 with the testing results we have. That's not really 11 reflected in the baseline. If I understand, right now 12 the baseline is simply what we get from 6850 right 13 now, which is based on one event. So, we wouldn't 14 really be comparing an appropriate baseline to an 15 appropriate CDF from the aluminum impacts.

16 MR. HYSLOP: So, does that mean that, if 17 you feel that the testing and the analysis truly 18 represented aluminum, that you couldn't use the 6850 19 baseline? Is that what you're saying?

20 MS. ANDERSON: I don't think that you 21 would be able to just take the 6850 baseline by 22 itself. I think you need to do a more nuanced 23 evaluation once you have a better understanding of the 24 phenomena that we're observing in those HEAF tests 25 that involve aluminum.

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

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

145 1 MR. HYSLOP: So, you're recommending, 2 after we understand the issue, just looking back at 3 6850 and verifying that's a suitable baseline or 4 making a decision --

5 MS. ANDERSON: Or adjusting it. I don't 6 think it's going to be a matter of verifying. I think 7 it's going to be a matter of adjusting it in some way 8 to make sure that you're uniquely reflecting just the 9 impacts of aluminum.

10 MR. HYSLOP: I guess that might be a 11 plant-specific analysis as well.

12 MS. ANDERSON: I think you could do it on 13 a generic basis or using representative plants.

14 MR. HYSLOP: Yes, that sort of gets to the 15 pilots --

16 MS. ANDERSON: Yes.

17 MR. HYSLOP: -- because, earlier in your 18 slide, you talked about having not just one PRA.

19 MS. ANDERSON: Correct.

20 MR. HYSLOP: And certainly we need several 21 pilots, so that we can look at this and not just have 22 to rely on one answer.

23 MS. ANDERSON: Yes, absolutely. But, as 24 I noted, I think we need to have a better idea of what 25 we're going to be asking the pilots to do before we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

146 1 get any volunteers.

2 MR. MELLY: Hi, Victoria. This is Nick 3 Melly in the Office of Research.

4 Can we go back to the slide with the dots, 5 the third slide?

6 MS. ANDERSON: Yes. Which, by the way, 7 are not representative of anything. This is a 8 pictorial -- yes, it's a cartoon.

9 MR. MELLY: All right. I was kind of 10 hoping that there were numbers behind them right now, 11 but --

12 MS. ANDERSON: No, no. It's a cartoon.

13 (Laughter.)

14 MR. MELLY: Okay.

15 MS. ANDERSON: I mean, aside from the fact 16 that we know that almost all the fire PRAs sit in that 17 CDF region.

18 MR. MELLY: Right. I wanted to speak to 19 this, just to reiterate the fact that the pilots are 20 really necessary as part of this process. The generic 21 issue assessment stage needs a risk evaluation. If we 22 do not have the pilots to do that, it falls back on 23 the NRC staff in order to make approximations, to make 24 assumptions on what that risk would be at various 25 plants.

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

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

147 1 Some ways that we've done this in the past 2 are the SPAR model, and there are conservatisms within 3 and there's some limitations on the NRC staff at what 4 resources we have available to make our assessment.

5 For instance, spatially in the room it's very 6 difficult to tell what other CCDP targets would be 7 involved in a SPAR model versus the wealth of 8 information that the plant PRA models actually have.

9 So, that's just I wanted to stress that the pilots are 10 very important if we're looking to get a realistic 11 representation of the increase or the delta risk.

12 MS. ANDERSON: And I'll say it again, that 13 we're not really going to be able to in good faith 14 encourage a licensee to volunteer as a pilot until we 15 have confidence that the evaluation that's being done 16 is based on realistic evaluation of the operating 17 experience and realistic testing.

18 MR. MELLY: And I hope that, with the 19 products that the EPRI Charter and the NRC Working 20 Group come up with in terms of frequencies, zone of 21 influence, operational experience to support that, 22 will bolster some confidence in that approach.

23 MR. GARDOCKI: Hey, Ashley, this is Stan 24 Gardocki. Oh, this Victoria. I was thinking that's 25 Ashley. Sorry.

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

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

148 1 (Laughter.)

2 MS. ANDERSON: That's Kelli.

3 MS. ANDERSON: Ashley is at home with her 4 baby. She's beat all of us today.

5 MR. GARDOCKI: Okay. Just one comment on 6 this graph. We've used it in the Generic Issues 7 Program, and people understand that it's not a go/no-8 go definitive line there.

9 MS. ANDERSON: Yes.

10 MR. GARDOCKI: It's a gradient --

11 MS. ANDERSON: It's soft line, yes.

12 MR. GARDOCKI: It's a gradient line. And 13 if you go back to the PRA Reg Guide 1.197, you'll see 14 that notation in there that says you take a gradient 15 approach when you get near that line. So, it gets a 16 little bit, you know --

17 MS. ANDERSON: Yes.

18 MR. GARDOCKI: It gives that go/no-go 19 approach. Okay?

20 MS. ANDERSON: Yes. I should take a 21 blurring tool to that and sort of make it nice and 22 fuzzy. If you're right on the line, you still have 23 further evaluation to do probably.

24 MR. GARDOCKI: You can try. I'm doing 25 that in updated documents that we have --

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

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

149 1 MS. ANDERSON: Yes.

2 MR. GARDOCKI: -- to try to show that.

3 MS. ANDERSON: Actually, I believe NEI put 4 in a comment to the effect that that should be done --

5 MR. GARDOCKI: Yes.

6 MS. ANDERSON: -- in NRC documents.

7 MR. GARDOCKI: One question I do have is, 8 we put an IN out about a year ago. And based on 9 operating in a plant, you have to take an IN that the 10 NRC puts out and evaluate for your plant to see if it 11 has impact. But the NRC doesn't have resources to go 12 out and see what reaction the plants had to the IN 13 that we put out. I don't know if your agency can 14 somehow get that information to see how the plants 15 reacted to that IN. If they put it in a corrective 16 action base, and they say, well, does it apply to our 17 plant, does it not apply to our plant, and what action 18 plan they have and correspond to that IN, that would 19 be a resource available maybe for your reach-out to 20 the plants that already out there. They don't have to 21 do anything. They already have done that, or 22 supposedly have done that.

23 MS. ANDERSON: Yes. From what I 24 understand, I believe most plants looked at that 25 testing and they believed that it was not NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

150 1 representative of their plant configurations and I 2 think determined that there was no action needed on 3 their part. So, are there any licensees with action 4 to the contrary? There are a couple of licensees in 5 the room, but I don't see any of them getting up to 6 correct me. So, I do not believe any plants made any 7 physical modifications as a result of that Information 8 Notice.

9 PARTICIPANT: I would say, because of the 10 prototypicality of the --

11 MS. ANDERSON: Yes, yes. I think that was 12 the determination, was that the tests that the 13 Information Notice was based on, most plants believed 14 that the data didn't have applicability to their 15 plant. So, they did not take any action to make 16 physical modifications. But, again, there are many 17 operating plants throughout the fleet, and I don't 18 have full knowledge of what all of them have done, and 19 no licensees have come to correct me yet.

20 MR. BOYCE: Hi. Tom Boyce. I'm the 21 Branch Chief for Regulatory Guidance and Generic 22 Issues.

23 These might be just feel-good comments, 24 but I'm observing this and I just wanted to share 25 that, from the program perspective, having the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

151 1 participation today is really very positive. The 2 worst scenario for the GI program is that we don't 3 even have public meetings and we don't get this input, 4 and we do operate with one-size-fits-all and perhaps 5 overly conservative type of outcomes. So, I think 6 this is very positive.

7 And the best thing for the program and for 8 safety, honestly, is if all those green dots and those 9 orange dots became blue dots. So, we're not in it 10 just to do something. We're in it to make sure safety 11 is accomplished.

12 MS. ANDERSON: Yes, we concur with that.

13 MR. HAMBURGER: Okay. Anything else for 14 Victoria?

15 (No response.)

16 MR. HAMBURGER: Okay. Thank you.

17 MS. ANDERSON: Thank you.

18 MR. HAMBURGER: Anything from the webinar?

19 (No response.)

20 MR. HAMBURGER: Okay. So, the floor is 21 now open for anyone who would like to make any 22 comments about any of the presentations we've given 23 today or anything tangentially related, HEAF-related.

24 MR. BOYCE: We just got a question from 25 Paul Gunther on the GI process. So, his question is, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

152 1 "Will the GI process look at any age-related 2 degradation mechanisms that might contribute to the 3 risk frequency for HEAF events?"

4 MR. GARDOCKI: This is Stan Gardocki.

5 As part of the Generic Issues Program 6 itself, we would look at the initiators of the event 7 and the contributors to the probability of the event 8 in the risk analysis. As far as aging management 9 specifically, the NRC has its own aging management 10 program for life extension, AMR process, and they have 11 specific engineers that work on cabling, submerged 12 cabling, and they put out the report on that as 13 influence to aging management. I don't know if that 14 addresses it correctly.

15 Do you want to add some more?

16 MR. MELLY: This is Nick Melly from the 17 Office of Research.

18 In addition to when we do the frequency 19 work, we also trend the frequencies of fire events 20 within power plants and the most significant events.

21 So, we're trying to capture any aging-related issues 22 when we're looking at the root cause of the events 23 themselves, as well as the frequency of occurrence in 24 the plants.

25 Currently, we have seen no increase in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

153 1 frequency of high energy arcing fault events that are 2 solely attributed to aging concerns. When we look at 3 the root cause, it's kind of all over the board of 4 different factors, such as human-related failures, 5 design specifications. Aging is also in there as well 6 as foreign material. It is an aspect that we're aware 7 of, and we look for it when we evaluate the frequency 8 of these events.

9 MR. TAYLOR: Yes, this is Gabe Taylor, 10 Office of Research. I'll just add one additional 11 piece of information that kind of goes on what Nick 12 said.

13 Back when we issued the Information Notice 14 last year, although we didn't do any type of formal 15 trending analysis, I did do an informal kind of back-16 of-the-envelope-type evaluation. And from the data 17 that we had, the operating experience that we had, we 18 didn't see any increase, any decreasing trend versus 19 time. So, it looks like it's somewhat of a constant 20 hazard or a constant failure rate type of event.

21 MR. HAMBURGER: Any other questions from 22 anybody in the room?

23 (No response.)

24 MR. HAMBURGER: You have one more?

25 MR. BOYCE: Yes. We have one more NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

154 1 question from Rob Cavado. His comment or question is, 2 "I want to reinforce the comment earlier that we need 3 a public meeting prior to finalizing the actual PRA 4 modeling to ensure the updated HEAF modeling is usable 5 and realistic." So, more of a comment.

6 MR. TAYLOR: Yes, I don't think we've 7 changed from this morning. I think we agree, agree 8 with that. And when we get to that stage, we'll 9 probably engage the stakeholders.

10 MR. HAMBURGER: Okay. And just for --

11 MR. CHEOK: Actually, I also have one 12 question. I said this morning that we need to get a 13 clarification of industry and other stakeholder 14 comments. I know that we have received an abundance 15 of comments, and I think the staff feels like we have 16 dispositioned them. I get the sense that a lot of the 17 disposition is not satisfying. Put it this way.

18 So, I'm going to offer again -- I heard it 19 from Victoria again, which is what reminded me of 20 this. That one of the steps that was proposed by NEI 21 was that we resolve all the comments before we start 22 on the next set of testing. So, I don't want to do 23 the next set of testing and hear that we did not 24 resolve all the comments. I think that would not be 25 fair to the staff or to all of our stakeholders.

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

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

155 1 So, again, I think I want to work with the 2 staff and work with NEI and EPRI on maybe a date as to 3 what we can do in terms of resolving all the remaining 4 comments, all the comments that need to be resolved 5 before we start our next set of testing.

6 MR. HAMBURGER: Go ahead.

7 MS. UHLE: Okay. Jennifer Uhle from NEI.

8 So, thanks, Mike. We are very interested 9 in working together collegially here to address the 10 issue. And we recognize the importance of a pilot in 11 determining the delta CDF. And so, yes, we hear you 12 and appreciate that.

13 But there's also the element of the 14 modeling that has to be, I think, discussed because I 15 think at least what I've heard is that we'll be given 16 the results of the modeling. And that, to us, is not 17 pleasing, considering the modeling -- having a 18 modeling background myself -- can go in a very 19 conservative direction. It can go in a very non-20 conservative direction.

21 And so, if we are going to be as an 22 industry -- you know, the generic issue disposition is 23 going to be impacted by the modeling. We would like, 24 the industry would like to play a role in at least 25 commenting on the modeling approach, rather than just NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

156 1 the data or the results that get generated through the 2 use of the model.

3 So, the more cooperation, and cooperation 4 -- obviously, NRC still maintains its independence --

5 cooperation is what I mean when I say "communication" 6 and justifying a variety of assumptions. The more 7 that occurs, I think the more confidence that people 8 will have in participating in a pilot, which is going 9 to be a critical part of this.

10 But, if we are just given a model and it's 11 used, and we haven't participated or aware of how it 12 was developed, that's going to be problematic from our 13 side of the issue.

14 MR. TAYLOR: Yes, I'll go ahead and 15 respond to that. I think I agree with what you said.

16 We need to get involvement early on. Probably I 17 misspoke earlier about not getting involvement early 18 on.

19 MS. UHLE: Okay.

20 MR. TAYLOR: As I said earlier, we're just 21 changing approaches. Just recently, that happened.

22 So, I think once we get moving, we'll obviously be 23 working with EPRI through that format. And then, as 24 we move along, we'll identify points and try to bring 25 in other stakeholders via public meetings or other NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

157 1 forums.

2 MS. UHLE: Okay. Great. I appreciate 3 that, Gabe.

4 And so, if I can just continue and say 5 that, getting industry information to give to the NRC 6 is obviously work on our part as well as your 7 licensees. And so, having assurance that this 8 information is going to be used in a realistic manner 9 would also help in encouraging the industry to be more 10 willing to participate in the process.

11 Thanks.

12 MS. VOELSING: This is Kelli Voelsing from 13 EPRI again.

14 Mike, I really appreciate your commitment 15 to go back and make sure that we addressed and 16 resolved all comments. And at least speaking for 17 EPRI, we'll make sure that we get you any that we 18 think are still outstanding in a concise format.

19 I would, however, say that I think there's 20 a strong probability that there will be additional 21 comments that we want to take into account in planning 22 future test phases based on the results of previous 23 tests. And so, until we have those test results and 24 the ability to spend the time and look at them, we 25 probably don't have a complete set of comments. But NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

158 1 we'll certainly get you the previous ones in a concise 2 and consolidated fashion.

3 MR. CHEOK: So, at lunch today, Stan 4 Gardocki came over to my office, and coming over to 5 this meeting, Tom Boyce was also talking to me. And 6 they looked at the generic issue timeline, right? And 7 we saw that this morning. And rightly so, the generic 8 issues staff would like to see things move on, so they 9 do not drag on forever. And if it's something that's 10 a safety case, we should not take forever to resolve 11 something that potentially could be safety.

12 But, on the other hand, we also have got 13 to get it right, right? So, I think when we talk to 14 Stan and Tom, we understand that this process maybe 15 may take a little longer because we want to get it 16 right, but we still cannot push it out too far. But 17 we understand the sequence of events that you laid 18 out, and we'll talk, talk over this with the staff on 19 the effect of the test schedules, the availability of 20 the lab, and things like that.

21 But those are comments that are good 22 comments. We will take that into account.

23 MR. GARDOCKI: To point out -- this is 24 Stan Gardocki -- point out exactly what this 25 conversation was, we had the pilot plants. You could NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

159 1 see on the screen the assessment plant timeline. And 2 the people on the phone, if they want to go to that in 3 my presentation?

4 We plan on completing the assessment by 5 the end of 2019, and then, rolling it all up and have 6 the assessment report done in the first quarter or two 7 of 2020. It's a very aggressive schedule based on 8 getting all this testing complete.

9 Based on the comments we are having today, 10 we see the pilot plant study is being done right after 11 June, July, and completed by August. And that's in 12 the summer of this year is when we're going to do that 13 additional testing with the bus ducts and the low 14 voltage. To complete the pilot plant study prior to 15 that testing is going to be a challenge.

16 MR. CHEOK: Again, we will work this out.

17 I think it's more important to get this right. I 18 think everyone wants to get this correct. So, we will 19 take everything, all the comments we heard today, into 20 consideration, and we will work out a schedule that is 21 not going to drag this thing out too much. But we 22 will work to do the right, to get the right 23 information, so that we can resolve the issue that has 24 a recommendation that could be defendable and that is 25 robust.

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

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

160 1 MR. HAMBURGER: I think we have one more 2 question on the webinar?

3 MR. BOYCE: Yes, we have a brief question 4 from Mohammad Mustafa, Entergy.

5 His question is, "Do existing Appendix R 6 and NFPA 805 separation criteria remain adequate?"

7 MR. HYSLOP: I don't know. We'll have to 8 see what the testing does. Certainly, this generic 9 issue is looking at a PRA and looking at risk results, 10 but I don't think it's limited to that. I think we 11 talked about Appendix R, the impact on that as well.

12 So, we just have to see, and I can't speculate any 13 further than that.

14 MR. HAMBURGER: I just want to do one 15 final sweep of the webinar for anyone who may have 16 called-in only. Can we open up all the phone lines on 17 the webinar to see if anybody has any questions they 18 need to ask? And if you don't have a question, now is 19 a great time to put yourself on mute.

20 Just give us a second to open up all the 21 lines.

22 Anybody on the webinar with a question?

23 (No response.)

24 MR. HAMBURGER: Okay. And again, if 25 anything occurs to you later, you can get in touch NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

161 1 with anybody on the meeting notice. We will sure the 2 comment or question gets to the right place, and we'll 3 get back to you.

4 Any last questions from anyone in the 5 room?

6 (No response.)

7 MR. HAMBURGER: Okay. And just a note.

8 I have two actions here for me. One is to circulate 9 the Draft Charter.

10 MR. BOYCE: We have, I think, one question 11 from Rodney Pletz. If you have your question, you can 12 ask it now. I unmuted you. Go ahead, Rodney.

13 (No response.)

14 MR. HAMBURGER: We have to open up the 15 phone lines again in the control booth. No, he says 16 we're good.

17 Rodney, do you have a question for us?

18 (No response.)

19 MR. HAMBURGER: Okay. Well, if you can 20 type it, we'll answer it.

21 So, I'm going to make sure that the Draft 22 Charter gets circulated and placed in public ADAMS for 23 industry comment before that becomes finalized.

24 And I will also publish today's meeting 25 package that includes all the presentations, the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

162 1 transcripts, and everything that was presented today.

2 So, if nobody has anything else, I want to 3 turn it over to Mr. Michael Franovich. He has some 4 closing remarks.

5 MR. FRANOVICH: Good afternoon. I'm Mike 6 Franovich. I'm the Director of Risk Assessment in 7 NRR.

8 I am Co-Chair of the GIRP. Mike and I co-9 chair the GIRP. There's a reason we've structured it 10 that way. We want to make sure that when or if this 11 issue migrates to NRR in terms of a transition, that 12 the expectations are really met.

13 Stan this morning described in detail the 14 process for the GI program. An issue migrating over 15 to NRR in the Program Office, there are a series of 16 processes that we have in the Program Office.

17 Certainly, if there's a viewpoint that there's an 18 imposition of a new requirement or a new position, we 19 will go through a very structured process. Even if 20 it's a generic communication such as a Regulatory 21 Information Summary, we do have processes to ensure 22 that we're not imposing undue burdens.

23 We have revamped those processes, 24 actually, as many of us have gone through, or, 25 actually, all of us have gone through backfit reset NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

163 1 training. And I want to emphasize that it is not 2 success for us to have incomplete information moving 3 in the regulatory program space and trying to move 4 forward with some types of expectations for licensees 5 to do more work.

6 Certainly, some of the issues discussed 7 today about risk modeling, fire modeling, the testing 8 and how representative it is, those are the types of 9 issues that would come up in our internal review in 10 NRR. Because when we look at it, we actually bring 11 those types of issues to our leadership team. The 12 leadership team is looking at it from a resource 13 impact, not just a safety perspective, but a resource 14 impact, as we have to approve resources to be able to 15 move forward. And that looks at both how long those 16 processes can involve and what will be the ultimate 17 end state.

18 We've talked about a lot of testing today 19 and the need for more communication. And I won't 20 reiterate that. There are some very good 21 characterizations of that.

22 I think it was recognized early on in the 23 GIRP when we did the Assessment Plan Report in the 24 summer timeframe that this is living plan; there would 25 be give-and-take back and forth.

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

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

164 1 If you notice in that plan, there are a 2 number of highlighted engagement points, in 3 particular, with ACRS. ACRS is a vital stakeholder 4 for us and they will scrutinize the pedigree of the 5 work and the fidelity, any kind of outcomes that come 6 from the research activity. And also, they want to 7 know which direction we'll be moving in. Those 8 processes engaging with the ACRS are very public, and 9 information is released in advance of the meetings at 10 ACRS. So, there are other engagement points where 11 information will be made available to other 12 stakeholders, not just the ACRS.

13 But point well taken about more 14 interaction. I think the interest we've had both here 15 in the room and on the webinar illustrates the need 16 for more dialog.

17 The other thing I want to really point to 18 is, ultimately, this comes down to a risk management 19 equation. And in risk management equation space 20 -- set aside the regulatory piece -- there's both 21 prevention and mitigation. And I think we need to not 22 lose perspective, that if more can be done with 23 prevention in terms of looking at predictive 24 maintenance and preventative maintenance practices, at 25 least that can help shape the risk management approach NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

165 1 to the issue and resolution.

2 We had some dialog. Kelli shared some 3 information about parsing more of the data and looking 4 more closely at the OpE. But we may not be able to do 5 statistical changes based on some qualitative 6 analysis, but if there are better practices that are 7 going to be promoted, we can take those matters into 8 consideration.

9 Victoria had some graphs up with the lines 10 that aren't really fuzzy, at least drawn on yours. We 11 do have some documents that do have some fuzzy lines, 12 by the way. NUREG-1855 on the treatment uncertainties 13 actually some illustrative points there, how to weigh 14 those in decisionmakings when you're not clearly on 15 one side or one regime versus another regime in terms 16 of thresholds.

17 But, to that point, we would consider 18 qualitative type of information there to say which 19 direction are we taking. The other thing we need to 20 keep in mind is, if there are other improvements in 21 the fleet, and we talk about FLEX, although FLEX was 22 not designed specifically for this type of hazard, 23 there is an optimization process that the industry has 24 promoted that we're well aware of and we're engaged.

25 And that's in NEI 1606 and 1608.

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

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

166 1 Some plants have gone forward and actually 2 done some modifications for other hazards to mitigate 3 those risks by modifying FLEX. With very subtle mods, 4 they've been showing -- granted, it's plant-specific 5 type of risk -- that there can be significant 6 reductions in risk.

7 So, irrespective of the regulatory piece 8 for a moment, just keep those in mind. If you're out 9 looking at doing modifications of the plants and you 10 see somewhere where there's an opportunity here to do 11 that, it may be a great opportunity in the span of 12 doing other work at the plant.

13 But I do want to note again, when we get 14 to the transition point, there will be open questions 15 that we need to make sure we have a smooth transition 16 of the issue to the Program Office.

17 There will be a risk-informed resolution 18 of the issue one way or another. That's the climate 19 we're working in. It's a smart climate.

20 And with that -- we are well ahead of 21 schedule, actually -- I will open it up for any last 22 questions here in the room and on the line.

23 MR. HAMBURGER: All right. We have one 24 question in the room, and then, we'll get to the last 25 question on the line.

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

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

167 1 MS. VOELSING: Again, sorry not a 2 question. Kelli Voelsing with EPRI.

3 Mike, I did want to give you confidence 4 regarding the holistic treatment, that at least from 5 an EPRI perspective, we've gone back and looked at is 6 there more guidance needed around these maintenance 7 and operational best practices. We found that there's 8 plenty of historical documents on that research, and 9 we've engaged with our advisors in both the 10 maintenance organization and the engineering 11 organizations, discussed it with executives, and we 12 highlight those in our pointers to those best 13 practices. So, that is definitely work that is out 14 there in the industry and ongoing. And the holistic 15 treatment at least is not lost on EPRI, and I think 16 the industry is well aware of that through our 17 advisory structure and their interactions with NEI as 18 well.

19 MR. FRANOVICH: Okay. Appreciate that.

20 MR. BOYCE: We have one question from the 21 webinar. Rodney managed to get his question in. So, 22 here's his question. He says, "This question is more 23 for participants from the industry. As we work to 24 refine and define the impact of HEAF, are any sites 25 taking action with their operators and fire responders NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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

168 1 to train them in the potential for increased ZOI from 2 these events? In other words, the survey that we did 3 over a year ago gave at least preliminary 4 identification of where HEAFs with aluminum are 5 located and could happen. Are any sites improving 6 their fire response training or fire preplan 7 information to account for these events? Do we need 8 any or want any industry alignment on this during the 9 interim period while the GI process continues?"

10 MR. HAMBURGER: I can't make anybody field 11 that question, but if anybody would like to?

12 MS. UHLE: This is Jennifer Uhle from NEI.

13 I don't have an answer to that question.

14 As with any issue, there's plant interdependency -- or 15 excuse me -- plant-specific dependencies of the impact 16 of the current knowledge. But, going forward, we 17 certainly, as an industry, are getting our views 18 together and ensuring that there is a common 19 understanding at both the staff and management level.

20 And we can certainly bring that up for consideration 21 on the industry side.

22 MR. BOYCE: We have a comment from Rob.

23 He says that, "We already have procedures in place 24 that address full compartment losses. This bounds the 25 issue."

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

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

169 1 MR. HAMBURGER: Okay. Any last questions 2 or comments before we close the meeting?

3 All right. Thank you all for your 4 participation.

5 (Whereupon, the above-entitled matter went 6 off the record at 2:31 p.m.)

7 8

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

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