ML20129A663
| ML20129A663 | |
| Person / Time | |
|---|---|
| Issue date: | 09/12/1996 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-2073, NUDOCS 9609180398 | |
| Download: ML20129A663 (300) | |
Text
w-.-~. ._. .-
. .. -A
@ffiolot Transcript of Pecoccdings 6R5t-2i?5 O NUCLEAR REGULATORY COMMISSION
Title:
434TH ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS)
- u.n. -
TRO4 (ACRS)
RETURN ORIGINAL TO BJWHITE M/S T-2E2s Docket Number: (NOT ASSIGNED) 415-7130 THANKS!
Location: ROCKVILLE, MARYLAND ;
- O i Date
- SEPTEMBER 12,1996 '
4
)
l O Di p [l* I :iu C Cil h _
! Work Order No.: NRC-840 Pages 1-260
-l 9609180398 960912 207 PDR NEAL R. GROSS AND CO., INC. C Court Reporters and Transcribers l 1323 Rhode Island Avenue, N.W.
w niastaa. a c 2'aot O
AAn S a, g u 2 3 4-4 m _ce Copy ae c
ny a ss C D 'Me Jte.cf tj 8 00mftiec l
I 1
1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION
.l )'
-s' + + + + +
3 4 434TH MEETING 5 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 6 (ACRS) 7 + + + + +
8 THURSDAY 9 SEPTEMBER 12, 1996 10 + + + + +
11 ROCKVILLE, MARYLAND 12 + + + + +
13 The Advisory Committee met at the Nuclear p..
's 14 Regulatory Commission, Two White Flint North, Room T2B3, 15 11545 Rockville Pike, at 8:30 a.m., Thomas S. Kress, 16 Chairman, presiding.
17 18 COMMITTEE MEMBERS:
19 THOMAS S. KRESS CHAIRMAN 20 ROBERT L. SEALE VICE CHAIRMAN 21 GEORGE E. APOSTOLAKIS MEMBER 22 JOHN J. BARTON MEMBER 23 IVAN CATTON MEMBER 24 MARIO H. FONTANA MEMBER
! / 25 DON W. MILLER MEMBER NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 l
2 1 COMMITTEE MEMBERS: (CONTINUED)
,- g 2 DANA A. POWERS MEMBER l )
x"/ 3 WILLIAM J. SHACK MEMBER 4 :
5 ACRS STAFF PRESENT:
6 JOHN T. LARKINS EXECUTIVE DIRECTOR 7 MICHELE KELTON TECHNICAL SECRETARY 8 ROXANNE SUMMERS TECHNICAL SECRETARY 9 SAM DURAISWAMY 10 CAROL A. HARRIS 11 RICHARD P. SAVIO 12 PAUL BOEHNERT 13 NOEL DUDLEY i A l
(>)
I 14 MEDHAT M. EL-ZEFLAWY l I
15 MICHAEL MARKLEY ,
l 16 AMARJIT SINGH 17 18 ACRS CONSULTANT PRESENT: j I
19 NOVAK ZUBER j 20 21 ALSO PRESENT:
l 22 WAYNE HODGES 23 CHARLES TINKLER 24 RICHARD LEE m
25 CHARLES ADER 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 ALSO PRESENT: (CONTINUED)
, m.,
2 HARRY SALMON 3 WILLIAM CAHILL 4 JOHN KELLY 5 ROBERT BARRETT 6 MARC PEARSON 7
8 9
10 11 12 13 t
(/ 14 15 16 17 18 19 20 21 22 23 24
,a i L
'q ,] 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
4 1 INDEX 2 AGENDA ITEM PAGE (s
O 3 Opening Remarks by Chairman Kress 5 4 Adequacy of the Code to Analyze Steam 5 Generator Tube Temperature Distributions 6 During Severe Accidents 19 7 Reconciliation of ACRS Comments and 8 Recommendations 125 9 Report of the Planning Procedures Subcommittee 10 Future ACRS Activities 11 LUNCH 12 Indian Point Unit 3 176 13 O
14 15 16 17 18 19 20 21 22 23 24 f')
) 25 NEAL R. GROSS 1 COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVd., N W.
(202) 234 4433 WASHINGTON. D C. 20005 3701 (202) 234-4433
5 1 P-R-O-C-E-E-D-I-N-G-S j g 2 (8:33 a.m.) l
,f t_,/ The meeting will no's please 3 CHAIRMAN KRESS: i 4 come to order. This is the first day of the 434th meeting 5 of the Advisory Committee on Reactor Safeguards. During 6 today's meeting the committee will consider:
7 (1) The adequacy of the NRC Codes to 8 analyze steam generator tube temperature distributions 9 during severe accidents.
10 (2) We'll do a reconciliation of our 11 past ACRS comments and recommendations.
12 (3) We'll have a report of the planning 13 and procedures subcommittee.
[,_ )
\/ 14 (4) Then we'll do some of our future 15 ACRS activities.
16 (5) After that will be our briefing on the 17 restart of Indian Point 3. l 18 (6) And then we'll work on our reports.
19 (7) If there's time we may do some l 20 strategic planning.
21 A portion of today's meeting may be closed to I l
22 discuss organizational and personnel matters that relate 23 solely to the internal personnel rules and practices of 24 this advisory committee, and matters, the release of which 25 would constitute a clearly unwarranted invasion 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
l 6
1 personal privacy.
i
,, 2 This meeting is being conducted in 3 accordance with the provisions of the Federal Advisory 4 Committee Act.
5 Dr. John T Larkins is the designated Federal 6 Official for the initial portion of the meeting.
7 We have received no written statements or 8 requests for time to make oral statement from members of 9 the public regarding today's sessions. A transcript of 10 portions of the meeting is being kept and it is requested 11 that the speakers use a microphone, identify themselves, 12 and speak with sufficient clarity and volume so that they 13 can be readily heard.
Q b 14 Before we get into the form- oart of the 15 agenda I do have a couple of items of current 16 interest. Some of these are listed for the members on 17 this little compendia you have before you. I'll just note 18 that a couple of those are flagged as special interest.
19 One, we do have a full complement of five 20 Commissioners now on board. You'll also note there's an 21 announcement of the retirement of William Russell, the 22 Director of ONRR. And there's a couple of very 23 interesting speeches made by Chairman Jackson regarding 24 the Millstone problem. So I would recommend you read O
tj 25 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 I'd also like to remind members that tomorrow g- 2 we're having a luncheon for outgoing retiring member,
\, I
'~'
3 Charlie Wylie. That will be here in the conference room 4 next door. And I'd also like to remind members that we 5 plan to take Charlie out to dinner somewhere tonight.
6 I'd also like to remind members that they're 7 scheduled today to do some interviewing for possible 8 replacements for Charlie. I think you have before you a 9 little schedule for this. Looks like this. Please note 10 the time that you're supposed to do this and react 11 accordingly, please. There's two of these we're 12 interviewing so be sure you note that.
13 MEMBER SHACK: Where is room 2E2?
O '
k_- 14 MR. DURAISWAMY: That's the conference room.
15 CHAIRMAN KRESS: You also have a set of 16 resumes for these people, 17 (Pause.)
18 CHAIRMAN KRESS: We hope today at least, to 19 get a first reading of our letter that we'll write on the 20 codes Adequacy and Mario, that I think is yours, so by the 21 end of the day you need to have some sort of a rough draft 22 first letter.
23 MEMBER CATTON: Good luck.
24 CHAIRMAN KRESS: There's some good O
( j 25 background in the little write-up 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
8 1 MEMBER FONTANA: Yes, I read that.
2 CHAIRMAN KRESS: -- Ivan put together. For 73 N. 3 the members' interest on this for the Thermal Hydraulic 4 Code Adequacy Hearing, there's some additional information 5 that you may or may not have seen in this pink -- two l l
6 reportings. It would be worth reading sometime today. l 7 Before we move on to the agenda, are there 8 other items that other members would like to bring up, l 9 before we start? I would note with pleasure that we have 10 Novak Zuber here with us. Any items anyone wishes to )
11 bring up before we start?
l 12 VICE CHAIRMAN SEALE: I would remind the 13 members that most of you have recently gotten quarterly
/ ,, \
-/ 14 reports from the two fellows on their activities. I'd 15 appreciate any feedback you might be able to give me on 1
1 16 those, and also if you have issues that you'd like for us 17 to have the fellows take a look at I'd like to hear from 18 those on you as well.
19 CHAIRMAN KRESS: Thank you. With that then, 20 I'll start with the first agenda item on the adequacy of 21 the codes. And Mario, that's yours, so I'll turn the 22 floor over to you.
23 MEMBER FONTANA: Okay. Before we proceed, I 24 want to emphasize that the Committee agrees with
- D 25 practically all of the NRC staff's efforts with respect to it ,/
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
9 1 resolution of severe accident issues, and we so stated in f-~g 2 a letter to Chairman Jackson of June 28th, 1996.
O 3 Now, as an aside I want to clarify one 4 statement in the letter that says, "There is no threat of 5 prompt containment failure proposed by direct containment 6 heating in the Westinghouse large dry containments." This 7 business, there's no such thing as no, what is meant is 8 that the risk from DCH for those containments is 9 sufficiently low, and additional investigation for that 10 particular issue was unwarranted.
11 The key issue with our letter however, was 12 that we had concerns regarding the capability of computer 13 codes for calculating steam generator tube heating, a
/9 2 14 potential rupture consequential or coincident with severe 15 accidents. Since this event potentially provides a direct 16 path from the core to the environment, it is a very 17 important issue, and it is so recognized by the NRC staff 18 and by the industry. And we appreciate that the staff has 19 agreed to return to present their case for ensuring the 20 adequacy of their codes for analyzing steam generator tube 21 heating during severe accidents.
22 As Tom mentioned there are two reports from a 23 review committee that the staff had, on August 18th and 24 19th, from Dr. Ishii and Dr. Griffith, and they're on this
^Q
(_,/ 25 pink -- and this too. Also as Tom said, we're pleased 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
10 1 have Dr. Zuber here participating in the discussions as a 2 consultant.
73
\_) At this point let's proceed and we'll start 3
4 with Mr. Wayne --
5 MEMBER CATTON: Before you start, can I make a 6 comment, Mario?
7 MEMBER FONTANA: Certainly.
8 MEMBER CATTON: I think the question that we 9 have before us is how to deal with uncertainty in a risk 10 evaluation. And that's really the bottom line. And the 11 first thing you need to do is to state the problem, and in 12 this case its probability of failure of a steam generator 13 tube during a severe accident.
A
' 14 We've had a lot of experience in how this can 15 be done and Novak was the leader in one of the groups with 16 a very complicated problem where this was done. It's not 17 uncommon for us to say that the probability of some 18 occurrence is less than 100, with some confidence. The 19 difference here is that in doing so, we're using a code to 20 try to marry phenomenological behavior to a risk 21 perspective that you might get from something like 1150, 22 and just how do you do that?
23 Rolling phenomenological behavior into a risk 24 perspective requires a sea change in the way you think.
?S Normally, we come at a problem as a designer, and when you NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERG 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
1 11 1 design something each step along the way you put in a l
,s 2 little just to be sure. When you get to the endpoint t i
3 you've got something on the table that you know will work, 1
4 everybody agrees with you, yes, it will work because 5 you've been there before. l l
6 In this case we have something in front of us ;
7 and the requirement is to make a statement about it. What l
8 do you do? I think you have to go back and test those 1 9 tools in a certain way. What we're going to hear about l
10 this morning is the adequacy of a code. Any code could do l 11 the job providing you tune it properly. And it's my 12 feeling, is that RELAP5, SCDAP, in whatever version they l 13 have on the table before us, has been so tuned.
s 14 CHAIRMAN KRESS: When you say it's been tuned, 15 has it been tuned strictly to a specific geometry? In 16 particular, the Westinghouse type --
17 MEMBER CATTON: Well, it's been tuned to the l 18 Westinghouse data.
l 19 CHAIRMAN KRESS: Well, are there other steam '
20 generator geometries out there that are different?
21 MEMBER CATTON: I think so but I don't know 22 how much difference they are. The point is that the --
23 you have countercurrent, single phase flow. Somehow you 24 have to fool the code into believing that it has. There
()
O 25 are complex mixing. There's a lot of things going on, and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 2000 5 3701 (202) 234-4433
12 1 you sort of know how to deal with them and you can put
, ,-s 2 them together. The matrix that you put them together in
! I 3 could be any. I mean, I could do it with --
4 CHAIRMAN KRESS: Or you could actually have a 5 lookup table if you wanted to.
6 MEMBER CATTON: Almost. Almost. It's a 7 little more complicated than that. And I would not 4
8 recommend that that be the approach, d
9 MR. ZUBER: May I make a comment on that?
10 MEMBER CATTON: Sure.
11 MR. ZUBER: You have based your comment, and 12 they are really correct, provided that facility was 13 properly scaled for the particular phenomena. If the p
(_ 14 facility was not well scaled the argument cannot hold.
15 The proof in the pudding is really, if you use that 16 approach is, how well is, how well is it scaled for that 17 phenomena.
18 MEMBER CATTON: And it turns out that in this 19 case, for the most part, it was well scaled. There's 20 still questions. In the core the fuel pins were -- the 21 simulated fuel pins were much larger, and also in the 22 steam generator the tubes were much larger. The flow 23 areas and friction factors, they matched. But the 24 question here is, what is most important and was it well l) 25 scaled or is it well understood, and when you look at the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
13 1 results, it's iffy.
2 I think that the scaling was good enough. If
.es's (G 3 you were a designer you would go in, you could use the 4 information and you would pick one side or the other in ,
l 5 order to be sure that everything was all right. What we l
l 6 want to do now is to take this information, take the code, .
7 and answer a question. And the question is, what is the 8 conditional probability of steam generator tube rupture?
9 What is it, given the information that you have, and how 10 do you get from there to here?
11 I don't believe that question is going to be 12 addressed this morning. But it's a separate question. I 13 think what you're going to hear about this morning is the O
'd 14 adequacy of the code to do the job. I think our letter 15 was overly strong in that regard. I don't think there was 16 ever a question in my mind as to whether or not you could 17 use the code. Yca certainly could. There are a lot of 18 questions about whether or not it was the appropriate tool l
19 to use.
I 20 Anyway, with that --
21 CHAIRMAN KRESS: Were you aware of these 22 scaled tests and the tuning?
23 MEMBER CATTON: Yes.
24 CHAIRMAN KRESS: Aware of the --
n 25 MEMBER CATTON: Well, you know, and I've NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTCN. D C. 20005 3701 (202) 234-4433
14 1 looked at the tests pretty closely and the mixing is one I
'm 2 of the key issues. And the test data shows that this V) 3 particular parameter that they have to represent it is .7.
l 4 It's not unreasonable to expect plus-or-minus 10 percent l l
l 5 in the experimental results. There are some reasons to !
1 6 believe that the mixing may be a little bit less. There 1 7 are arguments on the table that -- or that you could put 8 on the table as to why the mixing is even better.
l l
9 CHAIRMAN KRESS: That's the mixing in the l 10 inlet plenum?
11 MEMBER CATTON: In the inlet plenum, because i 12 that's the key. There are a lot of other questions, too.
13 The heat transfer in the hot leg, for example. It's
. f3 5
14 countercurrent, single phase flow in a thick-walled pipe.
15 That's a complicated problem. There's been some work done 16 on mixed convection in horizontal pipes but nobody's ever, 17 that I know of, has looked at the countercurrent problem.
18 What does this do to the heat transfer? Maybe 19 it increases it. Maybe it doesn't. What do you do about 20 that when you have to make a probauilistic statement about 21 survival of the steam generator tube? When you look at 22 the steam generator tubes you have a number of questions 23 that come up.
24 on the inside of the tube, well, it's mixed
/~'
25 convection, we know how to treat it. You can get NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE (SLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234 4433
15 Jackson's correlations in there pretty damn good even at l 1
2 the erandtl numbers of a gas. If you look at the other 3 side -- because the inside of the tube really, is not all 1
4 that important. If you look at the other side of the tube l 1
5 you have an empty steam generator that's filled with 6 steam. It's a huge structure and it's natural convection.
7 What do we know about natural convection in a 8 tube bundle? Well, not a hell of a lot. You can make 9 estimates, you can guess. Probably plus-or-minus 20 or 30 10 percent. Now, what do you do with the 20 or 30 percent?
11 Somehow you have to roll this uncertainty into the 12 statement you make when you're done.
13 CHAIRMAN KRESS: Why not treat that as an O(3 14 adiabatic wall inside?
15 MEMBER CATTON: If you do that, that's fine.
16 CHAIRMAN KRESS: Then you're going to fill the 17 steam generator tubes up.
18 MEMBER CATTON: See, you could go all along 19 the line. You could say, okay I want a conservative 20 result. I'm going to make the heat transfer in the hot 21 leg zero. Well, now you know the tubes will fill first.
22 So you know that's not the right answer.
23 CHAIRMAN KRESS: So in this case, you're 24 really looking for something realistic. Best estimate?
C\
Q 25 MEMBER CATTON: That's right. And you need to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005 3701 (202) 234-4433
16 1 marry these distributions in some way that makes sense,
,_s 2 conserve momentum, conserve energy, and so forth. The b)
/
3 code is going to do that for you. My concern is how do l 4 you treat the next step? And you know, we're heading for 5 risk-informed regulation. At some point we've got to 6 learn to deal with these things.
7 How do you do it? I don't think you take 8 huge, complex code off the shelf and run with it. I think 9 you have to think a little bit about the problem, and in 10 particular in this case, the problem's relatively simple.
! 11 But I don't want to argue about the adequacy of the code.
12 Any code can be made adequate by proper tuning to the l 13 data.
T(s) 14 Then you have to ask yourself, what do I do to I
15 be able to make a statement with confidence? You need to <
16 somehow take all of the uncertainties associated with all 17 of the steps along the way and roll them into that answer.
18 And I think that a prescription for doing this is 19 evolving, and I think the first step along the way was 20 when this kind of a statement was required for the large-21 break LOCA in the best estimate sense.
22 And that was a very complicated problem and 23 there were a lot of unknowns in it. In this particular
, 24 case it's like three or four heat transfer coefficients m
) 25 and a mixing coefficient, and we know where they all 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
17 1 plus-or-minus some number. It's just now, what do we do f-s 2 to treat it?
3 MEMBER FONTANA: Thank you, Ivan.
4 VICE CHAIRMAN SEALE: Can I ask Ivan a l
5 question?
6 CHAIRMAN KRESS: Sure.
7 VICE CHAIRMAN SEALE: It's relevant I think, 8 to the immediate issue. You made the comment that you 9 could make any code --
10 MEMBER CATTON: Almost.
11 VICE CHAIRMAN SEALE: Well, okay, yes. But 1
12 now, do you have the impression that the codes in question 13 and the data that have been used to make these
(%,
k/# 14 reconciliations, if you will, are robust enough so that a 15 tuning of the code in resolving issues having to do with 16 the operating environment is adequate to assure the 17 appropriate code response when you get into the accident 18 environment?
19 MEMBER CATTON: You know, I feel like Clinton 20 in front of the Press. That question was so long I can't 21 remember the beginning of it, but let me try anyway. Yes, 22 I think you could use the code as it is in its present 23 tuned state to address this problem.
24 VICE CHAIRMAN SEALE: Okay.
O
(_j) 25 MEMBER FONTANA: Well, at this point, let'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 l
18 1 get on with the NRC chance to make their presentation.
2 And as you said Ivan, there's two steps here. One is the (g 1 V 3 adequacy of the codes and the other is, what do you do 4 after you have that.
5 MEMBER CATTON: That's right.
6 MEMBER FONTANA: And '. think what. we're going 7 to focus on today is the first part, is the adequcy of 8 the codes. And at the end of that we'll get into the next 9 discussion of what should you do next. So at this point 10 I'd like to turn it over to Mr Wayne Hodges of the NRC.
11 MR. HODGES: Yes, I'd l'.ke to make a few 12 comments before we start. Charlie Tinkler will be making 13 the presentations. And particularly in light of some of
,rh
! /
k/ 14 the comments that Ivan just made, I think it would 15 appropriate after the presentation, that we would like a 16 letter from the ACRS clarifying the statement that was 17 made in the June 28th letter. This caused us a fair 18 amount of grief. We think it's incorrect and we'll show 19 you today why we think it's incorrect, and we would like a 20 letter correcting that.
21 As far as some of the other questions, it may 22 well be that the largest uncertainty is not in the actual 23 thermal hydraulic analysis. It may well come from other 24 parts of the problem. And it's also true that we could A
25 have probably have developed a simpler code, as Ivan says, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D,C. 20005-3701 (202) 234-4433
19 1 to do this, but then we would have had to come in front of 2 you people and defend it and that would take another year 7
i 3 or two to the two we had.
4 MEMBER CATTON: Actually, such a simple code 5 was developed for you. And it was part of the accident 6 management work that was done at UCLA.
7 CHAIRMAN KRESS: Do you have a conflict of 8 interest on this?
9 MEMBER CATTON: No, because I couldn't 10 participate. I just got to watch. Charlie, no comments.
11 MR. TINKLER: Active watching. Good morning.
12 I'm Charlie Tinkler from the Accident Evaluation Branch in 13 the Office of Research. And this morning I'm going to
\ talk about our analysis of steam generator tube thermal 14 15 hydraulic conditions during severe accidents.
16 During the course of my presentation I'm going 17 to go through an introduction of the issue, a brief 18 discussion of the background, a summary of our views, and i
19 I'll then go into a little more detailed discussion on the !
20 specifics SCDAP/RELAP5 modeling that was developed a l 21 number of years ago to simulate countercurrent natural 22 circulation.
23 I'll then discuss the consideration of fission 24 product deposition and its incremental contribution to the O 25
(/ heating of the steam generator tubes since that issue has NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
20 1 also been raised. f 1
g- 2 I will then go back and discuss the original k/
3 benchmarking of the code, the scaling issues that were 4 considered and the development of the experimental data.
5 I'll then present for you our preliminary results that are 6 going into the overall assessment of steam generator tube 7 performance. I'm just going to show you some, of the 8 sensitivity studies that we've done to address some of the 9 phenomenological issues that have been raised.
10 Based on our understanding of questions 11 concerning the use of the code, we most recently undertook 12 another independent review of the use of SCDAP/RELAPS for 13 calculating these conditions and we'll report to you the O
k/ And then finally, I'll go over 14 results of that review.
15 again our conclusions on this issue.
16 As you all know, as part of the rulemaking on 17 steam generating tube integrity a major part of the 18 evaluation has been given to the -- has been directed i 19 towards the consideration of the performance of flawed 20 steam generator tubes and their likelihood of failure 21 during a severe accident, when the reactor cooling system l
22 remains at high pressure and the secondary system is l l
l 23 depressurized. That generally constitutes the most severe 24 challenge to the tubes.
O)
(_ 25 Concern arises because the core heats up, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHOOE ISLAND AVE., N W. I (202) 234 4433 WASHINGTON. D C. 20005 3701 (202) 234-4433 l
l
1 21 1 generates hot gases, and these gases circulate around the 1
I 2 loop. And the circulation of these hot gases can create
/,_T (m'/ 3 conditions which would promote the induced failure of a 4 steam generator tube due to the elevation of its 5 temperature.
6 You've heard presentations in the past about 7 creep rupture models, and what I will show you today are 8 again, calculations that address creep rupture failure of 9 the steam generator tubes using a Larson-Miller parameter 10 approach.
11 Fission product releases from the core may 12 also transit through the system and may end up being 13 deposited on the steam generator tubes. And these fission
(/
\_ 14 products have heat associated with them, either 15 radioactive or non-radioactive aerosols, and they can 16 contribute to the heating of the tubes. So we will 17 address that as part of this overall assessment.
18 I want to point out that this is really a 19 coordinated effort between the Office of Research and NRR.
I l
20 NRR has the lead for the overall rulemaking activity, but l l
21 we are coordinating to produce a set of thermal hydraulic 22 boundary conditions for use in evaluating the structural 23 integrity of the steam generator tubes.
24 Further background, the SCDAPS/RELAP5 code as
/"'
( ,)N 25 our second tier or mechanistic serax and a thermal NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
22 1 hydraulic code was selected early on. --
_ 2 This goes in the benchmark against the Westinghouse 170
\' ,] 3 scale test data and we'll talk about that a little more. -
4 -
The code was also benchmarked against the early 5 application of the COMMIX 3D, finite difference field code 6 for full-scale calculations and the one-seventh scale test 7 experiments. We think we have considerable experience in 8 using this code, albeit tuned. I don't consider that in a 9 pejorative sense because it's been used most recently to 10 assess unintentional depressurization of the reactor 11 cooling system during the issue resolution associated with 12 direct containment heating.
13 As I described before the committee some time
- 14 ago, we used SCDAP/RELAP5 to assess the probability that 15 the RCS would depressurize. Now, as I recall some 16 committee members took comfort that we didn't really need 17 that in that assessment, but we nonetheless used it for 18 that, and it was considered in part of the DCH issue 19 resolution peer review.
20 Since SCDAP/RELAP5 does not in model 21 efficient, probably transport in deposition, we used the 22 VICTORIA code to assess fission product heating of the 23 tubes. Now, SCDAP/RELAP5 has a release model from the !
24 fuel. It further decreases decay heating in the core due y ) 25 to the release efficient products, but it doesn't have a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(?O2) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
23 1 model which calculates the transport and ultimate !
,-s 2 deposition of those products. So we used VICTORIA for N 3 that purpose.
4 And I want to point out that SCDAP/RELAP5 5 analysis are actually used to assess just the thermal-6 induced failure of the steam generator tubes. Spontaneous 7 steam generator tube rupture and pressure-induced steam 8 generator tube rupture are separate elements of the 9 rulemaking analysis.
i 10 I want to remind the committee for perspective 11 that the thermal hydraulic analysis is really one part of 12 an overall assessment associated with the steam generator 13 tube performance. There are the issues associated with
- (~%
kY 14 thermal hydraulic modeling which you're going to hear 15 about today, but there are also assessments associated 16 with the event tree quantification, the consideration of 17 other weak points in the reactor coolant pressure 18 boundary, the consideration of what constitutes a 19 representative flaw distribution in various plants.
20 There's also the tube performance model, the 21 Larson-Miller model itself has uncertainties associated 22 with it, which are being considered. There's also the 23 severe accident radiological consequence assessment. I'm 24 not going to talk about that today but you'll hear about im
\ ,)
25 that as part of the overall rulemaking briefing that, I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINCTON. D C. 20005 3701 (202) 234 4433
24 1 guess the committee will hear sometime in the next month 7s 2 or so.
t \
O 3 But we have done calculations using MFLCORE 4 and the MAXX code to calculate the off-site dose 5 consequences. And that's all factored into this 6 assessment. And finally, there is the consolidation of 7 all these elements.
8 This is a copy of a slight variation on a 9 viewgraph that Mark Kenton showed you, I guess a month or 10 so ago. And this is to give you a sense of what's 11 involved here. This is a containment event tree for the 12 consideration of tube performance. And it seems pretty 13 clear that a number of things have to go wrong here before t s
\- / 1. 4 you really seriously start worrying about tube i
15 performance.
16 Not only do you have the severe accident but 17 other things have to work against you. The pressurizer 18 safeties have to main -- you have to maintain pressure in 19 the RCS. The most serious challenge is associated when 20 secondary relief valves fail open and depressurize the 1
21 steam generator. It increases the delta P across the 22 tubes and also raises tube temperature because of 23 decreased heat capacity of the steam generator when it's 24 depressurized.
A
() _
25 Furthermore, the tubes can't have already 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 failed due to just the delta P, because we're now just
- 2 concerned about temperature-induced failure. Furthermore, I
'/
3 then we're concerned about split fraction. So all the J l
l 4 analysis that you're hearing about are analyses that go 5 into determining the thermal hydraulic boundary conditions l
6 associated with establishing the split fraction for this !
7 path. So I point it out just for perspective when we 8 start talking about the necessary precision of these 9 calculations.
10 In summary, we've use SCDAP/RELAP5 for 11 representative plants. We've done calculations for Surrey 12 and ANO Two. ANO Two CE plant without a PORV. We thought 13 it would be a good candidate to pick since, you know, for O
k/# 14 probabilistic reasons it has less capability for 15 depressurizing the primary system.
16 And basically we've looked at station blackout 17 sequences, what we call short-term station blackout la sequences, where there's an immediate loss of aux.
19 feedwater, so the steam generators dry out quickly. The 20 SCDAP/RELAP5 analysis predict failure of the hot leg or 21 surge line before the steam generator tubes, for non-22 flawed tubes, typically in the timeframe of 20 to 40 23 minutes depending on the plant.
24 Sensitivities on the thermal hydraulic
) 25 modeling did not substantially alter our conclusion. We NEAL R. GROSS COURT REPORTERS AND TRANSCRtBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASH'NGTON. O C. 20005-3701 (202) 234-4433
26 1 changed the interval between tube failure and hot leg s
2 failure by a few minutes, but we don't seriously alter the t )
~#
3 fundamental conclusion.
4 MEMBER SHACK: Is that because everything just 5 shifts? I mean, if the surge line temperature goes up 6 then the steam generator temperature -- they sort of move 7 together as you change the thermal hydraulic temperatures.
8 MR. TINKLER: The tendency is for them to move 9 together. Now, you can squeeze the interval by changing 10 things like mixing fractions and other parts of the 11 calculation, and I'll talk about that. But if you just 12 impose a change in system modeling, it generally has the 13 effect of shifting the entire transient for both the surge in\
kl 14 line and the tubes.
15 As an example, we've looked at 16 depressurization of one loop, one atmospheric dump valve 17 sticks open on a steam generator. There was concern, what 18 happens if you fail more than one DV on more than one 19 steam generator? What it actually did was lower the 20 temperature of the tubes at the time of surge line failure 21 because you diminish the heat removal capacity on the 22 secondary side so the PORV cycled more times. And it 23 cycled more times and heated up the surge line faster. j I
24 So there are system feedbacks and compensating )
(Oj 25 feedbacks that tend to wash out individual sensitivities. l 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 The analysis using VICTORIA to calculate 2 fission product transport in deposition confirmed our 7l
\
\~' 3 first assessment that the fission product deposition is a 4 minimal effect on steam generator tube heating. And we've 5 done a good many analysis using VICTORIA, looked at a 6 number of sensitivities, but the overall conclusion is, 7 it's quite clear that fission product deposition is a 8 pretty small contribution. At least based on our 9 calculations, and you can take a look at those.
10 Turning to the development of the SCDAP/RELAPS 11 model. The background on this topic is that the integral 12 experiments conducted by Westinghouse at one-seventh scale 13 really represent that fundamental basis for describing the
/~N
- 14 phenomenon. These tests represented the reactor coolant 15 system and they did confirm the existence of natural 16 circulation flows within a PWR.
17 It had actually been postulated that 18 countercurrent natural circulation flow would exist during 19 a station blackout. And that actually had been 20 incorporated in a number of calculations, including 21 industry calculations. These tests were the first 22 rigorous confirmation of that phenomenon.
23 And basically what you see is, this 24 countercurrent natural circulation redistributes decay y ,
) 25 heat from the core to the hot leg, surge line, steam NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
l l
1 28 1
L
(
1 generators. The good news is the delay cores heat up,
- 2 melting increases the probability of an RCS failure and I 3 depressurization of the reactor coolant system.
, 4 So if you were worried about DCH this is one 5 way by which you can lessen the probability of a high-6 pressure melt ejection. But it also increases the 7 probability of a steam generator to rupture because these f
8 hot gases circulate through thin-walled tubes, 50 mil 9 tubes.
10 And basically what the test showed was that 11 you established three loops of natural circulation, one 12 within the RBP, get upflow through the center of the core, 13 the hottest part of the core, return flow down the bypass
()
\ 14 paths and the outer fuel assembly paths. You alsc-get a 15 countercurrent natural circulation flow path in the hot 16 leg, transferring heat from the vessel to the inlet plenum 17 of the steam generator.
18 And what they further observed was a 19 countercurrent natural circulation flow in the U tube 20 steam generators. Some fraction of tubes carry hot flow 21 down to the outlet plenum -- actually they will carry it 22 down a little bit into the pump suction leg, then cold 23 flow returns through the remaining fraction of tubes.
24 MR. ZUBER: So, just a question then. You b
> (_) 25 concur that this is an important process, the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISuND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
29 1 countercurrent flow and the mixing in the lower plenum.
2 MR. TINKLER: The inlet plenum.
U 3 MR. ZUBER: The inlet plenum, yes.
4 THE WITNESS: Yes sir.
5 MEMBER CATTON: Charlie, how much variation 6 can you get in the vessel upper head due to uncertainties 7 in the SCDAP calculations of core behavior?
8 MR. TINKLER: Well, all these calculations 9 predict hot leg or surge line failure early in the --
10 MEMBER CATTON: No, that's not the question.
11 I'm just interested in how much variation can you get in 12 the upper plenum of the vessel where it's well mixed, as a 13 result of uncertainties in the SCDAP part of the
( .
14 calculation? I mean, is it 15 plus-or-minus 50 degrees, plus-or-ml;1us 100 degrees, or 16 what?
17 MR. TINKLER: The variations in the upper 18 plenum --
19 MEMBER CATTON: Yes, because that's your 20 boundary conditions.
21 MR. TINKLER: Right. The sensitivity is 22 generally considered axial power profiles because there is 23 no relocation of material during these transients.
24 MEMBER CATTON: I understand. I'm just
(~)
(_,/ 25 interested in what the range is, not the reason.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
l 30 l 1 MR. TINKLER: I don't recall on an absolute p 2 scale, what the temperature variation is. Feedback on the t
v/ 3 timing of surge line failure was small. A minute or two, 4 in the sensitivity studies that were done when we first 5 developed this model.
6 MEMBER CATTON: So somewhere on a shelf 7 somewhere, somebody has this information?
8 MR. TINKLER: There were calculations to 9 address timing of surge line hot leg failure based on 10 variations of --
11 MEMBER CATTON: Of the core model?
12 MR. TINKLER: Of flow loss coefficients, axial 13 power profiles -- this actually is addressed in a later
"( V) 14 viewgraph. I'm trying to recall from -- assuming on how 15 well I've got it memorized. We varied loss coefficients 16 for the flow paths up through the core, we varied the 17 axial power profile. I think we varied one other 18 parameter, but basically we saw a change in the surge line 19 failure, time of a minute or two, for those kinds of 20 calculations.
21 Now, much has been said, I guess, in these 22 discussions about inlet plenum mixing. What the tests 23 show was that there was significant mixing in the inlet 24 plenum. And when we talk about things like mixing
/T k 25 fractions, recirculation ratios, mixi.lg fractions is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS i 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 ;
31 1 basically the fraction of flow which sees the mixing of
,,_s 2 this hot stream and this cold stream returning from the e s 3 cold tubes.
4 Recirculation ratio refers to this flow rate 5 in the tubo bundle relative to the flow rate in the hot 6 leg.
7 MEMBER CATTON: And this, by the way, is the 8 single most important process that's going on here with 9 respect to the impact on the tubes. And I just might 10 mention some of the data. I looked at the data in the 11 Westinghouse report -- can you leave that for a minute?
12 MR. TINKLER: Sure.
33 MEMBER CATTON: And if you look right up about p
\/ 14 where that three is, and nobody seemed to know -- when you 15 look at the temperatures which were measured, what you see 16 is that, for the transient tests that were run is that the 17 thermocouple that was inside the tube -- at least that's 18 what it looked like from the data and nobody seemed to 19 know that it wasn't -- was higher than the temperature 20 below.
21 And what that tells me is, do you see the 22 arrow -- if you track the arrow along the hot pipe and up 23 into the steam generator tube, that at least on some 24 occasions there was much less mixing than there were on
("x
. ( ,) 25 others. And to me that was really a strange result. 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
32 1 was unexplained. Why was the temperature just inside the 2 tube higher than the temperature just below the core O
\
'J 3 support plate -- or, the tube sheet.
4 MEMBER FONTANA: I have a question that tends 5 to relate to that, and that is, if you have a temperature 6 distribution at the inlet of the tube -- and if for some 7 reason a very small number of tubes see a higher 8 temperature, it be through bad mixing or whatever, and 9 those tubes will have a lower density to start with and 10 then -- low density and should have a higher head, and why 11 will more and more flow goes into those tubes till the 12 flow gets high enough for the friction to stop it?
13 Why, if you have asymmetry of the temperature l O ,
14 to flow entering a tube, few of them get a higher 15 temperature than the others, why don't those few gather 16 most of the flow and get a lot hotter, faster than the 17 other one? Have you modeled that? In other words, have l
18 you modeled a split to start with, on the number of tubes ;
19 which have outer flow as compared to the hotter flow? l l
20 MR. TINKLER: We have performed sensitivity 21 calculations to look at the impact of varying the 22 fractional tubes carrying hot flow versus cold flow. And l
23 typically what you see is, when you increase the number of 24 tubes carrying hot flow, you distribute the same amount of (n 25 flow to a greater number of tubes and the tube 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 temperatures go down. It's important to remember that
-- 2 this is --
V 3 MEMBER FONTANA: Is this a transient 4 calculation where you get the subsequent effects of an 5 initial boundary condition which could drive it on through 6 a smaller number of tubes getting higher --
7 MR. TINKLER: Well, the calculation would 8 allow more flow to go through those tubes should it want 9 to, okay. That's in there.
10 MEMBER FONTANA: And you varied the number of 11 tubes that get hot flow to start with?
12 MR. TINKLER: And we varied the number of 13 tubes to get hot flow to start with. Now, SCDAP/RELAPS is r'N iN ') 14 used to calculate average tube temperatures, okay. That's 15 understood. Unen we do the evaluation of tube performance 16 we consider peak to average variations. Okay. Local 17 effects, if you will, within the plume that enters the i
18 tube sheet.
19 But that's something, you know, you're talking 20 about 3,000 tubes. There's no code in the world that 21 could tell you with any accuracy, what one tube gets 22 relative to another. So you do an assessment to estimate l l
23 the variation within the plume and then consider that as )
l 24 part of your tube performance calculation. i
(~T i
( ,)
25 MEMBER FONTANA: Yes, the question relates to ;
NEAL R. CROSS COURT REPORTERS AND i.'ANSCA!3ERS 1323 RHODE ISLAND AVE., N W. l WASHINGTON. D.C. 20005-3701 (232) 234-4433 ;
(202) 234-4433 I
34 1 a positive feedback effect. If you have poor mixing at f3 2 the inlet and you get a few tubes seeing hotter
( i e/ 3 temperatures, does that tend to kind of run away? That's 4 the question --
5 MR. TINKLER: Well, I mean, you know, it 6 doesn't have a large, a large influence.
7 MEMBER FONTANA: That's what I want to know.
8 MR. TINKLER: But I mean, you know, these 9 codes use a flow area and they use hydraulic, you know, 10 parameters to describe the flow path. If delta P will 11 allow, or the buoyancy effects will allow more flow to go 12 into that tube it will allow more flow to go in.
13 MEMBER FONTANA: Yes, there's no doubt that IA)
\' 14 the codes could calculate it. I was just wondering the 15 effect of tue inlet kind of mixing on that effect.
16 MR. TINKLER: On that?
17 MEMBER FONTANA: But if you vary --
18 MR. TINKLER: We varied --
19 MEMBER FONTANA: -- the temperature 20 distribution across the inlet phase I think you've covered 21 my question. Which I think you did.
22 MR. TINKLER: I also want to point out, in 23 response to Dr. Catton's comments, it's important to 24 remember that when you look at temperatures reported that
(~h
(_,) 25 there's a -- they use a rotating rake in this lower plenum NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
I 35 1 to measure temperatures at various locations.
g~ 2 So we have tried, we've tracked down as many b) 3 people as we could to find out the discrepancy that 4 appears to exist in that figure where it looks as if the 5 temperature actually increases slightly from here to here.
6 It's not a big increase but it is, nonetheless, an 7 increase, and it's counter-intuitive.
8 MEMBER CATTON: No, it's not counter-9 intuitive. There are lots of examples in this kind of 10 mixing where you get penetration. One that's really, I 11 find intriguing is if you take two cylinders, a cold and a 12 hot, you put the cold cylinder in the top, hot cylinder on 13 the bottom and you get a boundary layer, and the boundary i t s l
'- # 14 layers literally cross. The rising fluid goes right 15 through the descending cold fluid with very little mixing.
16 Thermal siphon.
17 MEMBER FONTANA: It's fingers.
18 MEMBER CATTON: Fingers. It fingers right 19 through. So you get, the fluid comes down the cooling l 20 surface, goes to the center of the tube below it, down to 21 the bottom, comes up the wall, crosses, and they literally 22 finger. !
l 23 So the thing that, when I was looking at the 24 data, was to look for that. Well, there's an indication O)
\_ 25 of the possibility of fingering. Now, that may be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. O C. 20005-3701 (202) 234-4433
36 1 nonsense, it may well be these rakes. The point is that 2 there is some uncertainty.
/-'T
\ ') Well, one of the points I want 3 MR. TINKLER:
4 to make is when you use these rakes, okay, they typically 5 try to have one thermocouple not sitting directly over top 6 another because they don't want the flow stream to 7 interrupt the measurement on the thermocouple, so they're 8 etaggered somewhat. But you can't -- that's not depicted 9 well in some of the figures.
10 MR. ZUBER: I have several questions. Let me 11 ask you first, are you going to discuss how this is being 12 modeled in the code? This process of countercurrent and 13 mixing, how do you model it on the code?
O
( I
\_/ 14 MR. TINKLER: Yes.
15 MR. ZUBER: And did --
16 MR. TINKLER: I'll describe -- well, yes.
17 There are other viewgraphs that talk about how we do that 18 and I'll go back to this figure if it's not clear when I 19 get to that.
20 MEMBER POWERS: Can I ask just a couple of 21 questicns? It may not be terribly important but it's out 22 of curiosity. In that hot leg where you've got the 23 countercurrent flow, what do the aerosols do?
24 MR. TINKLER: I'm going to talk about it in a
(^)N
( 25 lot more detail coming up, but basically, you know, we 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
37 l
1 a couple of different mechanisms. We see vapor l l
1
-- 2 condensation, revolatilization, turbulent deposition, 3 gravitational settling, and it moves on down the path.
4 MEMBER POWERS: It doesn't -- there's not 5 enough thermophoretic driving force --
6 MR. TINKLER: No.
7 MEMBER POWERS: -- to move aerosols from the 8 hot stream into the cold stream?
9 MR. TINKLER: Well, thermophoresis was not a 10 big factor in our calculations, and we do have a pretty 11 big delta T between the hot fluid in the hot leg and the 12 walls of the hot leg. That's a pretty big delta T but 13 it's not enough to cause thermophoresis to dominate over 5
\ 14 the other mechanisms. That's different than the 15 temperature difference between the streams. But just to 16 give you a sense of --
17 CHAIRMAN KRESS: Are those flows turbulent in 18 there?
19 MR. TINKLER: Yes. Reynolds numbers are 20 typically 10 5 , 10' throughout the system.
21 MEMBER CATTON: Dana, are you interested in 22 the aerosols essentially falling out of the one steam into 23 the other?
24 CHAIRMAN KRESS: Yes. That would be one --
25 MEMBER POWERS: I just have no idea. I mean, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASH NGToN, D C. 20005-3701 (202) 234-4433 l l
38 1 I know about thermophoretic deposition on walls, but I 7~s 2 don't know about thermophoretic movement of aerosols from I $
\_) 3 one stream to another. And I've always been curious. I 4 have no --
5 CHAIRMAN KRESS: My feeling is, Dana, if you 6 actually had that situation and then it was turbulent, 7 that the aerosols, if you look at a little control volume 8 in there, would think the net flow was basically zero in a 9 stirred vessel, and would behave as if the net flow were 10 zero, so there wouldn't be a tendency to convect them.
11 MEMBER POWERS: Why woul'n't they sense the 12 fact that there was a temperature difference?
13 CHAIRMAN KRESS: They would. They would see a
\/ 14 temperature difference. You would have to input that.
15 But I don't think -- I think the turbulence would probably l
l 16 overwhelm the thermophoretic between the streams.
1 17 MEMBER POWERS: It would be unusual for 18 turbulence to do that, by the way.
19 MEMBER CATTON: But it's only turbulent when 20 you get away from the interface.
1 21 MEMBER POWERS: You see turbulence -- things ]
1 22 as a function of size, whereas thermophoresis --
23 CHAIRMAN KRESS: I'm suspecting the 24 temperature gradient between those two streams is not real
/~'
(,T) 25 sharp, quite frankly.
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 MEMBER POWERS: That may be --
2 CHAIRMAN KRESS: That's part of tny thinking on
(.) 3 it.
4 MEMBER POWERS: Okay.
5 MEMBER CATTON: The temperature gradient would 6 be quite steep at the interface.
7 CHAIRMAN KRESS: At the interface?
8 MEMBER CATTON: It's not going to be turbulent 9 at the interface, because this is stably stratified, and 10 the stronger the temperature gradient the more you cause 11 the turbulence to decay. People who live in Los Angeles 12 know all about this.
13 CHAIRMAN KRESS: At any rate, the effect would
'v) t 14 be to keep aerosols away from the steam generator.
15 MR. TINKLER: Right. It's going to keep them 16 up in there, right. ;
1 l
17 CHAIRMAN KRESS: Yes. ;
18 MR. TINKLER: They already said it's not a 19 major issue.
20 MEMBER POWERS: The other question I have is, 21 in that plenum region where you've got that, maybe when 22 you get to r. hose calculations you'll tell me how you 23 modeled that.
24 MR. TINKLER: Okay. The basic approach for
()
t' %
25 using a 1D code like SCDAP/RELAPS for this calculation, 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
40 1 to split the hot leg into top and bottom halves to the 73 2 amount of countercurrent flow. The obvious limitations of
\ l u
3 1D code, you couldn't put two opposing fluid streams in 4 the same control volume.
5 CHAIRMAN KRESS: How did you deal with the 6 friction between the two when you did that?
7 MR. TINKLER: Well, part of the benchmarking a tuning if you will, of the code, involves adjusting loss 9 coefficients for those flow paths to produce, you know, 10 the energy balance around the loop. Now, it turns out 11 that we have to increase the loss coefficients a little 12 bit above what you would normally expect on the order of 13 two, and that being the hot leg.
p j 14 It's not quite as bad as is often portrayed, i 15 but -- and you can think of that as actually a momentum 16 exchange between the two fluid streams. So in the sense, 17 the fact that we have to increase the loss coefficients 18 makes sense. But, you know, this was clear from the --
19 visualization experiments where they used a dye in the 20 water to clearly observe these two fluid streams.
21 CHAIRMAN KRESS: Did you treat the two halves 22 as if that was a solid boundary across the middle --
23 MR. TINKLER: Yes.
24 CHAIRMAN KRESS: -- in terms of determining n
/ \
() 25 the hydraulic diameter and the friction coefficients 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
i 41 j l
1 so forth?
l 7, 2 MR. TINKLER: Yes.
\ !
3 CHAIRMAN KRESS: And tune that to get that two 4 factor too --
5 MR. TINKLER: Yes.
6 MR. ZUBER: Then you just divide it in half 7 and you always get half hot fluid and the other half cold 8 fluid?
9 MR. TINKLER: Yes. It's about a 50-50 split 10 in the hot leg. Now, the tests show an incline of the 11 interface. You know, they don't show a perfectly flat 12 interface all along. But you know, the average over the 13 length of it is about 50-50.
O
-- 14 MR. ZUBER: If you're just drawing a line they 15 would say yes, but it's just one half. Doesn't mean that 16 this is really correct. But anyway, this also assumes 17 that the experiment's well scaled for this phenomenon in 18 order to draw this conclusion or use this approach?
19 MR. TINKLER: Right, right. The fundamental 20 conclusion is that that kind of distribution of flow 21 within the hot leg will result in the right amount of heat 22 being --
23 MR. ZUBER: Well, you missed my question.
24 Then you divide this and you use this approach based on (n,,,I 25 these experiments?
NEAL R. GROSS COURT REPORTERS AND TRANSCR:BERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
42 1 MR. TINKuER: Right.
. 2 MR. ZUBER: The basic assumption is that these b- 3 experiments are well scaled for this phenomenon, otherwise 4 you would not be able to use this approach in the code?
5 MR. TINKLER: Well, in actual fact, okay, the 6 calculations --
7 MR. ZUBER: Forget the calculations.
8 MEMBER CATTON: The answer is "yes", Charlie.
9 MR. TINKLER: Well, the answer is yes but I 10 want to qualify it a little bit, because the first 11 calculations that were done, okay, were an extrapolation 12 to full scale, and then using the same approach, confirmed 13 against the one-seventh scale test data. It wasn't that m.,
(V 14 we took the SCDAP/RELAP5 calculations and then tried to 15 tune them to the one-seventh scale.
'6 Actually, the SCDAP/RELAP5 model was based on 17 the earlier COMMIX calculations that were performed 18 against full scale and validated against the test, and 19 then the SCDAP/RELAP5 analysis -- I'll describe this --
20 was used, the same approach was used to go back and ;
i 4
21 calculate the one-seventa scale test results.
4 22 MEMBER CATTON: The only part where the scale 23 --
24 MR. TINKLER: It's kind of circular, so --
(3) 25 MR. ZUBER: Right. It's very circular.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3.'01 (202) 234 4433
43 1 MEMBER CATTON: The only place where the 3 2 scaling, in my view, was a little bit shaky, was the size
/ t V 3 of the steam generator tubes and also the size of the 4 simulated fuel pins.
5 MR. ZUBER: Okay, well --
6 MEMBER CATTON: And the reason that the steam 7 generator tubes are important -- it really would have been 8 a whole bunch of fine tubes, and they used a lesser number l l
9 of larger tubes, and this is an interface between the )
10 plenum and the steam generator tubes, and that was not 11 scaled, other than for friction factor and flow area.
12 MR. ZUBER: See, going back to one of these 13 viewgraphs you had, that mixing in the inlet plenum really
(
\ 14 depends on the velocity amount of fluid which comes out of 15 the hot leg. So that flow, that flow, that process of 16 flow through the hot leg, countercurrent, affects the i 17 mixing which affects then the flow through the tubes and l 1
18 temperature distribution in the tubes. So it is l 19 interconnected. So if one is well scaled and the other 20 one is also well scaled you have no problem. If one is 21 not well scaled then the question is, what is the effect 22 and what are the uncertainties?
23 Now, there is, and I'm putting in this l 24 question now, I was reading yesterday the handouts from t
in
(_)> 25 Mr. Dudley, and there was a handout, attachment 5 which !
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 was nresented at the Fauske meeting, this one here. And I 7y 2 looked at the way this was scaled and the countercurrent V 3 flow -- are you familiar with the scaling?
4 MR. TINKLER: I'm familiar with that.
5 MR. ZUBER: Okay. If I refer, this WHL has a 6 coefficient CFS. What is that?
8 MR. ZUBER: Yes, this?
9 MR. TINKLER: Well, okay. That's their tuning 10 parameter for --
11 MR. ZUSER: That's a tuning parameter, okay.
12 Now, can you put the other slide which has the figures, 13 the Mickey Nouse? I was really perplexed with this
/\
t t
\/ 14 scaling and the process which occurs there. And this 15 morning when I came here this was the same comment that 16 Dr. Ishii made in his -- so let me just go over it. Can 17 you describe what is the characteristic of the flow in the 18 hot leg?
19 MR. TINKLER: The characteristic of the flow?
20 MR. ZUBER: Yes, how many parameters 21 characterize that flow?
22 MR. TINKLER: Well, typically what they looked 23 at was a scaling with a Richardson number, and there was 24 some concern whether or not preserving the Richardson
/%
(_) 25 number in the hot leg was sufficient, 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 l
45l 1 MR. ZUBER: But the Richardson number is based g 2 on the gradient.
b 3 MR. TINKLER: Well, yes, and the delta rho --
4 it's Grashof of a Reynolds square but they got a pretty 5 good match on the Richardson number in the hot leg. Now 6 the question is, is that by itself enough?
7 MR. ZUBER: Well, let me -- there are two --
8 the characteristics of this countercurrent flow, you have 9 two fluxes, one opposing the other one. You have tvro 10 flows --
11 MR. TINKLER: Right. l l
12 MP. ZUBER: -- and you have an area split, so 13 you have really three parameters.
O s# 14 MR. TINKLER: Right.
15 MR. ZUBER: And if you go in the literature, I 16 think there was a fellow Coylan, who 30 years ago worked 17 in the Bureau of Standards and he analyzed the flow, 18 different c.e ns it ie s , two pipes, inclined pipes and 19 horizontal. And there have been many, in the last ten 20 years there were other papers in the International 21 Journal, multiphase flow, which looked at this problem.
22 MR. TINKLER: Right.
23 MR. ZUBER: And if you read them, that process 24 is characterized by two Fraude numbers. What you have
,O
(_) 25 here, what you have here, this is just one Fraude number.
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 Now, this is a mass flow. What is the mass flow rate, 1
2 WHL? l 7s k) 3 MR. TINKLER: What is it?
l 4 MR. ZUBER: Yes. I mean --
5 MR. TINKLER: Characteristic number --
6 MR. ZUBER: That's the pounds per hour. Which 7 one, the hot leg with?
8 MR. TINKLER: Yes.
9 MR. ZUBER: Okay, what were the cold fluid?
10 MR. TINKLER: The ratio of hot leg -- well, I 11 mean, you know, it's just conservation of flow. They're 12 going to be roughly about the same.
13 MR. ZUBER: No it is not. This is embedded --
O km 14 MR. TINKLER: But pounds per --
15 MR. ZUBER: This is in --
16 MR. TINKLER: Mass flow rate --
17 MR. ZUBER: This is embedded. The fluid 18 number depends on the velocities. This is how you obtain 19 the D Mh 20 MR. TINKLER: Right.
21 MR. ZUBER: Five halves?
22 MR. TINKLER: Right.
23 MR. ZUBER: When you square did you obtain W/D 2 24 which is the velocity and the square root of D? This is r
( _j 25 this fluid number.
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 l l
1 MR. TINKLER: Right.
2 MR. ZUBER: But if you look at scaling of i )
3 this, what is on the rows, you have two fluid numbers, one 2
4 based on the light fluid the other one based on the heavy 5 fluid, which are partitioned according to the flow areas.
6 MR. TINKLER: Okay.
7 MR. ZUBER: Your assumption is, you got rid of 8 that problem, or not even addressing this question, is by 9 saying, half of the area is occupied by the cold fluid, 10 the other one by the hot fluid.
11 MR. TINKLER: You're suggesting a comparison 12 of the fluid --
13 MR. ZUBER: I'm not suggesting --
/
(
14 MR. TINKLER: -- numbers for the two fluid 15 streams?
16 MR. ZUBER: There are two fluxes and two 17 cross-sectional areas to reach this fluid flow.
18 MR. TINKLER: Correct.
19 MR. ZUBER: And they have known this for 30 20 years, at least look at the work of Coylan, Bureau of 21 Standards and International Journal of Mass Transfer.
22 These processes cannot test by three parameters or two 23 fluid numbers.
24 MR. TINKLER: Okay.
f' Q) 25 MR. ZUBER: Okay. Now, what you have here is NEAL R. GROS 3 COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASH!NGTON. D C. 20005-3701 (202) 234-4433
48 1 only one fluid number, and you've got it by arm waving or 1
1 2 not even looking at the other one. So this was my q
G 3 question. How was this scaled? And I'm really pleased 4 that this was also picked up by Ishii. And if you look in 5 his letters he says he does not understand of 6 countercurrent flow -- it was not clear what kind of 7 scaling analysis has been carried out for the hot leg 8 countercurrent flow. And to me, what 1 see is, it may 9 have been carried, it may have not, but from your answer 10 and what I saw in this handout at Fauske, it was waved 11 away.
12 MR. TINKLER: Okay.
13 MR. ZUBER: Okay, that's number one. These O
t /
v 14 then affects also the mixing, because that mixing would 15 depend on the velocity of the fluid which comes out. Now 16 again, you somehow put this under the rug by saying half 17 the area is occupied by the cold fluid, half by the hot 18 liquid.
19 MR. TINKLER: Well, it's not put under - I 20 mean, these were observations from the experiments which l 21 were scaled using the Richardson number. They didn't 22 scale fluid --
23 MR. ZUBER: Forget -- that scale that you use l l
24 in this -- that Fauske -- is the fluid number. They G
Q 25 juggle those numbers based on the gradient, D rho DZ.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W. ,
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-*433 1
49 1 MR. TINKLER: Right.
2 MR. ZUBER: This is -- you don't here a D rho 73 3 DZ, you have just delta over something. It's a fluid 4 number scaling, but countercurrent flow of two fluids in 5 horizontal inclined is characterized by two fluid numbere.
6 MR. TINKLER: Right.
7 MR. ZUBER: You have enly one.
8 CHAIRMAN KRESS: What is the typical 9 difference between those two fluid numbers? Are they 10 markedly different?
11 MR. ZUBER: Just depends, I mean, they are to 12 a parameter on the right-hand side, and you can make this 13 parameter equal to one. But you have really essentially
,m.
N/
/ You have two fluxes, two fluid -- two 14 this problem.
15 fluxes flowing to these two different areas.
16 CHAIRMAN KRESS: I would expect the fluid 17 number to be not that far different between the two, so 18 that one average fluid number which is what I think he 19 has, would be a reasonable representation --
20 MR. ZUBER: See, I don't know --
21 CHAIRMAN KRESS: -- but I don't know.
22 MR. ZUBER: I don't know --
23 CHAIRMAN KRESS: It wc ,d be nice to know what 24 those two fluid numbers are.
/~~N ,
(_j 25 MR. ZUBER: The only thing is, yesterday I 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
50 1 reading this and this was the first question which came to s 2 my mind. Where is the other fluid number? I didn't know I )
U 3 that they partitioned the area one-half and one-half. So 4 this was the reason of my question. And then I was 5 reading Ishii's comments. He got the same question. And 6 they also conclude that this is really the most important 7 thing. This is how you scale this, this is the mixing in 8 of plenum and how is the mixing scaled.
9 MR. TINKLER: A little later on you'll see 10 that we -- using this approach we went back and compared i
11 the mass flow rates that we calculated versus measured, on '
12 a one --
13 MR. ZUBER: Look --
i f
\~ 14 MR. TINKLER: -- set of scale test. You may 15 not take much comfort but we were within 11 percent on l 16 mass flow rates.
17 MR. ZUBER: Well look, I know that you can 18 tune any code to whatever you want. This is my experience 19 of 20 years associated with this Agency --
20 MR. TINKLER: I said you might not take much 21 comfort from it.
22 MR. ZUBER: Well, look I --
23 MEMBER CATTON: And you're right.
24 MR. ZUBER: The only thing is that if you want s
I )
, (_/ 25 to say this is a well scaled experiment, you have to prove 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 it. What I have here in -- what I read yesterday was not 2 convincing, Ishii was not convinced, and therefore my 7_s
( )
\'
3 question, final question is, how can you use this to 4 verify a code and determine uncertainties?
5 MR. TINKLER: Well, a little later on I'm 6 going to show you some figures of merit that may make you 7 feel a little differently about the plant calculations and 8 the test data. It won't be a comparison of Fraude numbers 9 of the two fluid streams, okay.
10 MR. ZUDER: But, this is important process, 11 and the code cannot address it and I don't know how you do 12 it artificially --
13 MR. TINKLER: Well --
x
\- 14 MR. ZUBER: But let me just finish. If you 15 cannot do it -- at least I can tune a code. I would feel 16 very comfortable if I have a good set of experiments, well 17 scaled and I tuned the code to it. So if the code cannot 18 model it, then the burden is on good and well scaled 19 experiments, and this has to be shown that what I read 20 here yesterday from this, I was not convinced, you didn't 21 convince me now, and evidently you did not convince Ishii 22 either.
23 MR. TINKLER: Okay.
24 MR. ZUBER: So all the calculations you made A
( ,) 25 to me is very iffy and to even know how to determine the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. O C. 20005-3701 (202) 234-4433
i 52 1 uncertainty from iffy conditions.
2 MR. TINKLER: I guess I'm not quite clear why 7_.s i \
3 you don't think that preserving buoyancy in terms of delta 4 rho over a reference density would be sufficient for --
5 MR. ZUBER: Nobody argues about this, but you 6 have two --
7 MEMBER CATTON: Novak, what they do is, they 8 separately, from the hot leg they separately tune the 9 mixing to the experiment.
10 MR. ZUBER: But then --
11 MEMBER CATTON: And then the question comes, 12 well how well do you know it?
13 MR. ZUBER: But more than that --
A
'v]
14 MEMBER CATTON: They may think they know it 15 absolutely.
16 MR. ZUBER: But more than that, how well is 17 this scaled if you have this process? Now, look, it is 18 very long that you have two fluid numbers and you have two 19 cross-sectional areas to which this fluid flows. Now if i
20 you said it's only half-and-half, is this always good? I 21 don't know.
l 22 MEMBER CATTON: But in the experime.nt, they 23 ran these with SF., right?
24 MR. TINKLER: They ran with both water and SF..
, ,)
- Q' 25 MEMBER CATTON
- So they ran it with a gas so NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
I 53 1 the prandtl numbers are about right, they made the r~s 2 Reynolds number and Richardson number, it was two l )
~
3 parameters. I'm not as concerned as Ishii was about that 4 particular issue. The thing is, it's a coupling of the 5 outlet of the pipe to the small tubes in the steam 6 generator is where I have a major headache.
7 MR. ZUBER: But what you say, with due respect 8 to you, I'm concerned. Because that mixing, what you are 9 saying, your concern which affects inlet, really depends 10 on the mixing what comes in. That turning around that Il little vortex really depends -- it's a tangential velocity 12 which -- the velocity which comes out of the hot leg w'..ich 13 gives the spin.
l \
\/ 14 Okay, now if this is well scaled the spin is 15 well -- if they impulse, I mean, this is the driving force 16 that fluid comes out. This gives the vortex. If that is 17 well scaled, fine. And the core cannot predict, fine. If 18 I have good experiments and well scaled to address this, 19 we are okay.
20 MEMBER CATTON: Well, they have the analysis 21 that was done by Mike Epstein. Isn't he -- he presented 22 it?
23 MR. TINKLER: Right, right.
24 MEMBER CATTON: What they did is they said,
(/"j) 25 let's look at this as a buoyant plume at the exit point, 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 right where the hot leg enters into the lower plenum. And by comparing it with a buoyant plume they got reasonably
+g o 2
%/
3 good comparison.
4 CHAIRMAN KRESS: Does that mean the momentum 5 is negligible?
6 MEMBER CATTON: That doesn't -- it says that 7 the impact of momentum is small.
8 CHAIRMAN KRESS: Oh.
9 MEMBER CATTON: I don't know if it's 10 negligible, because all these things play a role in the 11 mixing process. But what it does is, it lesse..a the 12 importance of the two Fraude numbers.
13 CHAIRMAN KRESS: Yes, it would.
- O
\ '/ 14 MEMBER CATTON: But I think there's still --
15 but see, you don't have to vary that mixing very much. I 16 mean, if you change it from .7 to .5 you begin to see an 17 impact because the temperature of the fluid goes up. l l
18 MEMBER FONTANA: Well, did the Westinghouse i I
19 experiments -- and you say they had a smaller number of !
20 tubes -- l 21 MEMBER CATTON: They had big tubes.
But l 22 MEMBER FONTANA: Big, a smaller number.
23 the question that we're discussing here I think, is the 24 plume of hot gas going into the inlet plenum of the steam n
i %
( ,/ 25 generator, the geometry of that inlet plenum, the number NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 2'A-4433
55 1 of tubes, or the widths of the tubes along the top and the 7-~g 2 geometry and the size of that hot plume that goes up and t f Q/
3 impacts the inlet face of those tubes. That could affect 4 the temperature distribution entering these tubes. It ,
5 also affects the recirculation. J 6 MEMBER CATTON: Right.
7 MEMBER FONTANA: Now in your next slide you 8 say that the steam generator tubes are split 35/65 based 9 on Westinghouse experiments.
10 MR. TINKLER: Right.
11 MEMBER FONTANA: The question is, did the 12 Westinghouse experiments scale that inlet situation?
13 MEMBER CATTON: They do not.
I,_h
\ )
\/ 14 MR. TINKLER: Well, what they preserve is the 15 ratio of flow area in the tube bundle area of the hot leg, 16 okay. And they preserved geometric similitude in here.
l l
17 Now, in order to preserve this flow area, they can't use 18 one-seventh of a three-qtarter-inch diameter tube so they l
l 19 make the tubes a little bit and they use less of them.
20 But this flow area of the tube area relative to the flow 21 area of the hot leg is scaled. They do preserve that 22 ratio.
23 so if you think you back up flow here, because 24 the flow comes in, you can't get it up into the tubes fast !
/~'
(_)N 25 enough, it spreads out, creates a plume which is part of NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D.C. 20005-3701 (202) 234-4433
56 l 1 the -- this is not the SCDAP model, this is the FAI n 2 approach to that -- spreads out into a plume, and you get
( )
l 3 mixing within that. Using this backup of flow in the 4 inlet plenum promotes that mixing.
5 Now, what I want to say is that because they 6 see that as being dominated by the ratio of this flow area 7 to this flow area, and the fact that the ore-seventh scale 8 test preserved that flow area, they think they capture 9 largely the effective mixing. You know, that's one of the 10 principal reasons why you g.t this mixing. Is this flow 11 area up through the tube bundle is less than this flow 12 area.
13 So you get circulation in here. This rises, O
b 14 it can't enter all the tubes at the same rate that it 15 exits the hot leg, and it spreads out. Then you get cold 16 flow returning through some of these same tubes through l 17 some of the other tubes in the tube bundle, it mixes.
l 18 CHAIRMAN KRESS: Is the overall diameter of 19 that steam generator one-seventh also?
20 MR. TINKLER: I --
21 CHAIRMAN KRESS: Because this would depend on 22 the --
23 MR. TINKLER: Well, I think it in order --
24 CHAIRMAN KRESS: -- separation --
On 25 V MR. TINKLER: In order to preserve this flow 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 area relative to that, it would to be. I don't see how it 2 could not be.
-)
3 CHAIRMAN KRESS: Ivan, you were concerned 4 about this tube problem in the outlet plenum, weren't you?
5 MEMBER CATTON: Not so much.
6 CHAIRMAN KRESS: That wasn't where I -- I 7 originally thougnt that was where you were worried about 8 it.
9 MEMBER CATTON: Well, what you're really 10 concerned about is the tube temperature just a little bit 11 above that three. Because that's going to be where it's 12 the hottest. The flow very quickly gets to be almost the 13 same temperature as the tubes. Or when the tubes get to rs
> 1
(_/ 14 the same temperature of the flow, whatever.
15 MR. TINKLER: I'd also point out, too, that's 16 an important point but it's also important to remember 17 there's a tube sheet right here, half-a-meter thick, okay.
18 You're not going to fail the tubes in the tube sheet.
19 MEMBER CATTON: You fail it just above them.
20 MR. TINKLER: You've got to get up past that 21 tube sheet.
22 MEMBER SHACK: You've got an 18-inch mixing.
23 MR. TINKLER: An 18-inch mix -- that's a lot 24 of L/D's and that's a big -- that's 25 tons. That's a big
()
gs 25 heat sink there.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE !SLAND AVE , N W.
(202) 234-4433 WASHINGTON. O C. 20005 3701 (202) 234 4433
58 1 DR. LEE: Charlie, could I make a comment?
2 MR. TINKLER: Sure.
NY 3 DR. LEE: This is Richard Lee from the Staff.
4 Novak, you mentioned about Ishii letter. Ishii is the 5 least familiar with this experiment, out of the three peer 6 reviewers, that we conducted a few weeks ago. To him is 7 that this set of experiments as indicated in his letter, 8 it is well scaled. His concern is that more -- I'm not 9 sure whether you read it, whether ;i was this experiment, 10 but I think he's asking is that, because we spent the hot 11 leg in the SCDAP/RELAP modeling, he was raising that it 12 could be some interactions between the two, hot and the 13 cold stream.
rs 14 MR. ZUBER: I think --
15 DR. LEE: And then, let me remind you, in 1990 16 and '91, you were part of the four experts that reviewed 17 the Westinghouse data. Mike Griffith, Viskanta, and 18 Ostrach, and you. At that time --
19 MR. ZUBER: I didn't -- let me --
20 DR. LEE: -- this can only --
21 MR. ZUBER: No, I didn't really --
22 DR. LEE: Let me finish. The key things that 23 you all raised at the time was the relation heat transfer 24 between the two streams of hot and cold was the key rs 25 witness with respect to the experiment. That was the key NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
59 1 thing that came out from that review.
,g 2 MR. ZUBER: 0..a y . Now let me answer one 3 thing.
4 DR. LEE: Okay. So, I just wanted to remind 5 you.
6 MR. ZUBER: Now, I say, are you in charge and 7 using taxpayer's money to ask peer review, I would then 8 get peer review what familiarly the things that I can 9 trust their comments. Now if you have here a peer review 10 member who gives a question number, questioning the 11 validity of these experiments, don't then discredit that 12 peer review member. That's not fair.
13 DR. LEE: I'm -- no, no, no. Ishii is -- I'm (nl
\/ I'm telling you he's the least 14 not discrediting him.
15 familiar with this. Based on this very short presentation 16 he is still comfortable about the scaling of the 17 experiment. That's not the point.
18 MR. ZUBER: The point --
19 MR. TINKLER: Let me just say -- look, I take 20 your comments quite seriously. We'll take a look at the 21 Fraude numbers for the opposing streams. But I wanted to 22 point out that momentum exchange between the streams is in 23 part, again, compensated by the fact that we have to use a 24 little higher loss coefficient. But we can take a look at
/~
( ,T/ 25 that for the test data and the plant calculations. 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
i 60 l 1 you know, that's a -- we can look at that.
p 2 The original scaling considered Richardson 3 number and the delta rho against the reference density.
4 But I understand your comment and we'll consider that.
5 MR. ZUBER: Okay, let me just make a final 6 point --
7 MEMBER CATTON: Why don't we have Ray 8 Viskanta's report in this --
9 MR. TINKLER: Because we asked Ray to take a 10 look at some additional issues and we gave him a little 11 more time to do it, and we should have it -- actually, I 12 thought we would have it by this morning.
13 MEMBER CATTON: Because he's really the only
.O
~
14 one who's familiar in some sort of an expert sense, with 15 these kinds of problems. Single phase, natural convection 16 problem is what you're dealing with.
17 MR. TINKLER: I per --
18 MEMBER CATTON: And he has a lot of experience 19 in that --
20 MR. TINKLER: I think all these gentlemen have 21 considerable experience in scaling and this general issue, 22 okay.
23 MEMBER CT.fTON: Well, buoyancy driven flows l
24 are quite different than what you normally deal with in l (y
'q,) 25 the nuclear arena.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
61 1 MR. TINKLER: True, but we've been more and fs 2 more concerned with buoyancy driven flows and the ALWR's
. \m ,/
3 over the last couple of years because of their passive 4 nature. So we've been looking at these kinds of issues, 5 along with mixed convections --
6 MEMBER CATTON: Somehow I think a lifetime of 7 experience is probably a little more valuable.
8 MR. TINKLER: Agreed.
9 MR. ZUBER- May I just make a final comment 10 because I have to leave for another meeting. RELAP cannot 11 calculate this process. -- current and countercurrent 12 flow in horizontal -- and they're mixing very well. I 13 would be very comfortable with that statement if I would
(
\~ ' 14 hear that a good set of experiments, well scaled to which 15 I can tune. And then this problem can be put to rest. So 16 the onus, as far as I can see from here is, look at the 17 good and bad scale experiments. From information I have 18 here, it's questionable and it would be desirable for RES 19 to present more information in terms of how was this 20 scaled and arguments, how was it done. No.
21 MR. TINKLER: I'd at least like you to reflect 22 on the benchmarking that we did do when I get to that. So 23 you may not feel differently, but at least consider the --
24 MR. ZUBER: Yes, but the point is, the
-( 25 question is, I know the code cannot do it. The process is q,,/
NEAL R. GROSS COURT REPORTERS AND TRANSCRfBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
62 1 important, but how do we address it? If I have good s 2 experiments I can tune that code in. I can defend this.
t )
~' But what you are really using, using a code which cannot 3
4 model it to justify the experiments which would then 5 justify the code. That's a round, a round circle of 6 arguments. The argument would be, this is well scaled, I 7 tuned the code, we put this problem to rest.
8 MR. TINKLER: Well, I understand, and I've 9 told you the scaling numbers, the similitude that was 1
10 considered in the tests, okay. A rough preservation of 11 Reynolds numbers, Richardson number, similitude in terms 12 of ratio of flow areas, heat distribution in the system, 13 they scaled masses.
s x/ 14 I don't have the picture, but they scaled the 15 mass of the tube bundle relative to the tube bundle in the 16 plant and the major structural -- I mean, because that's 17 where much of the heat removal occurs, is in the heat-up 18 of structure. So they scaled those masses between the 19 one-seventh and full-scale. That's one reason why they 20 have -- end up with the tube diameter they end up with, in 21 order to get the right mass of the tube bundle.
22 As far as heat transfer modeling goes, 23 basically we've used the standard RELAP5 heat transfer 24 package. For forced convection and the turbulent regime
/-
( 25 we used Dittus Boelter. In the laminar regime, Sellars, q,
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
i 63 l
1 equivalent Nusselt Number 4. For free convection we used 2 Churchill-Chu and McAdams for horizontal configuration.
7-b 3 And consistent with the general recommendation in McAdams, 4 the code uses the maximum of free and forced convection 5 for doing a calculation.
6 MEMBER CATTON: This is another example of, I 7 think inappropriate choices. The free convection, you 8 really don't have free convection, you have mixed 9 convection, and the name that should be up there is 10 Jackson. Jackson has been doing -- he has studied the 11 horizontal pipe, he's studied vertical pipes, he's looked 12 at ranges of prandtl number and he's in the nuclear 13 business.
,m (k-) 14 MR. TINKLER: When I get to page 40 --
15 MEMBER CATTON: You'll come back --
16 MR. TINKLER: -- I'll talk about mixed 17 convection and our incorporation of Jackson correlation to 18 calculate mixed convection. It's recognized that there .
l 19 are times -- when the Richardson number again, is around 20 one, it's regime of mixed convection. So that's 21 understood.
22 This is the, for those that like to look at i
1 23 nodalization diagrams, this is the nodalization that was ;
1 24 deve]oped in the 1986 timeframe, I believe, for following
/~N 1 (m,) 25 countercurrent natural circulation.
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 This is where we simulate inlet plenum mixing 2 in these mixing volumes, is the inlet plenum steam b,f-~s 3 generator. The mixing fraction roughly relates to the 4 flow rate in these volumes here.
5 MR. ZUBER: How do you split the flow from 404 6 to 405, from 404 to 406?
7 MR. TINKLER: This is where the code model was 8 benchmarked against both the COMMIX calculations and the 9 one-seventh scale test.
10 MEMBER CATTON: They tune the volume --
11 MEMBER FONTANA: That's the 35 -- 35/65 --
12 MR. TINKLER: No, that's not the 35/65. This 13 is -- actually, we don't -- the code doesn't know what a 14 mixing fraction is. You have to tell it areas and loss 15 coefficients and it calculates an equivalent mixing 16 fraction. Now, it turns out that when we benchmarked the 17 code it produced the right amount of heat removal in the 18 lube. We came very c_'ose, might say suspiciously close, 19 to the mixing fractions that were seen in the test, but 20 nonetheless we did. It wasn't forced to produce mixing 21 fractions. If I want --
22 CHAIRMAN KRESS: Given the total flow and the 23 volumes and areas of those nodes and given a mixing 24 fraction, does that give you a closed system to calculate O)
(, 25 those splits in the flow? Is that how it works?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(20?) 234 4433 WASHtNGTON. D C. 20005-3701 (202) 234-4433
65 1 MR. TINKLER: No, no.
es 2 CHAIRMAN KRESS: And that's not closed?
O 3 MR. TINKLER: No, that's not closed.
4 CHAIRMAN KRESS: You have to input the splits 5 as well?
6 MR. TINKLER: You input the split of tubes.
7 CHAIRMAN KRESS: But you adjust them somehow 8 so that when the code calculates a mixing fraction it 9 calculates the one you want to have?
10 MR. TINKLER: Well, what we did was, when we 11 first benchmarked the model we determined the areas, 12 volumes, and loss coefficients, that produced the right 13 energy balance around the system.
<,_')
\' 14 CHAIRMAN KRESS: Okay, that's based on an --
15 MR. TINKLER: And we never change that, okay.
l 16 We freeze that. Now, we change the number of tubes. It 17 has a feedback effect, but not a very strong one, okay.
18 It will cause the mixing fraction to shift a little bit 19 because we've used these areas and loss coefficient.
20 MEMBER CATTON: So Charlie, do you thir.k if I 21 were to slowly decrease the diameter of the tubes in the 22 Westinghouse facility, and add more tubes to --
23 MR. TINKLER: And add more tubes --
24 MEMBER CATTON: And add more to keep the
(~'s
('_ 25 things -- which way would the mixing go and why?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
1 66 1 MR. TINKLER: I personally think that it would fx 2 cause the mixing to improve.
(O I 3 MEMBER CATTON: Why? Would you base --
4 MR. TINKLER: Because I think having larger 5 diameter tubes allows a flow to go directly up in it more 6 easily.
7 MEMBER CATTON: Now, do you think there's any 8 historesis? When this process starts up you're going to 9 be -- you have to cover a range of Reynolds numbers and 10 Rehle numbers and so forth, as you slowly bring it up 11 until it gets at some sort of steady flow. What do you 12 think that will do? There you see, I could have a nice, 13 almost laminar plume rise straight up and hit just a few k- 14 tubes.
15 Now, it wouldn't overheat them probably, 16 because the mass flows are too low, but nevertheless, what 17 do you think it will do in the long run? This is Mario's 18 question rephrased. If I heat a few tubes and get it 19 started, is it going to change? Well, this is an 20 uncertainty. What do you do with it? Just punt?
21 MR. TINKLER: Well, given the variation and 22 the fluid conditions over the transient, I can't foresee I 23 that it would a -- that's it's going to draw somehow, much j 24 more flow to those tubes. It's going to draw a little A
( ,) 25 more and then it's going to adjust.
NEAL R. GROSS .
COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
67 f 1 MEMBER CATTON: Now, the point that I was 7s 2 making earlier is that, somehow we have to figure out how i )
We don't know 3 to quantify the uncertainty we're hearing.
4 any of these things with absolute certainty. Now here, 5 I'm not talking about your code, I'm just trying to 6 condition the committee --
7 MR. TINKLER: You understand that the answer 8 that I gave to your question in that, in other 9 calculations, not SCDAP/RELAP5 calculations but in other 10 calculations, that they'll allow these conditions to float 11 more readily, and actually they argue, there is a scale 12 effect that it actually increases mixing. Beyond that 13 seen at one-seventh scale. l
(^ (
k-)) 14 Because the mixing is determined by the mixing i 15 fraction and the recirc ratio. How much of this hot flow )
16 gets diluted by how much of this cold flow? And if you 17 increase this recirc ratio you dilute this with more cold la flow. And the industry calculations are based on a model l 1
i 19 that actually increases comewhat. Our calculations don't.
20 They hang around a recirc ration of two.
21 MEMBER FONTANA: Was COMMIX used to calculate 22 variation of parameters? In other words, COMMIX was used 23 to calculate the recirculating flow patterns in the inlet 24 plenum, wasn't it?
(\ MR. TINKLER: Well, not in any detailed sense.
( ,/ ?5 NEAL R. GROSS t0URT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE.. N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
68 1 There was a pretty large nodalization in the inlet plenum.
2 MEMBER CATTON: You have to take COMMIX
%.J 3 calculations with a grain of salt. Unless you've seen the 4 convergence testing of the r. dalization, you don't know 5 what the answer means.
6 MEMBER FONTANA: What I was kind of driving is 7 kind of -- parameter can very easily can do a series of 8 calculations where you vary the plume effect and this can 9 be done fairly easily so you can see --
10 MR. TINKLER: We can vary the number of tubes, 11 but the code doesn't know there's a plume that 12 mechanistically spreads and mixes --
13 MEMBER FONTANA: I know, I understand. That's (D
V 14 why you have to vary it to see what you can effect to 15 that.
16 MR. TINKLER: We have done calculations where 17 we tried to preserve those mixing fractions. We've done 18 other calculations where we let them float, whatever the -
19 - would calculate, you know, it calculated. And I'll talk 20 about those sensitivities.
21 As if the thermal hydraulics wasn't I l
22 complicated enough there was the issue of fission product 23 deposition and heating of the tubes. Again, our approach 24 was to use the VICTORIA fission product code with I
Q 25 SCDAP/RELAP5, thermal hydraulic conditions as the driver.
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 Write a little code, subroutine, that you pick the (q'O
)
2 SCDAP/RELAP5 calculations to drive the VICTORIA chemistry 3 models.
4 CHAIRMAN KRESS: Does that mean you input all 5 the flows and temperatures --
6 MR. TINKLER: Yes.
7 CHAIRMAN KRESS: Versus time?
S MR. TINKLER: Versus time. And you know --
9 well, I'll show you a little later on the nodalization, 10 but we have pretty good correspondence so it made it a 11 little easier.
12 The conclusion was that the fission product 13 transport and deposition in the RCS had a negligible t J C' 14 effect on the steam generator tube integrity. And the 15 points to remember here is that the fission product 16 release, it's pretty small. This is early in the 17 sequence, really, and relative to the steam generator tube 18 heating it's -- most of the fission product release occurs 19 after steam generator tubes heat up, or after the surge 20 line fails, in any event.
21 CHAIRMAN KRESS: Do we deal with the noble 22 gases any in this? Their heat?
23 MR. TINKLER: We have noble gases transiting 24 around the system. A good portion of those get released jO V 25 out of the PORV.
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 ,
i 1 CHAIRMAN KRESS: Oh, yes, I guess you're f" 2 right.
O} 3 MR. TINKLER: A good portion of them do.
4 CHAIRMAN KRESS: So basically, for this heat 5 it's mostly cesium and iodine?
6 MR. TINKLER: It's mostly cesium iodine. Now, 7 that represents, I think, something like three-quarters of 8 the heat on the tubes, something like that. And the 9 fission products again, are spread amongst the upper 10 plenum, the hot leg, the steam generator plenum and the 11 tubes. The steam generator represents, I think, something 12 over 95 percent of the area for deposition, so it's the 13 big sink at the end of the loop.
I
\ /
14 MEMBER POWERS: That's one of the things you 15 have to be very cautious about in using a VICTORIA with 16 somebody else's thermal hydraulics, is that will 17 inherently tend to spread the aerosols to make it uniform, 18 because it just treats the things it knew, where aerosols 19 have a tendency to deposit locally, like at inlets on 20 tubes and things like that, and anyplace where theiv's a 21 temperature discontinuity like at support sheets.
22 So you want to be cautious about, you know, 23 spreading it uniformly. It's a small fraction of the 24 decay heat at this point so it may not make any k ,)
m 25 difference, but local deposition at inlets and at support l
NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 i
)
1 71 1 sheets would be the things that I would tend to do outside 2 the code. I would tend to do, you know, I'd use VICTORIA 73 3 -- VICTORIA would give you a good idea about what the 4 particular sizes are and vapor fractions and things like 5 that.
6 But I don't think it has the microscopic 7 models -- I think you just have to go in and look at these 8 things on a hand basis to make sure you're not 9 artificially spreading things. Because the code will 10 inherently do that. It wants to make everything uniform 11 because it only sees one sheet. And especially right at !
l 12 that inlet where you've got that whirligig business going 13 on.
A 14 The particles aren't coming in square.
15 They've got a momentum that is different than the momentum 16 of the gas. And if the stopping distances are large 17 compared to the tube diameters they will tend to deposit 18 right on the inside surfaces there. And if there's any 19 temperature difference along the length of the tube, like 20 where there is a support sheet -- and boy, it does not 21 have to be much. A degree is a lot -- you'll get local 22 thermophoretic deposition there.
23 MR. TINKLER: Well, in response to some 24 comments now that you expressed earlier, we went back and f3
'!] 25 took a look at some of those things --
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 MEMBER POWERS: Oh, okay. >
l g 2 MR. TINKLER: -- the vena contracta effect 3 entry in the tubes.
1 4 MEMBER POWERS: How about just the inlet 5 efficiency problem? The fact that particles are going --
6 MR. TINKLER: Well, we varied the flurry of 7 the path between the inlet plenum and the tube sheet, 8 okay, to see what kinds of sensitivities might exist 9 there. Again, remember that that tube sheet is right 10 there. That's a large, large heat sink.
11 MEMBER POWERS: Yes, and that's the concern of 12 course, is that you get localized thermophoretic 13 deposition.
(D U 14 MR. TINKLER: But -- well --
15 MEMBER POWERS: It's not important?
16 MR. TINKLER: I suspect you could heat the 17 tubes up quite a bit in the region of that tube sheet and 18 not have any effect on the tubes whatsoever.
19 MEMBER POWERS: No temperature change at all.
20 MR. TINKLER: Well, they're supported. You 21 know, they pass through this tube sheet. But, we did look 22 at things like a vena contracta effect to see if it would 23 have any influence on our results and it didn't.
24 CHAIRMAN KRESS: The particles are probably fO), 25 pretty small.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
73 :
I 1 MR. TINKLER: These are pretty -- yes.
2 Actually, we were just looking at that. They're, most of x~ them are, a number mean is below one micron, 40 percent --
3 4 MEMBER POWERS: Number means they're always 5 below one micron.
6 MR. TINKLER: Well, 50 percent of them --
7 CHAIRMAN KRESS: It's the mass --
8 MR. TINKLER: Fifty percent of them above one 9 micron -- mass, on a mass basis.
10 MEMBER POWERS: Those are big particles.
11 MR. TINKLER: That's only 40 percent. That 12 means 60 percent of them on a mass basis are below one 13 micron, so there's still a lot of small particles, k> 14 MEMBER FONTANA: With respect to the main 15 concern of this meeting, would we be missing much if we 16 jumped to benchmarking of the SCDAP/RELAP?
17 MR. TINKLER: No, but let me just -- I want 18 to, lest there be an residual concerns, just quickly say 19 this is the nodalization we use for VICTORIA. Actually, 20 we use the same nodalization in steam generator that we 21 did in SCDAP/RELAP calculation. And the numbers are 22 pretty small. There's only 400 kilograms. They're 1
23 spreading it out over a big surface area. Only 50 of it's 1 i
24 fission product stuff, 40 of that's noble gases. So it's l D
( 25 not a real -- these aren't real big numbers here. When NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 ,
l
I 74 1 you look at the diameter of the tubes and you look at the 2 positive thicknesses, these are all pretty small.
7-w L) 3 MEMBER FONTANA: That's a main point.
4 MEMBER POWERS: It's only 400 kilograms of 5 aerosol. I mean, it's a formidable amount of aerosol.
6 MR. TINKLER: Well, I know, but --
7 MEMBER POWERS: Your point is still valid.
8 MR. TINKLER: -- is 25 tons, the secondary 9 structures are 100 tons, you know, these are -- and 10 there's fission product heating relative to steam heating.
11 So we can change this number a lot.
12 MEMBER FONTANA: Yes, that's an important 13 point.
I \
5/ 14 MR. TINKLER: Objective of the benchmarking.
15 Again, back to the one-seventh scale test. The objective 16 of that program was to perform experiments to investigate 17 the phenomena -- natural circulation -. There was 18 geometric similitude. They did represent the RFP, the hot 19 leg, surge line, steam generator, preserve flow 20 resistances around the loop.
21 Again, they maintained the ratio of the flow 22 rate to the steam generator tubes, the hot leg. Because 23 of practical limitations they couldn't make the tube 24 diameter of the steam generator tubes correct on the one-A o 4
(_/ 25 seventh scale basis, so they're a little large. And these NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
1 75 f 1 tests used a range of fluids under a range of conditions.
2 They used water, they used SF.. S F, is a high-3 O 3 density fluid gas. In order to get the right delta rho 4 over rho. They can't run these tests at 2000 degrees 5 Kelvin. To get the right density they've got to use some 6 other fluid. As your experience with the ALWR's, a lot of 7 times people run full height tests in order to get the 8 height right, in order to preserve buoyancy effects.
9 Well, they didn't, they couldn't do that and fit it inside 10 the building, so they ran it with a different fluid. But 11 these things were considered.
12 The low pressure tests using dye in a fluid 13 showed that a stable countercurrent flow was present in O
- 14 the hot legs with little turbulence at the interface.
15 That doesn't mean zero turbulence but there wasn't rapid 1
16 mixing. They did have basically a hot stream and a cold 17 stream.
18 And basically the tests show that the I
19 phenomena was consistent. It prevailed over a range of f
20 conditions. This same behavior, this same flow pattern l 21 prevailed over a range of power inputs stimulated by 22 heater rods in the core, ranges of pressures, different 23 fluids.
24 CHAIRMAN KRESS: These low pressure tests were b
N ,/ 25 -- you had glass sections so you could see this dye -- l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005 3701 (202) 234-4433 I
1
76 1 MR. TINKLER: Yes, they were acrylic so they
-s 2 could see it. Then there was an interruption of
\j 3 activities while they built the high pressure facility.
4 MR. ZUBER: How many geometries did they 5 change?
6 MR. TINKLER: Well, they would shut off a 7 steam generator, or shut off flow in the hot legs just to 8 observe natural circulation in the core region. How many -
9 -
10 MR. ZUBER: The question is, did they perform 11 a -- the only information I have is this, which I obtained 12 from Mr. Dudley, so my comments are based on this. There 13 may be additional information which may answer my
.,rN4
's / 14 questions, but I don't have it. The point is, did they 15 run experiments with different geometries, diameters of 16 the hot legs and of the plenum, in acrylic?
17 MR. TINKLER: I'm not aware that they altered 18 the basic geometry of the system after it was constructed.
19 They would shut off parts of it to observe behavior in 20 parts, but once they built this I'm not --
21 MR. ZUBER: They could have done -- see, one 22 of the problem is, if RELAP cannot calculate it, 23 countercurrent flowing horizontal pipes, single phase, 24 what is the database to verify a model, I mean, on which n
'( ,) 25 to base their model and tune their code? And what 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
77 1 have done is to assume half of the area is occupied by hot i 2 liquid, the other half by cold liquid. Fine. Are there !
3 conditions, possibilities, that this is not the case? In 4 which, then this approach will not be valid. And if it's 5 always the case you are on sound grounds. If it's not, 6 what do we do?
7 MR. TINKLER: You could get a better handle of 8 it by comparing flow rates in the loop. But there was a 9 strong consistency of results seen in these tests. I 10 mean, you see some variation in some of the parameters, 11 but behavior isn't substantially altered or the range of 12 conditions that they investigated.
13 CHAIRMAN KRESS: Would it be relatively easy G
14 to vary that split and see what effect it had in the code, 15 just with the code itself calculating it?
16 MR. TINKLER: See, if you vary it without 17 preserving ~--
18 CHAIRMAN KRESS: I presume you'd have to l 19 preserve the continuity --
20 MR. TINKLER: Well, some other parts of the l
21 heat deposition -- I'm not so sure -- i l
22 MR. ZUBER: T!is is -- ,
l 23 MR. TINKLER: -- what you accomplish in -- )
24 MR. ZUBER: This is a good example of !
rs 25 compensating errors. You see, you have, you make an NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005 3701 (202) 234-4433
78 1 assumption here. You compensate it by another factor in 2 the mixing and therefore if you change it you have to i
'O 3 compensate it, and then the question is, how do we address 4 the uncertainties? Which is really the last question we 5 have to answer.
6 MR. TINKLER: Okay.
7 MR. ZUBER: I think this is also, the 8 compensating error was also brought -- Dr. Lee? The 9 compensating errors was also brought by Dr. Ishii, so I 10 think --
11 DR. LEE: Well, I just want to make sure when 12 you quote his letter it was -- the Westinghouse experiment 13 was written. And what I point out is that we didn't spend 14 much time in presenting the data in details during this 15 review. And Ishii is comfortable about it and he believes 16 that the experiment is well scaled. He brought up some of 17 the things of course, he knows that the RELAP cannot model 18 the countercurrent -- it's a 1D in the hot leg. i 19 MR. ZUBER: Well, that's -- i 20 DR. LEE: That's why he brought out those 21 points that you brought out. l l
22 MR. ZUBER: That's right. This is the only 23 reason I'm bringing it up. ;
1 l
24 DR. LEE: I just wanted to make sure it's !
O Q 25 taken in context. , l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE IS: AND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 l
.0
79 1 MR. ZUBER: Because if CCFL is important and 2 the mixing in inlet plenum is important and the code
-~)
3 cannot do it on a mechanistic basis, how do we do it?
4 Experimental tuning. And then we can amend the 5 coefficient to adjust. And we can leave this if that 6 thing is well scaled and we know what we are doing.
7 DR. LEE: But initially you were raising the 8 question about the experiments that -- experiments
, 9 completely out of --
10 MR. ZUBER: No, I'm asking how it is scaled, 11 because what I see here is not convincing. Maybe there is 12 information somewhere else which will answer my questions, 13 but based on the information which I got from Mr. Dudley, l' \
2 14 we are not there, at least not in my view. And I think 15 this is also a concerned by Dr. Ishii's question.
16 DR. LEE: Novak, these experiments are not new 17 to you, right, because you have seen this one back in I
18 1990/91 when you were working at NRC?
19 MR. TINKLER: Look, I understand your specific 20 concern, okay, and let us take it into consideration. You 21 know, we can --
1 l
22 MEMBER CATTON: I'd like to just -- .
)
23 MR. TINKLER: Your concern basically, is that 24 it's the partitional flow in the hot leg.
(~%
(_) 25 MR. ZUBER: See, the point 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 I l
80 ,
1 MEMBER CATTON: I'd like to just get an answer 7s 2 to a question that may bear on what Novak is talking
(
v
)
3 about. Did you guys do a -- or did anybody, do an 4 uncertainty evaluation of the data?
5 MR. TINKLEP: We have looked at the 6 uncertainty --
7 MEMBER CATTON: Not looked. Has somebody gone 8 through in some kind of detail, where they can say that 9 the bottom line is plus-or-minus "X" percent? I couldn't 10 find it.
11 MR. TINKLER: No.
12 MEMBER CATTON: No. Without that, your 13 comments are right on the mark, Novak.
\ -) 14 MR. ZUBER: Well, as I said, maybe they have 15 done it --
16 MEMBER CATTON: They have not.
17 MR. ZUBER: Well, as I said, starting from the 18 scaling and experiments to the uncertainties.
19 MEMBER CATTON: Not only that, nobody has 20 really done an error analysis of the data. That I know 21 about.
22 CHAIRMAN KRESS: Are you talking about the 23 error with respect to the one-saventh scale test, the CFRR 24 error and when you extrapolate it, you extrapolate --
f)
(_) 25 MEMBER CATTON: Just take one step at a 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
81 1 There are certain questions about the scalings, but put 1
2 them aside for the moment. I 7s N. 3 CHAIRMAN KRESS: Just the error with respect 1
4 to the --
5 MEMBER CATTON: What kind of confidence can I 6 place in the numbers that came from the experiment?
7 Because somehow all this has to be rolled together.
8 CHAIRMAN KRESS: And you're talking about the 9 experiments were used to determine these figures of merit, 10 like the mixing and the --
11 MEMBER CATTON: How good is the --
12 CHAIRMAN KRESS: -- and you're asking how much 13 confidence do you have in the actual measurement for the ex i )
k/ 14 experiments, on those.
15 MEMBER CATTON: Because you see, each time you 16 -- like the anomalies on the tube inlet temperature
. 17 relative to the plenum temperature, nobody seems to know 18 where the rakes were relative to that temperature, nobody 19 seems to be sure which tube it was in. So you really :
20 can't grab hold of it. And it's my view that you either 21 address the uncertainties or you just have to eat the 22 bottom line, l
23 MR. TINKLER: Again, the point of this i 24 viewgraph was to at least alert the committee that this is
/ \
( ,) 25 not the first time we've looked at this general issue of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTGN, D C. 20005 3701 (202) 234-4433 l
82 1 scaling for these tests. Actually, once upon a time
,3 2 everybody thought that the inability to simulate radiation f )
3 heat transfer was the driving shortcoming. Because these 4 tests are run at 150 degrees C. We have 200 degrees K.
5 Obviously, radiation effects at 150 degrees C 6 aren't the same as the tests at 1500 degrees K, though 7 there was concern the radiation heat transfer that was 8 ignored in the -- not ignored, but the fact that it 9 couldn't be simulated in the tests. And you would get 10 radiation between the two streams in the hot log, for 11 example, and you'd get radiation to the pipe in the hot 12 leg.
13 We did look at that and we did show that, at
,e~\
k least for the SF., the convective heat transfer was w/ 14 15 simulated well. And additional analysis were performed to 16 determine the impact of thermal radiation on the heat 17 transfer And we showed that it would increase the heat 18 transfer coefficient. But the principal effect is to 19 cause the hot leg, or surge line, to fail earlier, because 20 that's where the biggest effect would be seen, j 1
1 21 So ignoring radiation in the hot leg, surge 22 line, causes us to conservatively underestimate, or biases 23 our calculation to underestimate time interval between hot 24 leg and surge line, and steam generator tubes.
O)
(, 25 MEMBER FONTANA: And I guess radiation between 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 the steam generator tubes, because there's so many of
- 2 them.
3 MR. TINKLER: Well, the temperature difference 4 between the fluid and the tubes --
5 MEMBER FONTANA: No, I mean from the outer 6 surface of the tube.
7 MR. TINKLER: Right.
8 MEMBER FONTANA: And it will radiate the other 9 tube. The fact that there are so many of them, this self-10 sealing effect probably --
11 MR. TINKLER: Yeah. I mean, and the -- any 12 radiation effects on that side are going to be smaller 13 because the temperature differences are much smaller than f' s s- 14 they are in the hot leg. Again, the original development 15 of the model was based in part on COMMIX calculations that 16 were sponsored by EPRI. Because in the very beginning, we 17 didn't have -- the NRC didn't have access to some of the 18 data, so there were COMMIX calculations that were done for 19 some of the low pressure tests.
20 Using this same split as we described for the 21 hot leg, those sorts of things, a COMMIX Surry plant model 22 was developed.
23 MEMBER CATTON: Do you know anything about the 24 nodalization that was used in the COMMIX calculations?
(3
(_) 25 Was convergence testing done?
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 l
l
[ 1 MR. TINKLER: I don't know that it -- I don't
,m 2 know the answer to that question.
( )
3 MEMBER CATTON: Because the COMMIX code and 4 its use are infamous in their ignoring numerical diffusion i i
1 5 problems.
6 MR. TINKLER: I know that numerical diffusion 7 is always an issue for any of the finite difference codes.
8 MEMBER CATTON: That's right.
9 MR. TINKLER: And it was a concern at one 10 point for some of the calculations.
\
11 MEMBER CATTON: Well, in the PTS calculations 12 done using COMMIX, the numerical diffusion was greater 13 than turbulent diffusion. As a matter of fact, with or l lf p)
C 14 without turbulence, the answer was the same. And in this .
1 1
15 particular problem, your interest is mixing, which means 16 you need extremely fine nodalization. And that's 17 expensive.
18 MR. TINKLER: Well, the --
I l
19 MEMBER CATTON: And before you can base 20 anything on the results that came from COMMIX, I think 21 that question needs to be addressed.
22 MR. TINKLER: Again, the COMMIX calculations 23 weren't done to predict with precision mixing in the inlet 24 plenum, okay? They were done to calculate heat removal,
\. 25 heat deposition around the system.
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 MEMBER CATTON: Yeah, but see, you're making i 2 the point here that high powered CFD is used to benchmark 3 RELAP; therefore, we're to believe RELAP. But without 4 giving us a -- some arguments as to why these particular 5 calculations using COMMIX have some pedigree, I think the 6 comparison doesn't say anything.
7 MR. TINKLER: Well, let me -- let's just go a 8 little further, okay? In order to benchmark the code 9 against the COMMIX full plant calculations, it did require 10 an adjustment of inlet plenum mixing; but there's three 11 volumes that represent the inlet plenum in a RELAP model.
12 We did have to adjust those. We didn't distort them 13 beyond the total volume of the inlet plenum, but we had to
\- I 14 adjust their relative sizes, flow area split and loss 15 coefficients.
16 We didn't force a match of recirculation flow 17 rates for mixing fractions. What we got -- what we 18 matched was heat removals in -- heat removal, heat 19 deposition in the hot leg and steam generators from the 2C COMMIX calculations. When we got a reasonable agreement 21 on those values, then we froze these loss coefficients, 22 mixing volumes and flow area splits.
23 We then took that model that was frozen, if 24 you will, and went back and assessed one of the -- some of
(~N (s,) 25 the 1/7 scale test data, and it was assessed against two 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 of the high pressure SF6 tests, S-6 and S-7. And we n 2 believe that those calculations show a reasonable 3 agreement with the test data.
4 We calculated the hot leg and steam generator 5 mass flow rates within 11% of measured values, and we ,
1 6 calculated vapor temperatures in the steam generator tubes 7 within 5%.
8 I just have this in your package as some 9 indication of the calculation. We don't produce exact 10 match throughout the entire core region or in the hot let, 11 but in the inlet to the core. You can see it --
12 calculated the values of 137 and a measured value of 134.
13 We under predict a little bit the cold half of the hot
/%
14 leg.
15 And I would argue in part that if we weren't -
16 - if we're calculating these temperatures pretty closely, 17 we're calculating the split between the hot -- between the 18 cold stream in the neighborhood. We couldn't match these j 19 temperatures if we were way off of those things. )
l 20 Same sort of thing for S-7, and you see the 2 .' <3ame kind of thing. Now when this model was developed i
22 back in --
23 DR. ZUBER: Is the density ratio for the SF. l l
24 the same? )
73 h 25 MR. TINKLER: Density ratio?
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 i
1 DR. ZUBER: Rho hot minus rho cold over rho j
~ 2 cold. See, because it's the same --
(#~
3 MR. TINKLER: Pretty close, pretty close, 4 yeah. We did develop -- well, I don't have it on --
5 actually, I have it on -- I'll show you a little later on 6 a temperature basis. I don't have it on a density basis, 7 but it's pretty --
8 MEMBER CATTON: How well did you do in l 9 comparing to the transients that were run? This was a 10 steady run.
11 MR. TINKLER: We'll come clean and admit that 12 the calculations were benchmarked agains: steady state 13 data because it was -- they have much better heat
/~hi N/ 14 balances. The transient data -- if you go back and look 15 at the data, you're forced to pick points off a transient.
16 I don't want to point to it here, but you're forced to 17 pick points off a transient measurements.
18 And because the structures represent such a 19 large heat sink, your ability to calculate and measure the 20 heat losses to the heat sink in the transient test, the 1
21 margin of error uncertainty on the heat balance is much I 22 greater. So the steady state tests were used. !
23 MEMBER CATTON: Well, there's also the things 24 like where the thermal center is is a function of time b
y ,/ 25 because it's certainly going to shift and how this plays a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE. N W.
(202) 234-4433 WASHINGTON. D C 20005-3701 (202) 234-4433
88 1 role in feedback to the actual flow rates. There's just a
~ 2 whole myriad of kinds of questions that one could ask.
3 MEMBER POWERS: When I look at the comparisons 4 between experiment and the calculations, they're very good 5 at the hot leg. They seem to be quite well elsewhere. Is 6 that just because for this particular calculation there's 7 been no tuning to that elsewhere?
8 MR. TINKLER: Well, you know, we weren't 9 trying to match temperatures at various axial locations in 10 the core. That's just too much to -- I don't know what we 11 would achieve. We're just trying to get the general 12 deposition of energy around the system correct. Now, we 13 over predict in some parts of the down flow paths and G
-- 14 under predict in up flow paths in various places; but no, 1
15 there wasn't an attempt to try to match those.
16 MEMBER POWERS: And these parts of the problem 17 just aren't significant?
18 MR. TINKLER: No, not at the end of the steam i
19 generator.
20 MEMBER POWERS: The vessel is the boundary 21 condition on the problem, and the scaling of the vessel is 22 inappropriate. The tubes are large, which means that the l l
23 horizontal direction friction factor relative tot he 24 vertical friction factor's probably not matched. And when ;
tQ k_) 25 you -- what's the impact of that? You know, there's a lot l
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 l of work that's been done on natural convection and tube l 1
- 2 bundles.
3 CHAIRMAN KRESS: It depends on how much the 4 pressure drop in the core has compared to the pressure on l
5 the rest of the system.
6 MEMBER CATTON: Well, it turns out that the 7 recirculation in the vessel, at least I think from looking 8 at the data, is kind of uncoupled from the rest of the 9 problem because you have really good mixing in the upper 10 part of the head, you know, where these things connect.
11 It's pretty good mixing.
12 CHAIRMAN KRESS: But you still have to drive 13 the flow through it even though there's mixing there, so -
/
s 14 -
15 MEMBER CATTON: I understand. What drives the 16 flow through the rest of the system is relative 17 temperatures in the head and in the steam generator tubes.
18 Because the steam generator tube is the other end of the 19 problem. But in the vessel itself, the kind of 20 recirculation you're getting in the core is going to 21 determine the temperature in the upper regions.
22 Now, there's lots of other sources of data one 23 could use to see whether or not you've got things done 24 right in order to match this part of the problem, and I
(_j) 25 was just asking if this was done.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASH!NGToN. D C. 20005-3701 (202) 234-4433
90 1 MR. TINKLER: Well, as I said, the original
,3 2 scaling considered flow --
/
%>) 3 MEMBER CATTON: I understand the original 4 scaling. I'm asking did you make any comparisons of how 5 well RELAPS/SCDAP would deal with just recirculating flow 6 in an intact core? Has that comparison ever been made?
7 And you get this big cell. The center part is rising and 8 then the outer perimeter -- it's falling. And the -- what 9 determines the velocities is going to be the relative 10 friction factors as well as their actual --
11 MR. TINKLER: I don't know that that specific 12 analysis has been --
13 MEMBER CATTON: Well, see, then you have to
\~ ' 14 ask yourself okay, what do I do to the temperature in the 15 head that becomes the boundary condition to the rest of 16 the problem in order to be sure I'm treating the 17 uncertainty in what I have properly?
18 MP TINKLER: I understand. 1 19 MEMBER CATTON: Okay.
20 MR. TINKLER: But I mean, I'm just saying that 21 some of these temperatures are immediately below the 22 elevation of the hot leg nozzle. They're not right on, 23 but they're not that far off. And if you take some -- you 24 know, I haven't looked -- well, we haven't looked at r~~
k_,N) 25 combined averages of these temperatures above -- at NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHoDE ISL.AND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
91 1 different elevations above and below the upper plenum.
2 MEMBER CATTON: If I just reduce the
) i
[os 3 recirculation velocity a little bit, the temperature in 4 the head will go up.
5 MR. TINKLER: Sure, sure.
6 MEMBER CATTON: And then the question is, 7 okay, how close are you to -- and how well are these 8 things being modeled? How well they're being modeled is 9 going to depend on how well you have selected the various 10 friction factors, and in particular, channel to channel in 11 the core -- how well were they picked and how well were 12 they validated?
13 If you don't get them right, you don't get the
,/ 3 i s
\/ 14 recirculation right. If you don't get the recirculation 15 right, you don't get the temperature right. And that's 16 the heart of your problem.
17 MR. TINKLER: This is largely an integral 18 benchmarking.
19 MEMBER CATTON: I understand that. But you're 20 benchmarking it against data that wcs inappropriately 21 scaled. Now if you did that, how well does it scale?
22 MR. TINKLER: Well, I guess --
23 MEMBER CATTON: And you could answer that 24 question by comparing with other data to say yeah, what we
[~h 25 have in here is just fine.
(,)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 I
92 1 MR. TINKLER: We're not -- we don't conclude
,- s 2 that it was inappropriate to scale.
(
~'
)
3 MEMBER CATTON: I'm not suggesting you do 4 this. All I'm trying to do is to bring out where the 5 uncertainties lie. Because the question for the licensing 6 people is what to do with them.
7 MR. TINKLER: I understand. This isn't in 8 your package, but I want to put this up because we've 9 talked a lot about this. This is a description of some of 10 the test data, the test data that we primarily rely on.
Il These are the tests with the steam generator thermocouple 12 rakes. These are mixing fractions and recire ratios that 13 are inferred from the tests. I 3
-- 14 These -- this represents the fraction of tubes 15 carrying hot flow and cold flow. Now, there are steady 16 state tests and transient tests. The steady state tests I 17 actually have cooling on the secondary side of the tubes 18 in the steam generator. Transient test is a locked open 19 valve on the steam generator. It's dry.
20 It's actually quite comparable to the 21 situation we have in the station black out with a boiled 22 dry steam generator. And if you look at this data, you'll 23 see that the number of hot tubes relative to the total 24 number is about 35%. If you look at this data, the (N) 25 transient data, where there's diminished heat removal NEAL R. GROSS CoVRT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234 4433
l 93 l 1 capacity on the secondary side, the hot tubes -- the l
2 fraction of hot tubes carrying flow is on the average of i k'~'Sl 3 53%.
l 4 Now you heard other calculations that use a l
5 50/20 split in the number of tubes. Our calculations use i
I 6 35. That actually biases us to calculate higher tube l
7 temperatures. But we think -- and we agree to a certain 8 extent that this data here is actually closer to the TMLB !
9 prime accident scenario of interest. But recognizing I
10 there's uncertainty in all of this, we've benchmarked our 11 model against the steady state data.
12 Here's the range of mixing fractions. It's 13 .76 on the low end and .89. We think there's some trend p-14 where the transient data may be a little lower, but 15 somebody could look at those same numbers and conclude 16 they're all the same. Reasonable people could come to 17 that opinion. And recire ratios hover around two.
18 Well, I'm just saying -- and it's not clear 19 they change independently either, okay? But we've talked 20 about it --
21 DR. ZUBER: How would you convince somebody 22 really that using the steady state approach is kosher?
23 MR. TINKLER: I would argue that because the 24 steady state tests have a better energy balance (3
(,) 25 themselves, it's -- you can reduce some of the uncertainty 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 in the comparison -- fundamental comparison. I would g-ws 2 further argue that using 35% of the tubes carrying hot
() 3 flow is a bias pushing you towards higher tube 4 temperatures.
5 But if you want to keep considering extremes 6 on mixing fractions and recire ratios, then you need to 7 reconsider whether or not you use that 35% of the tubes 8 carrying hot flow or go up to the 50%.
9 DR. ZUBER: See, my argument is that you can 10 use the mixing and the recirculation based on steady 11 state. I mean, this is --
12 MR. TINKLER: Well, I understand. But you 13 know, look at these numbers. There's not -- these aren't p,
A. 14 profound differences in numbers.
15 DR. ZUBER: See, to answer a question from an 16 SOB like me would be -- I can determine that the time 17 responce of the different processes ensure steady state is 18 okay.
19 MR. TINKLER: Well, let me -- I'm going to 20 show you some stuff late in the presentation that gets to 21 some of these kinds of issues. I don't want to promise it 22 will address them, but --
23 DR. ZUBER: The question is -- see, the 24 question like this can be answered. And in a really
(%.
(,,) 25 technically defensible and rational way, you can say the NEAL R. GROSS )
COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
95 1 response of the system is for this moving of the center of g3 2 gravity, these --
wY 3 MR. TINKLER: Right.
4 DR. ZUBER: -- so many seconds, they started 5 responding so many seconds. You cannot use this. We know 6 -- convince me, yes. This is a reasonable, technical 7 approach. And you can defend it in front of this 8 committee or in front of intervenors.
9 MR. TINKLER: I understand. But also, keep in 10 mind that t he transient data in many cases, we're forced 11 to pick from a sparse number of data points that are 12 available to us, okay? So we're taking a many thousand 13 second transient and we have two or three data points.
O 4
(
\/ 14 And we know these things may vary a little bit. The 15 number of hot tubes carrying flow goes up, it goes down; I
16 it is constantly shifting. I 17 Not wildly, but it does up and down. These 18 are average kinds of numbers.
19 MEMBER CATTON: If you want to make a risk 20 kind of statement, you somehow have to roll this 21 uncertainty that you're expressing into your answer.
l 22 MR. TINKLER: Well, we think that this is one 23 of many.
24 MEMBER CATTON: Of course it's one of many.
(O_/ 25 Of course it 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
96 1 MR. TINKLER: And --
73 2 MEMBER CATTON: Absolutely.
'v)
I l l
> 3 MR. TINKLER: Well, let me -- let's go through 4 the studies that we did do. But I wanted to show that to l
5 you just to give you a sense of the variation in the 6 tests.
i 7 MEMBER FONTANA: Since we're going to run over 8 our allotted time anyway, wouldn't now be the time to take 9 about a ten minute break and get into what you were going 10 to just cover next -- discuss?
11 MR. TINKLER: It's up to the committee. I'm 12 prepared to go ahead, or if you would like a break, --
13 MEMBER CATTON: Let's take a break now.
s s- 14 MEMBER FONTANA: Let's take a break now and --
15 MR. TINKLER: I made sure I didn't challenge 16 it this morning.
17 MEMBER FONTANA: How about 15 minutes, ten 18 minutes?
19 (Whereupon, the foregoing matter went off the 20 record at 10:32 a.m. and went back on the 21 record at 10:46 a.m.) i 22 CHAIRMAN KRESS: We'll just continue on with 23 Charlie and he can --
24 MR. TINKLER: Did I -- I'm not sure I covered
( ) 25 this viewgraph, but -- l 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
97 1 MEMBER FONTANA: No, you hadn't. That was the 2 next one.
,7~
3 MR. TINKLER: After having developed the model 4 and benchmarked it against the -- or at least confirmed it i i
1 5 against the 1/7 scale test data, a number of sensitivity 6 calculations were done. We varied the axial power profile 7 to simulate end of life and beginning of life in the core. !
8 We varied core and upper plenum cross flow resistances, 9 hot leg and surge line piping heat losses, heat transfer 10 coefficients in the upper plenum hot leg and steam 11 generator tubes, and simulated radiation heat transfer 12 between the hot leg and -- hot leg flow streams -- hot and I
1 13 cold streams in the hot leg.
(%
\-- 14 These are described in NUREG/CR-5214. I think 15 they're also described in the 1993 NUREG from INEL. And 16 basically these comparisons were done for a TMLB prime 17 type calculation without the depressurized secondary side.
18 And in all cases, the ex-vessel piping was predicted. The 19 hot leg or surge line was predicted to curve early in the 20 transient. Usually what you see is the calculations 21 predict the hot leg or surge line after the onset of 22 significant oxidation of the cladding.
23 The effect of the sensitivity analysis was to 24 only marginally alter the service line or hot leg failure
.g
. \ ,) 25 time by one to two minutes for virtually all tnese NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE. N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
I 98 ,
I calculations shown here. It was very little effect. In !
2 some cases, when we increased heat transfer, we caused the 73 3 hot leg to fail a little sooner. But they were not 4 significant sensitivities.
5 Heat transfer coefficients were varied by 20%
6 -- 10-20%.
7 CHAIRMAN KRESS: How much did you vary the 8 heat loss on this hot leg piping?
9 MR. TINKLER: Well, basically a lot of our 10 calculations are done adiabatic. We varied it 11 substantially assuming no insulation.
12 CHAIRMAN KRESS: So you got down to no 13 insulation on it?
(m'-) 14 MR. TINKLER: Yeah, we got down -- now, that ,
1 15 slows it down, you know, on the order of about five l
16 minutes, something like that, I think. l 17 MEMBER FONTANA: Now this hot leg piping --
18 this piping failure occurred before the steam generator i 19 tubes failed?
20 MR. TINKLER: In these calculations, the tubes l l
21 don't even come close to --
22 MEMBER FONTANA: Not even close?
23 MR. TINKLER: Well, you know, if you do real 24 extreme calculations, you can; but in these calculations,
/
(m,\/ 25 tubes weren't close. But this is not a depressurized NEAL R. GROSS CoVRi REPORTERS AND TRANSCRIBERS t323 RHODE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
99 1 secondary side.
2 MEMBER FONTANA: And these were done in 1988?
f~)
w-l 3 MR. TINKLER: These go back to '88. Some of 4 these were actually a little later, I think.
5 MEMBER FONTANA: When were the SF. tests done 6 at Westinghouse?
7 MR. TINKLER: In that same time.
8 MEMBER FONTANA: About the same time frame?
9 Those results are fed into this?
10 MR. TINKLER: I think these analyses were done 11 simply to look at what kinds of uncertainties are 12 associated with these parameters, not explicitly 13 addressing uncertainty in tests.
,,.s
- T 14 MEMBER FONTANA: Yes.
i 15 MR. TINKLER: Heat transfer coefficients can l 16 vary easily by 20%.
l l
17 MEMBER FONTANA: So that's, you know, i 18 obviously a real important question to us if surge line or 19 something like that creak -- the upper parts of the system 20 break before the steam generator tubes.
21 MR. TINKLER: But these were done without the 22 additional aggravating assumption of the ADV sticking 1
23 open. I'm going to start talking about the more recent i 24 calculations. And just for your -- just for a point of s_j 25 reference, the calculations I'm going to describe 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
100 1 done for the Surry plant. TMLB prime transient station 1
g- 2 black out where the immediate loss of aux feed water --
b) 3 these calculations, for the most part, were done with 35%
4 of the tubes carrying hot flow.
5 It turns out that we predicted a mixing 6 fraction of about .87 and a recirc ratio of about 1.9.
7 The calculations were done with and without the secondary 8 depressurization, and we did sensitivity analysis to 9 address the inlet plenum mixing, the hot and cold tube 10 split.
11 MEMBER FONTANA: This is recent. When were 12 these done?
13 MR. TINKLER: These were done in the last
/'N
- ' i t
14 couple of months.
15 MEMBER FONTANA: Okay.
16 MR. TINKLER: And just for your -- there's a -
17 - this mixing fraction and recire ratio, you can see i
1 18 they're pretty close to the experimental data. It wasn't 19 tuned for those particular numbers. And this is the 1
20 nodalization that was adopted for the recent Surry 21 calculations, a split of the hot leg and nodalization up 22 through the pressurizer of the PORV.
23 I'll talk a little later about treatment on 24 the secondary side, because this issue was raised in j%
( )
(_/ 25 Chicago during our independent peer review. Our modeling 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 of the secondary side of the steam generator in a 1-D 73 2 fashion -- this nodalization won't allow for recirculation b 3 on the secondary side of the steam generator which in fact 4 you would expect to occur because there's lots of cold 5 steel up high in the steam generator.
6 All the separator equipment and everything up 7 there -- lots of tons of steel. So you could reasonably 8 expect that you would get a circulatory pattern on the 9 secondary side of the steam generator that would further 10 serve to cool the tubes. And this nodalization doesn't 11 really give you much credit for that because it stays 12 pretty hot.
13 You don't get much circulation into the upper I
\ 14 part. These are just general sorts of system pressure 15 curves. It doesn't take too -- you boil off -- you boil 16 dry the steam generators around in here. So then you 17 cycle at the PORV set points. You get water discharge 18 through the PORV here, and then you resume cycling through 19 the PORV, so you sit at about 16 MPa.
! 20 One of the reasons for showing this -- in 1 21 part, we're -- you know, we're limited by how much heat we 22 can get out of the core, and then part of that's due to i
23 the fact that we get the water level below the bottom of l
24 the fuel pretty soon in these calculations. So there is
'Q
( j 25 something of a shortage of steam to further aggravate 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
102 1 by the nature of the sequence.
2 Case one is the short term station black out
\
(vs 3 with the secondary side remaining at the -- at its relief 4 valve set point pressure. There is no depressurization of 5 the secondary side. We get surge line failure shortly 6 before 14,000 seconds. In this calculation, we never did 7 fail the steam generator tubes. We ran the calculation 8 out to 19,000 seconds, and they hadn't failed yet.
9 CHAIRMAN KRESS: What is your failure criteria 10 again?
11 MR. TINKLER: This is Larson-Miller.
1 12 CHAIRMAN KRESS: You add up the -- !
13 MR. TINKLER: Cumulative damage index. The A
> 14 code calculates the cumulative damage index. When it gets 15 to one, it seems to fail. Now we did these calculations 16 allowing the pipe to fail and ignoring the predicted 17 failure just to see what would happen.
18 CHAIRMAN KRESS: Just kept going?
1 19 MR. TINKLER: Just kept going. '
20 CHAIRMAN KRESS: Just didn't let --
21 MR. TINKLER: Well, we also did some 22 calculations where we allowed it to fail and allowed it to 23 depressurize just to see what would happen because there 24 was a question --
25 CHAIRMAN KRESS: This looks like it ignored 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 the failure here.
,m 2 MR. TINKLER: Yes, yes, this one ignored it.
( I L/
3 Case two, which you're not going to -- which you won't see 4 -- actually, you won't see a plot, I have it in a table --
5 allowed the failure to occur. Because there was some 6 concern, well, suppose it failed, would the tubes continue 7 to heat up?
8 CHAIRMAN KRESS: Yes.
9 MR. TlNKLER: Well, they don't because all the 10 flow goes out the hole.
11 MEMBER CATTON: Pressure drops and --
12 MR. TINKLER: Well, the pressure drops too.
13 CHAIRMAN KRESS: Is there any consideration of i .,
14 degrading the steam generator tubes, wall thinning, or l
15 anything like that in the --
16 MR. TINKLER: Well, as part of NRR's overall 17 assessment, they look at the effect of flaws. Okay, now 18 these calculations that I'm showing you are from non- )
i 19 flawed tubes. Okay, but then they -- NRR will use these 20 thermal hydraulic transients and do stand alone 21 calculations with different stress assumptions and flaw 22 distributions. l 23 These are non-flawed tubes. This one, as I 24 said -- this calculation was continued for another hour or
/ )
( ,/ 25 so and the tubes didn't -- when you depressurize the i l
NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
104 1 secondary side of the steam generator, you reduce its heat 73 2 capacity. And you'll see the steam generator tubes track (v) 3 the hot leg much more closely.
4 And steam generator tube failure for case 5 three, it's the same as case one, except the secondary 6 side on the pressurizer loop is allowed to depressurize.
7 Steam generator tubes are predicted to fail right about 8 here. So between here and here is about 20 minutes. And 9 what you see -- what happens here is this is when 10 oxidation really begins in earnest.
11 You get lots of hydrogen generation starting 12 about this time, so this drives this up.
13 CHAIRMAN KRESS: A lot of heat and a lot of --
\ 14 MR. TINKLER: Yes.
15 CHAIRMAN KRESS: What turns that steam 16 generator tube over there?
17 MR. TINKLER: Turn it over?
18 CHAIRMAN KRESS: Yes.
19 THE WITNESS: We oxidize all the available 20 cladding with the available steam we have.
21 CHAIRMAN KRESS: And why doesn't it continue l
l 22 to track the hot leg?
23 MR. TINKLER: Well, the hot leg's got a lag in l 24 it because the fluid temperature is much higher than the (D
(_) 25 wall temperature. So it continues to go up for a while.
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 The delta T between the fluid and the wall for the tubes e is very small because the tubes are so thin. So as soon 7-
%.) 3 as the fluid temperature starts to turn over, the wall
\
4 temperature turns over.
5 In the case of the hot leg, because the hot 6 leg is already lagging the fluid temperature by several 7 hundred degrees, --
8 CHAIRMAN KRESS: I see.
9 MR. TINKLER: -- it doesn't turn around right 10 away. ,
I 11 CHAIRMAN KRESS: It doesn't turn around as )
i 12 fast.
13 MR. TINKLER: It will -- it would eventually
' ( ~~\
14 turn around, but it won't turn around right away. Well, 1
1 15 basically the conclusions of this base case calculation !
1 16 was that the pressurizer surge on a creep rupture was the i l
17 first RCS pressure boundary. It fails early in core 18 damage space. Fuel melting doesn't occur for another l 19 several thousand seconds. So there isn't any fundamental 20 reconfiguration of the core.
21 If the pressurizer surge line hot leg failures ]
22 are ignored and a secondary side ADV fails open, then tuber 23 rupture occurs roughly 20 minutes later. This is a plot 24 just to show -- just to give you a feel for the fluid
[(_)h 25 temperatures around the loop. Actually, what's coming up NEAL R. GROSS CoJRT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
WASHINGTON D.C. 20005 3701 (202) 234-4433 l (202) 234 4433
106 1 through the center of the core is about 25, 26 hundred 2 degrees Kelvin.
p 3 Mixing in the upper head, upper plenum region 4 produces a hot stream of about 1700 degrees K, 1660. This 5 is at the time of surge line failure. That fluid 6 temperature is reduced another 100 degrees by the time it 7 gets down to the end of the hot leg. This 1600 degree 8 steam mixes with returning cold steam from the cold steam 9 generator tubes at about 800 degree K to produce a bulk 10 average temperature entering the hot tubes of about 1000 11 degrees.
12 And you can see the temperature drop around 13 the steam generator. But it's -- this also illustrates O
V 14 why the steam -- why the hot leg doesn't turn around quite 15 as quickly. You can see this is 1660 fluid temperature.
1 16 The hot leg itself is still at 1250. Hot leg and the 1
17 surge line fail and the cumulative damage index is one at 18 about the time the wall temperature reaches an average of 19 about 1250.
20 In the case of the tubes, it's about 1150. So 21 when you look at absolute temperatures in the steam 22 generator, it's important to remember that it's a lot 23 hotter up here. You know, if it wasn't for the fact that 24 this is two and a helt inches thick, this would fail a (q) 25 whole lot sooner.
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 But the fluid temperatures decrease 2 dramatically as you go down the loop. And that's -- now,
,S
(_/' 3 you can do calculations that will alter the tube 4 temperature, but you're still going to preserve these 5 large delta T's.
6 MEMBER CATTON: Did you connect the metal from 7 underneath the 1600 tube to the metal under the 933?
8 MR. TINKLER: No.
9 MEMBER CATTON: It's two and a half inches 10 thick. Three and a half inches thick? And a distance of 11 only like this? Do you think you could maintain those 12 temperatures? I don't think so.
13 MR. TINKLER: Well, you expect that conduction Y' 14 would tend to --
15 MEMBER CATTON: Would lower the high one and 16 raise the lower one.
17 MEMBER POWERS: Tom, with those kinds of 18 temperature differences, you probably get a photophoretic 19 force on the --
20 CHAIRMAN KRESS: Yeah, those are higher than I 21 thought.
22 MEMBER POWERS: Yeah, that's a huge 23 temperature differential.
24 CHAIRMAN KRESS: It's a lot higher than I was
/'N
(_,) 25 thinking. Of course, that's on the average. It's not NEAL R. GROSS COURT REPORTERS AND TRANSCR:BERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
108 1 right at --
f-m 2 MR. TINKLER: At the interface it's going to (s_/)
3 be a little lower.
4 MEMBER FONTANA: ju probably already answered 5 this question, but those three little boxes at the inlet 6 of the steam generator, one says 1595, one says 1022, one 7 says 795 -- again, what were the flow splits, what were 8 the assumptions that went into the --
9 MR. TINKLER: Roughly 87% of the flow is mixed 10 in here. And this flow rate, the 795 flow rate, is about 11 twice that of the 1594. The recirc ratio is about two.
12 You could do a hand calculation and predict this 13 temperature. You don't need a code really.
iQ
-- 14 MEMBER CATTON: That's right.
15 MR. TINKLER: I mean, if you take this number 16 times two and weight it against this number and give a 17 mixing fraction of .87, it's that number there to a large 18 degree.
19 DR. ZUBER: What is the meaning of mixing that 20 would really prevent the rupture of the tube?
21 MR. TINKLER: The minimum mixing?
22 DR. ZUBER: Yes. Because I think you had 23 before two.
24 MR. TINKLER: Well, you have to remember now
(~
(_j) 25 it depends on whether or not the tube's get a flaw in 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
109 1 It would be a whole family of curves given a flaw.
e 2 DR. ZUBER: Give me a range. I think you O 3 raise your argument it was two, right? One of these 4 previous viewgraphs, --
5 MR. TINKLER: Two?
6 DR. ZUBER: -- the mixing was two. Mixing 7 ratio.
8 MR. TINKLER: Yeah, recirc ratic about two.
9 Well, you know, it's the recirc ratio in conjunction with 10 the mixing fraction. Well, anything -- you know, these 11 tubes fail when they get to about 1150. You increase this 12 another 150 degrees and you'11 start getting close the 13 challenging the tubes.
(3 14 DR. ZUBER: My question is how much error I 15 must make in the recirculation in the mixing?
16 MR. TINKLER: Well, like I say, it's a 17 combination of the two. I haven't -- we haven't sought to 18 see wnat was the limiting case from that. What we do are :
1 19 calculations --
20 DR. ZUBER: See, but this gives you the 21 sensitivity really.
22 MR. TINKLER: We do sensitivity calculcitions. j 1
23 We don't do them to see what's the worst -- what's the l 24 worst condition we can tolerate and not fail --
f'h.
h 25 DR. ZUBER: See, but this is your input to the l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 rho 0E ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
110 1 code which depends on your experiments and scaling. And 2 if you say this is not very sensitive, fine. If it's very 7S U 3 sensitive, then you have to look into it. And my question 4 is how close you have to be in order to be at the 5 precipice.
6 MR. TINKLER: I understand your question.
7 DR. ZUBER: This is -- all your calculations 8 are based on these two tuning factors.
9 MR. TINKLER: But I showed you the 10 experimental data. Now, I know there are questions. I 11 know that there are questions that rely -- that are 12 related to that, but those parameters that we selected --
13 no, no, we didn't select them. That's the result of the
( )
14 calculation using these flow areas and loss coefficients 15 produces those kinds of -- produces that mixing fraction 16 and that recirc ratio, which is quite consistent with the 17 experimental data.
18 The other approaches would predict an even 19 higher recirc ratio. So -- but we haven't done the 20 calculation to see how bad could we make the combination 21 of the mixing fraction and recire ratio and still show -- l 1
I 22 because part of the problem is then you're chasing -- and l
23 then it depends on what flow distribution you assume for 24 the tubes.
C' '
(_)N 25 MEMBER CATTON: No, these are coupled. You NEAL R. GROSS COURT REPORTEiiS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
111 1 need to look at both. The people who look at the tubes
,_. 2 come up with a probability distribution of failure for a
3 given temperature. They also need to marry in the 4 probability of a certain temperature distribution.
5 They put the two together and they come to a 6 bottom line. If you just do one probablisticaly and the 7 other not, then I don't know what you've got. They 8 incorporate the uncertainty into the materials end of it.
9 You should also be incorporating the uncertainty into your lo estimates from the thermal hydraulic side.
11 MR. TINKLER: Well, we think we do incorporate 12 the uncertainties. Maybe not in the same fashion. This 13 is a more traditional approach to incorporating fs t i N/ 14 uncertainties.
l 15 MEMBER CATTON: How can you marry apples and 16 oranges? They to it one way, you do it another.
17 Shouldn't they both be done the same if it's to have la meaning?
19 MR. TINKLER: I would only suggest that this l l
20 is -- in this overall assessment, there are -- each and 21 every decision point is not accompanied by a full blown 22 probablistic treatment. Okay, there are -- you know, you 23 weigh where you think you need --
24 MEMBER CATTON: The number of parameters that n
() 25 are uncertain in this context are few. The implication 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 that there are -- that it full blown implies something I 2 think quite diffegent. There are a few.
4
\' ,/
3 MR. TiNKLER: Well, there are a few with 4 respect to inlet plenum mixing, but there are other 5 uncertainties in the sequence.
6 MEMBER CATTON: Well, if there are others and 7 they're large, then they somehow should be rolled into 8 this as well.
9 MR. TINKLER: Well, let me tell you --
10 CHAIRMAN KRESS: Charlie, before we take that j 11 slide off, I wanted to --
l 12 MR. TINKLER: Okay.
13 CHAIRMAN KRESS: You might want to aduress O
b 14 that. I'm still concerned about Ivan's thought that 15 c;rcumferential heat transfer wasn't included in the hot 16 leg which would shift your likelihood to failure towards 17 the steam generator tubes if it were considered. Is there
. 18 some way you can une time constants like those Novak 19 mentioned in terms of circumferential heat transfer versus 20 transport of the heat down the tubes to allay that 21 concern? I don't know what effect --
22 MR. TINKLER: Yes, I understand the question.
23 It should be relatively -- those time constants should be 24 relatively large actually.
Cm 25 MEMBER CATTON: You could do --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIDERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234 4433
113 1 MR. TINKLER: But again, we're ignoring the 2 radiation heat transfer in this calcu,lation.
73 s
~~ ) 3 MEMBER CATTON: But to address your question, 4 Tom, if you just assumed that it's a constant transfer 5 coefficient, you could ask yourself what's the heat 6 transfer and then compare it with the hez.t transfer from 7 the hot gas.
8 CHAIRMAN KRESS: Yes.
9 MEMBER CATTON: And if they're comparable, 10 then you know that you're going to reduce the temperature 11 significantly. If it's down in the noise, one, two, three 12 percent, then you forget it. But -- so it's a trivial 13 thing to do.
\
k-) 14 MR. TINKLER: There are also issues --
15 CHAIRMAN KRESS: That would be well worthwhile 16 because that's -- if you don't do that, you're biasing the 17 failure in favor of the hot leg.
18 MEMBER CATTON: Right.
19 MR. TINKLER: But the temperature difference 20 across the pipe would also create additional stresses in 21 the pipe.
22 CHAIRMAN KRESS: No, sir; not included in your 23 Larson-Miller though.
24 MR. TINKLER: No, no, not at all. And we do
,O
( ,/ 25 Larson-Miller calculation assuming that temperature. Or NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(?O2) 23<-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
114 1 not that temperature, but the average wall temperature.
,m 2 If we put these kinds of --
Ys 3 CHAIRMAN KRESS: Excuse me, you average the 4 wall temperature to do the Larson-Miller? You don't use 5 that 1660?
6 MR. TINKLER: No, we use the temperature in 7 the upper half. We use an average wall temperature of the 8 upper -- of the hot path.
9 CHAIRMAN KRESS: Okay, sorry. Okay.
10 MR. TINKLER: But we would -- you would 11 introduce additional stresses in the pipe by having such a 12 temperature gradient across it. So if you just --
13 MEMBER CATTON: So then the future gradient O
U 14 we're talking about is circumferential?
15 MR. TINKLER: Well, yes.
16 MEMBER CATTON: But if it introduces 17 additional -- somehow you have to come to come conclusion 18 about uncertainties associated with ignoring one versus 19 the other.
20 MR. TINKLER: Versus the other. It is 21 believed though that the temperature -- that that i
22 effective conduction -- I mean, what really drives thic (
23 fluid temperature up is the oxidation transient in the 24 core -- the oxidation of cladding, that fluid temperature.
(D
(_,) 25 There's no way conduction of this can offset -- you look NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W, l (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 l l
115 1 at it at temperature escalation associated with zircalloy j
,r g 2 oxidation.
t <
V 3 There's no way it can keep up with that. You 4 couldn't transfer the heat out by conducting it around 5 that pipe. So you may get some smearing of -- an average 6 of this, but then you'll reduce the driving force through 7 the loop and reduce temperatures downstream.
8 CHAIRMAN KRESS: There is an offset.
9 MR. TINKLER: There will be an offset. You 10 reduce this temperature, you reduce flow down through the 11 system.
12 MEMBER CATTON: Well, again, you can address 13 this by looking at the time constant associated with the
-' 14 rate of increase of the temperature of the gases. You 15 compare that with the time constant associated with three i l
16 inches of steel. These are easy things to do. It's a ;
l 17 nice linear ramp once you start the oxidation, and you 18 know what it is.
l 19 It's so many degrees F or K per second, and 20 you can -- all you need to know is the thermal 21 diffuseivity of the steel and you can determine a time 22 constant. You compare the two.
23 MR. TINKLER: We have other calculations where i
24 we end up in part changing these -- in part changing this i O
\m ,/ 25 number. And like I said, you don't see a -- there'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 l
116 1 always a dampening throughout the system when you s 2 introduce a change on these calculations because they're
+
\
~
3 constantly seeking to adjust for this --
4 CHAIRMAN KRESS: Actually, I think this would 5 be relatively easy just to add in the RELAP also.
6 MEMBER CATTON: That's not what Tom's asking, 7 I don't believe. What Tom's asking is have these kinds of 8 simple evaluations been done? Has somebody looked at the 9 time constants associated with the temperature rise of the 10 gas versus time constants of the metal? What can you 11 ignore --
12 MR. TINKLER: We'll take a look at that.
13 MEMBER CATTON: And then compare that --
(O
\ s/ 14 compare the heat load on the top with possible conduction 15 through the steel to the lower temperature.
16 MR. TINKLER: We think -- now, admittedly, 17 it's been qualitative. But there are also considerations.
18 The entrance -- this is not fully developed flow in the 19 hot leg. We expect heat transfer coefficients would be 20 higher.
21 MEMBER CATTON: Yes.
22 MR. TINKLER: That would cause the hot leg to 23 fail sooner.
24 CHAIRMAN KRESS: What did you use for heat
/'
(D,) 25 transfer coefficient?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 i
117 1 MR. TINKLER: Well, --
_ 2 CHAIRMAN KRESS: This Dittus-Boelter?
('~'/ 3 MR. TINKLER: Well, you know, in the surge 4 line it's going to Dittus-Boelter because it's pretty high 5 velocity stuff. But otherwise, we use the -- you know, 6 the maximum of free or forced. Okay, but those are for 7 fully developed flows.
8 MEMBER FONTANA: That's fully developed flows.
9 MR. TINKLER: And you would expect -- and 10 we're nowhere near the necessary L over D's in the system 11 to get to a fully developed flow.
12 CHAIRMAN KRESS: That would bias you in the --
13 MR. TINKLER: That would push the hot leg A
_ 14 failure to be sooner.
15 CHAIRMAN KRESS: Yes.
16 MR. TINKLER: Sensitivity analysis -- to look 17 at the effect of inlet plenum mixing sensitivities, we 18 developed a little single loop model that was extracted l
19 from the Surry plant calculation. This started at the 20 anset of counter current flow, roughly 9000 seconds. And 21 we varied the number of tubes participating in forward i 1
22 flow.
i 23 We varied the mixing fraction, and we varied 1 24 the recire ratio. For these sensitivities, one parameter
.(~~-
r) 25 was varied at a time. Basically what we saw was change of I l
NEAL R. GROSS COURT REPORTERS AND TRANGCRIBERS 1323 RHODE ISLAND AVE.. N W.
(202) 234 4433 WASHINGTON. D C 20005 3701 (202) 234-4433
118 1 less than 20 degrees K on the peak tube temperature at the f 1
l 2 time of surge line failure. The surge line still fails !
f
(' ~ '
3 first. Peak tube temperature varies up and down by 20 4 degrees.
5 MEMBER CATTON: But now these sensitivity 6 calculations were done individually?
7 MR. TINKLER: Yes.
8 MEMBER CATTON: And there was no combined 9 impact of heat transfer?
10 MR. TINKLER: No, no. And we'll talk about 11 that a little later. These are -- this is a summary of 12 some of the full plant calculations -- sensitivities that 13 we did. Case one is the case that I showed in you plots
(%
( )
k/ 14 four without a depressurization on the secondary side.
15 Well, you see the steam generator tube temperatures run a 16 lot cooler, 140 degrees.
17 These peak tube temperatures generally occur 8 int he pressurizer loop for these calculations. So you 19 see, temperature of the tubes at the time of the surge 20 line failure is 847. Case two was we allowed the surge 21 line failure to occur. We allowed the system to 22 depressurize just to see what would happen. Well, the 23 tubes don't get any hotter.
24 Because of the break and reduction and A
(_) 25 pressure in the primary system, that's the peak of the NEAL R. GROSS COURT REPORTERS AND TRANSCR:BERS I 1323 RHODE ISLAND AVE. N W. l (202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 i
i 119 1 conditions. Case three, I showed you earlier, the plots.
2 Case 4 is these same thing. We allowed to break to a 73
()
3 curve. Tubes don't get any hotter. This was their 4 hottest condition.
5 Case five is similar to case one. We 6 increased the number of tubes carrying hot flow to 53%.
7 And we see a reduction. Case six, same thing. It's the 8 same as case three.
9 MEMBER FONTANA: Can you tell me what case 10 three was?
11 MR. TINKLER: Case three is our base case, 12 depressurized steam generator. Maximum delta P across the 13 tubes, maximum tube temperature. The difference between l
)
\/ 14 case six and case three is case six allows 53% of the 15 tubes to carry hot flow.
16 MEMBER CATTON: And that's with the same 17 mixing as the other one?
18 MR. TINKLER: Actually, we allowed it to float 19 in that calculation. We let the code seek its -- we 20 didn't fix it. It's about the same.
21 MEMBER CATTON: Well, all the code knows is 22 what you gave it based on those cold tests, the 1/7 scale 23 tests.
24 MR. TINKLER: Right, but it could have 7m
( 25 adjusted.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (702) 234-4433
)
120 1 MEMBER CATTON: Sure, but you don't have a
, 2 clue as to whether it would adjust in the right direction U, -~g 3 or not.
4 MR. TINKLER: Well, the point I was going to 5 make was that the mixing fraction changed in a trivially 6 small way. It changed slightly, but not -- .87, '85.
7 Same number. Case seven, we allowed all the steam 8 generators to blow down. We saw a peak -- we saw an 9 increase, but not much, because it was all set by the fact 10 that the surge line failure was accelerated due to the 11 repeated challenges on the surge line -- greater 12 challenges on the surge line.
13 Yes, sir?
\
'- 14 DR. ZUBER: I'm coming back to the old 15 question on the hot leg. You have a -- the temperature of 16 your steel, is it uniform on the perimeter?
17 MR. TINKLER: Along the perimeter?
18 DR. ZUBER: Yeah, I mean circumference. )
19 MR. TINKLER: Yes.
20 DR. ZUBER: Okay, and then what -- you are 21 calculating heat transfer from the cold liquid. Which one j l
22 is then hotter, the steel is hotter than the cold liquid?
l 23 MR. TINKLER: In the hot leg?
24 DR. ZUBER: Yes. i f~%
(_,) 25 MR. TINKLER: The fluid is much hotter.
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 DR. ZUBER: Much higher.
- 2 MR. TINKLER: Much higher.
V 3 DR. ZUBER: You are transferring then from the l
4 -- you have a -- bot leg would be on the top and cold leg 5 would be low?
6 MR. TINKLER: Right.
7 DR. ZUBER: Can you describe how you modeled 8 the heat transfer from the ho'. leg to the cold liquid and 9 from the cold liquid from the steel?
10 MR. TINKLER: These are two separate paths.
11 The hot fluid transfers, energy to the steel of the 12 simulation of the upper half of the pipe. And that is 13 isolated from the fluid in the lower cold half of the r~S N'sl 14 pipe.
15 DR. ZUBER: Is it still the same in both 16 cases?
17 MR. TINKLER: No, no, it's going to be 18 different.
19 MEMBER CATTON: I think in both cases, as far 20 as their calculations are concerned, the steel will be 21 colder than the gas will --
22 MR. TINKLER: Oh, absolutely.
23 MEMBER CATTON: -- because they're separated.
24 Where in reality, what you'll have is that in the cold r~
(T) 25 side fluid, some of the steel may well be hotter than 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
122 1 gas. And what that will do is it will create a vortex in
,s 2 the flow headed for the plenum, and that -- you know, I 3 think you started to bring that up before.
4 As the flow exits into the plenum, it will 5 exit with the vortex. Whereas in the actual test, it 6 probably didn't. And will that vortex -- what will that 7 vortex flow do to the mixing?
8 DR. ZUBER: Yeah, they also make some energy 9 -- between the cold liquid and the hot liquid.
10 MEMBER CATTON: You know, it's interesting 11 that when you look at a transverse jet in a flow, if you 12 just assume that it's an inertial impact, you get one kind 13 of an answer. If you assume that the cross flow creates a N/ 14 vortex, you bend the vortex over and you significantly 15 reduce the mixing, what happens here? Does it go up or go 16 down, and what do you think and why?
17 Do you follow the argument?
18 MR. TINKLER: Yes, I understand the argument.
19 But --
20 MEMBER CATTON: Or does it matter?
21 MR. TINKLER: -- I don't know that we -- that 22 any data exists to develop a heat transfer coefficient 23 based on that configuration.
24 MEMBER CATTON: And I would agree with you
\
'q, / 25 wholeheartedly. It's not --
NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
123 l 1 MR. TINKLER: And I --
2 MEMBER CATTON: So what do I do to account for g-b 3 this in the time / temperature distributions that I want to 4 predict? How do I deal with it?
5 MR. TINKLER: Changes in the -- again, changes 6 in the heat transfer coefficient aren't likely to alter 7 your conclusion within --
8 MEMBER CATTON: Well, I don't agree.
9 MR. TINKLER: Any changes will be to increase 10 the heat transfer coefficient to increase the time 11 interval.
12 MEMBER CATTON: Not necessarily. If you 13 reduce the mixing and you create this vortex flow, you may
\ 14 well have a hotter zone in the central part that is less 15 affected by the walls.
16 MR. TINKLER: This thermal center?
17 MEMBER CATTON: I'm going to have circulation 18 in that stratified flow. And both up -- both above and 19 below, I'm going to form a vortex type flow because the 20 hot flow is cooled, so it runs down the wall and rises in 21 the center; the flow in the bottom is going to be being 22 heated at the upper edges by the wall, and on the bottom 23 it's cooled by the wall.
24 So it gets quite complicated. As a matter of
/
\ ,/ 25 fact, I would think that there would be significantly more NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASH!NGToN. D C. 20005 3701 (202) 234 4433 l
124 l l
1 stronger stratification in the lower flow than in the
,- 2 upper flaw for that reason. In any event, I understand we
\.__)
3 don't know the heat transfer coefficient. I'm just 4 raising the question as another uncertainty that needs to 5 be dealt with.
6 MR. TINKLER: Well, I'm not saying we don't --
7 I'm not saying we don't know well enough. I'm saying we 8 don't know with the kind of precision that's being 9 suggested that we need here without a treatment -- without 10 a probablistic treatment. But let me get on to the 11 sensitivity that we did do on mixed convection.
12 DR. ZUBER: And you're neglecting the heat 13 transfer between the two gases, cold and liquid -- cold
- 14 and hot?
15 MR. TINKLER: Yes, yes. Let me introduce this 16 subject. Because of -- you know, because we understood 17 there were questiens about the use of the code for this 18 kind of application, and while we had thought that, you 19 know, these issues had been considered sufficiently in the 20 past, we undertook and additional independent review most 21 recently to assess the general adequacy of the code for 22 these kinds of calculations to see if there was something 23 we were -- something dramatic we were overlooking.
24 Was there some showstopper we hadn't imagined?
O
(_) 25 As I said before, people had considered the neglect of NEAL R. GROSS I
COURT REPORTERS AND TFMNSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 I
1
125 1 radiation heat transfer, some other issues associated with g- 2 this modeling. Well, we undertook it again. We used k_.g ) 3 Professors Viskanta, Ishii, Griffith. This was a review 4 we conducted quite recently in Chicago on the 19th and 5 20th in a public meeting, and Professor Catton was present 6 for the meeting to participate.
7 The peer reviewers provide preliminary 8 comments, recommendations at the conclusion of the 9 meeting, but they've also provided some written comments, 10 some of which, I believe, have been provided to the 11 committee. And their fundamental -- you know, their basic 12 conclusion was that the experimental data is --
13 experimental program was well conceived, well designed, r~_x
- 14 well carried out. j 15 They generally thought that the SR5 code was 16 quite adequate for the job, certainly adequate in some --
17 in the words of some to calculate natural circulation.
18 Given that the path is defined and the path was defined by 19 the experimental data, they thought that the 20 implementation of the code was good and that the 21 constitutive relationships within the code were adequate.
22 It's understood that there are always 23 uncertainties about heat transfer correlations and so 24 forth, but the basic constitutive relationships in the
(_) 25 code were adequate. They did think we needed to improve 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 our demonstration of the adequacy of the modeling
,s 2 approach, and they urged us to give some thought to i T G 3 establishing figures of merit which would summarize and 4 compare the analytical data or the analytical results and 5 the experimental data to further give credence to the view 6 that the tests represent the basic behavior.
7 They also recommended that it would be 8 worthwhile to perform additional sensitivity studies for 9 parameters which varied widely. At least in the case of 10 Professor Griffith, he was only interested in 11 sensitivities where parameters vary in a significant way 12 because there was a thought -- there was the belief that 13 small variations in individual parameters are likely to be O( /' 14 lost in an interval calculation like this.
15 DR. ZUBER: Mr. Chairman, I would like to make 16 a comment. I feel -- as a technical man, I feel very 17 uncomfortable when a regulatory agency selectively selects 18 quotations and makes statement based on comments by a peer 19 review group. I would like to draw your attention to the 20 first bullet, "The experiments are good and well scaled."
21 The point is -- and on page -- if you read Ishii's letter 22 -- ,
l i
23 CHAIRMAN KRESS: In this? i 24 DR. ZUBER: Yes, in --
I r~
( ,S) 25 MEMBER BARTON: Number two. l 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
' DR. ZUBER: Number two. If you read Ishii's
, 2 letter, the last -- hot leg scaling analysis on the first
%,)
3 page and the first paragraph on the next page, he's 4 questioning the scaling. It's not clear, it's i
5 questionable. And I think negle-ving this statement and 6 selectively taking statements which are favorable doesn't 7 reflect well for the regulatory agency.
8 And I think this should be called and tried to 9 be discouraged.
10 CHAIRMAN KRESS: Which paragraph were you 11 looking at? I was still hunting.
12 DR. ZUBER: Yeah, hot leg scaling analysis. l 13 MEMBER BARTON: It's the last letter in that O -- last letter in that package? Yeah, bottom paragraph on
\ s' 14 15 the first page.
16 DR. ZUBER: And the next -- and the first 17 paragraph on the next page.
18 MR. TINKLER: What we have provided to you is 19 a summary of their general views. Okay, we are not -- we 20 would not maintain that when you get people who are expert 21 in scaling that they wouldn't obviously have questions 22 about scaling. Okay, but this needs to be considered in 23 the broader context of the overall remarks.
24 And I guess I object to the suggestion that we n
k ,) 25 have selectively identified or are trying to present 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
128 1 the committee favorable quotes out of their evaluation. I
14 this situation. We think they are. Now yes, there are a 15 lot of uncertainties, and we'll try to deal with those.
16 But I think a lot of their presentation and a lot of the 17 work was done to try to respond to that particular 18 statement, which I think was an inaccurate statement, in 19 an ACRS letter.
20 MEMBER CATTON: And I think that I initiated 21 this discussion by indicating that the adequacy really 22 isn't the question. It's what you do with what you've got 23 that is. Now, I think I would be the first to admit that 24 the statement in the ACRS letter was probably too strong.
) Right, but what I'm saying --
x ,/ 25 MR. HODGES:
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l
?323 RHoDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
131 1 MEMBER CATTON: And I've told you what the 2 intent was. We had a meeting several weeks ago to discuss I~si
\J I want to 3 that. I told you then what the intent was.
4 know -- and this is a safety issue. This is not -- no 5 longer a code issue. You've tuned your code. What are 6 you going to do? And we haven't heard the answer to that 7 question.
8 MR. TINKLER: Can I just -- since I missed 45 9 minutes, my question might well have been addressed. On 10 the recirc ratio question which seemed to be a fundamental 11 difference between the industry calculation and yours, you 12 took the ratio from the tests and they had a different 13 number that they got from a model. This doesn't seem to
(^h
\I be one of the sensitivity parameters that gets addressed, 14 15 and yet somehow it seemed to me to drive it or to --
16 again, as it -- you know, in coming down -- the question 17 is is not how well you predict the temperatures, but what 18 difference it makes in sort of relative difference between 19 the surge line and the tubes.
20 You can be off on both of them as long as you 21 get them relatively right.
22 MR. TINKLER: Right, right. It's difficult to 23 make a direct comparison with the other calculations, 24 okay, because they're done for a different plant in a (G,) 25 little different configuration. But if we vary -- if 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
132 1 increased it by a factor of two -- we made it from two to g, 2 four, for example, the recirc ratio; and we're going to 3 decrease tube temperatures another -- I hate to stand up 4 here and -- I'm just going through the numbers in my head, 5 ratio -- probably on the order of something like, you 6 know, 50 degrees, something like that -- and calculations 7 have been done for an average plant with an average flaw 8 distribution using case three numbers.
9 And those calculations show a probability of 10 one tube rupturing of about 6%. If you increase the tube 11 temperature transient by 50%, you increase the probability 12 of a single tube rupture up to about 20%. Now if you 13 further increase the recirc ratio, you'll drop that 14 temperature.
15 MEMBER POWERS: So in the --
16 MR. TINKLER: And I'll describe some 17 additional sensitivity studies that will also drop that 18 temperature. i l
19 MEMBER CATTON: So what are the chances of it 20 going up 50%?
21 MR. TINKLER: Up 50% or up 50 degrees?
22 MEMBER CATTON: Oh , I mean 50 degrees. I'm 23 trying to use your numbers because you showed us where -- ,
I 24 or just told us how you go from 6% failure to 20% failure, (m) 25 MR. TINKLER: Well, we are still looking at NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W. j (202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 l
133 1 these ranges of sensitivities and looking at how --
2 MEMBER CATTON: I understand. But this, I V 3 think, is really the heart of the problem. If the 4 temperature of the steam generator tube increases 50 5 degrees and the failure goes up from 6% to 20%, there's an 6 extremely strong relationship between the two. Whatever 7 caused it go up 50% in the failure rate? When you have 8 steep slopes like that --
9 MR. TINKLER: See, I maintain that that's not 10 that steep a slope though. I've changed one temperature 11 relative to the others, okay, when it doesn't work that 12 way. I've showed you all these sensitivities where the 13 temperatures change less than that. Okay, I showed you --
( /
14 I repeated the -- I reported the maximum of --
15 MEMBER FONTANA: I don't think that's the 16 question.
17 MEMBER CATTON: No, you're --
18 MR. TINKLER: The 20 degrees, in many cases, 19 it's ten degrees; and when you combine them, they're not 20 additive. And I showed you a calculation where all I have 21 to do is increase the number of tubes consistent with 22 what's seen in the transient data, and I decrease the tube 23 temperatures by 50 degrees.
24 MEMBER CATTON: So where are you in all this th j 25 mush?
NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 l
134 1 MR. TINKLER: We're still using --
2 MEMBER CATTON: I mean, things are going up,
,f-~
V 3 things are coing down. How does somebody who's sitting at 4 the other side of the table develop some confidence in 5 what you're putting out? I mean, you can arrange these 6 things all over the place. What do you do with it? How 7 do you come to grips with it in a risk kind of arena?
8 How?
9 MR. TINKLER: Well, I'm not going to suggest 10 that we're going to establish a distribution on the 11 temperature transient. Okay, we're going to look at the 12 sensitivities. We're going to -- we'll probably do some 13 more sensitivity calculations between now and when we
(
N- 14 finish this --
15 MEMBER CATTON: Well, but you see, sensitivity l 16 by itself is not enough. I think you need to look at any l
l 17 heat transfer coefficient you want out of the three or l 18 four that are important and ask yourself what's it going 19 to be? How well do I know it? What things can cause it 20 to be different? Could it be different than the 21 experiment? And before you know it, you can say gee, I 22 think it's between this and this.
23 It could be a little bit lower, but the 24 probability is low, and then you roll that into some sort
(]_j 25 of a final statement.
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 l l
1 MR. TINKLER: I understand.
2 MEMBER CATTON: Without that, I don't know rx
(
3 what to do with all these numbers.
4 MEMBER FONTANA: I think one of the problems 5 that we're seeing here - there are a lot of factors that 6 affect this answer, if you want to call it that. And 7 you've addressed quite a few of them. There's system 8 factors, the configuration of the plant, what fails before 9 all that other stuff. There are a lot of things that go 10 on.
11 The thing that bothers me a little bit about 12 this change in the temperature by 50 degrees isn't so much 13 that change of 50 degrees; it's that you're in a region 7,
'- 14 where the steam generator tubes are being challenged.
15 And what we're talking about, age and recirculation and 16 all that, might be not so significant compared to a lot of 17 other things that affect that.
18 And the question that I think he's driving at 19 is is there an overall analysis of the various factors 20 that affect the temperature tubes, see, and try to 21 identify which are big and which are smaller and, you 22 know, not have to get into the gnat's eyebrow calculations 23 that we appear to be talking about.
24 MR. TINKLER: Well, in the -- you know, in the es
(_,-) 25 analysis or in the safety analysis report, we will report 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 the individual precise numbers. I've tried to give you a
, 2 sense of them here, but most of the sensitivities I've
\
' '/ 3 shown -- most of the sensitivities that push tube 4 temperatures up are all less. Every one of them was less 5 than merely increasing the fraction of tubes carrying hot 6 flow.
7 Every one of them. And I -- we haven't done 8 the calculation, but we're pretty sure that combined it's 9 going to be less than increasing the number of tubes 10 carrying hot flow. And I'll how you another sensitivity 11 to look at the secondary side of heat removal. But, you 12 know, we take seriously the comments of reviewers. And we 13 are looking to develop a figure of merit.
13 k_) 14 I'm going to show you briefly some of the 15 results that we have so far.
16 CHAIRMAN KRESS: Would that figure of merit 17 address the issues, concerns on scaling? I'm not quite 18 sure I know what figure of merit is.
19 MR. TINKLER: Well, I don't want to say that 20 it will. You know, he hasn't seen it yet because we've 21 just done these calculations in the last week or so. So, 22 you know, we can get back with them, see how they feel 23 about these additional comparisons. And I would expect 24 that you'll hear more about it in upcoming briefings A
25 before the committee.
(w-)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. O C. 20005-3701 (202) 234 4433
137 1 MR. HODGES: I'm just not sure how to
,_ 2 interpret your last statement. If you're telling me that 3 the choice of the fraction of the tubes overwhelms all the 4 other factors that we're talking about here, that's 5 interesting. But then what do I do to address that I 6 don't really know the fraction of tubes?
7 MR. TINKLER: Well, what I was suggesting was 8 right now we're using a fraction that is biased to elevate 9 the tube temperature. We're using --
10 MR. HODGES: Okay, you're convinced that you 11 have a conservative --
12 MR. TINKLER: Yes, yes, we're convinced that 13 we're using the 35% which is the small end to look at the (h
\_) 14 transient data. It's --
15 MEMBER POWERS: Could I ask a question about 16 you putting 400 kilograms of aerosol into the circuit some 17 place and some fraction of it's going to get down into the 18 plenum region, and some fraction of that then is going to 19 go into the tubes and deposit, and you've calculated that?
20 MR. TINKLER: Well, actually, 400 kilograms 21 released. 60 of it goes out to PORV, 40 of it's noble 22 gas. It's about 300; but, yes.
23 MEMBER POWERS: We get a lot of mass down 24 there. And if you have a sensitivity to localized (n) 25 deposition, and in a tube situation, sometimes you do 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
138 1 sometimes you don't; but because you do have that large 2 heat sink right at the inlet and because you have what 3 amounts to kind of a well mixed plenum connected to the 4 tubes, it looks like you might have a chance of having 5 localized deposition, would such localized deposition 6 produce enough flow resistance to alter that fraction of 7 tubes carrying hot flow?
8 MR. TINKLER: Well, I -- if you recall the 9 numbers -- I don't know how bad you want to make this 10 localized deposition. We have a fair amount of localized 11 deposition at the entrance to the -- we calculate -- the 12 numbers we report, 30 micron film thickness, is quite high 13 because we've reported the VICTORIA assessment of film C\
V 14 thickness which was based on a much lower density for this 15 material than it actually is.
16 So we're talking -- as I recall, the cross 17 section -- we can change that number by order of magnitude 18 and it's quite small. I mean, the number -- the diameter 19 of the tube is 20,000 microns. We're calculating a film 20 thickness right now of 30. We're probably overestimating 21 that because we're using a density of the material based 22 on 1,000 kilograms per cubic meter which is the VICTORIA 23 number.
24 MEMBER POWERS: Yeah, it's about right for an fG 25 aerosol.
()
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE1SLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 ,
i
139 1 MR. TINKLER: Well, this stuff is heavier than 2 that.
(l 3 MEMBER POWERS: Well, the material density can 4 be quite high, but the particle density itself is -- I 5 mean, most of the particles --
6 MR. TINKLER: Oh, this is liquid. I mean, 7 this is -- a lot of this is liquid.
8 MEMBER POWERS: If it's -- well, that's not 9 the material I would worry about. The fission products 10 that could be molten are a trivial part of your aerosol.
11 It is the non-radioactive component that's going to make 12 up the biggest amount of your mass.
13 MR. TINKLER: Okay, okay. Well, even at 30
~s 14 microns relative to 20,000 microns, I guess --
15 MEMBER POWERS: I would agree with you if that 16 -- but I got the impression that that was --
17 MR. TINKLER: You know, we would expect some 18 of that stuff to drip off the bottom of the tube sheet and 19 fall back down to the bottom of the inlet plenum.
20 MEMBER POWERS: And in fact, that was observed 21 in the Marviken tests. Exactly that phenomena was ,
1 22 observed. l 23 MR. TINKLER: So, you know, it's just -- I 24 wouldn't expect it to all stay up there.
A
(_,) 25 MEMBER POWERS: But I wonder, is your 30 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234 4433
i 140 1 micron number -- that's because it's smeared out over an
-, 2 entire surface?
\_/ Well, that's 30 microns at that 3 MR. TINKLER:
4 node. I mean, we nodalized around the tube bundle in the 5 same detail that we nodalized SCDAP/RELAP. It'e the same 6 correspondence and nodalization. So we model the first i i
7 couple -- it's not the whole tube bundle. It's the first 8 node up to maybe the first support past the tube sheet.
9 MEMBER POWERS: So maybe you just don't have 10 enough mass for even a localized deposit to affect your 11 flows?
12 MR. TINKLER: I don't think so. And you know 13 --
)
\- 14 MEMBER POWERS: Is that an area where you have 15 some sort of validation equivalent or cimilar to that kind 16 of validation you had for the thermal hydraulics analysis?
17 MR. TINKLER: Well, the validation of VICTORIA 18 is not as extensive as the validation of SCDAP/RELAPS in a 19 general sense. Okay, but, you know, we're currently 20 undergoing a peer review for VICTORIA. The code itself is 21 undergoing peer review. And you know, we asked them to 22 look at the deposition models and there's a general sense 23 that the deposition models in the code are reasonable 24 given your understanding of the thermal hydraulics.
- (j
/} 25 You have to adequately characterize the flow 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 paths before --
2 MEMBER POWERS: The correlations that are
'~# 3 typically used for deposition of aerosols in tubes in a 4 paper written by a relatively famous aerosol scientist, 5 and he concludes his paper by pointing out how difficult 6 it was for him to do the studies. And he has a -- just a 7 wonderful list of all the things that he says complicate 8 these kinds of development of these correlations.
9 And the things he cites in there are the kinds 10 of things I would think you'd have to worry about. He 11 cites gee, if your tubes aren't perfectly smooth, my 12 correlations don't work; if I get deposition -- as soon as 13 I get deposition, my correlations don't work; if I get any s- 14 kind of electrostatic charging on the aerosols, my 15 correlations don't work.
16 He has a list. It's about ten. It's a 17 marvelous admission of the difficulties of doing aerosol 18 experiments. And I'm wondering if in the face of that 19 kind of admission from the guy that developed the 20 correlation if he wasn't saying gee, if you're -- if it's 21 really important, you ought to do an experiment.
22 MR. TINKLER: Well, like I said, we take some 23 comfort in the fact that it appears te us that we could 1
24 change our number by an order of magnitude and it wouldn't 25 be any different.
1 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 MEMBER POWERS: That should be a comforting 2 thing all right.
(-)
+
\'
3 MR. TINKLER: It is. Or is has been.
4 (Laughte r. )
5 I hope it continues to be. But you know, 6 actually we've learned quite a bit about the application 1
7 of the codes for these kinds of analyses, and we've, you l l
8 know, discovered a lot of the pitfalls in trying to match 9 up a thermal hydraulic calculation and a chemistry l
10 calculation. l 11 MEMBER POWERS: It's a bit like mixing oil and 12 water.
13 MR. TINKLER: Like I said, I know 400
/
m I
(_) 14 kilograms in a reat sense is a lot of particles, but 15 relative to the masses of structures involved here and so 16 forth and the flow raise, it's still relatively small.
17 MEMBER CATTON: Dana, do the aerosols always 18 go to the coldest surface?
19 MEMBER POWERS: Well, it depends on what your 20 flow velocities and things like that are. But yes, the 21 cold surfaces are like magnetic attractions to aerosols.
22 MEMBER CATTON: So that means like in the hot 23 leg, the bottom most point is going to be the coldest, so 24 that's where I would expect the aerosols?
q
() 25 MEMBER POWERS: Well, you've got an NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
i202) 234-4433 WASHINGTON. D C 20005 3701 (202) 234-4433
143 1 interesting situation. I honestly do not understand what 2 happens in these counter current flow situations. But 7_
N' ,)
3 you've got the aerosol laden gas coming in the hot, 4 MEMEER CATTON: Right.
5 MEMBER POWERS: -- okay, because the cold gas 6 is going through the equivalent of a filter, so it's 7 coming in kind of aerosol free. There's a driving force 8 for the aerosol particles, first gravity; and there's a 9 thermophoretic term. But I have no idea how to evaluate 10 it. And then there's a turbulent inertia term. It's 11 trying to push that aerosols down into the cold material.
12 At the same time, they've got a velocity that 13 direction, and it's strictly a balance between all those (3
\_) 14 forces and the time they have available whether you're 15 going to get any significant deposition. Typically the 16 hot leg is not an area of a great deal of deposition.
17 What you typically get in the hot leg is just a 18 thermophoretic.
19 Because I think on one of these plots they 20 showed that the gas coming in was like astronomical 21 temperature, some 1700 degrees Kelvin. Well, obviously i
22 steel -- everywhere around it is a much, much lower 23 temperature or it would be creeping like crazy. And that 24 cracks some aerocols, but it's a big pipe. You know, r%
() 25 there's not much time for the particle to move through NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
144 I
1 that wall before it is moved down into that --
2 MEMBER CATTON: And you have a nice vortex 73 3 type motion set up so that the hot gases are continuing to 4 moca up.
5 MEMBER FONTANA: Are we concerned here on the 5 heating effect or are we concerned on decontamination 7 effect once you do --
8 MEMBER POWERS: I was just wondering -- and I 9 don't know at all, but if I put up an array of tubes and 10 try to push an aerosol through it from a plenum, where 11 would I get local deposition that creates some flow 12 resistance? And would that cause the flow to 13 preferentially go away from that tube to another one, (h
\ >' 14 where again you'd have some preferential deposition on one 15 of them and that it would cut down this fraction of tubes 16 carrying the hot flow?
17 Now the argument's made that it's a mass 18 problem. There just isn't enough mass to make a big 19 enough difference for it to even show up in the 20 calculations. I mean, yes, a little plot where it's a --
21 where it's barely off the thing. And I don't know. I 22 have not done a.y calculations.
23 CHAIRMAN KRESS: If you put all that mass on 24 the inlet area, would it be enough?
((3,) 25 MEMBER FONTANA: Yes, it's a boundary --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(20?) 234-4433 WASHINGTON. D C. 20005 3701 (202) 234-4433
145 1 CHAIRMAN KRESS: Well, kilograms is a lot of 2 mass to put --
73 3 MEMBER POWERS: It is a hell of a lot of 4 aerosol mass.
5 CHAIRMAN KRESS: You could play -- if at all, 6 deposit it in those inlets.
7 MEMBER POWERS: I don't know that that's the 8 case. And it is a kind of thing -- you know, my reaction 9 to the question if somebody forced me to answer it would 10 not be to turn VICTORIA on. I would want to look at the 11 VICTORIA calculations to get forces. But I think I'd do 12 it by hand.
13 CHAIRMAN KRESS: I would use the old EPRI O
sI 14 correlation of plugging.
l 15 MEMBER POWERS: Well, that's gone -- that's l
i 16 the problem with that -- using that correlation is that 17 they're very careful about saying it's smooth tubes with 18 no -- none of these inlet problems. I would use an inlet 19 efficiency calculation.
20 CHAIRMAN KRESS: No, I'm talking about the 21 plugging of small pipes by aerosols at the inlet, that 22 correlation.
23 MEMBER POWERS: Yeah, and that's -- that one 24 assumes a certain flow on things that may not be had here.
\
('~/
(, 25 I would tend to do it by hand, l
j l
NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
I 1 146 1 CHAIRMAN KRESS: But what may give me some s 2 comfort in thinking about this is that even if you plugged
\_) 3 those inlets to the steam generator plenum, you're doing 4 two things. One, you're increasing the resistance to the 1
5 flow going through the tubes, and you're shifting the 6 number of tubes to a better ratio so that the effect of it 7 in my mind -- and that's an area where you have a high 8 heat capacity so that the overall effect of the plug in 9 would be to lessen the likelihood of a steam generator 10 tube failure is the way I would --
11 MEMBER FONTANA: If you have more tubes 12 carrying hotter flow, as you --
13 CHAIRMAN KRESS: Right, you have more tubes.
I
\/ 14 MEMBER CATTON: Well, I was thinking that 15 you'd have fewer tubes carrying the hot gas if you created 16 flow resistances in some of --
17 MEMBER FONTANA: Plug a --
18 MEMBER CATTON: You just increase the velocity 19 to the few.
20 MR. TINKLER: Again, let me remind you that 21 the positive thickness at your inlet -- at the inlet is 30 22 microns. The tube diameter is 20,000.
23 CHAIRMAN KRESS: Yeah, but that doesn't --
24 MR. TINKLER: Okay, now if I take all of the rw
( 25 mass, I increase this by a factor 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 1
147 1 CHAIRMAN KRESS: You put it all in that first 2 node.
7 V' 3 MR. TINKLER: -- 40. I increase this by a 4 factor of 40, 5 CHAIRMAN KRESS: You're putting it all in the 6 first node.
7 MR. TINKLER: First node.
8 CHAIRMAN KRESS: But I'm going to put it all 9 right there at the inlet. I'm making that area 10 effectively zero.
11 MEMBER CATTON: Charlie, you don't remember 12 those tests that EPRI did when they were trying to argue 13 that the containment can plug itself up?
/3
- 14 MR. TINKLER: I understand. But this stuff is 15 still -- there's still going to be -- this is not going to 16 be all -- I don't think it's all going to be solid -- all 17 this control rod material is going to be solid.
18 CHAIRMAN KRESS: But my point was, even if you 19 did that, I think you're going to a better condition.
20 MR. TINKLER: It's molten. It drips off down 21 to the bottom of the inlet plenum.
22 MEMBER POWERS: And that very definitely 23 happened in the Marviken tests that they would get 24 deposits of these materials that came in as droplets and p]
t 25 they would coalesce and drip. Yeah, that will definitely NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W. f (202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
148 1 happen.
2 MR. TINKLER: In worse case, you won't be able r
73 i
V 3 to continue to --
4 CHAIRMAN KRESS: And my point was, Charlie, 5 that the more plugging you got of those tubes, the better 6 off you probably are in terms of this problem, and that 7 was my point.
8 MR. TINKLER: It may increase mixing.
9 MEMBER FONTANA: Strange.
10 MR. TINKLER: Let me just quickly -- Dr.
11 Catton raised the issue of mixed convection, I guess --
12 has raised it a number of times. Actually, the issue of 13 mixed convection came up in consideration of ALWR's for me
['
N- 14 a couple of years ago looking at something. But in 15 response to questions that we received that we don't model 16 mixed convection, we had INEL do a couple calculations 17 using an update version of the code with a mixed 18 convection correlation implemented.
19 For those of you that don't know, mixed 20 convection refers to the super position of free and forced 21 convection either assisted or opposed. For turbulent 22 conditions, opposed free and forced convection actually 23 increases heat transfer. So in the tubes, the dominant 24 direction of flow is upward. But the tubes are colder (n,) 25 than the fluid, so the tubes impose a free convection NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
l 149 1 1
1 1 down. l 2 So that's opposed. And that actually O 3 increases heat transfer. Basically we implemented the 4 Cotton-Jackson correlation for mixed convection that was 5 published in the International Journal of Heat and Mass 6 Transfer. The calculations we did showed that the steam 7 generator tube temperature increase was insignificant.
8 We raised the tube temperatures by four 9 degrees Kelvin. Earlier I referred to the issue of heat 10 removal on the secondary side. Basically our nodalization 11 of the secondary side of the steam generator doesn't allow 12 us to take much credit for some of the structures in the 13 upper part of the steam generator. So we did a S
/'b
\ 14 calculation where we actually -- basically we went to 15 coarser nodalization so we'd get something closer to an 16 average temperature.
17 And when we did that, we decreased the tube 18 temperatures by 65 degrees at the time of surge line 19 failure, which was the expected result. We thought we 20 would increase the heat sink on the secondary side.
21 MEMBER POWERS: Now when you look at the heat 22 transfer in the secondary side, the tubes going up are 23 going to be hotter than the tubes coming down. That's 24 going to give you preferential circulation in that large Q
Q 25 volume which says that the tubes that are furtherest away NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433 i
150 1 from the cold side are going to have the lowest heat
,s 2 transfer to the secondary side.
3 So somehow this should be accommodated in what 4 you're doing. I don't think you can just do this by 5 itself. Do you follow me? The cold side is going to be 6 different than the hot side on the secondary side. So the 7 result is going to be a circulation pattern gets set up.
8 The colder tubes are going to feed the steam to the hotter 9 tubes, and the further you are away, the lower the heat 10 transfel's going to be. .
l 11 MR. TINKLER: I'm not suggesting that we were j 12 in any way --
13 MEMBER POWERS: I'm not either.
/O
'/ 14 MR. TINKLER: -- a convective pattern in the 15 secondary side.
16 MEMBER POWERS: I think this is a good step.
17 But somehow, this needs to be rolled into what you do when 18 you come to a conclusion. I would really hate to have to 19 defend the convective pattern we predicted using any code.
20 MEMBER POWERS: I would too. But there are 21 other ways it can be treated.
22 MEMBER FONTANA: I may have missed something, 23 but what was it that you did here that improved the heat 24 sink on the secondary side?
/
' 1
\ ,/ 25 MR. TINKLER: We basically went to a coarse NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C, 20005-3701 (202) 234-4433
1 151 1 nodalization on the secondary side. The original
,s 2 nodalization -- well, we had -- there was no -- it allows I \
~' 3 no circulation from the top to the bottom of the steam l
4 generator on the secondary side. So the heat rises, but 5 you don't allow any further cooling once it reaches cold 6 structures up there to circulate.
7 So what we did was we basically used one large 8 node for the steam generator on the secondary side, and 9 that allows us to better transfer heat to the structures 10 in the upper part of the steam generator.
11 MEMBER CATTON: And this keeps the secondary 12 side cooler?
13 MR. TINKLER: Keeps the secondary side cooler C's #
km/ 14 and keeps the tubes cool.
15 MEMBER CATTON: And that's real. The question 16 is then, is the local fact within -- on the bottom plane 17 of the steam generator because you're going to feed f.om 18 the cold to the hot? And how do you take advantage of all l
19 that metal up there? I think if you want to, this other 20 aspect needs to incorporated.
l 21 MR. TINKLER: All right, I mentioned to you 22 that we started -- by the way, if you're interested, ,
1 1
23 there's a Richardson's number in the steam generator, so 24 we were around one, so it was mixed convection. But we
'(,)
( 25 just didn't get much of an increase.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234 4433
I 152 )
l 1 MEMBER CATTON: Well, I'm not surprised. 1
,-s 2 We've done similar kinds of studies and decided that the
('~' /
3 internal heat transfer coefficient in the steam generator 4 tubes was probably not the parameter you needed to worry 5 about.
6 MR. TINKLER: When looking at figures of 7 merit, we were searching for some data from the tests that 8 we thought had some relationship to the overall issue of 9 heat transfer. One of the figures of merit we -- this is 10 not in your package. This is a very recent comparison of 11 the delta T between hot tubes and cold tubes relative to 12 hot and cold streams in the hot leg. It's a heat exchange 13 or efficiency kind of measure.
O t s x/ 14 And we took all the steady state and all the 15 transient tests and they fell within this. You know, we 16 could narrow it down by selectively choosing favorable 17 tests, but we didn't. We took all the tests that we had 18 data on -- and remember that the tube heating transient's 19 really through here. And what you see is the delta T 20 between the hot and cold tubes is about half of the delta ,
i l
21 T in the hot leg.
22 That's the driving force. And this actually 23 shows the effects of a recirc ratio of about two. So the 24 temperature difference of the hot tubes and the cold tubes (A)_
25 is about half of what it is in the tubes. So you're NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W (202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 l
l
153 1 getting about the same sort of efficiency from a heat 2 exchanger point of view.
7~s t )
'~' This is the hot 3 We tried other comparisons.
4 and cold Tnt and T,1 or the inlet temperatures at the hot 5 leg. You know, where the hot flow is entering from the 6 core, cold flow is exiting from that -- so this is the 7 delta T. That's the maximum delta T across the system. H 8 and C, Tn - T, are the hot and cold temperatures at the hot 9 leg -- exit into the steam generator inlet plenum. So 10 it's the delta T across the hot leg down to the steam 11 generator divided by the delta T at the entrance from the 12 vessel.
13 This was data taken from 14 tests. And that's s
14 a pretty expanded scale, I might point out. So that's a -
15 - we think that's a pretty good measure that there is at 16 least a correspondence of delta T's across the system 17 between the tests and the plant calculations. We're also 18 looking to compare structure temperatures.
19 This is really, really preliminary, but this 20 is a measure of this hot leg temperature -- structural hot 21 leg temperature minus the steam generator tube temperature 22 divided by the driving temperature -- the hot leg 23 temperature in the inlet. The difference between these 24 two plots is it depends -- the test data isn't a real --
("%
(_) 25 isn't abundantly clear on the measurement, the temperature NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. O C. 20005-3701 (202) 234-4433
l 154 1 measurement.
2 So we were -- we used both the average of our
%-) two SCDAP streams. This is the average of the two. We 3
4 took the average pipe temperature from the hot path and 5 the cold path, and this is when we just take the hot path.
6 But we need to -- we're going to look at this some more, 7 but I'll just give you a feel that we're looking at some 8 of the figures of merit.
9 I guess I'd like to conclude now if I can find 10 my conclusion viewgraph. Basically what we saw was that 11 the biggest influence on the calculation was whether or 12 not the secondary side depressurizes. That was worth 150 13 degrees on the tubes. When the secondary side relief I
\ 14 valves remain intact, there's a very, very small challenge 15 to the tubes.
16 The tubes weren't predicted to fail when we
~
17 stretched out the calculation 19,000 seconds. We do see 18 that the thermal hydraulic boundary conditions are 19 influenced by some of the phenomenological issues, whether 20 or not 30 versus 50% of the tubes carry hot flow. To a 21 lesser extent, inlet plenum mixing conditions. There is 22 some influence of heat transfer modeling.
23 We think we under predict heat transfer in the 24 hot leg for the most part. We're probably biasing the n
\_,) 25 calculation towards the tubes in that respect. There 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
155 1 local effects beyond -- I mean, this is a general es 2 statement, but remember that the code calculates an
(
1 N_/ !
3 average hot tube temperature. There are variations with 4 the plume.
5 And in the overall assessment of tube 6 performance that you'll hear described I guess by NRR in 7 the next month or so, we'll talk about how they vary the l
8 SCDAP numbers to account for local variations. But j i
9 generally speaking, what we see is the variations in these 10 issues -- variations in the tube temperatures due to these i
l 11 issues is less than the sensitivity associated with some j 12 of the sequence variations.
l 13 And that within reasonable ranges, or what we j
/N l
' 14 think are reasonable ranges as evidenced from the test I i
15 data, that there is no significant influence on the steam 16 generator tubes. We're still looking at multiple 17 censitivity studies, looking at ranges of the data and l
18 95/5% confidence limits, looking to perhaps combine some 19 of those.
20 But if we do look at that, we will also look 21 at some of these other issues at the same time. You know, l
I 22 we see no reason to push these calculations to the extreme 23 in any one direction. So if you take a 5% confidence 24 limit, it doesn't take too many of those before you get to s
ss 25 such a low probability of the overall sequence that you'll NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS )
1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. O C. 20005-3701 (202) 234-4433
156 1 be in a generally low risk space anyway.
2 MEMBER FONTANA: Okay, well -
t7-~')
3 MR. TINKLER: That's what I have.
4 MEMBER FONTANA: Thank you very much. Any 5 additional comments from the committee?
6 MEMBER POWERS: Just a question maybe. This 7 particular code calculation exercise is looking at the 8 possibility that ordinary severe accidents that happen 9 every day could turn into a bypass accident. What is the 10 status on codes for looking at the bypass accidents that 11 do get predicted to occur, the ones that are initiated by 12 steam generator ruptures with subsequent other failures of 13 these systems?
f3 1
\
N/ 14 MEMBER CATTON: I'll bet you already know the 15 answer to that.
16 MEMBER POWERS: Those sorts of things get 17 called risk dominant accidents, but there's a great deal 18 of uncertainty it seems to me associated with the risk 19 posed by those accidents, and it seems like a lot of the 20 issues that were discussed here today about aerosol 21 deposition have a direct bearing on the risk significance 22 of those accidents.
23 And I wonder what our -- are we going to look 24 at that -- the adequacy of our predictions of the risk
) 25 significance of accidents that do involve steam generator 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 tube ruptures?
2 MEMBER CATTON: I would guess that if you're 7-N_.I 3 able to demonstrate that the conditional probability of !
4 failure of the steam generator tube is less than 10%, you 5 don't have to. )
I 6 MEMBER POWERS: Yeah, it shows up. Steam 1
7 generator tube ruptures do happen. And if they're 8 followed by --
9 MEMBER CATTON: But the way that they're 10 treated or have been treated in the past is that that's 11 just considered an early failure and you go with your risk l
l 12 evaluation. So the numbers are far too high. You really i l
l 13 ought to treat it for what it is, which is you have to 73 k- 14 feed all this stuff into the steam generator up through 15 the separators and dryers and out through the steam line 16 and finally through the valve -- the relief valve.
17 And by the time you do that, I think you're 18 going to get quite a bit of attenuation and you'll have a 19 significantly lessened impact on the public. As a result, !
i 20 maybe you ought to allow the steam generator tube l 21 probability of failure to go up under these circumstances l 22 if what you're looking for is risk neutrality.
23 I don't think that that's been done, but it 24 seems to me that this is a sensible approach is to say, i (n,) 25 "Okay. Look, if the conditional probability of failure 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
158 1 10 percent, I know I am well within the ballpark that I've ,
2 been everywhere else." I don't need to look at it any (g) 3 further. If it's greater, then I think you have to take a 4 look at it.
5 MEMBER POWERS: I think you're coming from a 6 -- you're saying the same thing, I think, which is that 7 the steam generator tube rupture as a bypass accident 8 sequence is overplayed. It's unfairly treated that the 9 risk significance is much lower.
10 MEMBER CATTON: Yes. But if you haven't 11 addressed that question, then I think you have to deal 12 with this one.
13 MR. ADER: Dana, this is Charles Ader with the !
14 staff. Dana, I'm not sure that I fully understood, l l
15 because 1150 looked at steam generator spontaneous rupture l l
16 as a bypass, and when it happens it was one of the risk 17 dominant sequences. !
18 My understanding, although I was not involved i
19 in some of the analysis, but my understanding is the 20 deposition in the transport in that process was maybe --
21 it didn't take credit for all of the areas you could have )
22 deposition in the secondary or in the primary, based on 23 what we might do today. And the numbers that were found 24 in the safety goal space is well below the safety goal, so 0
V 25 I don't think there was the need to go back and try to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASH!NGTON. D C. 20005-3701 (202) 234-4433
159 4
1 1 refine them to take credit for every possible path, i l
s 2 because of the other things they're looking at.
3 Are you asking, are we going back now and 4 seeing if we can do a better estimate based on what we 5 know of the releases from the spontaneous rupture?
6 MEMBER POWERS: I think that's exactly what I 7 was asking, Charlie, was that in 1150, when they did the 8 calculations with the source term code package, they came 9 up with truly horrific numbers, and so they convened an 10 expert panel to get horrific numbers down to just really 11 bad numbers. And it came in as risk dominant, even though 12 its frequency, its not very high, accident sequence.
13 And I guess what I'm asking is is that 14 accident sequence really in this risk base at all, or is i
15 it an overplayed accident? Because I know what was done 16 for the -- to get the really awful numbers down to just 17 bad numbers, and those were pretty -- I mean, best 18 characterized as hand calculations at best, and maybe top l 19 of the head guesses at worst, and as far as I know took no 20 credit for the secondary site whatsoever, and minimal 21 credit for some of the deposition processes that were i i
22 talked about here in connection with intact tubes. 1 23 I wonder if it is -- if anyone is ever looking i
24 at those accidents, or if -- if they have a role to play l i
("')
(, 25 in this new steam generator rules at all. l NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS l 1323 RHoDE ISLAND AVE., N W. )
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
160 1 MR. ADER: We had done in VICTORIA space, in 2 part of the peer review, was trying to use VICTORIA which G/
3 would be, you know, today's better code to do that, but we 4 were not doing it in the context of supporting our support 5 for the rule, because VICTORIA is being peer reviewed, so 6 it's probably a little premature to say the numbers you 7 get out of that should be taken as gospel.
8 The numbers before, as she said, they were
> 9 high when it happened, but it was risk -- the overall risk 10 was low enough that I think people accepted it at the 11 time, and this was trying to look to see is there a 12 scenario that had not been considered before.
13 MEMBER POWERS: This is looking at --
'- 14 MR. ADER: Even though, you know, you have to 15 have station blackout to start it, which is a low 16 probability, and then you have to have all of these other 17 failures. Is there a possibility for a tube bypass 18 induced thermally that maybe we hadn't looked at to be a 19 comparable contribute?
20 MEMBER POWERS: Yes. I think I understood why 21 these calculations were done.
1 22 MR. ADER: I think the answer is probably no, 23 not right now.
l 24 MEMBER FONTANA: Novak, do you have any '
/~N.
( ,) 25 additional comments or questions?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
161 1 DR. ZUBER: Well, if I may summarize. Based 2 on the information I have received from Mr. Dudley and
('-) 3 what I heard today, I have to make the following 4 statements. First, the CCFL and mixing in the inlet 5 plenum are important processes. That's number one.
6 Number two, neither one can be calculated by 7 RELAP on a mechanistic basis. It doesn't have that 8 capability for single phase flow. That's a fact. So the 9 code cannot calculate that. The only way it can -- how do 10 we address the problem?
11 Three, we can do it by tuning the code to 12 well-scaled and well-run experiments. I think this 13 argument can be defended in any technical forum and
S i
14 provide a rational basis for disposing with this issue.
15 Based on the information I have seen, I am not j I
16 convinced that the scaling of the CCFL in the hot leg is '
17 convincing. Information which has been available for 30 18 years shows that this process is determined by two Fraude 19 numbers. The information which is presented at Fauske has 20 only one Fraude number for the hot legs. I don't know the l
21 detail and the basis for just having one Fraude number.
22 Maybe this 2nformation is somewhere else, but it was not 23 presented neither in the -- either in the handout of 24 Fauske, nor at this meeting.
/~N
(_) 25 Consequently, I am not convinced that 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
162 1 scaling is complete or defensible. It may be, but I
. 2 didn't see it. I didn't see the -- I think the mixing 3 fraction, the recirculation ratios, which really govern 4 the process, really, and that are used in the codes are 5 based on steady state. I didn't see that this has been 6 addressed for transient and for different conditions.
7 Again, I think these are very important 8 factors which should be factored in the codes if this code 9 is going to be used for this purpose.
10 And, finally, the approach to help to 11 determine -- a road plan to determine the uncertainties.
12 I don't see how this is going to be done and on what 13 basis. And I think this is the final issue. I think this (3
(-) 14 should be provided by RES to this committee, in a road 15 map, how to do it. Otherwise, this exercise is really 16 just an exercise in -- I don't want to say futility, but I 17 don't see how to resolve the problem without putting the i
18 arms out on the uncertainty and the road map for it. And 19 this was not presented.
20 Oh, the other one is the tube ratio. I mean, 21 the ratio, how it is partitioned. Is this always the 22 case? On what basis? You know, everything is based on 23 one set of experiments. What happens if the scale is 24 changed? And I didn't see that either.
r~s
( ,) 25 MR. TINKLER: I just want to make sure I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS ,
1323 RHODE ISLAND AVE., N W. f (202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
163 1 understand. I don't want to go into this, but do you
,s 2 think the variation in the mixing fractions from the --
3 these here to these over in here, do you think that is 4 sufficient to cause --
5 DR. ZUBER: Well, you see -- wait, wait, wait, 6 wait. You see, that's the reason I asked you, what is the 7 effect? What is the -- at what point does this mixing 8 fraction and recirculation ratio affect the rupture? How 9 close I have to predict it? Am I close to a precipice, or 10 I'm so far I can have a variation of 50 percent, 100 11 percent? You know, this is the basic thing, and this has 12 not been addressed.
13 MR. TINKLER: Well, from here to here I think Q
- 14 was 15 degrees.
15 DR. SUBER: Well, how close do I have to 16 predict what ir, the variation of this fraction? What is 17 the minimum one which will cause, then, the rupture of the 18 tube?
19 MR. TINKLER: Okay. Well, I just want to let 20 you know that we -- you identified that one and you 21 specified the difference between transient and steady 22 state. And I just wanted to --
23 DR. ZUBER: As I said, these questions, some 24 of them, can be a result of scaling the transient versus f'%
(_,) 25 steady state. All can be a result in addressing 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
164 1 rational way it can be done, but I didn't see it.
fx 2 MEMBER CATTON: Mario, let me use the k_' 3 viewgraph for a few minutes, and then we'll let them 4 close. That way they can comment on what I have to say.
5 MEMBER FONTANA: Okay.
6 MEMBER CATTON: I just want to show a couple 7 of viewgraphs before we run and eat lunch. The reason I 8 happen to have them is Mario is going to present my paper 9 in a few weeks.
10 CHAIRMAN KRESS: Speak clearly and distinctly 11 so we can --
12 (Laughter.)
13 MEMBER CATTON: Yes.
I
\/ 14 CHAIRMAN KRESS: Identify yourself and tell us 15 why you're qualified to talk on this subject.
16 MEMBER FONTANA: Yes. What makes you think 17 you can talk about --
18 (Laughter.)
19 MEMBER CATTON: I'm up here with no 20 qualifications, other than that I have some viewgraphs.
21 That's it.
22 (Laughter.)
23 MEMBER POWERS: Have some viewgraphs.
24 MEMBER CATTON: And given that I have
,~n
(,,) 25 viewgraphs, I thought, gee, maybe I ought to just show NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
165 1 them.
,o 2 Okay. The first one -- this is a response
- \
U 3 surface, and after listening to the presentations here I l 4 can see we left a few things out of it. The parameters 5 that were felt to be important were the mixing, the heat 6 transfer from the tube to the secondary side, the transfer !
7 from the hot leg to the outside world, and internal. We 1 8 didn't treat the surge line as a part of this.
9 All of these coefficients are normalized, but 10 they vary from minus one to plus one. So you just need to 11 look at the coefficients to see what is important.
12 It turns out for the steam generator whatever l
13 we did in the hot leg was relatively unimportant. So all i I
[_2')
\ - 14 you see here is mixing and the parameter F and the ,
l 15 external heat transfer coefficient. Internal didn't make ,
l 16 a hell of a lot of difference.
17 CHAIRMAN KRESS: Where did you get those 18 response surfaces?
19 MEMBER CATTON: Created them.
20 CHAIRMAN KRESS: From the data, the calculated 21 data?
l 22 MEMBER CATTON: It was a different kind of l 23 model. Actually, it was a model that was generated first 24 by EPRI, generated by Wassell and Denny and others for
,r 3
) 25 EPRI. We took their model and then made 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
166 1 CHAIRMAN KRESS: Okay.
g 2 MEMBER CATTON: -- everything was balanced out l
(J 3 right. It uses, actually, single phase, stratified 4 countercurrent flow, and, you know, it accounts for the 5 variation in the areas. You get a nice set of equations.
6 Anyway, that's not the point. The point is, 7 what's important? What do these things look like? Just 8 look across. You can see which of these parameters are 9 important very quickly.
10 X1 is the mixing, and the mixing dominates 11 everything. There are even some no linear ones that come 12 in -- the mixing and the external heat transfer 13 coefficient, which again is not surprising, (o
V 2
14 This is only part of the picture. A lot of 15 this has already been -- a lot of the calculations that 16 would be needed to do this have been done. At least 17 that's what I gather in listening to the presentation.
18 What do you do with all of the numbers when you've got 19 them?
20 Now --
21 CHAIRMAN KRESS: These are response surfaces 22 based on a model?
23 MEMBER CATTON: Yes. And this parameter here 24 is the time to failure. Now, it is based on the Larson-b
\ 25 Miller --
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
\
I 1 CHAIRMAN KRESS: Yes.
~ 2 MEMBER CATTON: But that, again, is relatively i 3 unimportant. You pick your model, generate your response 4 surface, you calculate your time. They have a model. I l
5 mean, I wouldn't have chosen that model, but you can't 6 argue with what was put up here. It's certainly a hell of 7 a lot more complete than what we've done.
8 The question in my mind really becomes, what 9 do you do with it once you've got it?
10 CHAIRMAN KRESS: Okay.
11 MEMBER CATTON: Well, I'm not going to go into 12 a lot of detail with this, just to show you part of the 13 process, if I can ever get this thing on right.
p k_/ 14 CHAIRMAN KRESS: Put it down as if you were l
15 reading it.
16 MEMBER CATTON: You can sort of walk through j 1
17 these things now. What this mixing parameter actually is l
18 depends on how you interpret the data. If you do it in a 19 very simplistic way that you have a constant temperature 20 plenum, you would calculate one value. If you have this 21 recirculation, and many streams mixing in different boxes, 22 you get different numbers. This was done in the simplest 23 way possible.
24 When you look at the data you figure, gee,
?n (j 25 plus or minus 10 percent is not unreasonable. In a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234 4433
1 168 l 1 rotated thermalcouple rakes there's a lot of reasons that 2 I've come to the conclusion plus or minus 10 percent is
,7- ,
4
\',/ l 3 probably a reasonable choice. j l
4 The chances that are higher -- are gret.ter 5 than it being lower, and by a significant amount.
6 CHAIRMAN KRESS: What are you plotting there?
7 MEMBER CATTON: This is the chance that you're 8 going to be .7, plus or minus .1.
9 CHAIRMAN KRESS: Okay.
10 MEMBER CATTON: On the upper side, this is the 11 chance t hat you're going to be higher.
12 CHAIRMAN KRESS: Okay.
13 MEMBER CATTON: This is the chance that you're l
N-- 14 going to be lower.
15 CHAIRMAN KRESS: Okay.
16 MEMBER CATTON: And you do this with, for 17 example, the heat transfer coefficient.
18 CHAIRMAN KRESS: Bottom left? l l
10 MEMBER CATTON: This one. This is the next 20 one.
1 21 In mixed geometries, if you just take a oox )
1 22 full of a bunch of stuff and you look at one thing in that 23 box, if it's natural convection, typically you get a 24 number on the order of one in Christian units. If you l (O) 25 convert it to the metric, and I loose touch with it NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE., N W. l (202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
169 1 somewhere around 10, plus or minus, that's sort of a guess s 2 with the typical kind of ranges for this kind of data.
/ 6
And I'd just give it an equal distribution outside that 3
4 range. You can do the same thing on the hot leg on the 5 outer wall.
6 CHAIRMAN KRESS: This looks like the ROAM 7 process to me.
8 MEMBER CATTON: Except it came long before 9 ROAM, 10 (Laughter.)
11 This is just a --
12 MEMBER FONTANA: Don't rise to that. I mean, 13 just --
b 14 (Laughter.)
15 MEMBER CATTON: Tom, you're baiting me, and ,
1 16 that's good bait.
17 VICE CHAIRMAN SEALE: You found the bait.
18 MEMBER CATTON: If you look at the hot leg on !
19 the outer surface, if there's no insulation, you know what 20 the heat tranrfer coefficient is pretty well. If you put 1
21 the insulation on it, you kind of know what it is. The I 22 question is one of degradation at high temperatures and 23 all sorts of things.
24 So what do you do? Give it a -- it doesn't ]
('
( ,N/ 25 look right, but Mario can explain that when he's in Utah.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
170 i 1 The probability of the low number is the insulation. The probability of it being -- you kind of g- 2 V You can do the same thing 3 rationalize the numbers.
4 inside. Inside the tube it's a complicated problem. You 5 really don't know what the heat transfer coefficient is.
6 You know that it's not free convection. It's some kind of 7 mixed convection, but it's mixed convection in a conjugate 8 sense in that you've got to treat the pipe, too. So what 9 the hell do you do with it?
10 Well, what we did for this is we just used the 11 central zone here, the Jackson horizontal pipe, mixed l 12 convection. You know that it's not right, and then you 13 weight it.
(O
/
14 Now, it probably should have been weighted 15 higher above, but this number went up to 200 percent, or 16 twice Jackson's correlations. On the bottom side it only 17 went down to half. So there is a weighting to the high 18 side, rather than to the low.
l 19 Now, my view is that if you don't like these, 20 as we've discussed, you can change them. I sound just 21 like Theofanous.
22 (Laughter.)
23 It's true. And if you're bothered by the 24 bottom line that you get, then you go do the research to
/^
k,)\ 25 fix it. The more ignorance you have, the flatter 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 ,
i i
171 1 curve should be, and there's a direct relationship between 2 what I know and what I get. And if I want to fix it, the 7-f )
~'
3 path is clear.
4 And what does the result look like? Well, you 5 take the --
6 CHAIRMAN KRESS: Bottom line, is that a Monte 7 Carlo?
8 MEMBER CATTON: Yes, it was Monte Carlo with 9 100,000 samples. It's done on a PC. Once you've got the 10 response surface, this is fairly obvious.
11 And here it was just a copy of what Lalouche 12 did for the CSAU. This kind is really easy. There's only 13 four parameters.
10
\~ / 14 Probably should -- I really think after 15 listening to Charlie one ,. the -- a key parameter is this 16 fraction of hot tubes versus the others. That's very 17 important, and we didn't include that. But the goal was I 18 different. Here it was just really to try to exercise the 19 process where you incorporate phenomenological uncertainty .
I 20 into the PRA matrix. l l
21 Once this result was in hand, we just took 22 1150 for Surry, the loss of off-site power accident, and 23 made the right connections. So it's NUREG-1150, plus the 24 steam generator tube rupture with this kind of an j
(
(3,/ 25 uncertainty associated with it. And then it just tracks NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
1 172 1 right through to the bottom line, early fatalities, late 2 fatalities, the whole mess.
7_
U 3 Unfortunately, the steam generator tube l
4 rupture, you assume that it's exposed directly to the J l
5 outside world, and we know that's not the case. You've 6 got dryers and separators and everything else in this 1
7 volume that are going to filter a lot.
l 8 But look at the results. You can see the 9 relief valve closed, relief valve open, and you can see 10 the distribution. If you look at this, absolutely no ,
i 11 question of probability. The probability that it's a hot 12 leg nozzle that failed -- significantly higher.
l 13 Nevertheless, there's a finite chance that it '
)
r~%
k~ 14 could be steam generator tubes. With this, I could make a 15 statement that the conditional probability of failure is 16 less than or greater than some kind of complement. If you 17 don't like the answer, you can go back and improve on the 18 distributions. And to me, this is how you marry whatever 19 code you choose to use with the needs of risk-informed 20 regulation.
21 CHAIRMAIT KRESS: But my question is, can the 22 RELAP S model, as tuned to those experiments, be used to 23 develop a similar response surface that you have here?
24 MEMBER CATTON: Yes, I have no problem with
(_,) 25 that. I think -- again, I think it's inappropriate NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
173 1 because when you do that you carry a lot of excess 2 baggage. There's a lot of things that are done in the n
3 code that you really don't need to address these 4 questions.
5 On the other hand, if you choose to do that, 6 there are a few other things that you can incorporate that 7 would improve and would shrink some of these 8 distributions. It just depends on what you want to do.
9 The simpler your model, the more careful you have to be in 10 selecting the ranges. You can broaden them to account for 11 the simplicity. In the long run, I think it pays.
12 Anyway --
13 MEMBER FONTANA: Thank you. That might be an em I
V) 14 ide whose time has finally come.
15 CHAIRMAN KRESS: Theofanous would have been 16 proud of you.
17 MEMBER CATTON: Yes, he would.
18 (Laughter.)
19 But see, this is really a takeoff on CSAU.
20 CHAIRMAN KRESS: Yes.
21 MR. HODGES: Well, just in summary, I think 22 our major goal for being here today was to demonstrate to 1
23 you that the model we're using is an appropriate model. !
1 24 We're not done with our analyses yet. We've got a lot A
) 25 more work to do, and some of that is coming out from 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 ,
I
174 1 discussion.
- 2 I think we clearly will try to respond to the t
i 3 questions that have been raised by Dr. Zuber and others.
4 Some of the answers we already have; some of them we don't 5 yet. Whether we do a full response surface, or do it 6 another way, I mean, there's more than one way to skin a 7 cat. We will be looking at this for uncertainties. And 8 other uncertainties outside the thermal hydraulic model 9 have to be factored into this, so it's not just the 10 parameter. It's looking at the flow distributions and 11 other things as well, so there's a lot more yet to go into 12 this.
13 We're not done with the answer, and I'm sure
/ )
'NJ 14 we'll have to get back to you at least one more time, 15 talking about our results. Our primary purpose today was 16 to talk about the model we're using to do the analysis.
17 MEMBER FONTANA: Okay. Thank you. ,
1 la Well, I think this was a very illuminating 19 discussion, and we should have had -- the discussion 20 should have occurred regardless of the word " engineering" 21 in the letter. We can agree that the letter stated it 22 probably a little stronger than it should have been. But 23 the result would be the same.
24 I think this was an excellent discussion, and (m
(_,) 25 I think -- I think where you're going, at least to 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
175 1 looks like the right direction. You guys are doing a good n 2 job, in general.
\.j 3 At this point, I'd like to turn it over to the 4 Chairman. Tom?
5 CHAIRMAN KRESS: The Chairman is going to 6 declare a break for lunch at this point, and I ask that --
7 the next agenda item is supposed to start at 1:00.
8 MEMBER BARTON: 1:15.
9 CHAIRMAN KRESS: 1:15.
10 VICE CHAIRMAN SEALE: And it's the 11 reconciliation.
12 CHAIRMAN KRESS: Okay. We can push that.
13 Indian Point starts at 1:15.
I kl 14 VICE CHAIRMAN SEALE: And you're going to be 15 interviewing.
16 CHAIRMAN KRESS: Right. I suspect we ought to I
17 try to be back about 1:15 for the Indian Point, if we can. l 18 Maybe make that 1:30. We'll start 15 minutes late. 1:30.
19 (Whereupon, at 12:29 p.m., the proceedings in 20 the foregoing matter went off the record for a lunch 21 break.)
22 23 24 (O
's.j 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
176 1 A-F-T-E-R-N-O-O-N S-E-S-S-I-O-N
-s 2 (1:32 p.m.)
3 VICE CHAIRMAN SEALE: We're going to jump 4 ahead in the agenda that the members have to item 6, which 5 is properly at this time, or actually a little bit earlier 6 this afternoon, 15 minutes, on our schedule. But there 7 are a couple of things that we should have been able to do 8 this morning, or expected to, that we weren't able to, but 9 we'll get to that later.
10 We're going to hear from the Indian Point 11 Unit 3 restart group. That's John Barton's subcommittee.
12 And, John, do you want to take it from here?
13 MEMBER BARTON: Thank you, Mr. Chairman.
fs
'\_/) 14 The ACRS normally hears from licensees that 15 have had their plants in extended shutdowns for more than 16 a year. IP3 certainly fits into that category. It was 17 shut down by the licensee in February of '93. It was put 18 on the NRC problem plant list as a Category 2 facility in 19 the summer of '93. And it restarted with NRC's comments 20 in the summer of '95, and then had an extended shutdown 21 from September '95 until the spring of '96.
22 This is an informational briefing, and the 23 ACRS does not plan to prepare a letter on this session 24 this afternoon.
O
(_), 25 The licensee has agreed to provide a briefing NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE. N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
I 177 1 on the resolution of the issues that led to the 1993
,_ 2 shutdown and the status of new issues since the plant V 3 restarted in June of 1995.
4 NRC staff will have some representatives here 5 at the meeting, and they have not yet arrived. But they 6 will --
7 MR. SINGH: They're here.
8 MEMBER BARTON: Okay. They are here. They 9 are here.
10 We will be listening to and make comment on 11 the licensee's presentation.
12 At this time, I'd like to turn the meeting 13 over to Mr. Cahill to introduce your staff and who is A
14 going to speak on which issues.
15 MR. CAHILL: Thank you. Good afternoon. l l
16 I'm William J. Cahill, the Chief Nuclear 17 Officer of the New York Power Authority. We're here, as 18 Mr. Barton indicated, to make a presentation on the 19 circumstances that led to the Indian Point 3 being placed 20 on the watch list, and the measures that we've taken to 21 bring this plant from a down condition to one in which 22 it's running quite well recently. And we're looking 23 forward to it being taken off the watch list next year.
24 I promised my staff that I would make the f~^
(v 25 introductory remarks very brief, because I do have 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
l 178 1 extensive formal presentation. I would like to say,
, 2 though, that the New York Power Authority has two nuclear
(
3 plants. And in the period 1993 to 1994, part of both of 4 those years both of those plants were on the watch list.
5 And this led the Power Authority to address the 6 circumstances, including the involvement of top management 7 and the nuclear management, as to what could be done to 8 avoid any recurrence of that.
9 And I just want to say they did take 10 significant measures. There have been many changes in the 11 makeup of the top management of the company and in the 12 resources available to them, chief among them being that 13 -- and this was urged by the NRC and INPO, as well as O
5-) 14 initiated by the Power Authority -- a nuclear advisory 15 committee was appointed to serve to advise the Board of 16 Trustees of the New York Power Authority.
17 And the first committee was constituted with 18 Jim Asselstine and Leon Eliason and Bob -- Mr. Harris. I 19 think his name is Bob.
20 When Eliason went to serve on Public Service 21 of New Jersey, and Harris went to serve with TVA, we 22 reconstituted that nuclear advisory committee, and it now 23 consists of Jim Asselstine, Bill Conway, formerly of 24 Arizona Public Service and INPO and Vermont Yankee, and (3
( ,) 25 Gregory Kane, who was plant manager at North Anna and NEAL R. GROSS COURT REPORTERS AND TRANSCR:BERS 1323 RHoDE ISLAND AVE.. N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
179 l 1 prior to that at Surry.
fs 2 So we cover, in the advisory capacity, to the
( )
3 Board the top governing body of the Power Authority, the 4 regulatory and legal and financial background of Jim 5 Asselstine, the operation, the construction, design, 6 experience of Bill Conway, and the detailed operating 7 experience of Gregory Kane.
8 We have also made many changes in the 9 management of the nuclear power plants. Each has 10 contributed to address the underlying problems, which 11 largely are an inertia to adapt to the changes in the 12 industry which call for a more formal approach to 13 operation, adherence to procedures, procedures therefore O
s- 14 that have to work and be followable, a lag in that -- an 15 acceptance of status quo on the part of operators and 16 everybody above, a lack of a questioning attitude.
17 The succession of changes that we have 18 instituted, with some success and some pauses, but over 19 the last several years a continuous trend of improvement, 20 I think have achieved success.
21 One of the things we have now is at Indian 22 Point, and also at JAF, is an operations organization that 23 is in charge, that is concerned, over their plant and that 24 it is operable, maintainable, and when it isn't that they
/~
k,,S) 25 surface those issues so that they can be corrected. 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
180 1 has resulted in a restart.
2 We started this plant twice, once in '95 and 73
(
U 3 then again this year. In a continuous run, that's now 4 over five months. And we don't judge just by the 5 continuance or continuity of the run but by the way that 6 it is run, and the way that our people have responded to 7 the circumstances that arise in operation.
8 With that, I'd like to ask Harry Salmon, who 9 is our Vice President of Operations, and who heads up and 10 coordinates this team, to chair the presentation.
11 MR. SALMON: I'm Harry Salmon, Vice President 12 of Nuclear Operations for the New York Power Authority.
13 In our team this morning, Bob Barrett is the Plant Manager 14 of IP3, Marc Pearson is the Operations Manager at IP3, 15 John Kelly is our Director of Regulatory Affairs and 16 Special Projects, and he is here in the slide position 17 this morning.
18 And to assist us in the back we have Bob 19 Deasy, the Vice President of Appraisal and Compliance 20 Services; Bill Josiger, Vice President of Nuclear i
21 Engineering and Project Control; George Grochowski, our 22 Director of Reactor Engineering; and Muzafer Karasulu, who l 23 is the Reactor Engineering Supervisor for the Power 24 Authority.
O) i s , 25 MEMBER POWERS: Can I just ask, who of these i
l NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS I 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 1
181 1 gentlemen would have charge over health physics in your f.
2 organization?
t )
3 MR. SALMON: The plant managers and John Kelly 4 and the Director of --
5 MEMBER POWERS: It's not so much an issue as 6 it seems to be an area where you continuously get superior 7 performance. So it's more congratulations.
8 MR. CAHILL: We're going to brag a little 9 about that.
10 MEMBER POWERS: Okay. It sounds like you 11 deserve to.
12 MR. SALMON: The agenda for today will cover 13 what was requested, which is to discuss the reasons we got
/%
14 on the watch list, the process of getting started up 15 again, and what has occurred since our startup in July of 16 1995.
17 The Power Authority was established in 1931 as 18 a public benefit corporation. We pay for capital 19 improvements to the plants through bond issues and our 20 revenues for O&M funding. We get no tax revenue. We 21 provide about 25 percent of the electric power used in New 22 York State on a full load day, and on a low load day we 23 provide up to 35 percent of power. Our capacity total is 24 7,000 megawatts from 12 facilities.
(' , 25 In our organization, we have a Chairman and NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 J l
l 182 1 Chief Executive Officer, who is the lead of a five-man 1
2 Board of Trustees appointed by the Governor. The trustees
,7_
3 represent the various areas of New York State. Bob 4 Schoenberger is our President and Chief Operating Officer, 5 and you've already met Bill Cahill, the Chief Nuclear 6 Officer.
7 A couple of years ago we organized into 8 business units in anticipation of the coming deregulation 9 of the industry, so you'll knew we have transmission and 10 generation. And nuc. gen. is currently a separate 11 business unit.
12 Within the nuclear generation organization, 13 I'm the Vice President of Nuclear Operations. You've met
[\ l
\ms/ 14 Bill Josiger, the Vice President of Nuclear Engineering i 15 and Project Control. And Bob Deasy, who reports directly i
16 to the President and Chief Operating Officer, and for 17 oversight and administration and advising Bill Cahill, the ;
1 18 Vice President of Appraisal and Compliance. Both of the 19 plant managers report to me.
l 20 Indian Point is a pressurized water reactor 21 Westinghouse 4 loop. It was acquired from Consolidated 22 Edison in December 1975, and on August 26th we celebrated 23 the 20th anniversary of the first commercial generation at 24 Indian Point 3.
(A_) 25 There are a lot of significant improvements NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 !
l
183 1 put into the plant over the life of the plant -- new steam 73 2 generators and a blowdown recovery system in 1989. The
~
l 3 significance is that we used low cobalt tubes, and we have 1
4 very low radiation levels in the plant. New main 5 condenser with titanium tubes and tube sheets, full flow 6 compensate polisher system, new feedwater heaters and MSR 7 . tubes, where we went from copper nickel to stainless, and 8 eliminated copper, and the concurrent contribution to tube 9 corrosion.
10 New variable speed circ. water pumps for fish 11 protection and thermal efficiency, and service water pumps 12 for reliability, new low pressure turbines, both fixed and 13 rotating parts. We went to ABB turbines. And new k/ 14 containment fan cooling coils, again going to stainless 15 steel from admiralty brass to fix some corrosion problems 16 that we were having in the coils.
17 How did IP3 get on the watch list? To do that 18 I want to go over this chronology. In January of '91, we 19 recognized decreasing performance at the FitzPatrick 20 plant. The previous year's capacity factor had been 80 21 some percent, and it was down to 48 percent in 1991.
22 There were a series of events that had operational 23 connotations and one significant one where we spread 24 contamination outside of the reactor building. And as a A
'(_j) 25 result of that, in January 1992, FitzPatrick was placed on NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE , N W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234 4433 l ,
184 1 the watch list.
- 2 In October of '92, while FitzPatrick was in
3 the midst of a one-year outage, there was recognition by 4 SALP performance and five escalated enforcement actions 5 during the SALP period that IP3's performance was also 6 declining. We implemented a performance improvement plan 7 in '93, just the time that Fitzpatrick was getting back on 8 line.
9 In February of '93, there was a voluntary 10 shutdown at IP3 because of problems in testing of some 11 safety systems. In May of '93, an NRC special inspection 12 team visit, a confirmatory action letter in June, and 13 continued effort to solve the plant problems with a f
\/
) In January of '94, 14 performance improvement plan.
15 FitzPatrick had run for a year reliably and was removed 16 from the watch list.
17 In April at '94, it was -- we started a more 18 comprehensive restart and continuous improvement plan for 19 IP3. In March of '95, we conducted our own startup 20 readiness evaluation, determined we had made the 21 improvements necessary to start the plan up. That was 22 confirmed by the NRC readiness assessment team inspection 23 in April. The confirmatory action letter was modified in 24 June, requiring that we pause at 40 percent power and do n
k.,j) 25 an assessment of the plant operation at that point, 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 !
185 1 brief the NRC before we continued, and that we pause again s 2 after that. When we reached 100 percent power, do another
\_/ 3 assessment of the plant operation, and present that 4 material, and that was done.
5 The presentation of that material in october 6 removed the -- completed the actions under the 7 confirmatory action letter.
8 In July, we started up and operated for two 9 months, shut down for equipment problems in September.
10 During the effort to heat up again after repairs, we ran 11 into some human performance and procedural errors, which 12 caused us to evaluate and initiate a significant procedure 13 upgrade and operator upgrade.
.O
'-] 14 That effort was complete. We made additional 15 material improvements and successfully started up in April 16 of '96 and have continued in operation since then.
17 What were the issues exactly on that watch 18 list? The SALP scores for the period ending August '92 19 had declined in five of seven areas. We had had the 20 escalated enforcement actions. There were continuing 21 concerns with performance of engineering and technical 22 services. There had been recent changes in the corporate 23 staff, significant changes. And because FitzPatrick was 24 already on the watch list, and the stress on the corporate
('T
( ,/ 25 staff of overseeing two plants with declining performance, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
186 1 Indian Point was placed on the watch list with
,- 2 FitzPatrick.
3 Contributing factors that were cited by the l l
4 NRC were inadequate management controls, untimely and 5 ineffective review and correction of problems, an 6 increasing number of plant deficiencies, and two 7 significant events in the spring of '93. In February, it 8 was identified that the ATWS mitigation system activation 9 circuity -- AMSAC -- had not been properly tested since it 10 was installed in 1989, during that period, and it affected 11 its operability during that entire period.
12 Also, in March of '93, while doing maintenance 13 and getting into midloop operation, the operators
/^N O 14 improperly isolated the reactor coolant system level 15 indication that was operable at the time. They saw an 16 anomaly, tried to troubleshoot it without a formal 17 procedure, and isolated the operable level system.
18 The NRC inspection team, special inspection 19 team, reported out in May, and the root causes that they 20 identified were weak management processes and control, 21 weak management skills and vigor, and weak independent l
22 oversight.
23 MEMBER APOSTOLAKIS: Excuse me. ,
1 24 MR. SALMON: Yes?
O 25 MEMBER APOSTOLAKIS: What does " weak V
NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS i I
1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433 l
187 1 management skills" mean?
2 MR. SALMON: In this case, they felt that
/-~\
t t
~#
3 managers were -- we had a lot of people who had come 4 through operations or through engineering, and although 5 good technical people, they were not being successful at 6 setting high standards, at holding people accountable for 7 meeting those standards, at taking a whole. lot of problems 8 at once and prioritizing and getting solutions made in a 9 timely manner.
10 So it was a sense that the managers, many of 11 whom had grown up through the plant organization over 15 12 years, needed more management and supervisory training, an 13 opportunity to rotate, and some experience from outside in
(~% l l
k -} 14 benchmarking against the industry.
15 MEMBER APOSTOLAKIS: Thank you.
1 16 MR. SALMON: When this SALP report came out in i
1 17 the fall of 1992, the Power Authority put into place a l l
18 performance improvement plan. It looked at the issues and 19 tried to put in place corrective actions that would solve 20 those issues.
21 After going on the watch list, the NRC and the 22 Power Authority sat down and modified this plan and added 23 a whole series of issues that had to be resolved prior to 24 startup to the plan. And by December of '93, it became
,a
() 25 obvious that the plan that had been put in place and the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 PHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C 20005 3701 (202) 234-4433
188 1 actions that were being taken were not being fully 2 successful at solving the issues and getting us to a point 73 i/ 3 of being ready to start up.
4 We stepped back, brought in additional 5 resources, did a thorough root cause again, and put into 6 place the restart and continuous improvement plan. This 7 was presented to the NRC in May of '94. It started B showing improvements. Additions were made to it in 9 November, and by the spring of '95 this plant had been 10 successful in bringing the plant and the material 11 conditions and the operations of the plant to a condition 12 that we were ready for startup.
13 The Power Authority conducted its own
,O 14 evaluation, and the restart and assessment team came in 15 and confirmed that we were ready to restart.
16 MEMBER POWERS: You indicate that the 17 difference between PIP and RCIP is a more thorough root 18 cause analysis.
19 MR. SALMON: Yes.
20 MEMBER POWERS: Is that really the only thing 21 there? Was there more to the difference between these two 22 programs?
23 MR. SALMON: I think there were a number of 24 factors. One is we brought in a very active new plant g
( ,/
25 manager in the spring of '93. We looked at the issues 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
189 1 we saw them in the spring of '93, and the performance c 2 improvement plan was not aimed -- had not been aimed at V)
(
3 restart in the plant. Remember, it was put out before the 4 plant shut down. It was aimed at fixing programmatic 5 issues on a long-term continuous improvement basis.
6 When we shut down in February of '93, because 7 of the testing problem, we started to look at extent of 8 condition and the formality of the processes and found 9 there were a number of things we needed to fix. We tried 10 to patch the performance improvement plan but really 11 didn't get to the core issues.
12 At FitzPatrick we had had a significant event 13 that got everybody's attention. We put contamination
( )
C/ 14 outside the building, and that was evident to everybody.
15 And so the denial that you were in trouble was --
16 (Laughter.)
17 -- quickly washed away. Okay? Here, because ;
18 it turned on a test in a specific system, I believe that I
19 we had a period of denial for almost that period of time 20 of people saying, "Well, that's just over there. It 21 really doesn't apply to me." And by going through the 22 more in-depth review, we got down into what are the core 23 issues? What are the personnel performance issues? What 24 are the weak skills? How much is management turnover?
A I
(/ 25 Will a lot of new people in the jobs hurt us in the NEAL R. GROSS COU3T REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
190 1 conduct of business? And we got down past the denial part 2 of it and got into the core of we needed to make 3 fundamental improvements here.
4 MEMBER POWERS: I guess what I'm really 1
5 interested in is your view on whether having these 6 programs be successful is very, very dependent on the 1
7 person in charge of it or the organization of it.
8 MR. SALMON: I think it -- if you have core 9 buy in by the line management, who has to make the 10 improvements, it takes hold quicker and probably solves 11 the problem more deeply.
12 In the case of the performance improvement 13 plan, it was almost a classical TOM approach. It was very O
U 14 heavily chart and graph intensive, and there were times 15 when, initially when they got started, some of the 16 managers spent a couple hours a day getting the data 17 together to plot the graphs and make the run charts and 18 try to make the improvements.
19 And going down to Indian Point at the time, 20 from FitzPatrick, you were impressed by the -- all of the 21 graphs. And there was also an effort to schedule 22 everything. And so to make sure there was good use of 23 resources they had a very detailed schedule, brought in a 24 lot of -- the new plant manager brought in a lot of
() 25 people.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASH!NGTON, D C. 20005 3701 (202) 234-4433
l l
191 1 l
1 We got to the point at one point where we were
- 2 scheduling when the trash got emptied. Everything was
(
3 going to be on there. It quickly became clear that that 4 was too detailed a level, but we had brought in a new guy 5 with a lot of experience, and he had used these techniques 6 elsewhere and was working with those. And I think it 7 masked -- there was so much effort put into the process 8 that -- and long with the denial, that it prevented us 9 from getting to the core issues.
10 MR. CAHILL: The short answer is its people.
11 You need the programs in the organization, but the perfect 12 program, the perfect organization, without the drive and 13 initiative of the individuals -- and it's not just one, A I
(
k_/ 14 but a group, and often a succession -- has changes 15 initiated, sometimes not vectored exactly in the -- what 16 ultimately was the right direction. But at least to 17 change and get -- overcome the initial frictions of 18 change. l 19 Without that kind of skill, drive, initiative 20 that people provide, the programs in the organization are 21 barren.
22 MEMBER POWERS: The interesting question is 23 when you have an imperfect organization, and imperfect 24 plans, can that be overcome by superior people?
/ \
Yes, of course.
(_) 25 MR. CAHILL: Yes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
192 ,
l l
1 MEMBER POWERS: So that's your experience
,S 2 here, is that correct?
.]
3 MR. CAHILL: That's my experience. ;
4 MEMBER POWERS: It has a real impact, because 5 you cannot regulate people's personality and drive. You 6 can regulate organizations and things like that, and what 7 you're telling me is that's not the vulnerable item here.
8 MR. CAHILL: Good programs, good organization 9 help. It's an obstacle to overcome poor programs and 10 organization. But the right skills and drive can even do j 11 that, and it's -- that's the reality of everything.
12 Nothing is perfect.
13 MEMBER APOSTOLAKIS: When you developed these O)
\- 14 programs, did you have outside help, or was it done 15 internally? l 16 MR. CAHILL: There was outside help, more i
17 before -- or just as I was coming there was consultants l 18 and -- l 19 MR. SALMON: There was significant outside 20 help in the development of the restart and continuous 21 improvement plan from December of '93 through March or 22 April of '94.
23 MEMBER APOSTOLAKIS: Now, the more thorough 24 root cause analysis, how thorough is it? How -- I mean,
.C'\ 25
'q_) it seems to me that you can go on forever trying 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
193 .
i 1 for causes. Eventually, say, that somebody failed his 1 l
.s 2 course working as an undergraduate, I mean, that could be
! )
%J Where do you draw the line?
3 a root cause.
4 MR. BARRETT: Can I try, Harry?
5 MR. SALMON: Go ahead, Bob.
6 MR. BARRETT: You're right. You can ask --
7 I've heard a lot of theories on this. You can ask the 8 question, why five times? You can go to 10 times. But, 9 in reality, the effectiveness of a comprehensive and 10 successful root cause program is do your actions to 11 prevent recurrence really prevent recurrence? That's the 12 true litmus test. And that is the measure.
13 If you continue to have recurrence of an x- 14 issue, you haven't fixed it. You didn't get to the root 15 cause. I think that's about as simple and most practical 16 answer there is to that question.
17 MEMBER APOSTOLAKIS: So -- I don't think 18 you're saying that, but the true test, then, is to see l
19 whether the thing happens again. l 1
I 2C MR. BARRETT: No. That is one way of doing 21 it. The other way to do it is to have experienced people ;
I 22 that understand the technology, understand the process, l 23 have been on the front line before, and have had 24 experience and have success in solving similar problems. ;
(~s \
(_,) 25 And you know what doesn't work. Unfortunately, that's NEAL R. GROSS l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
194 1 part of a learning process is making mistakes. You put a 2 corrective action in place and it misses target, but you 73 U 3 need to be flexible and to go back and revisit that and 4 recognize that that didn't fix it, we need to do more, and 5 you reopen that issue.
6 There's a lot of methodologies that have been 7 in place. INPO has been a big proponent of the human 8 performance root cause analysis program, and I subscribe 9 to that. And that gives you some insight.
10 But the real measure, after you're done and 11 you think you have the answer, you're only human, you may 12 have made a mistake, is to continue to monitor. And if it 13 resurfaces, you need to go revisit that. So in a sense, tO
-- 14 you're right, something that has never occurred before, 15 you're in virgin territory. You may miss the target on 16 your corrective action. You may find you have to go a 17 little further. But it's to have that dynamic a program 18 and open a program that is so critical that allows 19 reopening issues that you believe initially are closed is 20 one aspect of it.
21 MEMBER APOSTOLAKIS: So when something happens 22 that requires root cause analysis, is it part of the 23 standard analysis to worry about human error, obviously --
24 MR. BARRETT: Yes.
(D
(_) 25 MEMBER APOSTOLAKIS: -- but also perhaps 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
i 195 l 1 maybe other things have contributed, the context that led I
~ 2 to the human error, perhaps organizational issues, and so O
3 on? So is that something that is the exception or 4 something that routinely is considered and dismissed 5 sometimes, or most of the time perhaps?
6 MR. BARRETT: It's routinely considered. I 7 would say -- I've been in this business a while. I would 8 cay in the '70s it was typically you would give the person 9 a couple of days off and that would solve the problem, 10 those of us who have been around that long, and that used 11 to be the way to solve problems. If a human was involved, 12 that in reality it could have been your procedure was 13 poorly written, it wasn't clear, you may have had poor i \
2 14 lighting, the wrong tools, inappropriate guidelines, and 15 the like.
16 And one of the ways to more directly get to 17 the effective correction action is to present the root 18 cause to experienced people and have them listen to it, 19 and judge it, and comment on it before you initially 20 attempt to close your root cause analyses, to take 21 advantage of that wealth of experience that's out there in 22 the organization.
23 MEMBER APOSTOLAKIS: Okay. Thank you.
24 MR. SALMON: We made some improvements. The
(~h ,
\ ) 25 principal improvements that we made to -- that got us to NEAL R. GROSS COURT REPORTERS AND TRANSCRl8ERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASH:NGTON. D.C. 20005-3701 (202) 234-4433 l
196 I
1 restart in '95, in the areas that the special inspection 2 team had noted, in management processes and controls we
[~~b
~
3 clearly communicated the management expectations to the 4 staff. We instituted an effective corrective action 5 program. We made significant improvements to the 6 surveillance and preventive maintenance programs, to 7 address the issue of -- that we found in AMSAC and 8 additional issues that were found in the surveillance 9 r gram.
10 We moved engineering and licensing to the site 11 from the corporate headquarters to get them closer to the 12 plant and be able to provide more direct assistance as 13 issues came up. And we improved plant material condition b
k/ 14 to an acceptable level to support restart.
15 MEMBER BARTON: Harry, how many corrective 16 action systems do you have at the station?
17 MR. SALMON: We now have two. We have one 18 that identifies material items, so those are problem 39 identification sheets and work requests. And the other is 20 a single system called the deficiency event review system, 21 which gathered up all of the other systems that we 22 previously used and goes to a single system with a single 23 count.
24 This is one we imported from -- it came from k)s 25 Virginia Power through Detroit Edison through Niagara NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
197 1 Mohawk, as we benchmarked the industry. And Ralph Sylvia 2 had arrived at Niagara Mohawk, they had had a watch list
[- 3 experience come off, and the DER process looked like one l
4 that would be helpful.
5 MEMBER BARTON: Thank you.
6 MR. SALMON: In the management skills and 1
7 vigor, we established a full management team. Remember, there had been concerns about management turnover when we I 8
9 were placed on the watch list. Management training in i l
l 10 development issues were undertaken. There were a number i
11 of short courses for supervisors and managers, to assist
]
4 12 them with management skills and communication skills.
13 And we had achieved a questioning attitude and I
\-- 14 attention to detail on the part of the staff. They were 15 beyond denial and they were looking at helping identify i 16 the problems, and routinely identifying 90 percent of the i l
l 17 DERs that were written -- were being written by the staff l 18 in the field, not by QA, and not by outside agencies.
1 19 In independent oversight, we enhanced the 20 effectiveness of the quality assurance organization and 21 its acceptance by line management. That had been a 22 weakness when we went on the watch list.
23 We expanded the membership and the charter of 24 the Safety Review Committee. We established the Nuclear
,a 25 Advisory Committee for the trustees, as Bill Cahill had i]
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON,0 C. 20005-3701 (202) 234-4433
198 1 mentioned. We also established an independent safety
- s. 2 engineering group, which we had not had previously and is 7
3 not required by our current tech. specs. In review, the 4 Review Committees were now providing effective oversight.
5 MR. MARKLEY: Harry, just one question before 6 you move along too far. With your old improvement program 7 and your new one, what kinds of things were you measuring 8 then as opposed to later? You said you had a tremendous 9 volume of things, and especially things like human 10 performance, which is kind of a hard thing to measure a 11 lot of the time anymore.
12 MR. SALMON: We recently -- and this was just 13 as we started up in '95 essentially -- help me with that,
/ 14 Bob. Were -- I think that was -- no, it was '94. In94, ,
15 as we started the restart and continuous improvement l 16 program, we started -- we imported a process of measuring 17 human performance errors from Texas Utilities. Bill 18 Cahill came from Texas Utilities.
19 And we measured human performance errors per 20 10,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> and keep -- and we used the same goals that 21 TU was using, so that we could benchmark against them.
22 And we've watched those human performance error rates come 23 down over the last year, so that we're down into the very 24 solid range that represents good industry performance at
/~N
(._,) 25 Indian Point 3.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
199 1 We started up in June of '95, and we delayed f3 2 startup in June to replace the reactor vessel head 0-3 rings. Just as we were getting ready to start up we 4 realized that one of them was leaking. And rather than 5 start up after that period of time, delayed and replaced 6 the head 0-rings.
7 During the startup, or when we were partially 8 up in power, we had leakage from the safety relief valves.
9 And during that period, we reduced pressure to try to 10 reseat the relief valves. We operated that way for two 11 days, and then in review realized that we didn't have a 12 plant-specific analysis that allowed us to be in that 13 operating region.
(,, l
\/ 14 We shut the plant down, reseated them, shut 15 down, started up again, and then did a root cause ,
16 evaluation of the -- of why we got into that position.
17 And the reasons that we found were that we had -- one of 18 our procedures didn't specify the low pressure limit in 19 that condition, and we misapplied one of the alarm i
20 response procedures that would infer you could use --
21 operate at low pressure, but didn't have a design basis 22 behind that statement.
23 Once we got into it, and intended to only be 24 there for a couple of hours, we stayed there for two days.
q
(,) 25 That's an inadequate adherence to the procedure.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234 4433
200 1 We misapplied the technical specifications.
,w 2 There are statements in the technical specifications that
' )
%.) 3 say, "You are analyzed for all departure from nuclear 4 boiling safety limit if you are below these curves." But 5 that was a safety limit curve; it wasn't an operating 6 curve. And, again, we misapplied and misinterpreted that.
7 Part of that was due to improper -- what we 8 later figured was improper management influence. We had 9 several people in our organization who had come from 10 plants who did have a plant-specific evaluation and had 11 operated their plants in that condition before. So when 12 presented with that, they thought that they forgot what 13 the basis of doing it at their previous plant was and
(~
(h '
i 14 thought that with the words and the interpretation of our 15 tech. specs. we were on solid ground in that point.
16 We found that because we lad a strong 17 management team in place that they ha.d -- that they were 18 making the decisions as operational decisions and weren't 19 using the plant Safety Review Committee as frequently as 20 they might. And so we modified the guidance and made sure 21 people understood what was appropriate for line management 22 to do, and at what points -- take a more robust look at 23 using the Safety Review Committee as a backup to say, 24 "Okay. Now, let's go look at this reparately."
(,,/ 25 We identified poor communications, and there NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234 4433
201 1 were people in the organization who knew that this was not
- 2 an analyzed place. We didn't get to the right people.
b 3 Reactor engineering was one, and we took an easy answer 4 rather than getting all the way down to all of the people 5 in the organization who could have helped us with the 6 proper decision.
7 And it demonstrated that with that many new 8 people in the organization that haven't customized tech.
9 specs. that there is a continuing challenge to assure that 10 the people operating and making decisions are pulling all 11 of the strings on the tech, spec. and the basis, and the 12 FSAR and the basis, to make appropriate decisions for our 13 plant with our tech. specs.
p k_ / 14 We did additional training, and we have made 15 some changes. We have put in a tech. spec. change to put 16 pressure limits into the tech. specs. and taken additional 17 actions. We also analyzed that place where we were, and 18 getting a specific analysis te. were -- were not at risk at 19 that point. We just didn't have the analysis to prove it 20 when we went there.
21 Are there any other questions on that?
22 We have a project underway to go to the 23 standard tech. specs. What we put into our -- what we 24 have tried to put in is a line item improvement -- is put
, -m 25 in the words from the standard tech. specs. in our current NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
202 i i
1 one, and the limits, so that we'll -- as we submit 2 standard tech. specs. for approval next year, we will t
,_ s !
3 already be feeding them in in places where they can help 4 our operators in the current timeframe.
5 In September, we shut down to repair a 6 generator hydrogen cooler. And during heat up, during the 7 mode change, to come out of cold shutdown after that 8 repair, we left several safety-related pump switches in 9 pull-to-lock, which made them inoperable for their )
10 function. And, therefore, when we realized that, we went ;
i l
11 back to cold shutdown to reassess that operation. It i 1
l 12 happened that that was the week we were talking with the i 13 NRC on our assessment of our previous operating period. l
/ ,
_ 14 Later on, there was another event which was of l
15 interest, and that was during heat up. The closed cooling 16 isolation valve was left closed. We were having trouble j l
17 maintaining temperature, so we shut the discharge valve on 18 the service water side of the system. It locked pressure 1
19 in between the discharge valve and the check valve in the 20 system, lifted the relief valve, and the relief valve 21 stayed open, didn't reseat fully. So we leaked water out 22 through that relief valve for three shifts without 23 catching it on shift turnover.
24 That was another event, and so there were l
()
/~N 25 these three events which all resulted in enforcement NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS 1323 RHODE ISLAND AVE., N W. .
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
i l
203
/ 1 action, but which all told us we needed to spend more time ,
1 2 with operators and procedures. So we extended the outage 7s i )
~ 3 to improve operations, procedures, and personnel 4 performance.
5 The next slide slows the self-assessment !
i 6 conclusions from our 100 percent power self-assessment as !
7 we briefed them to the NRC, and this is -- this was having 8 had the low pressure event, having left the switches in 9 the wrong position the weekend before this presentation, 10 and before the issue with the closed loop cooling.
11 We said that the plant systems had performed 1 12 well, and the plant staff responded effectively to 13 emergent work. The plant staff continued to use good use j 1
/~N a l
\/' 14 of the corrective action program. However, we need to fI l
1 15 improve the timeliness and effectiveness of the corrective 16 actions. The threshold was good. Staff was identifying 17 the items. We were following up, but we weren't as timely 18 as we needed to be.
l l
19 Conservative decisionmaking was evident at all l 20 levels, but we needed to improve methods and use of the 21 resources for making informed decisions. Reflects not 22 getting to the people in reactor engineering who could 1
23 have answered the question properly for low pressure 24 event.
((_,)/ 25 Management expectations were clearly l
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
204 1
1 communicated, but additional emphasis on individual ,
j 2 accountability was required, and that's still one of the
(
~
3 keys to continued success is building ownership and 4 individual accountability in individuals throughout the j l
51 .
organization. l 6 Personnel qualifications were adequate, but we 7 needed increased training on the importance of the j 8 licensing and design basis considerations. Again, l 9 reflecting the low pressure event. And there has been 10 significant training in that.
11 As a result of those events and that 12 assessment, we went into a program of procedure upgrade 13 and operator upgrade, and Marc Pearson came aboard as the vs 14 Operations Manager at IP3. Marc will tell you what he 15 found and what were the actions we took to resolve these 16 issues that were identified in the fall.
17 Marc?
18 MR. PEARSON: Thank you, Harry.
19 Good afternoon. When I arrived at Indian 20 Point 3 in the autumn of last year, I conducted a review 21 of the documents regarding operatior.a, including critiques 22 of the three events Mr. Salmon discussed. I also 23 conducted discussions with personnel from the Nuclear 24 Regulatory Commission. the Institute of Nuclear Power (O
() 25 Operations, and, of course, New York Power Authority.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234 4433
205 1 Four major problem areas were identified and 2 are listed here. A low level of ownership was causing 7-~
i 4
%J 3 reduced operator performance. A lack of planning was \
i 4 reducing the probability for success. Inadequate training
{
i 5 effectiveness was preventing training being used to l Last, the lack of effective ,
6 improve operator performance.
7 self-assessment reduced the opportunity to improve.
8 So what did we do? In a rapid manner, we 9 created and communicated expectations in some fundamental 10 areas. First, was the need to focus on every job and to 11 perform at a high quality level. Second, that a low 12 threshold for reporting problems was imperative, coupled 13 with the need to be willing to accept criticism. Third,
[D
- 14 and one we stress very strongly, was strict adherence to 15 procedures, including requiring that all procedures used 16 to operate the plant be at the operating location and be 17 followed step by step. Fourth, we increased personal 18 accountability and coupled that with openness in i 19 communications.
l 20 Several other corrective actions were also 21 implemented. We commenced the performance of the i
22 immediate investigations for operator error events. We 1 23 were supported by corporate office personnel in doing k
24 that,
( )
25 We acquired resources from several station and
\
l
( ,,f NEAL R. GROSS COURT REPORTERS AND TRANSCR!BERS f 1323 RHoDE ISLAND AVE.. N W.
(202) 234-4433 WASHINGTON. D.C. 20005 3701 (202) 234-4433
i 1
206 1 corporate organizations to support procedure improvements I l
2 weekly, simulator examinations and training, evaluated by i
. (cs ,
3 operations management, as well as weakly ops management 4 discussions with crews in training.
5 A personnel resource plan was developed, and 6 resources were provided to support it. Contract personnel 7 with extensive operating experience were placed on-shift 8 as mentors, to provide improvement feedback to on-shift 9 personnel as well as to manage them.
10 Similar to, but completely separate from the 11 mentors, other personnel were assigned to shift to provide 12 a broad safety oversight function, and report to station 13 management.
(
(_) 14 Overall, the station organization was 15 realigned culturally, to support safe operation of Indian 16 Point 3.
17 MEMBER BARTON: Can we go off this line, and 18 back to procedural compliance? How do you continuously 19 reinforce -- since there's been a long history compliance 20 issues at IP-3, how do you continue to reinforce that 21 message?
22 It's nice to have a policy that says "you 23 shall comply with procedure verbatim," or whatever your 24 words are, but, you know, people have short memories, (Q_j 25 half-lives are short, whatever. And, since you've had NEAL R. GROSS i COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 l
! 207 1 this problem for many years, what are you doing different 2 now?
7 i
3 MR. PEARSON: Last fall and early winter, we 4 actually re-educated the operators to an understanding of 5 why procedure compliance was necessary.
6 Those operators who were not on-board with 7 that philosophy have either changed, and are on-board with 8 the philosophy, or were placed in other positions while we 9 continued mediation for them.
10 Ene-itic actions that we did, and continue to 11 do was we hao a et of procedures called system operating 12 procedures, which were, up until that time, categorized as 13 reference use procedures. I made them all continuous use t'%
- 14 procedures, and the operators understood very clearly what 15 that was.
16 In addition, up in the weekly simulator 17 examinations that we give, which are done at the exact 18 same level of rigor that the annual requalification exams 19 are given in the simulator.
20 In fact, we use the NRC initial license 21 operator exam criteria, where you get grades of three, 22 two, and one. We do that every, every week.
23 I normally -- if not I, then one of my office 24 management personnel supporting, me go up there, along
(~~).
(, 25 with training management, do those evaluations, and there NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON D C. 20005-3701 (202) 234 A433
208 ,
l 1 were numerous remediation plans that came out of that.
2 In addition, the shift mentors were 7x k
3 specifically directed to be monitoring them, and they 4 report weekly. At the end of each week, they write a 5 report, and I also sit down once a week with one of them, 6 along with the tactical assessment group, which is the 7 other oversight group. They also give written reports.
8 There's a tremendous amount of feedback.
9 But to go back to the first thing I said. Re-10 education of the operators, so that the underlying need 11 and reason to follow procedures was accomplished. Thank 12 you.
13 MR. CAHILL: Let me just reiterate, though.
O
'NY 14 The procedures have to be followable. So we, where they 15 couldn't be, there was this problem with compliance, with 16 what doesn't work, we rewrote the procedures and we held 17 up work, as necessary, until the procedures were right.
18 MR. PEARSON: Mr. Cahill himself, along with 19 Mr. Schaumberger and other members of the station and 20 corporate management, came and discussed with the 1
21 operators, along with, while I was talking to them, and 22 sometimes separately, about the need, and no matter what i l
23 the impact on schedule was, we had to meet the fundamental 24 need of following procedures.
,y
(_,) 25 MEMBER BARTON: Thank you. I understand.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
209 1 MR. PEARSON: Significant improvemente. And 2 what were our results? Significant improvements and real 7-(_) 3 changes occurred, and continue in the areas of procedural l
4 compliance, shift turnovero, training effectiveness, 5 attitude and professionalism of the operators, as well as 6 three-point communications on-shift.
7 We acquired significant additional resources, 8 including eight new, non-licensed operators presently in 9 training. We completed two initial license exams, one in 10 December of '95, and one in April of '96, resulting in all 11 11 candidates receiving licenses.
12 This, by itself, supported significant 13 improvements in operations by reducing overtime and p)
's L
14 providing developmental opportunities. We also commenced 15 two more license classes, one in January and one in April 16 of this year.
17 We completed several actions to strengthen 18 operations management, experience, and capability. For 19 example, a former shift manager from Indian Point 3 was 20 rehired and requalified as a shift manager.
21 Indian Point 3's SALP period ended on March 22 second. Operations was evaluated as a Category 3, with 23 positive comments concerning management oversight and 24 initiatives taken during the latter part of the period.
1 (0) 35 In early April, the Indian Point 3 plant was NEAL R. GROSS COURT REPORTERS AND TRANSCR;BERS j 1323 RHODE ISLAND AVE., N W. i (202) 234 4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
1 210 1 successfully started up. Overall, we were successful in e-s 2 focussing the station organization on safe operations.
L/
3 This completes my part of the presentation. I'd like to 4 introduce Mr. Robert Barrett.
5 MEMBER BARTON: Some of the members might not 6 understand three-point communications. Would you --
7 MR. PEARSON: Give an example?
8 MEMBER BARTON: -- explain what you mean by 9 that?
10 MR. PEARSON: Sure. Okay, I'll be the board 11 operator. That means you've got to give me the order.
12 MR. BARRETT: Okay, Marc, I'd like you to go 13 over there and start Bravo reactor coolant pump.
O 14 MR. PEARSON: Start Bravo reactor coolant 15 pump. I understand.
16 MR. BARRETT: That is correct.
17 MR. PEARSON: Then I go over, and I start the 18 Bravo reactor coolant pump. And oh, by the way, we're 19 teaching self-checking very animated, so we can see that 20 it's being done. And then I will report " Bob, the Bravo 21 reactor coolant pump has been started."
l 22 MR. BARRETT: Understand Bravo reactor coolant j i
23 pump is started.
24 MR. PEARSON: That's correct.
(n)
(_/ 25 MEMBER BARTON: Thank you.
I l
l NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
211 1 MR. BARRETT: Now, I'd like to refer you to 7s 2 the calendar you have in the, it was in the front of the 1
() I 3 book that was before you. ,
l J
4 From time to time, I will refer to, to events 5 and times, and I think that will be helpful from a time 6 sequence, to tie the, the discussion about the results 7 together for you.
8 I'm going to first present the results of the 9 SALP report that covered the period of, that ended in 10 March of '96, this year, and then I'll talk about the 11 period thereafter.
12 So first, I'll start, I'll talk about the, the 13 most recent SALP report. The period covered by our most A
i \
\_/ 14 recent SALP went from August of '92 to March of '96, but a 15 significant period in time was not included in the SALP 16 report because the plant was under a confirmatory action 17 letter. That was from March of '93 to May of '95.
18 So the SALP really only covers from May of '95 19 to the March '96 period. The scores are as depicted 20 before you. Operations received a score of three; 21 maintenance, a score of two; engineering, three; and plant 22 support, one.
23 In the operations area, as Marc had mentioned, 24 in the management oversight, significant noteworthy (Aj) 25 efforts were directed by management to improving operator NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON. D.C. 20005 3701 (202) 234-4433
212 1 performance during the last three months of the period.
2 It noted that material conditions in the plant
[-
3 significantly challenged the operators.
4 In the area of procedures, it noted 5 significant progress had been made in the procedure 6 upgrade program, and procedure adherence was found to have 7 improved.
8 And, in the tech spec and design basis 9 knowledge area, it noted that weak understanding of the 10 licensing and design basis of the plant existed.
11 In the engineering area, the SALP score of 12 three in the engineering, as you recall. Response, I
13 response to emergent problems impacted long-term O
t
\_ / 14 engineering programs. Engineering work backlogs 15 increased.
16 Routine performance varied significantly, 17 sometimes poor. Engineering was not successful in 18 improving work prioritization and backlog controls, the 19 set point control program and configuration management.
20 It noted that poor communications contributed 21 to low, to the low pressure of that, and noted that 22 management was aware of these problems, but these problems 23 were not effectively addressed.
24 And maintenance. Maintenance received a SALP t'~\
score of two.
(_) 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234 4433
213 1 MEMBER BARTON: Bob, before you go on to 2 maintenance, let me ask a question in the engineering.
y 3 And the SALP also noted communications problems between 4 engineering and operations, two key areas --
5 MR. BARRETT: Yes.
6 MEMBER BARTON: -- of any plant's overall 7 performance. Are you going to talk about what corrective 8 actions you've taken, to improve that?
9 MR. BARRETT: Yes.
10 MEMBER BARTON: Later on?
11 MR. BARRETT: Yes, I will.
12 MEMBER BARTON: Thank you.
13 MR. BARRETT: In the maintenance area of SALP,
,a
( )
's / 14 management oversight of maintenance was good. They noted 15 conservative decision-making.
16 They noted a broad review of extent of 17 condition of equipment problems. Emphasis by workers on 18 procedure adherence was noted, and there was good 19 supervisory oversight.
20 However, work planning was not fully 21 effective. There was a need for better questioning 22 attitude during tests. Again, the increasing maintenance 23 backlog was noted. And also observed the declining 24 material conditions since the restart in July of 1995.
/m 1
(_) 25 In the plant support area of SALP, and I've NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N W.
(202) 234 4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
214 1 got it broken up into the five areas, because, really, the 2 five major groups make up the plant support area. Plant g
U 3 support received a grade of one.
4 Radiological protection. They noted continued 5 strong performance. That excellent information was made 6 available to workers, ALARA performance was very good, and 7 a lot of these were the results of improvement, successful 8 improvement initiatives.
9 In the security area, improvements were due to 10 management attention. There were few low-significant 11 security events during the period. Overtime and 12 compensatory measures were minimized.
13 Emergency preparedness. They noted a well-U 14 qualified and -trained staff. They noted a very good off-15 year exercise. Dagraded conditions in the emergency 16 preparedness area were promptly identified. And it 17 commented that the transition to the NUMARC emergency 18 action levels, EALs, were effectively managed.
19 We should note that NYPA took the, the lead in 20 New York State, for all three, three plants, over the 21 three nuclear sites in the, in the state, for the, for the 22 conversion, and the combined efforts of the NUMARC EALs.
23 Fire protection. NYPA effectively gained and 24 maintained control of their program. They noted initial (m
V 25 resolution of the emergency light issue was weak, but that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
i I
215 i 1 success in this area resulted from a task force that was 2 formed, and oversight by management.
7-s !
(_s In the area of housekeeping, it noted mixed 3
4 results. The high traffic areas were well-maintained.
5 However, when you got into the out-of-the-way, less-6 travelled areas, management expectations did not appear to 7 be fully developed.
8 I covered the SALP period. I'm going to now 9 talk since that period. And the principle improvement 10 areas are noted on the next slide. There's really five of 11 them. That doesn't mean we aren't focussing on a broad 12 spectrum of areas, but I wanted to particularly talk about 13 these. These are the more significant.
(O N '/ 14 Accountability, both individual and 15 departmental. A material condition, it was referred to 16 several times in SALP, and, you know, we do 7 ave a large 17 backlog.
18 Teamwork. Conservative operations. And one 19 thing I want to say here, and this is really a, operations 20 being in the driver's seat, an operations mentality here, 21 too.
22 Historically, at Indian Point 3, you'll learn 23 that, that operations wasn't in the driver's seat. It l i
24 was, it was maintenance, and we have put operations back 7-w
(_ 25 in the driver's seat, as Mr. Cahill has, has already NEAL R. GROSS ,
l COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
l 216 1 mentioned.
2 And communications. In the area of 7-( 'l 3 accountability, there are several examples that talk about 4 improvement in accountability at Indian Point. Let's take 5 the management observation program. This program was 6 identified as weak in an NRC report that was issued in 7 January of 1996.
8 We have made significant improvements since 9 that period. Over 2,900 individual management 10 observations have been performed since the beginning of 11 February of this year.
12 Every deficiency is documented on the 13 management observation form, and correct actions are
(~N.,
\/ 14 typically corrected during the observation, or, or shortly 15 thereafter, or they are tracked to resolution.
16 This program has, has increased management i
17 presence in the field and coaching for the workers.
18 Managers, therefore, are communicating expectations I 19 directly with the workers.
20 Schedule adherence. You're talking, 21 typically, a 12-week rolling schedule, your weekly, weekly 22 schedules. Our performance in April of '96 was in the 50 23 tt 60 percent range. It has now improved to the 70 to 90 24 percent range, and is improving on a weekly basis.
/^
(_)g 25 MEMBER BARTON: Is that across the board, 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
217 1 is that any one area far exceeds the others, and kind of 2 brings your overall numbers up?
7s
( )
3 MR. BARRETT: Across, it's across the board, 4 John, and the next one is, is, really explains why, 5 because managers are accountable. What we have is we have 6 the overall site score, if you would, for the preceding 7 week, and we discuss this in our Wednesday POD, plan of 8 the day meeting.
9 And each department is, receives a score, and 10 then in a bar chart. And then each manager is accountable 11 to discuss and explain the causes for why they did or did 12 not meet the goal for the week.
13 And what we invariably find, almost 100 e~s
( )
\/ 14 percent of the time, is that the reason Department A 15 wasn't successful is because Department B didn't provide 16 some necessary support, whether it be operations, 17 maintenance, engineering, radiological controls.
18 And that is a very good interactive 19 discussion. And that is, I think, is one of the reasons, 20 through that communication, why we're starting to improve.
21 It was interesting that one of the reasons 22 for, why maintenance was, did poorly three or four weeks 23 ago was because two 50.59 evaluations were not performed 24 by engineering.
r%
( 25 But what we found was that was not even on the NEAL R. GROSS CoVRT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
218 1 schedule, and co that went back to scheduling. They
,s
- 2 failed to put that into the schedule, and therefore that 3 caused a burp in the maintenance area.
4 Another area for accountability is, if you 5 look at the cover sheet for the plan of the day handout, 6 they, they now have individual names against each of the 7 priority items listed there, things such as the station 8 priorities, the limiting conditions for operations that 9 are scheduled.
10 And we typically do not take voluntary LCOs.
11 These are just LCOs that are required, due to the, due to 12 the surveillance test program, or a preventive maintenance 13 program. And any plant special evolutions, such as 2
0 2 14 grounded containment entry, major plant maneuvers, 15 etcetera.
16 overdue assignments have been reduced at 17 Indian Point. Our action commitment tracking system, 18 ACTS, we call it. Back in April, there were 138 overdue 19 items. This, yesterday's result -- and we review these on 20 a Wednesday, 23 are now overdue.
21 Deviation event reports, DERs, that have been 22 open graded in 60 days. And this is where you have an 23 event, you need to evaluate it, and, and it's a part of 24 our corrective action program. Graded in 60 days. And we
-[ '\
(_/ 25 had 38 of them graded in 60 days, back in April. It's NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234,4433
219 1 seven of them today.
gx 2 MR. MARKLEY: Bob, one quick question. You've
( ,
)
3 got as lot of things you're checking, and schedules, and 4 measuring, and items, like you can count 2,900 here, but 5 how do you ensure that things have improved, as opposed to 6 being completed?
7 I mean, what, what levels of accountability 8 assurance are you doing, with regard to verifying 9 something? You know, if you had a problem, something's 10 actually improved, as opposed to just, you know, do you 11 check that off your list?
12 MR. BARRETT: Okay. That's a good, good 13 question, Mike. We have -- in our DER process, we have l' '/ 14 what we call trend DERs, and part of our, our analysis is, 15 when we code the deviation event reports, we put them in 16 the various categories, classifications, and we analyze 17 that on a regular basis, weekly.
18- And, where we see an event recurring, we will 19 generate another DER that, based upon the fact that it is 20 now a trend, and now -- and we elevate the priority of 1
21 that.
22 All trend DERs must be closed in a 23 presentation to the plant leadership teau. So that it 24 elevates that to a higher level. And I -- from, from an (3
\_) 25 overall standpoint, I think that's the best answer I can NEAL R. GROSS COURT REPORTERS AND TRANSCR:BERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASH!NGTON. 0.C. 20005-3701 (232) 234-4433
220 1 give, for that.
<x 2 MR. SALMON: We also have a set of indicators (v) 3 that we track, which say here are goals that reflect top 4 quartile performance in the indu" ry, and these are the 5 ones we want to' achieve in our 1c..q-term improvement of 6 the plant.
7 And we manage those, and look at them monthly, 8 and determine whether we' re achieving our goals in the 9 areas, and if not, then I have to brief the President and 10 Chief Operating Officer on why we didn't achieve our goals 11 in the certain areas we wanted.
12 So we go back into them, look at is it 13 process, is it individual performance, is it some support 14 issue? And we, we look at each one of those, and, and 15 assess why it is we're not achieving our goals on a 16 monthly basis.
17 MR. CAHILL: Also, if things don't work, the 18 operator, operator's in control mentality that we have 19 now. They are the user.
20 And that's coupled with the new attitude of 21 willingness to surface and complain about deficiencies, so 22 that it's heard, and, and, in tracking systems, but also, 23 literally heard, and people complain to each other, which 24 is a healthy and, and effective interaction.
b(_/ 25 MR. BARRETT: Okay. On the next page, when 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
221 1 talk about material condition. Our corrective maintenance 7s 2 backlog is high.
3 Our, our non-outage backlog totalled 1,600 in 4 April. We're working to reduce that. We're currently 5 slightly less than 1,250, and we have a plan in place to 6 reduce this to less than 700 by the year's end.
7 Before we restarted the plant, earlier this 8 year, the entire backlog was assessed, and this is both 9 from an individual standpoint and from an aggregate 10 standpoint.
11 Additionally, we, we paid particular attention 12 to each of the, our 23 risk-significant systems. And we 13 tabulated, for every deficiency in the plant, if there was
?
1
\/ 14 an impact on the operations departmen_ in the area of off-15 normal operating procedur's or on emergency operating i 16 procedures.
17 And, based upon that, we're satisfied that it.
18 was safe to, to return the Indian Point 3 to operation.
19 This was earlier thic year.
l 20 MEMBER BARTON: On your maintenance, I was l
21 just counting work week costs, what's your assessment of 1 22 how many months of backlog work does that represent?
23 Because, you know, the work week cost can be --
24 MR. BARRETT: Right.
f#%
( ,) 25 MEMBER BARTON: -- tightly screwed 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 j
222 1 maintenance overall on the pump.
-s 2 MR. BARRETT: There's, there's a lot of those,
/\
U 3 John. We can't give you a good number on that right now.
4 We're improving our work control process. From a 5 productivity standpoint, we just put in place a FIN team.
6 I would say that the numbers we could give you right now 7 would not be representative.
8 MEMBER BARTON: It's something that you're I
9 aware of, and --
10 MR. BARRETT: Yes, we are.
1 1
11 MEMBER BARTON: -- are going to pay attention l 12 to, just not counting.
13 MR. SALMON: Backlog based on the numbers, f .
i ; .
'V 14 John, is about, is about six months backlog.
15 MEMBER BARTON: Okay.
l 16 MR, SALMON: But we're, we're working towards 17 estimates, putting man-hour estimates on these jobs, and 18 then we can, we can do an, an estimate from a resource 19 basis.
20 MEMBER BARTON: Okay.
21 MR. SALMON: We have that capability in our 22 system. We're filling out the database, to be able to 23 make that an ongoing part of our planning process.
24 MR. BARRETT: Our high priority work requests.
. f~;
- V 25 We're at less than a hundred today, as compared to, they NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234 4433
223 1 were over 150 in June. But we are focussing on the higher 7s 2 priority ones, and the ones that impact operations is the f )
\_/
3 next bullet.
4 Our focus is on eliminating operative 5 impairments, operative work-arounds, control room 6 deficiencies, etcetera. There are -- you walk in Indian 7 Point's control room, you'll find no locked-in control 8 alarms at 100 percent power.
9 Only one area of the plant requires an 10 operator, during the routine performance of his rounds, to 11 put on protective clothing, and we're working on 12 eliminating that area.
13 Operator work-arounds receive high priority.
I,,h
\/ 14 There are 18 non-outage operator work-arounds today.
15 There's a total of 39 in the plant. Control room 16 deficiencies are under control. There are 31 of them that 17 are non-outage. A total of 51.
18 operations performs a quarterly aggregate 19 impact assessment of all impairments. If they, if they 20 see a trend developing, then that will be accelerated to 21 monthly.
22 One of the things that can come out of this is 23 a decision to shut the plant down, and make, to make 24 repairs, if the, if any one warrants that. That's, we're m
/ \
(,) 25 not only looking from -- each, any one individual item NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON D.C. 20005-3701 (202) 234-4433
224 1 could be totally unacceptable, and, and this process will,
)
fw 2 will help identify that.
I 1 U' 3 We recently established a Fix-It-Now team, and 4 productivity is steadily improving in this area. We have 5 a senior reactor operator, a non-licensed operator, 6 several maintenance mechanics, an electrician, I &C 7 support, and two planners, and a maintenance supervisor.
8 This is our Fix-It-Now team.
9 Our engineering backlog continues to increase,
! 10 and we've taken steps to address this. Staff augmentation 11 is underway, to reduce this backlog, following an action 12 plan that is in place.
13 And temporary r.odifications in the plant, the (n)
\/ 14 number of them do remain high, but we feel they are under 15 control. Several of these are -- for example, each leak 16 repair to be installed will be counted as a temporary 17 modification. Now, that's, that's typically, primarily 18 why our temporary modifications are, have gone up 19 recently.
20 In the area of teamwork. Marc discussed 21 briefly, did discuss the procedure adherence issue 22 earlier, right after he became the manager of operations.
23 Prior to and during initial several months of operation, I l
24 we, we staff a 24-hour-a-day safety reviewer, and support
,m
/
(_j) 25 the operating shifts.
NEAL R. GROSS I COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234 4433 l
l
225 1 As a result of implementing strict procedural i
n <
2 adherence, when you come across a procedure that can't be bl l 3 followed as written, you need to fix it in order to 4 implement that program. And that's how we supported that. I 1
l 5 We have an effective emergency plan, and we !
6 have a, a severe weather action plan that we developed 7 this past, past winter, and that has been successful l
8 during the several storms both in the winter and in the l 9 summer of this year.
I 10 Twelve-week schedule support is focussed, and i 11 that, and that requires site-wide teamwork. And we're 12 improving significantly in this area, as noted before.
13 Tech spec improvements. Our plan is to i I
t )
\J 14 implement standard tech specs in 1998. We're going to 15 submit them in 1997, but, for implementation, 1998.
16 In addition to that, we have over a dozen, 12 17 to 15 line item changes to the technical specifications, 18 to improve, and to bring our tech specs into the current 19 age, if you would. The Indian Point 3 tech specs hadn't 20 been changed, line by line, as things evolved over the 21 years, and kept up with the industry.
22 VICE CHAIRMAN SEALE: This is probably a good 23 place to ask a question I've been brooding about over 24 here.
(m )
25 Earlier, in your operation SALP report, there
(_/
NEAL R. GROSS C'iURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
226 1 was a comment that tech spec and design basis knowledge
,3 2 was weak. And then, a little bit later, there was a
( )
3 concern about configuration management, and also 4 communications.
5 I'm curious. Do you, who owns -- we talk 6 about ownership in plant operations. Who owns your design 7 basis?
8 MR. SALMON: The director of design 9 engineering at IP-3 owns the design basis.
10 VICE CHAIRMAN SEALE: Okay. Does he have a 11 process, and have you examined the process by which that 12 desi.gn basis is superimposed on all engineering changes, 13 procedure rewrites, and all of those other things, to
/3 i
is) 14 maintain your confidence that you are living within the 15 design basis requirements?
16 MR. SALMON: We have, we are making 17 improvements that address of all these issues.
I8 VICE CHAIRMAN SEALE: Yes.
19 MR. SALMON: In those very areas. One of the 20 things that we, we have recently completed a, a partial 21 review of FSAR alignment with procedures. We find some 22 improvements that can be made, go back into our process, 23 and put additional hooks in the processes, --
24 VICE CHAIRMAN SEALE: Yes.
(A) _,
25 MR. SALMON: -- that will assure that we make NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
1 227 1
1 those connections every time.
fs 2 VICE CHAIRMAN SEALE: Well, as you do these
( \
() 3 tech spec improvements, that's clearly one of the things 4 that -- ,
l 5 MR. SALMON: Right.
l 6 VICE CHAIRMAN SEALE: -- you'll get a chance ;
7 to exercise.
8 MR. SALMON: Yes.
9 VICE CHAIRMAN SEALE: And it might be very 10 worthwhile to be very sensitive to that as you, as you use 11 them, and employ them to review the process. !
I 1
12 MR. SALMON: Yes. We're hoping -- yes. j l
13 VICE CHAIRMAN SEALE: As well as to change the n ,
I \
V 14 tech specs.
1 15 MR. SALMON: We're going forward with an FSAR l 16 review of representative systems. i 17 VICE CHAIRMAN SEALE: Yes.
18 MR. SALMON: And we're hoping to coordinate 19 that closely with this tech spec improvement program.
20 VICE CHAIRMAN SEALE: Yes.
l 21 MR. SALMON: We're also talking with, closely l 22 with -- Indian Point Unit 2 is also in the process of 23 this, and we have a very tightly aligned design basis, so 24 we're, we're working with them to try to leverage all our
\
(')
V 25 efforts, and get the best product 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
228 1 VICE CHAIRMAN SEALE: Okay.
7 ~g 2 MR. BARRETT: We have weekly system health
( )
3 presentations by our system engineers. Our plant 4 leadership team attends these, as well as other key 5 department managers.
6 And, during these meetings, the system 7 engineer receives recognition. He interacts directly with 8 the senior managers on-site. Decisions are made, and 9 priorities are established.
10 With regard to training support, line staff is 11 assigned to training as adjunct instructors. People such 12 as non-licensed operators, maintenance supervisors, and 13 engineers.
-- 14 Training attendance. First-time offered 15 training attendance is graded at 98 percent, year-to-date.
16 This is a big team effort, because this reduces the amount 17 of time the training department has to focus on make-up 18 sessions.
19 We've achieved superior plant chemistry 20 performance. And this is not just the chemistry 21 department building. This is site-wide. You heard Mr.
22 Salmon talk about some of the improvements.
23 But, for those of you familiar with the 24 chemistry indicator. The best you can achieve in (3 For each of the last four
() 25 chemistry indicator is 1.00.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
229 1 weeks, the numbers are 1.03, 1.00, 1.01, 1.01. And this i is the result of across the board focus by the entire site 73 b 3 i on chemistry.
4 Conservative operations.
5 MR. SALMON: Let me, let me interrupt for a 6 minute, Bob. The -- you had asked the question before, 7 about the SALP period, that the operations and engineering 8 weren't communicating properly.
9 VICE CHAIRMAN SEALE: Yes.
10 MR. SALMON: This, these system engineering 11 presentations have, have really made an improvement there.
12 The operators have confidence that the system engineers 13 are, are paying attention to their issues. And, by having
14 switched to this operations focus, we've really made a 15 bridge for the engineering department.
16 MR. PEARSON: In addition, the, we've 17 commenced integrating some of the engineering and operator I
18 training. So that the engineers and operations people are 19 sitting in the same classroom sometimes, going through the 20 same training.
21 VICE CHAIRMAN SEALE: Excellent.
22 MR. PEARSON: Okay.
23 VICE CHAIRMAN SEALE: Excellent.
34 MR. PEARSON: Another thing. Going back to (p_) 25 the procedural compliance question, also. One 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
230 1 things I got across to my people was, if the schedule said fS 2 seven o' clock in the morning, they were supposed to be in 3 the classroom and commence training, I viewed their not 4 showing up on time as indicative of their procedure 5 compliance mentality.
6 If they all showed up on time, and the 7 instructors show up or. time. And that has nothing to do 8 with procedures, but I connected it.
9 VICE CHAIRMAN SEALE: By the way, is the tech 10 spec revision, are you going to the new tech specs?
11 MR. BARRETT: Yes.
12 MR. PEARSON: Yes, we are.
13 VICE CHAIRMAN SEALE: And now, so is Indian
( )
14 Point 2 also?
15 MR. BARRETT: Yes.
16 MR. PEARSON: Yes.
17 VICE CHAIRMAN SEALE: And Fitzpatrick is in 18 this?
19 MR. BARRETT: Yes.
20 VICE CHAIRMAN SEALE: The boat with you, too?
21 MR. BARRETT: Yes.
22 VICE CHAIRMAN SEALE: Good.
23 MR. BARRETT: We will submit in 1997.
24 VICE CHAIRMAN SEALE: Okay.
O
() 25 MR. BARRETT: And hope to implement in 1998.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005 3701 (202) 234-4433
231 1 VICE CHAIRMAN SEALE: Okay.
2 MR. BARRETT: We intend to implement in 1998.
(,_s) 3 MR. KELLY: We have project teams assigned to 4 go to the plants. They are in place, with some contractor 5 support, have been for several months. And it's all been 6 fully funded, and, and the schedule's already out, and now 7 being complied with.
8 MR. BARRETT: Conservative operations. As I 9 said before, we've put operations back in the driver's 10 seat. Management walks the talk, and demonstrates its 11 focus on, on safety, and has done so this year.
12 For example, on,'on March first, when we had 13 problems, and had to take some instrument air compressors r~N, (s_/i 14 out of service and do piping mods, it was required that a 15 back-up air compressor be brought in as a contingency. On 16 March 12th, the plant was at normal operating temperature 17 and pressure.
18 We were attempting repairs to recharging water 19 header. And repair -- and it, we believed it, through the 20 job planning, that that job should have been able to be j 21 performed under those conditions. It presented challenge j 22 to the operators, and the direction was given, return the 23 plant to cold shutdown to effect repairs.
24 On July first, when we had a need to replace 1 ,) 25 one of our travelling water screens, the initial focus was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4430 WASHINGTON. D C, 20005-3701 (202) 234-4433
232 1 to use a barge-mounted crane. Management direction was to
,3 2 bring in a land-based crane, rather than to bring in a
('~') 3 barge, with the plant at full power, and park it close to 4 the intake structure.
5 Now, this was done at a considerable 6 additional expense. Somewhere in the neighborhood of 7 500,000, versus 125,000 dollars it would have cost for the 8 barge crane.
9 On July 13th, we reduced power to 45 percent, 10 to recalibrate a reactor coolant pump flow transmitter, in 11 order to avoid a challenge that it might present to the 12 operators, above that power level.
13 On August 25th, we reduced power to 60
/~N
? I
\_/ 14 percent, in order to reduce the dose rate in the area of a 15 pressurizer spray valve that required repair. That, that 16 reduction in power reduced the dose by over two person rem 17 for the entire job.
18 Risk-significant activities at the station are 19 reviewed twice a day by the plant leadership team. Safety 20 system unavailability is minimized by focus. Focus on the 21 LCO duration, as you saw before. Earlier, I referenced 22 the LCO coordinators. This is a new initiative we put in 23 place at Indian Point that helps us to minimize LCO 24 durations.
r
(\) 25 MEMBER POWERS: When you say that the risk-NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON, D C. 20005 3701 (202) 234-4433
233 1 significant activities are reviewed twice daily, that
-s 2 means the day shift and the swing shift?
's. J 3 MR. BARRETT: The plant leadership team meets 4 twice a day. And, in the morning meeting, we discuss the 5 risk-significant activities that are coming up during 6 today, and in the near future, and, in the afternoon, we 7 review what's coming up this evening, and tomorrow.
8 And, if it's a weekend, then we'll go through 9 to Monday. And we make decisions on that. We change, we 10 change, we change activities as a result of those 11 discussions.
12 MR. PEARSON: In addition, it also provides 13 calibration to the people who are reporting to us, because (r)
N- ' 14 we may think that something's risk, which they don't, and 15 that improves communications and expectations there.
16 MR. BARRETT: We put this in place. If you 17 look at the calendar. The blue period is, I failed to i
18 mention, the blue color is the plant on-line. If you look i 19 at May 20th, there's a two-day shutdown.
20 There's a control shutdown on May 20th. That i
21 was a result of a work activity that had risk-significance 22 associated with it. A suage lock fitting associated with ,
I 23 an operator for a main steam isolation valve was being 24 tightened, to fix an air leak.
A
( ,) 25 It was noted in the job package as a plant NEAL R. GROSS COURT REPORTERS AND TRANSChedERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433 j l
234 1 trip risk, but that was not, that risk was not i 73 2 communicated to upper management of the station. This is
?
G 3 one of the corrective actions we're putting into place.
4 That two-day period was an extremely positive 5 experience for Indian Point 3. That was one. If you look 6 at the blue, you think the blue is where we wanted to 7 celebrate.
8 I tell you, we need to celebrate that two-day 9 period. Not only the lessons that we learned from it, but 10 the proof to the plant that they can turn the plant around 11 in a short period of time, and that was what this plant 12 needed, the staff needed to learn, that their own 13 abilities. So I think that was a very positive period.
\~- 14 We've, we've dedicated an individual, we've 15 assigned an individual shift technical advisor to each of 16 the operating shifts. Previously, this was a dual role 17 function of additional, of addition, of one of the other 18 watch standers on shift.
19 Operations performs an additional control 20 panel walk-down, after shift turnover, and after the 21 turnover meetings.
1 22 During, sometime during the shift, an l 23 individual will walk-down the entire panel, control room 24 panels, looking for any problems. This is another lesson A
( ,) 25 learned, as a result of some events at Indian Point. And NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHoDE ISLAND AVE., N.W. i (202) 234 4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
235 1 this has been very successful.
73 2 The fire brigade leadership has been taken
! a b 3 away from the control room operator, and we've assigned it 4 to an individual on-shift, but that does not have any 5 operating shift responsibilities, so.
6 We will likely achieve the lowest three-year 7 average radiation exposure at the end of this year. We're 8 presently -- in the industry. We're presently at 18.24 9 person rem at Indian Point, for year-to-date. Our goal 10 for the year was to be less than 34. We'll likely come in 11 somewhere around 25 for the year.
1 12 MR. PEARSON: You really need to repeat that !
I 13 for Mr. Powers. l
("y l
(-) 14 (Laughter.)
15 MEMBER POWERS: You guys don't want to come 16 around Department of Energy reactors.
17 MR. BARRETT. John, how would you like to have 18 been able to say that?
19 MEMBER BARTON: Yes, that is, that's a good, 20 that is quite an achievement.
21 MR. BARRETT: In the area of communication.
22 Effective, we believe there are effective operations pre-23 and post-shift turnover meetings. These have been a 24 significant upgrade at Indian Point, under Marc's
.O
) 25 leadership.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
f 236 1 Plant-wide meetings are structured and well-fs 2 coordinated. Plant-wide tailgate meetings are held (o) 3 weekly, using an established, issued agenda. We've re-4 established the monthly Indian Point 3 project and Indian 5 Point 3 engineering meetings.
6 MEMBER POWERS: I guess I'm not too sure what 7 you mean by a tailgate meeting.
8 MR. BARRETT: Tailgate meeting is you stop 9 work, and you take approximately an hour out of the day, 10 and you discuss. It's a communication tool for, to talk 11 with each of the shops, each of the workers, and each of 12 the department managers are required to be at the tailgate 13 meetings, and these are led by the supervisors.
s
- 14 And there's an, an agenda. Typically, you 15 talk about what's going on at the station. What 16 challenges are we facing, what are successes or problems 17 that we have. Human performance, safety issues, and so 18 forth.
19 MR. SALMON: Tailgate, as I, my understanding 20 is, comes from the line crews, who daily do it on the 21 tailgate of the pick-up truck.
22 MEMBER POWERS: I thought maybe you were out 23 at Shea Stadium or something.
24 MR. SALMON: No.
rs
( ,) 25 (Laughter.)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N W.
(202) 234-4433 WASHINGTON D C. 20005-3701 (202) 234-4433
237 1 MR. BARRETT: And it's two ways. And input is 2 obtained from --
7-N-l 3 MEMBER POWERS: Tailgate meeting, that's 4 right.
5 MR. BARRETT: Managerc, department managers 6 present quarterly self-assessments of cheir departments 7 during the regularly scheduled plan of the day meeting.
8 MEMBER BARTON: How is that coming? That's a 9 new program, right?
10 MR. BARRETT: Yes. It's a new program. We 11 started, it was introduced at Fitzpatrick. We're doing it 12 at, at IP-3.
l 13 And it's very instructive and informative, I)
\- 14 because departments realize what other departments are l 1
15 struggling with, and why, why it's important for them to 16 support, when asked from another department to lend a 17 hand, how they can see their action goes toward that.
18 This includes everybody from public relations, 19 to finance, to operations, to engineering. Every 20 department. Security. We believe we have an effective 21 speak-out program.
22 A performance enhancement review committee 23 meets weekly, to review human performance events. Human 24 performance events receive PLT, plant leadership team
,m
( ,) 25 attention daily, and are discussed weekly 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
l 238 l 1 tailgates.
7\ 2 I want to talk a little bit about human 3 performance, because there's been a little, there's some 4 interest in it. I want to talk about operations, 5 operations department.
6 In the January, February, March time frame of 7 this year, their human performance rate was three and a 8 half times the site average. Three and a half times the 9 average for the site, per 10,000 hours0 days <br />0 hours <br />0 weeks <br />0 months <br />.
10 The last three months, June through August, 11 their rate is only 47 percent of the site average.
12 They're less than half of the site. That is due to a 13 tremendous effort in that area.
/ \
t i
\ '/ 14 And the whole site is coming down. The whole 1
15 started at a rate of 1.8. At the beginning of those same 16 three months, the site average was at 1.8 rate, per 10,000 17 hours1.967593e-4 days <br />0.00472 hours <br />2.810847e-5 weeks <br />6.4685e-6 months <br />. Now, the site is at .63. So the site rate has 18 come down.
19 MEMBER BARTON: What's changed in that period 20 of time? If, that's effective since when, data since 21 when?
22 MR. BARRETT: This year. January, February, 23 March, compared to the latest three months.
24 MEMBER BARTON: Now, is that operating versus
(~n
( ,)
25 outage period?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE.. N W.
WASHINGTON, D.C. 20005-3701 (202) 234-4433 (202) 234 4433
239 1 MR. BARRETT: It is operating versus outage.
2 MR. PEARSON: That's, but that's part of it.
,S 3 However, when I took over late last year, as I talked 4 about, I said lessons learned was a big, fundamental area 5 we talked about.
6 And, just as we educated the operators on the 7 need for procedure compliance, we also educated the 8 operators on the need to report problems, and, through 9 reporting of problems, we can then fix them.
10 And, you know, high capacity factor and high 11 SALP scores are the result of certain fundamental things, 12 and that was one of them. And the fact is, there are 13 certain operators that have to be massaged, be brought k/ 14 back into the fold.
15 And there was one case where one operator had 16 been causing more than his share of him, and we removed 17 him from shift for two and a half months, remediated him, 18 he regained the use of his license, and he was restored to 19 being able to go on-shift.
20 MR. BARRETT: I think it's the result of, in 21 the operations, specifically what Marc is talking about, 22 and a tremendous focus on identifying problems. From, 23 from a site-wide standpoint, it's doing thorough root 24 causes and effective corrective action. So we have had O)
(_ 25 several stand-downs this year.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005 3701 (202) 234-4433
l 240 l 1 When we see a trend -- our most recent trend, g3 2 we saw a slight increase, and we decided to hold a stand-
'v) 3 down, site-wide stand-down, to discuss human performance 4 and the focus that needed to be redirected. And we found 5 that to be quite effective.
6 Our, our -- I'm going to talk, I started 7 talking communications. But we have a pre- and post-job 8 briefing. The pre-job briefing offered has been 9 tremendous in reducing human performance events, for, for i
I 10 plant maintenance activities, both in operations and in 11 maintenance.
12 The management observation program. The 13 people being out in the field, the supervisors, coaching, O
'- 14 communicating expectations.
15 I don't think there's any one cause for this 16 improvement. I think it's an across the board focus. I l l
17 think it's operations being in the driver's seat. I think 18 it's the support from engineering, for, for procedure 19 changes. There's a lot of reasons.
20 VICE CHAIRMAN SEALE: Have you identified the 21 number of times when you have, there has been a 22 requirement for a supplemental training module, or 23 something of that sort, that's been generated because of 24 an operations requirement?
/~%
i i
(_/ 25 That is, where the people have said "we need NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234 4433
241 1 to have some more training in how to do this process,
,_s 2 before we go out and try to do it in the plant."
( .)
'~# 3 MR. PEARSON: Yes. We call that just in time 4 training.
5 VICE CHAIRMAN SEALE: Okay.
6 MR. PEARSON: And we instituted pretty 7 rigorous just in time training from the aspect of -- if I 8 could give you the theory.
9 VICE CHAIRMAN SEALE: Yes.
10 MR. PEARSON: It is that let's make believe 11 we're going to be doing an evolution tonight. And we get 12 together in operations and say "okay. Let's assume we do 13 the evolution, and we make a mistake. What would be our C\
t
's) i 14 corrective action from doing, making the mistake?" And 15 we'd say "oh, let's get some training." Great. Let's do 16 that training first.
17 VICE CHAIRMAN SEALE: Yes.
18 MR. PEARSON: Then do the evolution, then we 19 won't make a mistake. In fact, it's gotten to the point 20 where we were doing, about three months ago, turbine stop 21 and control valve testing. And the reactor operators on 22 the shift, the night before -- we try to plan and tell 23 people what's going on, they asked for adjustment time 24 training.
(/^N
,) 25 (Laughter.)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433 J
242 1 VICE CHAIRMAN SEALE: That's exactly what I
- 2 was asking.
4 k#~/ 3 MR. PEARSON: And you might be surprised by 4 this, but we gave it to them, along with the whole group, 5 and it went off great.
6 MR. BARRETT: That's a real good anecdotal 7 story. I have a slightly different one. Earlier, when 8 Marc had started that, I recall coming down the hill, at 9 Indian Point, walking along, and I passed some operators 10 who were heading up to training and that. And they were 11 kind of mumbling, complaining about having to go to just 12 in time training.
13 About three hours later, I saw them in the I
( 14 control room, and I asked them how it went, and the story x-15 was dramatically different. They said that was the 16 greatest. "Now, I really appreciate the fact that they 17 took the time to do that."
18 MR. PEARSON: In fact, we even have the non-19 licensed operators show up for the just in time training 20 that occurs with the simulator, for the licensed crews, 21 and they've come back and greatly appreciated that.
22 MR. MARKLEY: What kind of changes did you 23 make in your work control process, to give control back 24 over to operations?
25 MR. PEARSON: Now, back to education. 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
243 1 education, by Bob and the rest of the management, to the
,f g 2 work control personnel that we had to, operations and the
- \
t '
\_) 3 rest of the organizations had to work together to control 4 risk, and to plan things so we don't trip in place, in l l
5 front of the operator's challenges. For example, too many 6 activities going on simultaneously. That type of thing.
7 We also have transferred some operations 8 people into the work control organization, and we want to 9 do more. But we put some good people in there, and have 10 created a lot of discussions.
11 The risk, the risk discussions between work 12 control that Bob's talked about, helps out an awful lot.
13 I spend a lot of time talking with the work control people !
i
/~% <
l
\~ 2 14 myself.
l 15 MR. BARRETT: Operations sets the priorities.
16 MR. CAHILL: On the highest level, I think top 17 management can emphasize what's important, in many ways.
18 But one way is to put -- if you're in the operating phase, 19 which we are in now, and we hope to stay in this for the 20 next two decades, put operating-experienced, operating-21 oriented top management in.
22 Harry was the plant manager of an operating 23 plant, which he got out of the watch list. And Bob was 24 the operations general manager of that plant, and we
("s
( ,) 25 brought him to Indian Point, first as operations manager NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
l 244 1 and then as plant manager.
s 2 And, and we put Marc in a position where he
! \
~
3 could assert the operations viewpoint. And we told, I 4 told engineering, and maintenance, and all the other 5 supportive functions that they are support of operations.
6 So management can emphasize, and, of course, 7 it changes with the phase. In the, in the, in early 8 design of a project, engineering properly is in control.
9 No construction, no operation.
10 It shifts to construction, and then to start-11 up, and then finally, into operation, which changes from, 12 from a transient, which starts out, and peaks, and 13 disappears, to a continuum of ever-increasing search for,
[/)
\- 14 for excellence in operation.
j 15 Now, we're still pursuing that, but we've made 16 a tremendous investment in that. You see the, the chart 17 there. The first part is red, and the quarter prior to 18 that was red. As we shifted that orientation.
19 And Bob mentioned one time, that we were 20 leaving cold shutdown, and enter this hot shutdown and 21 start-up, that we turned around and cooled down. That 22 takes time. And it's a disappointment to, to those who 23 are just rooting for the plant to start up.
24 And we did that several times during that
/~ l
(_,N) 25 time, once for a leak in the steam generator handle. In a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE. N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
245 1 prior culture, they said "well, that's not so bad. That s 2 will go until we refuel."
{ !
i/
3 MR. BARRETT: It takes --
4 MR. CAHILL: We don't accept that anymore.
5 MR. BARRETT: Exactly. It takes setting the 6 example, walking the talk, if you would. There were two, 7 two examples that I talked about, that, that were not 8 popular decisions from, from a maintenance standpoint, if 9 you would, this year. And perhaps in the old days, when 10 maintenance was in the driver's seat, these wouldn't have 11 happened.
12 But, specifically, on March 12th, the plant 13 was returned to cold shutdown to effect repairs.
5
'- 14 Everybody at this table supported that fully, but I can 15 tell you that, deep in the maintenance organization, there 16 was a lot, and in the operations organization, there was a 17 lot of disappointment and question as to why we made that 18 decision.
19 The decision to reduce power to less than 45 20 percent, to calibrate the flow transmitter. That was not 21 a popular decision among certain departments. Those are 22 hard decisions to make, but you have to set the example, 23 if you're going to convince people that what you say is 24 what you mean. i l
/~T
!w_) 25 And you expect that shift manager, at three l
NEAL R. GROSS COURT REPoRTEPS AND TRANSCRIBERS 1323 RHUDE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON, D C. 20005-3701 (202) 234-4433
246 1 o' clock Sunday morning, to follow the guidance and the g3 2 words that you put out, because he's going, he's going to
( )
3 believe what you do, not what, you know, that's going to 4 carry a whole lot more weight than what you say.
5 MEMBER BARTON: It's encouraging to see you're 6 taking those steps, but I think you'll find that the top-7 performing plants do that as courses, do that as course of 8 normal business. So, it's good, it's encouraging to see 9 that you have now those expectations --
10 MR. BARRETT: Exactly.
11 MEMBER BARTON: -- of performance, as, as 12 heading in that direction.
13 MR. BARRETT: I'm not saying it's unique in 14 the industry. It's a change.
15 MEMBER APOSTOLAKIS: Let me ask a --
16 MR. BARRETT: A turning point.
17 MEMBER APOSTOLAKIS: Let me ask a question 18 regarding communications, before we go on. Do you have a 19 formal mechanism in place, to communicate with outside 20 organizations? I think, like. of course, INPO is one, but 21 also other organizations, perhaps. !
22 MR. BARRET1'. Yes. We have a number of, of l
organization -- there are industry organizations that ;
23 l i
24 exist. The Northeast of the country has quite a few of I r%
y)
( 25 them.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W. l l
(202) 234-4433 WASHINGTON. D C. 20005 3701 (202) 234-4433 I
l 247 1 There's training, there's engineering g 2 procedures. The emergency operating, the NUMARC procedure U 3 upgrade I talked about. We literally, there are literally 4 hundreds of visits, other site visits made by people from 5 Indian Point, in participating in this kind of activity.
6 MR. PEARSON: All of the, all the five major 7 areas, maintenacce, engineering, operations, health 8 physics, and I forget what the last one was, had 9 benchmarking efforts with NEI.
10 We all attended, I believe all attended the 11 summer meetings. I know I did, down in Richmond, where 12 the best of the best gave benchmarking results of the 13 industry.
/~~~N 14 We have Indian Point 2 right next door, and we 15 sent people over, when, if they have a problem. They send 16 people. In fact, they sent people to help us do a self-17 assessment, in operations, a month and a half ago.
18 MEMBER BARTON: I don't know if you mentioned, 19 the other, there was an area in training where they did 20 benchmarking.
21 MR. PEARSON: Training. Yes. You're right.
22 Yes.
23 MEMBER APOSTOLAKIS: Now, as an example.
24 Under teamwork, the first bullet says management of fm
() 25 emergent issues. Let's say you do identify an issue, from NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
248 1 your experience. Do you communicate that to the outside
,m 2 world?
i \
3 MR. BARRETT: Yes. We use the industry 4 network. That's one of the things we consider, whether to 5 send a message out over the network. And I think the part 6 30 wide uses the network pretty effectively. But, during 7 our DER process, when we learn something new, that we want 8 to share it.
9 We had a recent -- I'm trying to think of the 10 one where we unfortunately led the industry in the, which 11 was that?
12 MR. PEARSON: Was that the cross-connection of 13 SI accumulators?
,O f
\ ') 14 MR. BARRETT: Yes. The SI accumula*. ors.
15 There is a, there is a, there was a requirement in our 16 tech specs that, if you had the level indicator out of 17 service, you -- )
18 MR. PEARSON: One of the two.
1 1
19 MR. BARRETT: One of the two level indicators, l 20 on the SI accumulator and the containment. If that 21 indicator was out of service, the tech specs actually told 22 you could cross-connect two accumulators, let them l
23 equalize in level, and then read the level off the other 1 24 accumulator and confirm the level in the, in the first !
\ ,/ 25 accumulator.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS l 1323 RHODE ISLAND AVE.. N W.
(202) 234-4433 WASHINGTON. D.C. 20005 3701 (202) 234 4433
249 1 Well, we discovered, through the review of
.s 2 that, and this is part of our 50.59 upgrade effort, we C 3 discovered that that was not supported by the FSAR, 4 because the FSAR, we're taking credit for one possibly 5 being out of service, but the other three being available.
6 Well, -- operable.
7 Well, if you've got two cross-connected, and 8 one of those two is inoperable, you de facto have two of 9 them inoperable.
10 That was in our, it's our tech specs, and it 11 was there. It was incorrect. We shouldn't have had it 12 there. And we put a change in, to revise the tech specs, 13 and we went out over the network.
/^\
! i
- 14 There are, I would say conservatively, half a 15 dozen other industry items have been noted on, on the 16 weekly, or the daily reports, as a result of that.
17 MR. KELLY: Let me make a note on that, too.
18 MR. BARRETT: Sure.
19 MR. KELLY: Although no other plants had a 20 tech spec requiring them to cross-connect the 21 accumulators, a number of plants, by procedural process, 22 were doing precisely that, and it was not known, 23 generally, that this was an issue of, relative to a deign 24 basis concern.
C
(_)h 25 MR. BARRETT: Plus, we're, we're watching 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
250 1 network, too. I think that we have a --
+
ps 2 MEMBER APOSTOLAKIS: It's both ways.
V)
(
3 MR. BARRETT: Right. Unit 2, for example, had 4 a problem with a certain kind of a switch and, and this 5 was early, earlier this year, when we're getting close to 6 our start-up.
7 And, of course, we visited next door, found 8 out what the problems were, and devised a scheme and a 9 methodology to test ours before start-up. So it is two-10 way.
11 MR. CAHILL: This is a change. We're talking 12 about the way we are now. Several years ago, that was 13 part of the problem that we, the NYPA organization, 7s
( ) 14 particularly Indian Point 3, tended to isolate, and 15 internalize, and not communicate with neighboring units, 16 or within the industry.
17 MEMBER APOSTOLAKIS: Yes. I 18 MR. CAHILL: That's really different now.
l 19 MEMBER APOSTOLAKIS: I mean, if you look at 20 past incidents of some significance, say from the LERs. l 21 MR. CAHILL: Yes.
22 MEMBER APOSTOLAKIS: There are two things that 23 stand out, as causes, contributing causes, because it's 24 never one thing. And one is the issue of, of so-called
/~%
( ,)
25 crganizational learning, which is related to what we're NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005 3701 (202) 234-4433
251 1 talking about. Learning from the experience of other 2 organizations, and acting from lessons.
7-
'" And the other one is the so-called, most of 3
4 the human errors have been circumventions, which means --
5 or violations, which means routine violations of 6 procedures, of whatever management says, just to make 7 things happen, because, you know, this is an industrial 8 facility.
9 If you really go by the letter of the book, it 10 may take forever to do certain things. And it turns out 11 that the major accidents or incidents, that was a 12 contributing factor.
13 Is that something you worry about, or you're n
I\ -) 14 pretty confident that what you think is happening out 15 there is, in fact, happening? That there are no routine 16 shortcuts, or violations that might. I mean, it's very 17 hard to. That's the so-called informal culture.
18 MR. SALMON: Well, one of the things --
19 MEMBER APOSTOLAKIS: Okay. There is a 20 separate culture that management says we should have, and 21 then you, you have something else that's a little 22 different. That's the formal ching.
23 MR. SALMON: One of the things is to try to 24 publicize the proper action, and celebrate that, and
( 25 reward that, as well as penalize people who 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
252 1 MEMBER APOSTOLAKIS: So, you give incentives
, 2 to people.
3 MR. SALMON: To we try to --
4 MEMBER APOSTOLAKIS: -- -- who actually do 5 what the management wants?
6 MR. SALMON: Every time, you know, you can 7 hand them the free dinner.
8 MEMBER APOSTOLAKIS: Yes.
9 MR. SALMON: Or, you know, you go out, and you 10 catch them doing the right thing.
11 MEMBER APOSTOLAKIS: Right.
12 MR. SALMON: You do a free dinner, you put it 13 in the --
/%
! )
\/ 14 MR. BARRETT: We just had an operator discover l
15 a, a leak on a service water pipe in the, in the plant.
16 And there was no wetness. How he discovered it was he 17 noticed a quarter-sized stain on a painted floor.
18 There was no liquid. And he searched, and 19 searched, and, after awhile, found that there was a 20 pinhole on the, on a weld, in the service water pipe, and 21 that was identified.
22 He received a, what we call a bravo card, and 23 recognitior from Marc as a result of that. We need to 24 celebrate that.
rx 25 MEMBER BARTON: We need to move along with b\_-)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234 4433 WASHINGTON. D.C. 20005 3701 (202) 234-4433
253 1 this.
2 MR. BARRETT: In summary, we feel that the !
k, )
3 competence of the plant's staff is evidenced by l 1
4 conservative decision-making, effective teamwork, in 5 managing evolving issues, and that ownership of the plant 1
6 is evident.
7 We have a strong lessons learned focus, as 8 evidenced by the plant leadership team. We're doing 9 deviation event reports on a daily basis. Extensive 10 condition assessments are continuously improving, and 11 comprehensive.
12 We have a strong industry experience program.
13 Our self-assessment program is improving. We, we took the ts i 6 1
\2 14 successful post-transient review process from Fitzpatrick, i
15 and implemented it in Indian Point 3. I think that 16 originally came from a plant on the Jersey shore.
17 And we have a continual improvement in human 18 performance, as noted earlier. And not the primary one, 19 but the result, is plant reliability is improving, as 20 evidenced by our ongoing 113-day continuous run, and the l
21 fact that the plant has been on-line, essentially at full l l
22 power, for 157 of the last 159 days.
23 MEMBER APOSTOLAKIS: Just, just for the 24 record. I don't think conservative decision-making is p
(,,) 25 something we should be striving for. I should, I think --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
254 1 I think it should be risk-informed, realistic decision-fm 2 making.
I I
'm ) (Laughter.)
3 4 MR. BARRETT: This is the PRA.
5 MEMBER APOSTOLAKIS: There is a big 6 difference. No, there's a big difference. And I'm sure, 7 I'm sure that's what you mean.
8 MR. BARRETT: I'm sure, I'm sure we're in 9 violent agreement.
10 (Laughter.)
11 MEMBER APOSTOLARIS: Conservative decision-12 making is a different thing. I think what you mean is 13 that --
p
(,
14 MR. BARRETT: Well.
15 MEMBER APOSTOLAKIS: -- when in doubt, you 16 want to be conservative.
l 17 MR. BARRETT: Right. The conservative --
18 MEMBER APOSTOLAKIS: That's risk-informed.
19 MR. BARRETT: The conservative decision not to 20 bring in a barge-mounted crane, to bring in a land-based 21 crane, was based not on conserving finances, but on 22 reducing risk, the potential for risk.
23 MEMBER APOSTOLAKIS: So it was risk-informed.
24 MR. BARRETT: Yes.
r's
(_) 25 MEMBER BARTON: I think you are in violent NEAL R. GROSS '
COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
I 255 l I
1 agreement. Yes.
)
s 2 MR. SALMON: Yes, we are. I'd like to wrap up
(
3 by just saying that we have had a number of improvement 4 plans, as we've briefed you. We also have a business 5 plan.
6 We have melded those two into the site-7 specific business plan, because, to really achieve the 8 fundamental improvements, we need to resource, and 9 probably we need focus.
10 And so our, by combining our continuous 11 improvement plan with the business planning process, we've i
12 zero-base the budget and make rure we're putting the 13 proper resources on the issues that need to be solved.
r\
14 We've set goals for ourselves, so that we 15 don't get complacent because we haven't run well for a 16 long while, and now the plant is running, and we don't 17 want people to think ah, we're successful.
18 So we set breathtaking goals. We want to 19 achieve superior safety and regulatory performance. We 20 want to perform at the SALP 1 level. l l
21 We want to achieve excellence in operations, 22 as measured by top quartile in any measured category that )
23 the industry keeps. The NRC indicators and the INPO 24 indicators.
(~\
\ ,l 25 We want to control costs, to assure that we m
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHoDE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D.C. 20005-3701 (202) 234-4433
}
256 1 use our resources wisely and put them in the right place,
- 2 to assure plant safety and reliability. And we want to
,]
3 continue to build an effective organization that allows us 4 to achieve those things.
5 A short view of the future here. We're moving 6 as, with all our energy, to achieving these goals of SALP 7 1 level performance and top quartile. We're investigating 8 some options.
9 One is to partner with an advanced nuclear 10 enterprise, a company that is currently running their 11 plants well, to bring additional resources and techniques 12 into play, to achieve long-term performance goals.
13 We are, currently have Entergy Operations in-
)
\ms' 14 house, doing a due diligence, to see what it would take 15 for them to participate in the management services, and 16 the management of the Power Authority plants.
17 CHAIRMAN KRESS: Has any other plant done this 18 before? Or, is this -- this is the first time I've heard l
19 of it. l l
20 VICE CHAIRMAN SEALE: I think, if this comes l
21 about, it will be the first time, although Bechtell and, l 22 and Consumers explored, in depth, the taking over of 23 Palisades. But that, that didn't come about. It's very 24 complicated, in every, every angle of the thing.
n k,) 25 MR. SALMON: And this, this is, this is a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234-4433 WASHINGTON. D C. 20005-3701 (202) 234-4433
257 1 management services kind of contract. The Power Authority 2 would retain the license, the liability, the
/, s\
\
'~J 3 responsibility, and the ownership of the plant, and bring 4 the Entergy processes, management information services, 5 and, and management talent to bear on, and make even more 6 rapid improvements in our performance so we can meet this 7 goals.
8 MEMBER BARTON: So this is, is this anything 9 different than what Entergy has done down south, with some 10 of the other plants?
11 MR. SALMON: Yes. 1 I
12 MEMBER BARTON: This is different?
13 MR. SALMON: There's a limitation, because O
k-s 14 we're a not-for-profit organization, that the contracts 15 have to be limited to five years, and there's an automatic 16 bail-out on either person's part for three, so it causes 17 us to, to look at an alliance, as opposed to a, to a, them 18 taking over the license.
19 MR. CAHILL: And South Entergy's other 20 operations, they own every one of them as a part of 21 Entergy, including River Bend, which was absorbed through 22 a merger.
23 MR. SALMON: The other thing that has been 24 discussed is whether there could be an alliance or a joint
/
(3) ,
25 operating company formed, within New York State, for 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
258 1 nuclear utilities in New York State, to gain economies of
<~s 2 scale in a, in a new, competitive environment, and to take
/
V) 3 advantage of the combined management strengths of the 4 participating companies. And that is being evaluated, but 5 it's not a near-term, it doesn't appear to be a near-term 6 option.
7 Finally, we've talked a lot about SALP 1. We 8 understand it's a pretty simple thing to look at the 9 definitions, and the difference between SALP 1 and SALP 2.
10 SALP 1 is consistent superior performance.
11 Not just good, and not just most of the time, 12 but reaching those levels that allow you to, to have
,s-13 comprehensive corrective actions, and eliminate recurring
( )
\/ 14 problems. That's our goal, and we're moving in that 15 direction.
16 MEMBER BARTON: Does that complete your 17 presentation?
18 MR. SALMON: Yes.
19 MEMBER BARTON: Are there any further 20 questions from any members of the committee?
21 (No response.)
22 Any comments?
23 (No response.)
24 Well, we thank you for the presentation.
o
( ,) 25 Thank you for your attention to the, to the issue. One NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W.
(202) 234 4433 WASHINGTON D C. 20005-3701 (202) 234 4433
1 259 l 1 quick question I'd like to ask is have you looked back, p 2 and I think we all need to learn from a process like i
C/ 3 you've been through, including ACRS, plants were designed 4 to run, not be shut down for a long period of time.
5 Where, in, in your soul-searching of this 6 whole several past years, where did the process break 7 down? You had, you know, in-house oversight, INPO was 8 supposed to be helping the industry, the NRC is supposed 9 to be regulating the industry.
10 You know, where do you see the, the process 11 having failed? So maybe we can all learn from this, and 12 preclude it from, from happening to someone else.
13 MR. CAHILL: Well, offhand, my impression is
/3 (Vl 14 openness, openness and communication amongst all of the 15 elements in this industry. The indivihal licensees, and 16 their relationship to each other, and the regulator and 17 INPO. If there is open communication, an objective, 18 professional approach to things, the deficiencies surface.
19 If there is isolation, the, if there is 20 impaired communication, or no communication, or partial 21 communication, a lot of things that, are just not noticed.
22 They're just not realized.
23 These plants are rugged. They have a lot of, 24 lot of resilience to them, and they can go for years,
)
(Q 25 deteriorating, both in the material condition and in the, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N W. 1 (202) 234-4433 WASHINGTON, D C. 2000 5 3701 (202) 234-4433
260 1 in the skills, qualities, and attitudes of the people,
,s 2 which is the most important. And everybody's got to be
- )
V 3 vigilant to resist that.
4 Now, I think we found a right formula here, 5 and I've seen this in other plants that needed to find the 6 spark to restart, and start right. And so it's openness, 7 this questioning attitude, this, this freedom of, of 8 communication, and professional approach to the work.
9 MEMBER BARTON: Thank you.
10 MR. CAHILL: Thank you very much.
1 11 MEMBER BARTON: I'd turn the meeting back over 12 to the Chairman.
13 CHAIRMAN KRESS: Thank you. I also want to
( )
k' 14 second the thanks of the Committee. We, we get a lot out 15 of this kind of presentation, in terms of, of things that l
16 can be done to improve operations, so we really appreciate l
17 you willing to come in and talk to us.
18 MR. CAHILL: Thank you.
19 MR. SALMON: Thank you very much, sir.
20 CHAIRMAN KRESS: At this time, I propose a 15 21 minute break.
22 (whereupon, the foregoing matter went off the 23 record at 3:29 p.m.)
24
('N N, 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
I CERTIFICATE This is to certify that the attached proceedings before the United States Nuclear Regulatory Commission in the matter of:
Name of Proceeding: 434TH ADVISORY COMMITTEE ON REACTOR SAFEGUARDS (ACRS)
Docket Number: (NOT ASSIGNED)
Place of Proceeding: ROCKVILLE, MARYIAND were held as herein appears, and that this is the original transcript thereof for the file of the United States Nuclear Regulatory Commission taken by me and, thereafter reduced to typewriting by me or under the direction of the court reporting company, and that the transcript is a true and accurate record of the foregoing proceedings.
. 'l ~
1&fj CORBETT WINER Official Reporter Neal R. Gross and Co., Inc.
O NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1 1323 RHODE ISLAND AVENUE. NW (202) 234-4433 WASHINGTON. D.C. 20005 (202) 234-4433
i
{ INTRODUCTORY STATEMENT BY THE ACRS CHAIRMAN 434TH ACRS MEETING, SEPTEMBER 12-13, 1996 THE MEETING WILL NOW COME TO ORDER. THIS IS THE FIRST DAY OF THE 434TH MEETING OF THE ADVISORY COMMITTEE ON REACTOR j SAFEGUARDS. DURING TODAY'S MEETING, THE COMMITTEE WILL CONSIDER THE FOLLOWING:
(1) ADEQUACY OF THE NRC CODES TO ANALYZE STEAM GENERATOR
- (6) PROPOSED ACRS REPORT (7) STRATEGIC PLANNING i
1 A PORTION OF TODAY'S MEETING MAY BE CLOSED TO DISCUSS i
ORGANIZATIONAL AND PERSONNEL MATTERS THAT RELATE SOLELY TO THE
- INTERNAL PERSONNEL RULES AND PRACTICES OF THIS ADVISORY
- COMMITTEE, AND MATTERS THE RELEASE OF WHICH WOULD CONSTITUTE A CLEARLY UNWARRANTED INVASION OF PERSONAL PRIVACY.
- THIS MEETING IS BEING CONDUCTED IN ACCORDANCE WITH THE 4
- PROVISIONS OF THE FEDERAL ADVISORY COMMITTEE ACT.
DR. JOHN T. LARKINS IS THE DESIGNATED FEDERAL OFFICIAL
- FOR THE INITIAL PORTION OF THE MEETING.
WE HAVE RECEIVED NO WRITTEN STATEMENTS OR REQUESTS FOR TIME TO MAKE ORAL STATEMENTS FROM MEMBERS OF THE PUBLIC REGARDING TODAY'S SESSIONS. A TRAL! SCRIPT OF PORTIONS OF THE MEETING IS i BEING KEPT, AND IT IS Rt%ASTED fMAT THE SPEAKERS USE ONE OF THE MICROPHONES, IDENTIFY THEMSELVES AND SPEAK WITH SUFFICIENT CLARITY AND VOLUME SO THAT THEY CAN BE READILY HEARD.
I WILL BEGIN WITH SOME ITEMS OF CURRENT INTEREST.
1 i
J
O O O W M
Analysis of Steam Generator Tube Conditions During Severe Accidents Presentation to the Advisory Committee .
on Reactor Safeguards September 12,1996 i
Charles G. Tinkler Accident Evaluation Branch Division of Systems Technology Office of Nuclear Regulatory Research
O o o ~!
Outline ,
O Introduction / Summary 9 SCDAP/RELAP5 Modelling 9 Fission Product Deposition i
e Benchmarking of Codes / Scaling Issues G Preliminary Results 1
O Independent Review G Conclusions i
2
~
O O O -
Introduction
Background
O As part of the rulemaking on steam generator tube integrity, consideration has been given to the performance of flawed tubes and their likelihood of failure during a severe accident when the RCS remains at high pressure in relation to .
the secondary system. Concern over tube performance arises due to hot gases, as a result of core heat up and oxidation, circulating through the steam generator tubes.
O Circulation of high temperature gases may induce tube failure due to elevated temperature of the SG tubes.
9 Fission product releases from the core may deposit in the steam generator tubes further elevating the heat up of the SG tubes.
i O RES and NRR are coordinating the analysis of SG tube thermal-hydraulic boundary conditions and their structural response in support of the proposed SG tube rulemaking.
3 i
~
O O O .
Introduction
Background
O SCDAP/RELAP5, as the NRC mechanistic severe accident / thermal-hydraulic code, was used to assess fluid heating of SG tubes.
- Benchmarked against H 1/7-scale test data and 3D code (COMMIX)
- Considerable experience developed in similar applications (DCH, unintentional depressurization) over nearly 10 years
- Since SCDAP/RELAP5 does not model fission product transport and deposition, VICTORIA was used to assess fission product heating of the tubes.
9 SCDAP/RELAP5 analyses are to assist in addressing the thermally induced SGTR. Spontaneous SGTR and pressure-induced SGTR are examined separately.
4
~
O O O .
Introduction
Background
9 Thermal hydraulic analysis is one part of the overall assessment of tube performance during a severe accident. Range of areas being addressed includes:
l
- Thermal-Hydraulic Modeling Event Tree Quantification RCPB Weak Points
- Representative Flaw Distribution
- Tube Performance Model
- Severe Accident Radiological Consequence Assessment
- Result Consolidation for Safety Analysis
mesen amm. Pzn e unen smen so Tees se Tees de ussve unhamin Seq Prob. End Snesse on ce,. Menemas sessa e Remmie nessi nei non nees.e nesse. Mahasia =th ptm t het RCS Fehn.
commes one = Pt*""' d'88 8'8 #9888 Laos of TDARU SooTonFW PER SAFETIES MESWM PRESS 41UPT NO4tMPT Msgy40 R1 R2 I
' m 1
M I
R5 g Rs
,a I
RF Re R9 I Rio I
R11 w ga @ Pressure Station Blackout, Quantified to Account for Degraded Steam Generator Tubes 6
O
~
O O -
Summary l
l 9 Analyses using SCDAP/RELAPS have been performed for representative plants for scenarios of interest (high pressure TMLB' sequences with depressurized secondary side) to estimate effects of high temp fluid circulation
- SR5 analyses predict failure of hot leg or surge line before SG tubes
- Sensitivities on T-H modeling did not alter conclusion on tube integrity 9 Analyses using VICTORIA to calculate fission product transport and deposition confirm fission product deposition is a minimal effect on SG tube heating 7
~
O O O .
SCDAP/RELAP5 Modelling
Background
O Integral experiments conducted by W at 1/7 scale, representing the reactor coolant system confirmed the existence of natural circulation flows within a PW~R primary system during a severe accident.
9 The redistribution of decay heat from the core to the hot legs, surge line and steam generators both delays core heat up and melting and increases the probability of RCS failure and depressurization. Lessens the probability of HPME, but increases the probability of SGTR given a core melt.
8
~
O -
O O .
Pressurizer Steam Steam p generator generator n n
'^ ^ ^ ^ "
l l In-vessel ' "
i circulation i
j ()
OiO -
w Loop natural '
11 circulation i Countercurrent I , s.6v
> Hot led M j f- -
4 :
i [C = - flow y gj]
/ j _
k, M hl g i
f _
4r : <
r l
- Water )
h r Stech :n hjV
! SEE hEfolLS ON INLET PLFt0004 Mi%IN&
P394-LN87017 g
Figure 1. Severe accident natural circulation flows. .
o O O
~
/ 0o n
I g... . . . .d r ;_ m o
/
Reactor vessel inlet = bT V Outlet plenum
\ plenum
@ @' j W^
_= =
- b Hot leg P431 LN87031-3 Figure 2. Hot leg natural circulation stream flowx 18 1
~
O O O .
SCDAP/RELAPS Modelling (continued)
SCDAP/RELAPS (SR5) model development for Surry 9 Hot leg was split into top and bottom halves to model the countercurrent now (now cannot be split within a control volume of a 1-dimensional code)
- Observation from Westinghouse visuuraation experiments
- Nodalization preserved character of Gow pattern 9 Steam generator tubes were split 35%/65% to simulate hot vapor flowing from '
the inlet plenum to the outlet plenum through some of the steam generator tubes, and cooler vapor returning to the inlet plenum through the remaining tubes.
- Based on Westinghouse experiments 9 Three radial rings were used in the core and upper plenum, connected by i cross 00w junctions. More recent SR5 calculations used five radial rings to better represent core.
4 11
O O O
~
SCDAP/RELAP5 Modelling G Standard RELAP5 heat transfer package is used Forced Convection
- Turbulent: Dittus-Boelter e Laminar: Sellars Free Convection: Churchill-Chu (vertical)
McAdams (horizontal)
O Code uses maximum of the forced and free convection heat transfer coefficients 12
- t l
O O O .
1 Hot Leg and Steam Generator Nodalization for Modeling Hot Leg Countercurrent Flow 4
s 5 h /
p ...-- y ---- p ,
6 3 3, 6 33 6 7 ,
4 .....
/ ....
p, ---M
$ 2 3 2 s, 7 S p, ..... S,, .7
/,
4 3
/
i 2 e
8 1 3, o
8 ( ,
Reactor y u}.. &.,
p I 0 g
17 vessel 406 407yg 410
/ r soe j 465 T -
400 T 412
+ 402 + 404 ' 1 1 ! 2 .
17 2 7musum wmum/}w/ws7m 430 I m' -
4 l 3 l 2 l g
rms/uins>unisisis//i>ww/////wr7m777 i L
171 o i Q P431 LN87031-2A 13
~
O O O .
Fission Product Deposition ,
9 Objective of the analysis To determine the effect of fission product transport and deposition in the RCS on steam generator tube integrity 9 Summary of approach i Used the VICTORIA fission product code with SCDAP/RELAPS calculated thermal-hydraulic conditions as input e Conclusion .
Fission product transport and deposition in the RCS have a negligible effect on steam generator tube integrity, because the fission product release is relatively small and late in the transient O Volatile fission product release represents 5-10% of decay heat S Fission products spread among upper plenum, hot leg, SG plena and tubes 14 I l I I
- - _. w -- -- - ---_
~
O O O .
Fission Product Deposition (continued)
VICTORIA t
Used VICTORIA to analyze fission product behavior under station blackout conditions l
M4or VICTORIA models: fuel release i chemistry and condensation /revaporization t aerosol deposition (laminar deposition, turbulent deposition, settling, thermophoresis, pipe bend)
M4er VICTORIA inputs: temperatures (fuel, gas, heat structure) from SCDAP/RELAPS flow rates ;
geometric data (cell size, flow area, hydraulic diameter, surface area) !
i 15
VICTORIA Nodalization
, & h 11 28 Steam 31 14 Region of Greatest Generator Steam-Heat Loading 10 27 Tubes 32 15 PORV Upper Plenum j 8
\-
9 26 33 16 6 -
f7 ";4 z y c 25 437
- 23 24 Inlet Outlet 2L v Hot Leg Plenum Plenum n
5 = -22 . .
Core 4 e 21 ] ,
3 e 20"j n
t 2 c 19 t JL 1 g 18 &
Steam Source m
o o o .
Fission Product Deposition (continued)
Results ,
9 Fission product heating of SG tubes is minimal ~.15 MW/ steam generator -
Total Decay Heat ~20 MW G Relatively small mass of deposited material Total release ~ 400 Kg Fission product irelease - 50 Kg Deposited thickness is small ~30 p Tube thickness ~1270 p Tube diameter ~20,000 y O Dominant mechanisms are vapor condensation, settling, turbulent deposition i
9 Revelatization occurs in upper plenum, hot leg and inlet plenum 17 l
- ; ,. :j:' '
i ! l !l ! l
~
0 o it g
n a
0 0,
6 1 b e
e u e ft lz H z t
c L o u n d
o 0 g o
P r 0 l t
n b, h o
io 5
\ s s
i 1
F e
n n
[
i l i o hl 0 u e
g r
s t
b, aec l
g i
t n 4 r a 1 Wdd i H
t e
m a
e y 0 er S 0 b , )c
_ baw 3 1
es
_ uro 0
(
e
_oTf , 0i m 8 s b, T l
_ f e 2 obu
[
1 gt f O O
no b, 1
_ ie t 1 ado en b, 0
0 Ht s i r
0 1
F 0
0 0,
0 0 0 0 0 0 0 09
_ 0 0 0 0 0 0 0 2 4 2 0 8 6 4
_ 1 1 1
_o i x l$1 i
,lll!l,I
O
~
O O .
Benchmarking of SCDAP/RELAP5 Model and Scaling EPRI E Natural Circulation experiments e Objective was to perform experiments to investigate phenomena associated with l natural circulation flows under severe accident conditions in PWRs with U-tube steam generators e ne experiment scaled to preserve the geometry at 1/7 scale (need to represent RPV, hot leg, surge line, and SG)
- maintain the ratio of flow area through the steam generator tubes to the hot N
4 e 1/7-scale model of a reactor vessel and two hot legs with U-tube steam generators.
De 1/7-scale model: hot leg dia. = 5.14", SG id = 0.305", number of SG tubes
= 216 9 Conducted several series of tests using water, low-pressure SF. and high-pressure
- SF.
19
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ._______\
O O o -
Benchmarking of SCDAP/RELAPS Model and Scaling (continued)
EPRI E Natural Circulation experiments (continued)
Findings:
9 Low pressure tests showed (by using dye in the Guid), that a stable countercurrent Gow was present in the hot legs, with little turbulence at the interface between the two flow streams. The test results display the basic nature of the natural circulation phenomena. Flow patterns were consistent over a wide range of conditions.
O High pressure SF. tests provided data for validation of codes. Five series of '
experiments (total of 36 tests) with high pressure SF. were conducted.
Temperature measurements in the steam generator inlet plenum and tube inlets indicatcd that the Guid in the inlet plenum was well mixed. In none of the ;
experiments was there any indication that Guid from the hot leg Dowed directly
' into any of the steam generator tubes.
20
O O O
~
Benchmarking of SCDAP/RELAP5 Model and Scaling (continued)
EPRI E Natural Circulation experiments (continued)
Review of experiment scaling:
G Initial scaling review was performed by INEL
- Radiation heat transfer was not taken into account in the experiments (Iow
. temperature of simulant used). But at higher Re numbers (105) typical of high pressure SF., convective heat transfer dominated and tests scaled well.
High pressure SF experiments are adequate for code assessment.
O Independent review of INEL scaling evaluation
- P. Griffith (MIT), R. Viskanta (Purdue Univ.) and S. Ostrach (Case Western Reserve Univ.)
- confirmed INEL findings about the adequacy of the experiment scaling.
- Additional analyses were performed to determine the impact of thermal ,
radiation on the overall heat transfer. With radiation heat transfer, the heat transfer coefficient was about 25% higher than without. '
21
O O O l; Benchmarking of SCDAP/RELAP5 Model and Scaling (continued)
SCDAP/RELAPS (SR5) model development for Surry 9 EPRI sponsored COMMIX calculations of Westinghouse low pressure tests, with good agreement between the calculated and measured responses.
I S Using the same modeling approach, a COMMIX Surry plant model was developed.
9 SR5 Surry model was benchmarked against COMMIX Surry plant model.
- Altered the SG inlet plenum mixing volumes /Gow area split, loss coefrwients through the now path
- Did not force a match of recirculation flow rates or mixing fractions I 9 When reasonable agreement was reached, the hot lag and steam generator input model was " frozen" and used for natural circulation calculations.
22
O O O Benchmarking of SCDAP/RELAP5 Model and Scaling (continued)
Further assessment of SR5 ,
t G Developed SRS model for E 1/7 scale experiment. Same modeling approach used in Surry calculations was used to model the facility.
O Assessed against two high pressure SF tests (S-6 and S-7).
9 SR5 calculations were in reasonable agreement with the measured data (e.g., flow patterns were the same as in the experiment, temperature profiles were the same as in the experiment)
- Calculated hot leg and SG mass flow rates within 11% of measured values
- Calculated vapor temps in the SG tubes were within 5% of the measured Values 23
n .
REACTOR VESSEL VAPOR TEMPERATURES FOR TEST S-6.
CALCULATED MEASURED top Of 54 7 90 4 .
tpper plenum x
-o 145.7 143.5 136.9 165.0 15 t0 158.8 -
, 173.7 15,.. 5 ........ ........ ............. - g3_4
........ ............- - 115'1 * *
- C 123.7 - 136.9 155.6 148.2 I2c -
148.8 145.0 167.4 157 2 147.1 137.2 i 137.2 . 137 2 i 134.3 Hot I 158.3 146.5 126.0 i 66 8 67.7 i 70.3 ..................................
- . :. 119.0 113.6
~592 : 120.4 l .
187.2 : 137.0 :
..........i.........i......... 194.3 i 132 2 i......... 122.8 Tcp of 207.3 i 152.9 i 119.3
..........;........4.......... . . . . . . . . . .;. . . . . . . . z. 12 6.3 Core . 190.0 : 128.5 :
- : ..........:........ 4.........
..........:.........z........ 186.6 i 145.3 i 129.8 204.7 i 165.1 i 123.8 ......... .........g...... ..
..........:.........<......... 186.7 j 154.6 i 133.9 l
l i i ..........s... ....,,,,,..,,,
........ 9...................
1872 i 175.9 ! 138.6 178.5 i 167.3 i 132.1
. . ~82.9 : 1862 : 141.0 172.5 3 169.4 ! 143.5 150.3 j 1536 i 154.7 ......... ........:.........
161.4 ! 163.3 3 146.4 I
bOC o.
g g ,_ - - - - - - - - - - - - - - - " - - - ' - - - - - - 146.4 114.5 139.1 24 Vessel G
p w ~g .
R ~ >
REACTOR VESSEL VAPOR TEMPERATURES FOR TEST S-7.
MEASURED CALCULATED Top Of eg.3 W P eritsri i %, 8i.6 ,
125.3 o 136.6 133.0 123.0
, 137.5 135.2 143.5 129.8 ........ ............. - 75_g
.~.....
........ ............- - g3_g "
116.4 --
106.4 - 141.3 136.2 123.0 134.7 124.0 122.6 14 2.4 135.3 123.3 i 128.3 . 128.3
- 314.5 ........ .. .... ............ . . . . . . . _ . . g'
........ ........ .............;............. 143.1 134.6 117.5 i 64.5 . 66.2
- 63.8 -
- : 112.7 148.8 : 115.0 : 97.4 143.8 i. 113.7 !.
. . -..-....?........*.........
........ 4........:......... 170.4 i 122.7 i 115.6 IOp Of 163.7 i 125.7 i 101.2 --------.j...-.-...j.........
g 167 2 : 127.4 : 118.3
- . . ---------j.--------j---.....
..........:........ 4......... 164.3 : 132.5 121.0 152.9 :: 133.8 :* 103.9
................... ......... 164.3 j 139.3 i 124.0 j j --------? -~~--*4---------
.................. 1......... 164.7 j 153.5 j 127.6 141.3 ! 134.7 1.........! 110.3 -"----t"-----i--------
129.5
..........:........ 162.1 : 166.7 :
......... ......... ......... 153.5 j 151.0 j 33i,3 121.5 i 123.7 i. ii1.1
. . . . . . . . . _ ..........,........ i45. i i46.6 i316
............. 3 g.........
Of Cor ......______...__. 127.9 95.6 26 Vessel Q
. - - . . - - - - - _ . . . - - - . . - - . - _ . _ _ _ . - - - . . . . - . . ~ . - .
O O O .
Benchmarking of SCDAP/RELAP5 Model and Scaling (continued)
Sensitivity calculations for earlier Surry TMLB' analysis (NUREG/CR-5214,1988)
S Initial c=Iculations were performed to investigate if ex-vessel failures might occur, where they would occur, and when they would occur in relation to core damage progression.
9 Performed sensitivity calculations to consider:
axial power profile
- core and upper plenum crossflow resistance
- hot leg / surge line piping heat loss
- heat transfer coefficients in the upper plenum, hot leg, and steam generator tubes
- simulated radiation heat transfer between hot leg flow streams e Conclusions from sensitivity calculations:
- In all cases, ex-vessel piping failures were predicted to occur during the early phase of core heat up and melt progression.
- Effect of sensitivity analysis was to only marginally alter surge line or hot leg failure time.
26
O
~
O O Preliminary Results Recent Application of SR5 G Surry Plant calculations
- TMLB' transient
- Base case: # SG tubes participating in forward tiow - 35%
mixing fraction = 0.87 recirculation ratio = 1.9 e Calculations performed with and without SG secondary depressurization 9 Sensitivity analysis was performed to address inlet pienum mixing, hot / cold tubes split l
l i
(
i l
l l
! 27 t
l l
i
a ,u+-.r= . - . ---...u..
. ,, - - . . . -.--m.-- a s =====*--s~ = ~ --2--. ----"---'
I
' 4 g i
iO, i
i 1
l i
O se.m.
I &
- i @
- i
.N M
(L
)
f l
.., m l
i J
. ENu :
l 0 l' (:-
s Jei g
l z - -
s
~
D1 l
c2. g
!O j o _a ,
p"CO
- O , __
g -
I d
m f
,I.o
@ f I i2 i
.N y
e
.i_
t
=,
O -- ,
-3 g . - _
O WW
(/)
- - s e : :': q _
! b iEi MMM j , Q -1 - l a s , p-s es s b
i l 1
4 i
I l
- . : , l l ' ; , l 0
0
. ~
0 _
~
- ~ 0 0
2 1
e s
a i
MV 0
0 0
0 5
1
- C E '
t s
l 0. )s u '
0(
0 e s i 0
0i m e 1 T M
R '
t .
n a
l 0 _
P 0 -
- i 0 -
0
- y 5 r _
- r u .
- S _
0
- - e O
0 0 0 0 0
0 8 6 4 2' 1
1 1 2 1 n
km ,
o.
o o
i .
o
- : o i : :
N
- . l
- 1 d
T"' :
e3 i.
o
- o G ei si - o :
y) ai
-. e!. 8 O : O. .
O h b -
- a i
! M i !
i w
- o. ur
- . o*
oe 1
3 : : _ I i
m -
8E
-h g l l
@ : i
- E i i
l W -
- i C i i i
i
- o.
(1, -
i 8o l to bnum i
- 3 -
2 W !.
- . o.
o o o, o, o m d d O (w) ie^ei pinbii pesdeiioo b
O O O .
Surry Plant Results - Case 1 S 2000.0 .
i 2
a y C O Surge line -
g -
A A Hot leg E V 7 SG tube Hot leg failure S
$ 1500.0 -
13 Surge line failure r .
e s -
7 1000.0 -
G E
2 _
h _
_$ co+
1 ' ' ' '
h 500.0 0- 10000.0 12000.0 14000.0 16000.0 Time (s) 1 31 l
1
O O O i
Surry Plant Results - Case 3
$ 2000.0 .
i -
i e
s 3 C O Surge line -
1 A A Hot leg V V SG tube Hot leg failure
[ -
$ 1500.0 -
U
{ -
Surge line failure e
7 1000.0 -
e E
2 -
9 .
4 * ' ' ' '
h 500.0 14000.0 16000.0 a- 10000.0 12000.0 Time (s) sz
Preliminary Results (continued)
Recent Applications of SR5 (continued)
G Surry Plant calculations
Conclusions:
!
- pressurizer surge line creep rupture was the first RCS pressure boundary failure e pressurizer surge line failed in early phase of core damage (before the onset of fuel melting) e if pressurizer surge line and hot leg failures are ignored AND a SG secondary ADV fails open, SG tube rupture occurs ~20 minutes after the first RCS i pressure boundary failure i
33
O O o -
= .
l'ouv l
,sav
[ T ; ,
4 , , , , , ,
E I MstV ; Poav< ; : :savs o
i l'
ll 1
_ E E as 4 K 9M K (bee embe) i Presswaaer
--. , l l - - ..
l
. - . Sol K (eeld embe evaleti
, 4D K Costasesneat i 995 K thea sube intes)
E5M K I
/
Ikk 'M l arasam
~
l a_Kj74p24is*4 ,,, ,, , ,, !
sisx " 'I (SG -----i it tu l
i --- ------ ...__.............~..........r.__-b. .
i
,JL 14 i
l
~ 6 '
l RCP
^ ',-
.a
- =-: -=
)l, V
- .~ u. _
1 I
Surry Case 3 vapor temperatures at the time of surge line failure (14 460 s).
N
Preliminary Results (continued)
Recent Application of SRS i e Surry: Sensitivity calculations
- Single loop model extracted from Surry plant calculation for TMLB' sequence calculations started at the onset of countercurrent flow (t=9200s) and extend fi an interval of 5000s parameters varied -
e number of SG tubes participating in forward (hot) flow: 29% to 61%
mixing fraction: 0.76 to 0.89 recirculation ratio: 1.69 to 2.25 !
- for each sensitivity study, one parameter was varied at a time
Conclusions:
minimal impact (< 120K) on the peak average tube temperatui for each of the individual variations 35
'. O O O Table 3.4: Summary of SG tube results for Surry calculations Pressurizer Loop SG Tube Results Non-Pressurizer Loop SG Tube Results
'"* I @ Time of
@First Ex- @First Ex-Peak Peak Case Vessel Failure Loop Vessel Failure Loop Temperature Temperature 3,, , Seal Cleared Cleared Temp AP Temp AP Temp AP Tmp &
l (K) (MPa) (K) (MPa) (K) (MPa) (K) (MPa) l l 1 847 9.27 1094 9. I 1 no 851 9. I 8 1094 9.04 no 2 847 9.27 847 9.27 no 851 9.18 955 8.85 no 3 987 15.80 1147 16.20 no 863 8.73 953 9.11 no 4 987 15.80 987 15.80 no 863 8.73 956, (3.90)* no 5 825 9.10 825 9.10 no 832 9.01 832 9.01 no 6 933 16.03 933 16.03 no 829 9.05 829 9.05 no 7 943 16.07 1041 16.08 no 951 16.07 1145 16.08 no
- Calculations stopped at surge line failure - therefore the time of ex-vessel failure is the same as the time for the peak tube temperatures. % .
a
O O o '
Independent Peer Review e Obtain independent assessment of the adequacy of SCDAP/RELAPS modeling of natural circulation under severe accident conditions for the purpose of calculating the relative timing aad failure of RCS components in order to evaluate the risk associated with thermally-induced steam generator tube ruptures G Reviewers: Raymond Viskanta (Purdue University)
Mamoru Ishii (Purdue University)
Peter Griffith (Massachusetts Institute of Technology)
S Review on August 19-20, 1996 L
t 37 I
~
O O O Independent Peer Review (continued)
O Peer reviewers provided preliminary comments and recommendations at the conclusion of the meeting 9 Peer reviewers - comments:
- the experimental OY 1/7 scale experiments) data is good and the experiment '
is well designed, and additional experimental data is not needed
- the SR5 code is adequate for the job (i.e., to calculate natural circulation under severe accident conditions), the implementation of the code is good and the constitutive relationships used by SR5 are adequate i 9 Peer reviewers - recommendation:
e demonstrate more clearly the adequacy of the modeling to show that the experimental data and the analytical results give the same systematic overall behavior in the reactor system being studied. {i.e., establishing a "fq':ure of merit" summarizing and comparing the experimental data and the SCDAP/RELAP5 calculations}
- Also recommended that it would be worthwhile to perform sensitivity calculations for parameters which vary widely {i.e., a factor 2, affecting the heatup rates of different reactor components (e.g., surge line, hot leg, SG tubes)}
38
Independent Peer Review (continued) 4 Action taken by NRC:
- working with INEL to develop a " figure of merit" to demonstrate more clearly the adequacy of the modeling i
' - working wie.h INEL how best to perform additional sensitivity calculations ,
i l
i 39 l
i
i O O O ?
l i
l Independent Peer Review (continued)
G Investigation of effects of mixed convection on heat transfer
- INEL performed calculations to investigate this affect for the ANO-2 1 TMLB' scenario l
l -
Mixed convection refers to a combination of forced convection (Re number '
dependent) and free convection (Grashof number dependent)
- Focused mixed convection affects on vertical components (i.e., SG tube and surge line) l
- Implemented the mixed convection for vertical flow (Cotton-Jackson; l International Journal of Heat and Mass Transfer-1990) in SR5 ,
l
- Results showed that the calculated change in steam generator tube i temperature is insignificant - SG tube temperature increased 4'K l
40 l
l i
O O O .
Independent Peer Review (continued)
G Additional analysis for Surry:
- Case 10: Modify depressurized SG secondary to better account for available heat sink.
- Preliminary results: SG temperature is 65K lower at time of surge line failure Event t (surge line failure) Tsc (@ time of surge t (SG failure) line failure) _
Case 3 14460s 987K 15560s Case 10 14770s 922K 16190s 41
- O O O '
Conclusions e SG tube thermal hydraulic boundary conditions are most directly influenced by variations in the accident sequence which determine pressurization /depressurization of the primary and secondary system
- Failure of primary and secondary relief valves
- Intact secondary relief valves results in very small challenge to tubes e SG tube thermal hydraulic boundary conditions are also influenced by ~
phenomenological issues associated with counter current natural circulation
- SG inlet plenum
- heat transfer modelling ,
- local effects e Variations in the treatment of these issues within reasonable ranges did not significantly worsen SG tube boundary conditions 42
8\
4 o&
EYE NEW YORK POWER AUTHORITY INDIAN POINT 3 NUCLEAR POWER PLANT l
Meeting of the Advisory Committee on Reactor Safeguards Rockville, Maryland September 12,1996 l .s c as ir.3 3i... .. .. ,i2,,e i
New York Power Authority l Indian Point 3 Nuclear Power Plant I William J. Cahill, Jr. Chief Nuclear Officer d_ _
L AeL:
Harry Salmon, Jr. Vice President Nuclear Operations Robert J. Barrett IP3 Plant Manager Marc Pearson IP3 Operations Manager Meeting of the Advisory Committee on Reactor Safeguards Rockville, Maryland September 12,1996 ACRS IP-3 Meeting on 9/12/96 2 I
l
AGENDA 4
Opening Remarks LL _
W. Cahill i ---
- espear C 6.
NYPA Background H. Salmon IP3 Watchlist Issues H. Salmon Corrective Actions and Restart H. Salmon Improvements for Operation M. Pearson
, (after restart)
SALP Results R. Barrett PrincipalImprovement Areas R. Barrett View of the Future H. Salmon ACRS IP-3 Meeting on 9/12/96 3
ym c c c9 m NYPA BACKGROLND 4
o&
LY -
Power Authority of the State of New York, established 1931 l_)3pdl Not-for-profit public benefit cerporation Provides 25% of all electric power consumed in New York State l Electric capacity of 7000 megawatts from 12 facilities:
60% hydroelectric 25% nuclear 15% fossil ACRS IP-3 Meeting on 9/12/96 4
CORPORATE ORGANIZATION 4
i o&
I_k _ -
Lnb&
CHAIRMAN AND CHIEF EXECURVE OFFICER a
I I
VP Public Affairs President and Chief Operating Officer General Counsel Sr. VP Business Services --
Sr. VP Transmission VP Appraisal and Compliance Services --
Sr. VP Power Generation
- Sr. VP President Marketing and Development --
VP Human Resources Sr. VP Energy Efficiency and Technology --
Chief Nuclear OfTicer Nuclear Generation ACRS IP-3 Meeting on 9/12/96 5
\
l I I }\ll! '
. - 6
. - w N v . -
im -
m g
n O4%
i l
=~ r e ro rtn uo I
t w gC n
Et c T
I I
r aj e lw e ro l
. d l cP A
I
. r s Nnud ds u P a Z
I V
N r A
p m
a M
G s n
o l
l P
t n
a R
i t
a r F e
p J A
O
. O I i
r a
e r
t l
r e
g N a N r e
P V l Mt n
a O
I i
0 0
c r
a l
P 3
P n
a I
T l e
c u _
- N s n s
=.
- A f i
h C
e i
e c
n e
i t
o a
r e
- R S e
c n
O p
s E a s e
i n
lp i s
n u N o B l
~_
C r a
d e n
E a i l t r
a s N i
r G A P
a r
p p
i D
t c
r o
e R V
._ A E
- e e 6
- t 9 t /
i n 2
_ L u n
o 1
/
9 n
C C
- o _
l w g i
e n ve i t
e L R y M e
N t
fe 3-
- a P S I S
R C
A l I l 1\ll l l\l 1lllll
INDIAN POINT 3
__ 4 Pressurized Water Reactor / Westinghouse 4-Loop Plant Generator Electrical Output: 1020 MWe gross / 980 MWe Net Acquired from Con Ed: December 1975 -
1 I- - T Full Power License: April 1976 E
le
" **~"" ,h Major upgrades during 1980s:
e New steam generator and blowdown recovery system o New main condenser e Full flow condensate polishcr e New feedwater heaters and MSR tubes (eliminate copper) e Nev. variable speed circulating water pumps and service water pumps e New low pressure turbines e New containment fan cooler unit cooling coils ACRS IP-3 Meeting on 9/12/96 7
f
W W _
CHRONOLOGY OVERVIEW g Jan 1992 FitzPatrick plant placed on watchlist Oct 1992 Recognition of declining performance at IP3 (SALP and enforcement actions)
Jan 1993 Implement Performance Improvement Plan (PIP)\ - -l FitzPatrick restart ump Feb 1993 Voluntary shutdown of IP3 May 1993 NRC SpecialInspection Team at IP3 June 1993 - Confirmatory Action Letter /Watchlist Jan 1994 FitzPatrick off the Watchlist April 1994 Implement Restart and Continuous Improvement Plan (RCIP) for IP3 March 1995 NYPA Startup Readiness Evaluation completed April 1995 NRC Readiness Assessment Team Inspection June 1995 Confirmatory Action Letter modified,IP3 ready for restart July 1995 Startup and 2 months of operation Sept 1995 Shutdown for equipment problems Oct 1995 Human performance errors during restart l Initiate Operations Procedures and personnel upgrade effort April 1996 Successful restart and continued operation ACRS IP-3 Meeting on 9/12/96 8
IP3 WATCHLIST ISSL ES 4 o&
LL ! _
e SALP scores for period ending August 1992 L 1=d6 declined in 5 of 7 areas ,
e Escalated enforcement actions i e Continuing concerns with performance of Engineering and Technical Services e Recent changes in corporate staff e Significant management attention needed for both plants (FitzPatrick and Indian Point 3)
ACRS IP-3 Meeting on 9/12/96 9
I
CONTRIBLTING FACTORS 4
oW e Inadequate management controls I 0- ^^^
~
e Untimely and ineffective review and correction of problems e Increasing number of plant deficiencies e Two principal events:
- Incomplete testing of AMSAC (Feb 1993)
- Improper isolation of RCS level indication (Mar 1993) e NRC Special Inspection Team (May 1993)
- Weak management processes and control
- Weak management skills and vigor
- Weak independent oversight ACRS IP-3 Meeting on 9/12/96 to
CORRECTIVE ACTION PROGRAMS 4 e Performance Improvement Program (PIP) o&
-Issued Jan 1993, modified Aug 1993 to add NRC restart issues I !- -
- Did not produce required improvements (Dec 1993) ze o Restart and Continuous Improvement Program (RCIP)
- Expanded program reflected more thorough root cause analysis '
-Issued May 1994, updated Nov 1994 based on progress l e Start Up Readiness Evaluation (SURE)
- NYPA assessment of RCIP implementation
- Request NRC Readiness Assessment Team Inspection (RATI) April 1995 ACRS IP-3 Meeting on 9/12/96 II
PRINCIPAL IMPROVEMENTS FOR RESTART $
e Management Process and Controls A M
- Management expectations clearly communicated to plant staff f _
- Effective corrective action program
- Improvements to surveillance and preventive maintenance program
{p ,_ [
i
- Moved Engineering and Licensing to site ,
l
- Plant materiel condition acceptable to support plant restart e Management skills and vigor
- Established full management team
- Management training and development initiatives
- Emphasized questioning attitude and attention to detail I
e Independent Oversight l
- Enhanced effectiveness of QA and acceptance by line management
- Expanded merr.bership and charter of Safety Review Committee
- Nuclear Advisory Committee to Trustees established
- Independent Safety Engineering Group established
- Review committees provided effective oversight ACRS IP-3 Meeting on 9/12/96 12
l l
l RESTART AND INITIAL PERIOD OF OPERATION (1995) $
n&
e Conservative decision to delay startup for RV O-rings (June) g IE, n r
e Operation at reduced pressure to reseat leaking SRV (July) """~ ~"
e Plant shut down to repair Generator Hydrogen Cooler (September) '
e During heatup, safety-related pump switches left in ' pull-to-lock' (October) e NRC / NYPA public meeting on self-assessment results (October) o During heatup CCW isolation valves inappropriately closed resulting in relief valve-lifting and 1300 gal CCW leakage into V.C. (December) e Outage extended to improve operations procedures and personnel performance '
I ACRS IP-3 Meeting on 9/12/96 13
m-_-_ -
SELF-ASSESSME NT CONCLUSIO NS 4 e Plant systems performed well and plant staff responded effectively to emergent work.
e Plant staff continued good use of corrective action program; 4 _
Pg{pj]
improvements needed for timely implementation and effectiveness of corrective actions.
e Conservative decision-making evident at all levels; improve methods and use of resources for making informed decisions.
e Management expectations clearly communicated; additional emphasis on individual accountability required.
e Personnel qualifications are adequate; increased training needed on importance oflicensing and design basis considerations.
ACRS IP-3 Meeting on 9/12/96 14
IMPROVEMENTS FOR OPERATION (after restart experieriee)
M Observations: O LL _
Conducted:
e w g-Identified:
Review of NRC, INPO and QA Information Low level of ownership and accountability Reviews of Event Investigations-low pressure operation Lack of planning and preparation mispositioned switches Inadequate training effectiveness cooling water leak operator response Ineffective self-assessment Discussions with Personnel Personal Observations ACRs IP-3 Meeting on 9/12/96 15
IMPROVEMENTS FOR OPERATION O
Actions:
. oW LL _
e Simple, clear, consistent expectations b~5NrE
- Attention to detail
- Learning from mistakes
- Procedural compliance
- Accountability ,
o Rapid investigation of events e Resources for procedure improvements e Strong management involvement in Training e Long-range personnel plan e Shift mentor program o On-shift oversight group ACRS IP-3 Meeting on 9/12/96 16 I
L _ _-__ _ _ _ _ _ _ _ _ _ - _ - _ _ _ _ _ .
w w w IMPROVEMENTS FOR OPERATION l O Results: A*
e Significant Improvements -
L!_ e e
- Procedure compliance i
- Shift turnovers
- Training effectiveness
)
l
- Attitude and professionalism point communications e Operator " Pipeline" was started
- 8 non-licensed operators in training
- December and April license examinations, all eleven candidates successful
- Commenced new license classes in January and April,1996 e Strengthened Operations Management e SALP Period ended March 2,1996 with Operations evaluated as a Category 3 with positive comments concerning management oversight and initiatives taken during the latter part of the period e Successfully started up the plant in April,1996.
ACRS IP-3 Meeting on 9/12/96 17
SALP St M VIARY o&
Period 8/16/92-3/2/96
- f_g_ _
l
~~~
1 1m
^3/93-5/95 Excluded Due To C.A.L. " ' " ~ * " *
- l Overall Operations 3 Maintenance 2 Engineering 3 Plant Support 1 ACRS IP-3 Meeting on 9/12/96 3g
- - _ _ _l s
OPERATIONS SALP l 4 MANAGEMENT OVERSIGHT PROCEDURES Significant, noteworthy efforts were In the latter three months of the SALP directed by management to improving period, significant progress was made p y tLe w J---
%+=
operator performance in the last three )
in the procedure upgrade program and j months of the SALP period. operator adherence to procedures was found to have improved.
1 MATERIEL CONDITION l Materiel condition of the plant Weakness in administrative procedure significantly challenged the usage continued to be identified later operators. in the SALP period.
Reacted to emerging problems and TECII SPEC AND DESIGN BASIS delayed focus on planned long-term KNOWLEDGE performance improvements. Weak understanding of the licensing ,
1 and design basis of the plant.
ACRs IP-3 Meeting on 9/12/96 39
ENGINEERING SALP
- Response to emergent problems impacted long term programs e Engineering work backlogs have increased e Important technical problem response was good g i 5 Lud@iv e Routine performance varied significantly, sometimes poor e Not successful in improving:
e Work prioritization and backlog controls e Setpoint control program e Configuration management e Poor communications contributed to low pressure event e Safety evaluations were of good quality e System engineers provided aggressive and timely support to Operations I e System engineers did not perform trend analysis e Manag ACils IP-3 Meeting on 9/1266 20
VIAINTENANCE SALP e Management oversight of maintenance was good o Conservative decision making
.f _b, _
m e Good coordination between maintenance and other departments l Ji _g w we~ ==aa e Broad review of extent of condition of equipment problems e Good performance of maintenance activities e Emphasis on procedure adherence by workers e Good supervisory oversight e Effective identification of equipment deficiencies e Appropriate prioritization of work e Work planning not fully effective e Some surveillance procedure weaknesses e Need for better questioning attitude during tests e Increasing corrective maintenance backlog e Declining material condition since restart on July,1995 ACRs IP-3 Meeting on 9/12/96 21 1
Radiological Protection Emergency Preparedness O e Continued strong performance e Well qualified & trained staff gl e Excellent information available to workers e Very good off-year exercise 3 d__r,- !
__ ==
- Surveys & contamination controls very good e Degraded conditions are promptly identiried e ALARA performance very good e Transition to NUMARC EAL's effectively managed e Successful improvement initiatives Fire Protection Security e Effectively gained & maintained control of program o Improvements due to management Attention e Success resulted from task force and oversight e Few low-significant security events e Initial resolution to emergency light issue was weak e Overtime & compensatory measures minimized Housekeeping e Knowledgeable staff e Mixed results e High traffic areas are well maintained e Management expectations not fully developed ACRs IP-3 Meeting on 9/12/96 22
PRINCIPAL IMPROVEMENT AREAS LL e ACCOUNTABILITY L_ M E,_
e MATERIEL CONDITION e TEAMWORK e CONSERVATIVE OPERATIONS e COMMUNICATIONS ACRS IP-3 Meeting on 9/12/96 23
ACCOL NTABILITY Management Observation Program M e Identified as weak by NRC in 1/96, significant improvements made since e Over 2,900 performed since 2/96, deficiencies tracked to resolution 1_I _ m o Increased management presence and coaching in field bu,2 Od e Managers communicate expectations directly with workers Schedule Adherence e 50-60% In April'96 improved to 70-90% now e Managers accountable to identify and address causes of poor performance Accountabilities Are Assigned and Published in POD e Station priorities e LCO coordinators e Special plant evolutions Overdue Assignments Reduced e 138 Overdue action items (ACTS) on 4/1. Reduced to 23 Now e DER's Open >60 days was 38 on 4/29, Reduced to 7 Now ACRs IP-3 Meeting on 9/12/96 24
VIATERIAL CONDITION 4
e Corrective Maintenance Backlog is High p e Non-outage backlog was 1600 in April,96, reduced to <1250 now e Entire backlog was fully assessed prior to startup in early 1996 1-l g
wuM, E2cmN.
e High priority WR's total less than 100 now compared to over 150 in June e Management focus is on eliminating operator impairments i
l e No ' locked-in' control room alarms at 100% power i
e Only one area in plant requires operator to don PC's during routine rounds e ' Operator Work Arounds' receive priority (18 non-outage exist now) i e ' Control Room Deficiencies' under control (34 non-outage exist now) e Operation performs aggregate impact assessment quarterly, monthly if warranted e Fix-It-Now Team recently established, productivity is steadily improving i e Engineering backlog continues to increase e Staff augmentation underway to implement backlog reduction action plan e Temporary Modifications remain high, but are under control ACRS IP-3 Meeting on 9/12/96 25
l TEAMWORK l 4
e Management of emergent issues - new Plant Standard provides direction e Initial 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> / day safety reviewer support to operating shifts f f e Effective emergency plan and severe weather preparations b dgEEar b e LCO coordinators e 12 week schedule support e Tech Spec improvements e Upgraded contaminated tool issue room o Weekly system health presentations I
e Line staff assigned to Training l e Training attendance > 98% at regularly scheduled sessions e Superior plant chemistry performance l
ACRS IP-3 Meeting on 9/12/96 26
w w w 1
I CONSERVATIVE OPERATIONS e Management's safety focus demonstrated by example O e 3/1/96 Backup air compressor to support instrument air piping mods
.f_E _ -
e 3/12/96 Plant returned to CSD to avoid operator challenges J gu_JM _w During repairs to charging system piping e 7/1/96 Direction to use mobile land crane vs. barge crane to replace intake screen e 7/13/96 Power reduction to 45% to calibrate RCP flow transmitter o 8/25/96 Power reduction to 60 % to repair pressurizer spray valve l e Risk significant activities reviewed twice daily by Plant Leadership Team (PLT) e Safety system unavailability minimized by PLT Focus e Dedicated shift technical advisor e Operations additional control room panel walkdown following turnover e Fire Brigade leadership assigned to a non-shift-duty individual e Will likely achieve industry's lowest 3-year average radiation exposure at years end ACRS IP-3 Meeting on 9/12/96 27 p
.A - A
CO VIMENICATIONS l
e Effective pre and post shift turnover meetings nW 1 1
e Structured and coordinated plant meetings _l (
e Weekly plant-wide tailgate meetings - using established agenda Em e b e Plant manager re-established monthly IP3 project and engineering meetings e Risk significant activities reviewed twice daily by PLT e Managers present quarterly self-assessments at POD meeting e Department manager action item feature added to daily POD meeting e Upgraded pre-job briefing standard implemented e Oral instructions discouraged by PLT in favor of written instructions e Operations 3-point communications is the norm at IP-3 e Effective Speakout program o Performance Enhancement Review Committee meets weekly l e Human performance events receive PLT attention daily and are discussed weekly at tailgates ACRS IP-3 Meeting on 9/12/96 28
l
SUMMARY
4 e Competence of plant staff as evidenced by:
M l l
e Conservative decision making l l e Effective teamwork in managing evolving issues -I -
[
En"'"
e Ownership of plant issues is clearly evident e Strong lessons learned focus:
o PLT reviews DER's daily e Extent of condition assessments e A strong industry experience program is in place e Improving self assessment program o Successful FitzPatrick post transient review process implemented I
e Continual improvement in human performance o Improving plant reliability:
o 113 days of continuous operation e On line 157 of the past 159 days ACRS IP-3 Meeting on 9/12/96 29
CONTINL OL S IMPROVEMENT AND THE BL SINESS PLAN O
e Continuous Improvement Plan merged with Site Business Plan to zero base budget and link resources -
to required improvements e Focused Goals:
Superior Safety and Regulatory Performance (SALP 1)
Excellence in Operations (Top Quartile, NRC Indicators, INPO Indicators)
Cost Control (Top Quartile)
Organizational Effectiveness (Top Quartile)
ACRS IP-3 Meeting on 9/12/96 30 i
, e-1 s .
VIEW OF THE FUTURE 4 e NYPA achieves goals of SALP 1 and top quartile operation. -
e NYPA partners with an Advanced Nuclear Enterprise to bring additional resources and techniques into play to achieve long term performance goals.
e NYPA takes part in an alliance or a joint operating company with other New York State nuclear utilities to gain economies of scale and to take advantage of the combined management I
strengths of the participating companies.
ACRS IP-3 Meeting on 9/12/96 31
VIEW OF THE FL TL RE (SALP I vs.SALP 2)
A SALP 1: O e Superior level of safety performance '
e Programs and procedures provide effective controls
[
En e o Issues are a entified through self-assessment e Comprehensive corrective actions e Recurring problems are elimi; sated e Thorough root cause analysis SALP 2:
e Good level of safety performance o Programs and procedures normally effective, some deficiencies may exist e Self-assessment normally good, some issues may not be self-identified e Corrective actions usually effective, some my not be complete e Root cause analysis normally thorough.
ACRS IP-3 Meeting on .I' N 32
b INDIAN POINT #3 1996 PERFORMANCE W 8 W MAV JUNE 1 JULY AUGUS S M OC OBER NO MBER C MB R 1 2 3 4 1 2 1 2 3 4 5 6 7 11 12 13 14 6 7 8 9 10 11 12 3 4 5 6 7 8 9 8 9 10 11 12 13 14 1 16 1 18 19 20 21 13 14 1 16 17 18 19 10 11 12 13 14 15 16 15 16 17 18 19 20 21 l
22 33 24 25 26 27 28 20 21 22 23 24 25 26 17 18 19 20 21 22 23 22 23 24 25 26 27 28 1 4 31 l
On-hne Outage [ l Manual Shutdown Auto Rx Trip KEY EVENTS DURING 1996 2/24/96 Commenced Reactor Startup 5/2096 UnR Shut Down Manua#y - 34 MISV Broken Air une 3/2/96 SALP Penod Ends U21/96 Generator On Une 9 2242 Hrs.
3/11/96 New Vlce Pres. Nuclear Ops. H. Salmon, New Plant 6/14/96 Volt To IP3 By NRC Chairman S. Jackson Manager R. Barrett & Restored General Manager 7/6/96 Hub Mdier Vlsils Site I
h Of Support Poshion J. Comiotes 7/11/96 Unit Commences Pwr. Reduchon 0 0610 Hrs. To 600 j' 3/12/98 Proceeded To CSD - Repair Cheryng Heeder MWe 81150 Hrs. For Repair of Extract. Stm. Leek (Operator Challange) 7/13/96 Further Pwr Reduchon To 450 MWe g 0100 Hrs. For 3/33/96 Commenced Reactor Startup Support Of FT-416 Calibrabon 3/26/98 PORV 456 / Seisty Relief Valves Leak By - 7/15/96 Unit g 100% Pwr 9 2325 Hrs.
- Un6 Cooled To 510 Dog 11610 PSIG 7/22/96 Unit Commences Ma'sual Shut Down g 1430 Hrs.
3/37/96 Una e NOP!NOT For Reper of SI-AOV-6768 l 47196 Generator On Une @ 1456 Hrs. 7/23/96 UnN g 100% Pwr 0 0130 Hrs.
) 4/3/96 Reactor Manuany Tripped 01612 Hrs. - 6/1496 Turbine Rurdiack Tc 600 MWe 3 2020- Attnbuted To Hi Temp. Across H2 Coo 6ers instrument Bus Spike While installing Fuses in R13 l
! 4@96 Gene stor on une g 2306 Hrs. 6/15/96 UnN O 1000 MWe 81023 Hrs.
I 4/1C95 NRC/ FEMA Observed Exercise 6/25/96 Power Reduchon To 60% To Repair Pzr. Spray Valve f 5/1396 Restored I&C Manager And Maintenance Manager 4558 l Posthons. Created Gen. Supv. Electncal Maintenance 6/27/96 Unit Returned To 100% Pwr, Posibon 9/5/96 Una Pwr. Reduced To 930 MQ. ' ,Maintam Main
[ 5/12/96 SALF Moebng Generator Lead Box Vibs.
- At U,ec 5/16/96 E-Plan Drill NRC Observed l
m j
i KEY PERFORMANCE INDICATORS (resuns er or a/314e)
OAL RESULT GOAL RESULT j SALP <2.5 Collectrve Rad. Exposure (REM) <34
- Not Elec. Cap. Factor (YTD) >42.9% Personnel Contaminabon Events <10 j Thermal Performance >995% 90.0 Bdgt. Perf. (Yr. To Date in 3 Mils) <36.5
- NRC CMI Penalbee 0 Overtime (% Base Payroll) <16.9%
- l NRC Violabons ?evel 1.2.or 3) 0 Temp Mods (Open) <32 l Forced & Ptanned Cap. Loss <5% 3 DER Poformance (Av. Eval. Wks.)
j Unplanned Scrams O Pwr. O Personnel Error Rate j Unplanned Safety Syster, Actuabon 0 -Signi6 cant Errors <0.05 l Site Recordable triury Rate <2.14 :2.28 -All Errors <1.00 :1.0e I
les Lost Time Accident Rate <0.46 10.5_ Chemmtry indicator <1.17 g i
- Preliminary Figures EUm / l. l Fair .E Good
! IPS Detus Report 9446 1
i 1
-