ML20151T956
| ML20151T956 | |
| Person / Time | |
|---|---|
| Issue date: | 04/20/1988 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| References | |
| ACRS-T-1659, NUDOCS 8804290209 | |
| Download: ML20151T956 (250) | |
Text
,
/'
ORIGLNg g ,
NUCLEAR REGULATORY COMMISSION
.m=-----==--==es========me======= mss ===========================-
In the Matter of: )
)
THERMAL HYDRAULIC PHENOMENA )
SUBCOMMITTEE )
l l
O ;
4 *
- >n . , , ,
l, .
- i '
i m ,
.. t.
- g,
- ~ L !]5ip:ty;i [;l} j li n f
==================w=======================""""""""""""""""^ i r,
HERITAGE REPORTING CORPORATION O OJkWR,wter 1224 L Street, N.W., Sehe 640 Washbgton, D.C. 20005 ;
(282)628 6 eao42902co eeo4 o PDR ACR$
T-1659 DCD
1 PUBLIC NOTICE BY THE 2 UNITED STATES NUCLEAR REGULATORY COMMISSION'S-3 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 4
m 5
6 7 The contents of-this stenographic transcript of the 8 proceedings of the United States Nuclear Regulatory 9 Commission's Advisory Committee on Reactor Safeguards (ACRS),
10 as reported herein, is an uncorrected record of the discussions 11 recorded at the meeting held on the above date, t:
12- No member of the ACRS Staff and no participant at 13 this meeting accepts any responsibility for errors or
{} 14 15 inaccuracies of statement or data contained in this transcript.
16 17 18 ;
19 20 21 22 23 24 ;
25
() Heritage Reporting (202) 628-4888 Corporation
,1
- - . . _ ~ _ . . , _ _ _ . - - , _ _ . . _ . , -
221 7'
(_)N 1 UNITED STATES NUCLEAR REGULATORY COMMISSION 2 ADVISORY COMMITTEE ON NUCLEAR SAFEGUARDS 3
In the Matter of )
4 )
Thermal Hydraulic Phenomena )
5 Subcommittee )
6 Conference Room 7 Best Western Motel 700 Lindsay Blvd.
8 Idaho Falls, Idaho 9 Wednesday, April 20, 1988 10 11 The above-entitled matter came on for hearing, 12 pursuant to notice, at 8:30 a.m.
13 BEFORE: DAVID A. WARD *
, Chairman 14 Research and Manager on Special Assignment E. I. du-Pont de Nemours & Company 15 Savannah River Laboratory l Aiken, South Carolina 16 ACRS MEMBERS PRESENT:
17 DR. WILLIAM KERR 18 Professor of Nuclear Engineering Director, Office of Energy Research 19 University of Michigan Ann Arbor, Michigan 20 21 CONSULTANTS:
22 MR. I. CATTON 23 MR. M. PLESSET ,
24 MR. V. SCHROCK 25 O Heritage Reporting Corporation (202) 628-4888 ,
l
. _ _ _ . . _ , _ _ _ _ _ . _ . . _ _ _ _ _ _ _ _ - - . _ -_ _ . . _ _ - _ - . , _ - _ _ . _ - _ _ _ _ _ . . _ - , . _ . . _ . _- _ O
. . . . . - . . - - . - . . . .~ -
l 222
~
l' APPEARANCES: (Continued)
I 2 NRC COGNIZANT STAFF-MEMBER:
3~ MR. PAUL BOEHNERT 4
-NRC STAFF PRESENTERS: ,
5 L. SHOTKIN 6~
N. LAUBEN 7 i N. ZUBER {
8 i
- j. 9 -
10 '
11 l 12 r
- 13 !
14 L
i 15 16
- l. .
i i
17 !
r 18 !
l 19 !
e i
!- 20 ,
' ?
21 i i 22 '
1 I a 23 !
24 ;
i 25 I t
O Heritage Reporting Corporation l
(202) 628-4888 t
t l.i .. . - _ _ _ . _ , . _ _ _ _ _ _ . . _ _ _ __ _ . _ _ _ . _ _ _ _ _ _ _ _ _ . . _ _ _ _ . _ !.I
223 (f 1. PROCEEDINGS 2 MR. WARD: Good morning. The meeting will now come 3 to order. This is the second day of a meeting of the Advisory 4 Committee on Reactor Safeguards, Subcommittee on Thermal 5 Hydraulic Phenomena. I am David Ward, the subcommittee 6 chairman.
7 I apologize to the other committee members and 8 presenters that I was not able to attend yesterday. Sorry I 9 missed the presentations.
10 Today we are going to hear a review of the final 11 version of the proposed revision to the ECCS Rule, and then 12 after the noon break, we will begin subcommittee discussion of 13 a proposed report on the NRC research in the area of thermal
() 14 hydraulic phenomena.
15 And I believe first we are going to hear a brief 16 presentation, before getting into the discussion of the ECCS, 17 we are going to have a bricf presentation from Dr. Shotkin on 18 the -- well, his views of the, I guess, the direction of the 19 research program is taking and should be taking.
20 Also, I would like to work in some time, perhaps we 21 can do this after lunch, I would like to hear the reaction of 22 the subcommittee members and consultants to what they heard 23 yesterday about the RELAPS code.
24 Are there any comments now that anyone would like to 25 make before we go ahead with Dr. Shotkin.
O Heritage Reporting Corporation (202) 628-4888
224 tsj 1 (No response.)
2 MR. WARD: I should acknowledge for the record that
'3 Dr. Kerr is here also, and will be here for the rest of the 4 meeting.
5 Lou, go ahead, please.
6 (Slide.)
7 DR. SHOTKIN: Good morning, this presentation was 8 supposed to have gone on as the last presentation yesterday, '
9 but once we found out that Chairman Ward and Chairman Kerr were 10 going to be here this morning, I guess it was decided to just 11 delay it until this morning.
12 I guess I was asked to do a review of the future 13 plans for thermal hydraulic research within NRC. And I will go
/"')
(_/ 14 over these plans as we have them today. There will be two 15 things, two items that have not been factored into the 16 viewgraphs that I will mention as I go along.
- 17 The first item is the reorganization that the Office 18 of Research is going through right now and, in fact, even this 19 week the staffing level is being worked out. And as I have 20 knowledge on anything that will affect what I have on the 21 viewgraphs, I will point that out.
22 Another item that is not factored in us the 23 prioritization of research that is also going on, and there is -
24 not much I can say about that except to say that the criteria 25 that are being used to prioritize research within the agency.
' -( ) Heritage Reporting Corporation (202) 628-4888
~
225
( 1 Now'this.is one type of.research'prioritized against another 2 type.
3 Thermal hydraulic research is falling cut in one of
.4 the lowest categories, primarily because it is directed towards 5 solution of long-term problems rather than solution of 6 immediate problems. And for that factor alone it is coming out 7 rating, very subjectively but still coming out rating in the 8 lowest category.
9 The outline of my presentation, I will cover four 10 things.
11 First, I will present a viewgraph that I presented at 12 the last meeting in Los Alamos, slightly modified, and I will 13 show you where it is modified.
14 Second, I will talk about some new applications 15 projects.
16 Third, I will talk about this task group on future 17 thermal hydraulic research. That's an internal NRC task group 18 composed of members from Research, NRR and AEOD.
19 And finally, since we are going through a fiscal '90 20 budget exercise, I will just show you where we stand on that 21 exercise, because part of that exercise is looking into the 22 future.
23 (Slide.)
24 This was a viewgraph that I presented at the last 25 meeting, just showing projects and schedules. Planned O Heritage Reporting Corporation (202) 628-4888
226
(~)h s_ I resources I'll show on the last viewgraph of the presentation.
2 The new items since the review at Los Alamos that 3 ICAP and code consortia line was extended through '91. That 4 was just a mistake on my part at Los Alamos, and I had left out 5 the line on basis studies, and I just continued that.
6 The effect of the reorganization on this graph will 7 be where you see in-vessel accident management, just cross of 8 "in-vessel".
9 What this shows is that the kind of work we are doing 10 we are dividing into two areas -- research which is shown on 11 the top.
12 DR. KERR: Excuse me. Does crossing off "in-vessel",
13 is accident management going to include in-vessel, or that in-14 vessel is not going to be studied?
13 DR. SHOTKIN: It r.Lans that all accident management 16 is included rathei. than breaking off and saying just in-vessel.
17 DR. KERR: Thank you.
18 DR. SHOTKIN: What tnis shows is that in the research l 19 area, which includes our experituental facilities and our 20 computer codes, that we are planning to close out almost all 21 projects around fiscal year '91. And the task group that I 22 will tell you about, our mission is to define what acidotic 23 level thermal hydraulic research should reach within the 24 agency, what acidotic resource level, what code we should keep 25 and what should we be doing in research area.
}
Heritage Reporting Corporation (202) 628-4888
227 D
d 1 Now on the bottom part you see the applications area, 2 and that's where we see that as continuing into the future with 3 sv...e longer term activities, although some of them will come to 4 an end as needed.
5 And since I went through that at the last meeting, I 6 won't spend very much time on this viewgraph unless there are 7 any questions.
8 MR. WARCs Lou, I see there was a Federal Register 9 Notice that came out I think from the materials side of RES 10 recently that was asking for public comments on a proposal to 11 look at -- well, a redefinition of large break LOCA as it 12 applies to ECCS performance and equivalent qualification, not 13 the containment, but to those two as an extension of other 14 large break arguments.
15 Now do you have a role in that?
16 DR. SHOTKIN: Yes, we reviewed that before it went-17 out, but that is going along independent of our work on the 13 ECCS Rule if that is what you are talking about.
19 MR. WARD: No, I realize that. But in the longer 20 term, do you see thermal hydraulic research issue in that? I l 21 mean will that be part of that?
22 DR. SHOTKIN: It depends on the comments that come in l
23 and how we handle the comments.
i 24 Yes, there could be some longer-term role, but I 25 don't think it will be a large role.
l O Heritage Reporting Corporation (202) 628-4888
228
() 1 (Slide.)
2 Since we are stressing starting up or poing into 3 applications projects, I just thought I would talk about a new 4 of the projects that we have going on. And, again, the longer-5 term view is that as we phase out thermal hydraulic research or 6 bring it down to some appropriate acidotic level, we will be 7 concentrating on applying the results of that research to solve 8 issues of regulatory interest, and here are just a few 9 examples.
10 We have just received a draft user request from NRR 11 where they have defined certain containment issues that they 12 want us to be looking at. They primarily involve the 13 definition of the super heat temperature in the containment
() 14 locally as it affects local equipment after a main steamline 15 break, or I guess any other accident where you could get. super 16 heated steam in a container.
17 A lot of the Office of Research is working on things 18 called generic issues, and some of the branches that are 19 working on the generic issues, or at least one of the branches 20 as written us a couple user need letters where they want us to 21 help them out to resolve some thermal hydraulic aspects of 22 their generic issues.
23 One of them is steam generator tube rupture where 24 they are looking at the iodine transport to the environment 25 after steam generator tube rupture, and we have been asked to O Heritage Reporting Corporation (202) 628-4888
229
( 1 do some analyses of just the thermal hydraulic behavior of the 2 steam generator after steam generator tube rupture where the 3 tube rupture occurs near the top of the tubes, or near the 4 U-bend.
5 Another new issue that is being looked at in 6 connection --
7 MR. WARD: That's a U-tube generator is what you mean 8 or is it?
9 DR. SHOTKIN: No, actually -- I'm sorry, one is for 10 Kawaunee and one is for Calvert Cliffs. So one is a U-tube and 11 one is the once throtigh.
12 DR. KERR Are you describing things that you are 13 going to do or things you have been asked to do?
' ( 14 MR. WARD: Things that we have been asked to do, yes.
15 On the containment, we have been working with these 16 people for about a year to define the project, and now they 17 have got a draft in writing, and we are not going to iterate 18 with them. We haven't started work on that one.
1 19 On the steam generator tube rupture, we are again in 20 the process of defining what has to be done after we have 21 received the user need letter I would say about a month ago.
22 The next one, the turbine trip without reactor trip 23 came up as a result of the B&W safety assessment that agency 24 was doing. And one of our engineers came up with an over-25 cooling transient that we hadn't looked at pressurized thermal O Heritage Reporting Corporation (202) 628-4888
s l'
230 1 shock, and in fact Idaho has finished most of their 2 calculations on that that do show that when you have a turbine 3 trip --- actually it's worded backward. It's a reactor trip 4 without a turbine trip. Sorry about that.
5 When you have a reactor trip and the turbine does not 6 trip,-you do get an over-cooling that leads to much more over-7 cooling than we have seen in the worst cases we looked at on 8 the PTS which were things like main steamline break.
9 So this is. reactor trip without turbine trip.
10 Just to give ycu a -- now, tnese are 011 fairly small 11 projects with quick turnaround times associated with them, and 12 that is probably the kind of work we are going to be getting 13 into in the thermal hydraulics area in the future when we talk t')
\_/ 14 about applications for most of these.
15 The next one was a turnaround time of about two 16 weeks. AEOD, this LaSelle plant had lost --
17 DR. KERR: What you are describing now is important 18 work that needs to be done. It's hardly research.
19 DR. SHOTKIN: Okay, as we look -- it's being done -
20 within the Office of Research. ,
t 21 DR. KERR Okay.
22 Dr SHOTKIN: However, I agree with you that it would 23 hardly be called research in the traditional sense, and it is 24 well to keep in mind that we have broken up our work into 25 something called research, but I think everyone would
(
l Heritage Reporting Corporation l
(202) 628-4888
.m--w. -, . ,
P 231
(]f 1 understand as research in a traditional sense and we are 2 phasing that out, and we-are moving into the area of 3- applications, and it is still being done-within the Office of 4 Research.
5 MR. WARD: Are these things that-five years ago would 6 have been'done in NRR, I mean, some of these, or all of them, 7 or what?
8 DR. SHOTKIN: 'I think the -- I would say just about 9 everything except the dPA at the technical training center, 10 yes, and this in-vessel -- the accident management is a much 11 bigger thing, so I don't know if they would have done it.
12 But things like the steam tube rupture, reactor 13 triple out-turbine trip, the LOS analysis, to some e the
() 14' containment-issues, yes, would have been done in the past 15 probably under tech assistant contracts.
16- There was an incident at the LaSalle plant, a boiling
, 17 water reactor, where power oscillations resulted when some 18 malfunction in the feed water, and AEOD just wanted to do some 19 calculations to see what went on. And the Brookhaven plant 20 analyzer was set up. They responded quickly, and we had the 21 answers that AEOD .3nted within -- I'd say within about a week i
! 22 and a half.
23 I have talked about this before. The technical 24 training center has become a user of our nuclear plant 25 analyzer, and we are in the process of defining the exact scope O. lieritage Reporting Corporation (202) 628-4808 1
1
i 232
() 1 of that program. The global scope has been defined as what has j 2 to be done, but just what the hardware and tne software 3 requirements are will be defined. There will be a meeting next 4 week.
5 And, fina).ly, the-reorganization, you just cross of 6 "in-vessel", and that will be probably one of our major 7 activities in the future, is working, or at least in my 8 branch's area, it will be working on an accident management t 9 plan.
10 And again if you are asking is this traditional 11 thermal hydraulic research, I would have to answer no and yes 12 with the emphasis on no, because primarily we will be looking I 13 at what are the needs of the operating staff. So there will be O 14 a human element in this that is missing in most of our previous 15 thermal hydraulics research.
16 We will have to conclude our knowledge of plant 17 instrumentation, environmental factors. It is a much broader i
18 program. Some parts of it will involve traditional thermal ,
- 19 hydraulic research, although if we are talking of doing i 20 analyses of our experiments, that will probably be done for the l
21 most part in another branch, and again that is part of the 22 reorganization.
I 23 However, if anything does get done, it will have to 24 include analyses that go into core melt, or vessel breach, and l ,
l 25 what happens in the container. So all of these will have to be i ,
() Heritage Reporting Corporation l
(202) 628-4888 ;
F
. ~ ~ , . _
233
/~N
(_) 1 factored into the accident management.
2 And, again, this will be much beyond our traditional 3 -erea of thermal hydraulics.
4 Now let me return to --
5 MR. SCHROCK: Lou, that's a kind of contradictory 6 statement in the sense that whether you have core melt or not 7 depends on the evolution of all of the events leading up to the 8 incipient core melt. Unless you know how to do all of that, 9 you are not going to have any worthwhile information on core 10 melt.
11 So what does it mean exactly? It excludes things 12 that are not directly core felt?
13 DR. SHOTKIN: I'm glad you asked that.
14 MR. SCHROCK: Is it a base change process of the 15 solid materials to --
16 DR. SHOTKIN: Well, let me answer -- I'll give you 17 the real answer second. Let me answer what I think you are 18 asking first, and I know people that worked in thermal 19 hydraulics look at thermal hydraulics as being a much broader 20 field. That includes everything with this two phases or even 21 multiphase, and why couldn't it be a metal and liquid metal.
22 And in that definition of thermal hydraulics, yes, 23 that does involve thermal hydraulics.
24 I am using the term thermal hydraulics in the sense 25 of this subcommittee which is the thermal hydraulic phenomena O Heritage Reporting Corporation (202) 628-4888
N 234 m
\_) 1 . subcommittee that I believe is restricted to looking at thermal 2 hydraulics before there is any core damage. l 3 Once you get into core damage, there is another !
I 4 subcommittee that looks at that.
5 So when I say thermal hydraulic research, I am using f 6 it in the very narrow sense that's connected with the 7 definition of our work in the past and the work of the 8 subcommittee. But now give you a real answer to your question.
9 Where does accident management research. fit in and 10 why are vnt doing it, and let me talk in terms.of just the very 11 broad perspective of NRC regulating an industry.
12 What we are talking about at first are the emergency 13 operating procedures; let me first cover that. I will break my 14 answers -- a three-part answer. Let me do it that way.
15 First, there is emergency operating procedures that 16 we are covering. NRC is regulating the industry. We send out 17 the emergency procedure guidelines. The industry comes back 18 with their response to that. They thca develop their emergency 19 operating procedures which we don't -o~'ew, but which we do _
20 send investigating teams out to the plants, and that is now a 21 big area in NRR. And they do check out these EOPs as part of 7 22 exercases or drills that they do at a p.lant. And they have -
23 come book and found certain areas where improvement can be i
24 made. ,
25 However, from the NRC perspective, that is being
- (:) Heritage Reporting Corporation (202) 628-4888 t
, ,.-...-,wm-.-, -,--.,--__-,_._,~_y, - . ~ _ . - - - . - _ . . , , s.. -
J 235
.G
(-) I handled. That's being handled by NRR through their EPG 2 guidance and their EOP review through these teams.
3 Now, the second area is accident management, but let i
4 me skip that. Let me go to the emergency response. That's 5 after you get containment breakthrough. And, again, NRC is 6 handling that. ,
7 We have emergency response plans that we work with
- 8 the industry and the local governments, including 9 Massachusetts, to try to see how these plan? come into effect.
10 So in the regulatory sense, we have covered accidents 11 up to the time of inadequate core cooling, and accidents from 12 the time of containment failure afterwards.
13 Where the gap is is between those two times; between 14 the time when you have reached inadequate core cooling and you 15 start melting the vessel -- molting the core through the time 16 when you might get a vessel breach, and then through whatever -
17 happens in the containment until you get containment breach, a 18 possible containment breach. And that is the area that there 19 is a void in the regulatory sense.
20 There is also probably very large voids in the 21 research sense as well, but that's part of our task in accident 22 management is to try to define what research needs there are in 23 that voided area.
24 Did that cnswer your question?
25 MR. SCHROCK: I suppose kind of.
O Heritage Reporting Corporation -
, (202) 628-4888
236
() 1 DR. SHOTKIN: Can I continue?
i 2 MR. WARD: 'Please do.
3 DR. SHOTKIN: Now, having said that, let me return to 4 traditional thermal hydraulic research and what we are doing to 5 try to define what our future requirements are.
6 We have something called the senior program 7 management steering group which was composed of division i
8 directors. And one of their tasks is to look at things like 9 thermal hydraulics. This group got together and formed a task 10 group made up of branch chief where our job, and I am on that 11 committee or I'm on that task group, our job is to talk with 12 each other from the producer point of view, and the user need 13 point of view as to what should we be doing in the future as
\_/ 14 far as agency needs in thermal hydraulics in the traditional 15 sense. This is up until inadequate core cooling in this first '
16 area that I was talking about.
17 And what I would like to go over is something that
- 18 just the same documents we have been using within the task i
19 group to show you what we are doing. And we met a couple of .
20 weeks ago. We are exchanging correspondence. We intend to 21 meet next veek to finalize our recommendations and then go back I 22 to the senior management group in the early part of May.
23 (Slide.)
24 And the first aspect -- the first ground rule is Part 25 A, that there is a declining emphasis on thermal hydraulic O Heritage Reporting Corporation (202) 628-4888 L_
237
() 1 research', and again this is traditional thermal hydraulic 2 research, within the agency.
3 And as I showed on the bar chart, most of our --
4 almost all of our major programs are' coming to an end. 2D/3D, 5 ROSA-IV.are computer codes, we're all coming up with either 6 completion of projects or final versions of codes by '91. -That 7 covers everything. The final versions of the codes, as you 8 heard yesterday, are more like June or July of '89. Then we 9 leave a year or two for ICAP assessment, and by '91, we have 10 come up with last planned versions of code.
11 I am talking about planned versions, and what this 12 task group job is to look at what maintenance levels are we 13 talking about in the future.
OP 14 Now as part of the declining emphasis, we do plan to 15 have a solid base program that will include computer codes, 16 which is code maintenance primarily; basic studies to support 17 any modeling needs in the computer codes; maintenance of a 18 thermal hydraulic technical support center here at Idaho that 19 will help us apply the results of this research to priority 20 studies of interest to NRC; and the nuclear plant analyzer, 21 which we see as a use-oriented tool that can help us resolve 22 problems.
23 Now Part B, we just talk about what thermal hydraulic 24 codes we wanted this task group to look at, and for the next --
25 (Slide.)
O Heritage Reporting Corporation (202) 628-4888
4 2
238-() 1 In order to try to focus a decision by the task ;
2 group, we have asked them to'just-look at five computer codes 3 which are listed on the bottom, and the year for the final 4 version is also listed. And as you can see, in at least three, 5 the bottom three we finished, or at least we have told the task 6 group that we are finished. RAMONA-3B and COBRA-NC, there is 7 no --
8 MR. WARD: What is the difference?
9 DR. SHOTKIN: What?
10 MR. WARD: You said we have finished, or at least we 11 have told the task group that we are finished.
12 DR. SHOTKIN: Well, the purpose of this document was j 13 to generato discussion within the NRC, and we took what we
() 14 hoped were representative extreme positions to see who would 15 scream essentially.
16 We have given the task group the option to change any 17 of these plans or to como up with new recommendations, but we :
t t 18 have taken a definite stand so that they can shoot it down if 19 they want.
20 DR. KERR: Well, are these recommendations on your
! 21 part that these be -- !
22 DR. SHOTKIN: Okay, that's a good point. Let me 23 separate recommendations from final actions from preliminary I
i 24 actions.
l 25 What you see here on this table are final actions and O. Heritage Reporting Corporation (202) 628-4888 i r
t l
239
( 1 preliminary. actions. The final actions are on COBRA-NC and 2 RAMONA-3B where we have definitely said we are finished, and we 3 said that last year.
4 The preliminary action is on TRAC-BWR where we are 5 saying this is the final year, 1988, and then the others are 6 scheduled for 1989.
7 (Continued on next page.)
8 9
10 11 12 13 14 15 16 17 18 1 19
- 20 1
j 21 l, 22
! 23 24
[ 25 C)
Heritage Reporting Corporation (202) 628-4 8 f38 l
l
240
( )- 1 DR. SHOTKIN: Just to help the Task Group get 2 calibrated, we pointed out the fact that we either have closed 3 or are closing almost all of our test facilities. The main 4 one, the 2D/3D, we will finish in 1989. We have MIST 5 finishing this year, and there may be a follow-on program of 6 separate effects testing with the owners group.-
'7 ROSA-4 is the one that extends furthest out in time, j 8 But as I showed in my bar chart, our commitment to that program 9 will be a minimum commitment with just information as needed.
10 So let me go down to the bottom which is really the 11 main thing that the Task Group was supposed to comment on. Tnd 12 that is the maintenance plan. They could also comment on our 13 final version plans, if they wanted to. But primarily, we 14 wanted them to comment on maintenance plans.
15 We assume that we will finish codes as planned. But 16 then the question is which codes do they want us to maintain in 17 the future, and what is involved in code maintenance. And we
- 18 put up just the representative plan, a strong position, clear, 19 just to generate discussion, which by the way it did.
- 20 We put RELAP-5, a maintenance plan of two to three 21 man years per year. And we put the major effort on that, 22 because it is the most widely used.
- 23 On TRAC PWR, we put either zero or up to three man
(
24 years per year as dictates. For the last three, we said that 25 our plans right now are to put it on the shelf. And this was a
} Corporation Heritage Reporting (202) 628-4888
241'
) I concept that people had not been willing to face up until now, 2 our-saying tliat when we are finished ~with a code, that we will .
3- just put it in an air-conditioned room, _and that is it. And-4 when people want to use it, they will take it off the shelf and-
~
5 use it. And we are getting comments back now.
6 DR. KERR: Are you sure that you should put it in an 7 air-conditioned room?
8 (Laughter.)
9 DR. SHOTKIN: What we mean by that is that the 10 documentation is completed, and that the tape is in good enough 11 use. So that when somebody wants to use it a year from now or 12 a year and a half or two years from now, that they will-have at 13 their disposal everything that they need to use it. They may 14 not have continuous expertise, but they will have everything on 15 the shelf, that you just need to take it off the shelf and i
16 start it.
17 MR. CATTON: Does that mean a QA manual for each one
! .18 of them?
i 19 DR. SHOTKIN: That may not mean a QA manual. That 20 will mean a users guide on how to use it, and some necessary 21 information on the limitations of that code use. But not 22 necessarily a detailed QA manual that looks at every model in
! 23 correlation with the code. Although I think that we are i
24 getting very close to QA manuals on just about all of them.
25 MR. Sci! ROCK: Your statement that RELAP-5 is widely
! lieritage Reporting Corporation (202) 628-4888 l
r
[
m <
242 ,
O s_/ :1 used and that the TRAC PWR_is little used suggests that.there ->
2 .are not many. computations done on BWRs.
3 Is that a proper conclusion?
4 DR.-SHOTKIN: No.
5 MR. SCHROCK: -So what are used'for BWRs?
6 DR. SHOTKIN: RAMONA, and to some extent RELAP-5.
7 MR. SCHROCK: But we have not heard anything about 8, BWR applications of RELAP-5.
9 DR..SHOTKIN: Because it is'not supposed to be used j 10 for PWR. That has been the NRC position. Well, let me tell 11 you the comments that we have gotten back from our group. That-12 they disagree with our maintenance plan for TRAC PWR and for 13 RAMONA 3D. For RAMONA 3B, they felt that they wanted'a. code 14 that could look at BWRs, and that would include the feedback 4
15 from kinetics.
16 And to some extent, that is based on the view of what
- 17 our user offices have been faced with in the last few years and 18 what they see coming up in the next year or two. It is not a ,
19 very long-term view. [
4 20 On TRAC PWR right now, they are using it. They are 21 coming in with applications from the UPI plants. And they are
- 22 using that code to do the audit assessment of what the industry [
23 is cohing in with.
24 On TRAC BWR, they would agree that we could put it on 25 the shelf. And putting it on the shelf does not mean that we j Heritage Reporting Corporation (202) 628-4888
243
({[ 1 are abandoning it. It means that the needs are so infrequent 2 that it just does not pay to support any larger maintenance 3 level.
4 Now we will probably have to negotiate with the' labs 5 and with our management as to what level of maintenance that we 6 should keep on these codoc. But this is the extreme position, 7 of putting it on the shelf. And this Task Group-is going to 8 get together and really thrash out is this going to_be 9 acceptable.
10 Two things that did come out of the Task Group 11 meeting that probably might help you understanding the nature 12 of the discussions. They brought out two points. First is 13 that putting something on the shelf, that we also put it on the
() 14 shelf with a bunch of input decks. That if we just have the 15 code on the shelf with a manual or whatever you want to call 16 it, a QA or user manual, that if we do not have input decks 17 generated by the people who know how to use that code, then it 18 may make the code unusable.
19 So that may be something that comes out of this, 20 although it is very premature for me to say this right today.
21 But there may be a larger emphasis on generating input decks 22 for these codes while they go up on the shelf or before they go 23 up on the shelf.
24 The other thing that they brought up which keeps l
25 coming up occasionally at NRC is why does not NRC have its own l Heritage Reporting Corporation (202) 628-4888
244 1 in-house capability to do these analyses. And that is 2 something that we will discuss, but I do not think that will 3 get anywhere.
4 But what it may mean that if we decide to do that, 5 that if we put it up on the shelf, that we may want to put it 6 up on the shelf with the capability to use on a smaller machine 7 like a PC.
8 This is the Task Group charter of what they were 9 supposed to be commenting on. And we did comments primarily on 10 B. This is what we focused in on at the meeting. On A, we 11 said, well, those are our close-out plans, and we discussed it
- 12 in more detail at the meeting, and you can talk about it. You 13 can say are there any capabilities left out, what criterias 14 should be used to determine code improvement. We have told you 15 our criteria. It depends on use. What level of experimental 16 studies is appropriate to meet future needs.
17 There was very little discussion on Point A from our 18 user offices. And I guess that they are just leaving it up to 19 us to determine that.
20 The main discussion was on B, the code maintenance l
21 plans. Which computer codes should be maintained. And I think 22 that I gave you some flavor of what the responses where. What 23 activities should code maintenance cover. And again, what they i 24 added was a generation of input decks. What code user 1
25 conveniences should be provided, and that was the emphasis on
(} Heritage Reporting Corporation (202) 628-4888
245
') the input decks to go along with the codes as they go up on the
( I 2 shelf.
3 And at our next meeting, and we are scheduled to meet 4 next week, we are going to draft recommendations on thermal 5 hydraulic code plans, primarily on code maintenance, and we are 6 going to submit it.
7 Oh, here is the name of this group, Reactor System 8 Safety Senior Research Program Steering Group. That is why I 9 could not remember it. It has eight words to it.
10 And we are going to do that by mid-May. We are going 11 to prioritize recommendations and state the basis for the 12 recommendations. And I guess that we ought to factor in any 13 results that come out of this subcommittee, recommendations
() 14 based on this meeting and the previous meeting at Los Alamos.
15 I guess that by the end of tomorrow that we will have 16 some draft of your recommendations, so that we can use it next 17 week.
18 Is that fair to assume?
19 MR. WARD: Probably not. Because really, any 20 recommendations should be made by the full committee, and not 21 just the subcommittee. Even though we will have something 22 written, that will not stand as our recommendation.
23 DR. SHOTKIN: Okay. I guess that the way that we 1 24 could work it is that we could get our Task Group 25 recommendation, and we could even send it to you for comments.
O Heritage Reporting Corporation (202) 628-4888
i 246 (f 1 We could exchange literature. And maybe-we could' delay the 2 Senior Research Program Steering Group meeting until we have ,
3 your recommendations. :
4 MR. WARD: I think that sounds more reasonable. I 5 would like to try to get a letter to the Commission on the 6 subject from the full committee at ths May meeting, but I do 7 not know whether we will be able to do that or not. Bill.
8 DR. KERR: Whose recommendations or whose plan is 9 this that you have talked about; it is not all of research, but i
10 some subgroup of research, or is it all of research?
11 DR. SHOTKIN: Okay. The people who are involved in l 12 the Task Group, there are two members from research. I am a 13 member, because we are the producing group. We are producing f
( 14 these codes, and we are looking for users. And what we have on 15 there are the possible users within NRC. So there is one from 16 the Office of Research from another group that looked like the ,
t 17 main user. There is one from NRR representing the whole .
18 office. And one from AEOD, again representing the whole 19 office, and at the branch chief level. !
- 20. DR. KERR: So what we have been hearing is a r 21 recommendation of that Task Force group?
s 22 DR. SHOTKIN: No. It is a recommendation of the 23 producing group, my group, of what our plans are for the codes, i l
24 and bringing it before the NRC user group, and having them ;
i comment on it and make any changes that they want to it.
25 P
C:) :
Heritage Reporting Corporation '
(202) 628-4888 i
. _ _ . _. - _ - _. . _ ,,_ _ _. , .- m, - __ . - _ . _ . - - , , , . , _ , . - , - .
247 (2,,I 1 DR. KERR: Is what we have seen a result of that 2 process?
3 DR. SHOTKIN: No. What you have seen is a result of 4 internal discussions within the Office of Research.
5 DR. KERR: It is a smaller group than all of the 6 Office of Research, is it not?
7 DR. SHOTKIN: Well, it is a group within the Office 8 of Research that would be interested in this, possibly users of 9 this.
10 DR. KERR How many people?
11 DR. SHOTKIN: We get directions from our office 12 director and the deputy director. It goes through a division 13 director and down to a branch chief. So that is the chain.
O 14 DR. KERR: I am not making my question clear.
15 Is this in effect a recommendation of one person, 16 three people, ten people, or none of the above?
17 DR. SHOTKIN: Okay. I drafted it meeting with people 18 in my group. And I think that the people involved in drafting 19 this were myself and the section leaders. I think that we did 20 show it to a couple of other people in the group. We then 21 tasked it through --
22 DR. KERR: That is enough.
23 DR. SHOTKIN: But it was reviewed by our deputy 24 division director and our division director, who takes his 25 orders from our office director. The directions to develop O Heritage Reporting Corporation (202) 628-4888
- 1 248
.(_
O) I this came from our office director.
2 DR. KERRt The second question, has the National 3 Academy of Sciences report had any influence on this set of 4 recommendations?
- 5_ (Pause.)
6 DR. KERRt If you have to think that long, I have my 7 answer.
8 DR. SHOTKIN: Yes. I am trying to remember. I 9 remember that they said something about thermal hydraulic 10 research.
11 DR. KERRt Okay.
12 DR. SHOTKIN: But there is very-little. If I could 13 just try to picture the dynamics that go into this. Our 14 budget, the office budget is going down. There are new areas 15 of research that are coming up cumpeting for money in 16 traditional areas. One of the new areas is, for example, 17 accident management.
18 DR. KERR I am just trying to find out if that NAS 19 report was helpful, and I gather that it was not particularly 20 helpful to you.
21 DR. SHOTKIN: Not in justifying any continuation of 22 thermal hydraulic research beyond what you have seen, no.
23 DR. KERRt You know, the report might be interpreted 24 to say shut down all thermal hydraulic research, which would be 25 a helpful recommendation, maybe not a pleasant one. But it did O Heritage Reporting Corporation (202) 628-4888
i:
249
} 1 not have much impact on this set of recommendations.
2 DR. SHOTKIN: That is correct.
3 DR. KERR: Thank you.
4 DR. SHOTKIN: Now in this, we are going through a 5 1990 budget exercise. And these numbers for 1990, 1991, and 6 1992 have very little meaning, especially after the 7 reorganization. But this just does give you a snapshot of 8 where we were a couple of weeks ago in developing our fiscal 9 1990 plans.
10 What we have done is taken primarily the projects 11 that you saw on the bar chart, and tried to put resource 12 estimates in terms of millions of dollars by.each of those 13 areas.
14 And again, to show the reorganization, the bottom, 15 the in-vessel part should be crossed out, where it says 16 in-vessel accident management. So I will start from the 17 bottom, and show that we did break that out separately to show 18 that this is something that is connected to thermal hydraulic 19 research, but may be quite different, because it involves the 20 humaia element, instrumentation, and a lot of other things 21 besides just thermal hydraulics.
22 The main thing that I want to point out here is that 23 these are plans. If you want to know whose plans these are, 24 the detail is at the division level. But these are the office 25 plans as interpreted by the people who have to implement the O Heritage Reporting Corporation (202) 628-4888
250
() 1 plans.
First, we are showing thermal hydraulic research 2
3 going down with time. And I believe that that is an agency 4 recommendation. And thermal hydraulic research is what you see 5 defined here. It is the computer codes, the traditional 6 thermal hydraulic computer eodes and the traditional thermal 7 hydraulic experiments. And you can see that we show going down 8 to some level of about $4 million to $5 million by 1991 or 9 1992.
10 We show the regulatory applications part increasing 11 to another level starting at about $2 million this year, fiscal 12 year 1988, and going up again to between $4 and S5 million 13 around the year of fiscal 1990.
() 14 The project that are you see is primarily the 15 Technical Support Center here in Idaho that we show reaching a 16 level of about S2 million that would be looking at the 17 applications of the results of thermal hydraulic research to 18 priority studies.
19 We show advance light water reactors going up, 20 although we are very unsure of what work has to be done in that 21 area. And I think that I covered that at the last meeting.
22 And then finally at the bottom, we show accident 23 management. Although here, you have to cross off the in 24 vessel, and maybe those numbers would have to be higher if we 25 include all of accident management. But we do show that as O lieritage Reporting Corporation l (202) 628-4888
251 I~') 1 increasing, and then approaching another level, b
2 And that is all that I have, unless there are any 3 questions.
4 MR. WARD: Let's see, the bottom line is not included 5 in the next to the bottom line, or is it?
6 DR. SHOTKIN: No, it is not. The reason that it is 7 going to show up in the budget on a separate line item. And 8 for the other reason, that it is largely out>.de of traditional 9 thermal hydraulics.
10 MR. WARD: Yes.
11 DR. SHOTKIN: Although we are working on a plan, and 12 that mission will probably be more clearly defined after the 13 reorganization.
() 14 MR. WARD: Could you describe in just a couple of 15 words the broad brush of the Research Office reorganization?
16 DR. SHOTKIN: Yes. Let me give you the constraints.
17 Let me give you whatever information that you need to do your 18 job, if I can put it that way.
19 MR. WARD: How do you know that?
20 (Laughter.)
21 DR. SHOTKIN: Well, based c4n what you asked me, I l
22 guess, 23 MR. WARD: Oh, okay, i
24 DR. SHOTKIN: What I cannot give and this 25 is --
l Heritage Reporting Corporation l (202) 628-4888
252
(- 1 MR. HARD: 'Well, some of it is confidential. I I 2 understand that. Just tell us what you can.
3 DR. SHOTKIN: But let me define what it is. It has 4 to do with the individual staffing plan, because that is 5 individuals. The union, we have union that does represent f 6 these individuals, and the union is supposed to see the 7 individual staffing plan before anyone else knows about it.
8 And the staffing plan is still being worked on. -The union has 9 not seen it, so I cannot discuss the individual staffing plan.
10 I cannot name any individuals.
11 What I can tell you is that the broad brush, and 12 actually I did think that you might ask the question, so I did 13 prepare sort of an answer --
14 MR. WARD: You were not going to volunteer it though?
15 DR. SHOTKIN: What?
16 MR. WARD: You were not going to volunteer it.
17 DR. SHOTKIN: Well, I did volunteer the effect on our 10 plans. i 19 The broad brush of the organization, without getting 20 into names of divisions, or names of directors, or branch ,
21 chiefs, or anything liko that, is to put more emphasis in the 3 1
, 22 Office of Research on severe accident type research. l 1
23 And how it affects my own area is to get more 24 involved in accident management. There will be continuing 1
25 emphasis on severe accident research as traditionally -
, I L
($) Reporting Corporation ;
Heritage (202) 628-4888
v 253
) 1 understood, primarily on containment and on in vessel research.
2 Most of the office remains unchanged and unaffected 3 by_the reorganization. There are just a few branches and a few 4 divisions that are affected.
5 Another objective is to more clearly define the role 6 of the two deputy directors, one a deputy director on generic 7 issues and another deputy director on research. So the 8 structure has been realigned, so that they have a much clearer 9 definition of who they interact with. That is it.
10 Is there anything else?
~
11 MR. WARD: No, that is fine. Thank you very much.
12 Are there any questions or comments before we go on 13 to the next section of the agenda?
O 14 (No response.)
15 MR. WARD: Okay. Lou, it is you again. The final 2 16 version of the rule revision.
T
- 17 (Continued on next page.)
18 19 20 21
! 22 23 24 I
25 lieritage Reporting Corporation (202) 628-4888
p.
4 254 j } 1 MR. SHOTKIN: I would like to acknowledge the large 2 effort that was put into this by Harry Togmasian. Travel 3 restrictions prevented his being here. I would like to-also 4 acknowledge the large efforts put in previously by many people 5 but, primarily, people who aren't here: Bill Beckner and Jose 6 Rias.
7 With that as an introduction, I think we are finished 8 on this job. That's why I'm giving out the credits is we 9 believe our job on the ECCS rule is just about finished after 10 low these many years.
4 11 Now, there will be some dotting of "l's" and dotting 12 of "t's" but we think we have got all the right people on board 13 and I would like to just give an introduction and Norm Lauben
() 14 who has really taken hold of this in the last few months will 15 go through with all of the details of where we stand.
16 First, let me go through on several areas that are of i 17 a more schedule type thing that I think you would be interested 18 in. First, the rule package, itself: The rule package has 1
19 been signed by the RES Office Director. That was last week.
- 20 And the Deputy Director and Division Director. So, we have got
, 21 research finished.
l-22 Key NRR staff were also briefed, although, you know, 23 we have been briefing these people all along for quite a while,
, 24 but the final briefing, again, was held last week. And this 25 was prodded by our coming to this subcommittee meeting and i
1 l
l Heritage Reporting Corporation j (202) 628-4888 L__
255 k_m) I wanting to tell you that we're finished. And we don't want to 2 have to -- we don't feel we have to come back to the 3 subcommittee again. And we do want to go to the full committee 4 next month.
5 Draft comments have been -- now, where does it go 6 next? It goes to NRR. We brief their key staff, so we don't 7 expect major problems. It goes to ARM, which is Administrative 8 Resource Management. That's the rules and records. They have 9 already seen the draft. We have got their comments back and we 10 have incorporated them. So, we don't expect a problem there.
11 OGC, this is our legal people. There are some key 12 people. We have briefed them. That could be the hooker, but 13 they will more be interested in reporting requirements and we O 14 think we can handle that, if they do want to change it. We are 15 scheduling to issue the final rule in August of '88.
16 How, continuing the update on the status. The rule, 17 itself, you know went out to comments. We received the 18 comments. As a result of that and internal discussions, the 19 rule, itself, remains unchanged after the comment period. And 20 this ACRS Subcommittee and full committee have seen that rule.
21 The regulatory guide -- oh, here is my favorite. I 22 happen to have just personally got very interested in this on 23 what's confidence level and what statistics are we using. What 24 we have done is put explicit incorporation of language about 25 the non-use of statistical confidence -- and I am still using O Heritage Reporting Corporation (202) 628-4888 6
.k 256 3
i () 1 the wrong words. Probably "level" is the wrong word. It 2 should be "integral" or "limit" although level probably covers 3 everything.
'4 And let me tell you why we did that'and why it has 5 just been of some personal interest to me. I don't claim to be L 6 an expert, although my problem is I think I am an expert on it, 7 now, but I don't think I could really defend it.
8 There were commenta on the last ACRS letter on our 4
9 use of confidence or non-use of confidence integrals. We have 10 seen some recent articles in the literature --
11 DR. KERR I think one should be careful because it 12 is true that those comments were part of a letter, but they 13 were added remarks, the views of one person, I believe.
I.
14 MR. SHOTKIN: That's correct. But I look at that as 15 something that we didn't do right if there is still confusion.
16 DR. KERR I am not suggesting that these comments i
17 are good or bad. I am simply saying they did not in that i
18 letter represent the views of the committee. They represented l
l 19 the views of one person.
20 MR. SHOTKIN: My reason for bringing this up is that 21 I don't think we did a good enough job of explaining what we 1 22 meant in the literature that we put out because there's been a 23 good deal of confusion. And I take the confusion as being more 24 our fault than anyone else's.
I 25 MR. WARD: You mean from the pablic comments there O Heritage Reporting Corporation (202) 628-4888 l
l
257
(~)N, 1 has been confusion.
u.
2 MR. SHOTKIN: No. From -- well, I'll give you three 3 specific areas. One, the comment -- the attachment to the ACRS 4 letter, another recent article that we've seen in transaction 5 of ANS and then our own -- wi?ntn our own organization, our NRC 6 science advisor, have all misinterpreted or misquoted what is 7 in the rule package. And that's just because we haven't made 8 it clear enough.
9 Now, what we are trying to do is make it clear that 10 we are not requiring -- and here's the clear statement: We are 11 not requiring the use of a statistical confidence integral. I 12 would love to talk about it, but let me pass by if there are no 13 questions.
() 14 By the way, I use -- it's statistical confidence. It 15 doesn't mean that we don't have confidence in what we are 16 coming up with. It's the statistical definition of confidence.
17 So, we have tried to put in language that makes that clear 18 within the regulatory guide. The rule remains unchanged.
19 We have merged the appendices that were in the 20 regulatory guide that you have seen and that took quite a bit 21 of work. And there are some other minor changes and Norm 22 Lauben will go over what they were.
23 The ECCS compendium, that was a huge document. We 24 have been trying to get comments and I guess it's too big of a 25 document and nobody really wants to comment on it. But all the O' Heritage Reporting Corporation (202) 628-4888
258
( ') 1 comments that we have received, which really weren't that many, 2 we are incorporating now into a final version.
3 Now, you did review the CSAU, which was the Chapter 4 4. But rather than change the compendium, that Chapter 4 will 5 'be incorporated, or any changes to that will be in the final 6 report of Novak's TPG, which is going to come out after 7 August of '88.
8 Here are some items of interest for the ACRS, this 9 subcommittee. The ACRS brought up the need for grandfathering 10 the existing ECCS rule. The comments that we received were 11 overwhelmingly in favor of grandfathering it and that Js what 12 we are going ahead with. And we wil'. be glad to discuss the 13 reasons why.
) 14 Since the last time, some questions came up about the 15 2200 degree Fahrenheit limit, which -- and we are very lucky.to 16 have Norm Lauben involved in this because he was involved in 17 defining the initial limit. And he will be able to tell you 18 what's been going on and why we feel that the 2200 Fahrenheit 19 limit is still good.
20 One of the commentors came up with an idea that: Why 21 hadn't we included concurrent steam generator tube rupture with 1
22 LOCA. We did define that as a generic issue. It has a number:
23 GI-141. It is being prioritized. It is coming out very low or 24 dropped as a generic issue, as we suspected. But we have gone 25 t . rough the formal process of doing that, f Heritage Reporting Corporation l (202) 628-4888 l
259 1 And, finally, the ACRS had comments primarily on the 2 CSAU. This is -- Novak will talk about that later. The code 3 scaleability, applicability and uncertainty, they -- you 4 reserved judgment until we could demonstrate the reflood peak 5 which will now be demonstrated probably around August of '88.
6 And we are'not tying the schedule of the CSAU demonstration to 7 the final rule. In other words, we are going to come out with 8 a rule for industry use in..npendently of whether NRC can 9 demonstrate that they understand what they are doing.
10 MR. WARD: That's kind of a -- where does that. leave 11 you? I don't I:now if you said that flippantly or what.
12 MR. SHOTKIN: No, I didn't.
13 MR. WARD: You must have some confidence -- I mean 14 the rule in its present form depends on there being some 15 rational way of making the sort of definition that the CSAU 16 attempts to make. If you, indeed, conclude from the 17 demonstration on the reflood peak that it can't be done, that i 18 is going to leave you in sort of bad shape for applying the 19 rule. So, you must have some, you must be far enough along to 20 have gotten indications that it is going to work.
21 MR. SHOTKIN: In fact, I talked with the people here 22 yesterday. Yesterday or Monday. I came here early and just 23 sat down and talked about where we stand. And I am confident 24 that what Novak and his grcup are doing is really top rate, top 25 notch. Easily understandable that we can -- we will even come O Heritage Reporting Corporation (202) 620-4888
260
() 1 up with confidence integrals. We can come up with probability 2 estimates. What I don't know is how large they are going to 3 be. What is the actual probability limit and what-is the 4 temperature. That hasn't come out, yet. But I am personally 5 willing to defend what they are doing and how they are doing it 6 to industry. And I think they have got some very ingenuous 7 ideas of how to come up with response surfaces and to generate 8 the statistics that are needed that industry may very well want 9 to follow, although we are not forcing industry to follow it.
10 But, again, you wanted to see the numerical 11 demonstration, and that won't come out until probably August.
12 But I think everything else is in place.
13 These are just some final items. CRGR: As part of
) 14 our rule package, we are requesting to them that they don't 15 have to review this further. That they have seen it before.
16 They have okayed it and we haven't made any major changes to it 17 that would affect their review.
18 The ACRS full committee, we hope that we are still on 19 your schedule for next month with the final package, even 20 though the CSAU numbers won't be demonstrated. And as far as 21 the Commission goes, we had our meeting with our management.
22 If you want to know who that is, that would be our office 23 director, deputy director, division director and some other 24 people.
25 They want to check with the Commission to see if we Heritage Reporting Corporation (202) 628-4888
. m 261 p) (- 'l even have to go back to them since they have reviewed this 2 before and they have okayed it. So, we are trying to grease 3 the wheels to get this finished and out by August.
4 MR. WARD: Okay. Let's see. The grandfathering 5 thing. Is Norm going to talk about that or is the time to 6 bring that up? -
7 MR. WILSON: I'll talk about it.
8- Md. SCHROCK: Lou, would you reiterate when do you 9 think you will publish the rule? I mean was that August?
10 MR. SHOTKIN: Yes.
11 MR. SCHROCK: You mentioned an ANS transaction 12 reference with respect to statistical analysis.- Could you tell.
13 me what that is?
MR. SHOTKIN: Can I tell you privately what that is?
14 15 MR. SCHROCK: Sure.
16 MR. WARD: Thank you, Lou.
17 Norm Lauben will be the next speaker.
18 MR. LAUBEN: I am Norm Lauben from the Reactor Plant 19 Systems Branch, Division of Reactor and Plant Systems, Office 20 of Research.
21 (Slides shown.)
22 I guess the one slide that I didn't have here is the 23 slide of what I am going to tell you. And I am going to 24 include all the things that are in the agenda, but I am going 25 to restructure it a little bit differently.
%)
Heritage Reporting Corporation (202) 628-4888
262
'(> 'l The first thing I want to do is go through a 2 background of some history and how we got where we were. The 3 second thing would be reiterate what Lou spoke about in terms 4 of the status of the rule, and then go through the package 5 which I hope everybody either got a version of, which was last 6 week's version or the final version that went out of our office 7 that I gave to Paul last night.
8 And we will visit, as we go through that package, we 9 will visit each of those sections which are of most interest in 10 more detail. And by that process, we will cover all the things 11 that you requested in the agenda, 12 The first think I want to do is go through a little 13 ancient history which is all somewhat nostalgic for me because O 14 really it's been about 19 years since we've been working on 15 this rule or working or ECCS rules.
16 The first thing was that way back in about 1970-1971, 17 there was concern over the ability of ECCS to protect the 18 reactor during a large break LOCA peak after -- well, the 19 interest peak after a test in the semiscale resulted in all the 20 injected ECC liquid from being expelled from the vessel.
21 Well, I guess at that time we didn't know why; but, 22 quite obviously, if you just have one loop in your experimental 23 facility and you are putting in the water in exactly the place l 24 where the steam has to be expelled, I don't think there is any 25 question about what's going to happen now. Since we are 19 I
O Heritage Reporting Corporation (202) 628-4888
263 N-) 1 years' smarter, we know that that is what would happen.
2 And, as a matter of fact, I think it is only fitting 3 that we'now have the recent UPTF data that shows that just 4 about all the water goes down into the vessel and none of it is 5 ejected in comparison to what we learned from those initial 6 experiments.
7 But because that did get a lot of public attention, 8 we very hastily published the interim acceptance criteria in 9 1971 which specified a few things that boiling water and 10 pressurized water reactor analysis had to contain. But there 11 was very, very.little technical basis for that rule, for that 12 interim rule. And that caused the Commission to decide that we 13 should nave a rulemaking hearing which was extremely lengthy O 14 and resulted in -- I think it actually lasted from December of 15 '71 well into 1973. So, on an on-and-off basis, it was very 16 long.
17 And the final acceptance criteria was issued in 18 December '73 and became in force in December of '74 which was 19 the now famous or infamous 10 CFR 50.46 and Appendix K.
20 This next slide says, "Present rule," which means the 21 current rule, the Appendix K rule. Not the rule we are talking 22 about passing, but the current rule that is in force right now.
23 10CFR 50.46 requires calculation of ECCS performance 24 for a spectrum of break sizes using the methods of Appendix K 25 and Appendix A, the GDC, require consideration of a single O Heritage Reporting Corporation (202) 628-4888 1
264
- ('T
(_/- 1 failure, while Appendix A also requires consideration of loss 2 of either onsite or offsite power when you do the analysis.
3 The criterion 50.46 the peak cladding temperature
.4 must be less than 2200 degrees Fahrenheit. The local oxidation 5 cannot exceed 17 percent as a cladding thickness, equivalent
~6 cladding thickness. Hydrogen generated cannot exceed one 7 percent of the total possible hydrogen generated from all the 8 reaction of the cladding. The geometry must be coolable. We 9 have interpreted that to mean you must be able to calculate its 10 coolability. And there is also a provision that long-term 11 cooling must be established and demonstrated.
12 Appendix K contains some required features. In 13 particular the 1.2 times -- I think that should be 1971 ANS O 14 decay heat standard and also acceptable features such as most 15 heat transfer correlations and things like that that are 16 mentioned.
17 Appendix K requires documentation of evaluation 18 models of sufficient detail as to require an amendment for 19 significant changes, defined as a change that would result in a 20 20 degree change in peak cladding temperature. And change 21 includes the entire evaluation model, not just the code, but 22 the input and any process variables that go into the input.
23 And if it is a hand calculation, it means the hand calculation, 24 too.
25 This is how tradition has caused us to apply the rule O Heritage Reporting Corporation (202) 628-4888
265 I) 1 in review of models over a period of years. One thing is'that 2 there really are three sources of conservatism that exist in 3 any.of the models that are being used today. One is that there 4 are those things in Appendix K which are specifically required, 5 such as -the 1.2 times ANS decay heat.
6 Secondly, there are other. sort of staff traditions 7 which we always, you know, not having a method for evaluating 8 uncertainty, we are always looking for the conservative way 9 that a particular model could be applied. So, in very many 10 cases, we required correlations that either bounded all the 11 data or, traditionally, maybe 95 percent of the data with one 12 of the other standards that we used a lot. So, that became 13 just the way that we did business over the years.
14 And, also, then vendors, because the models were very 15 simplified models, very often, especially early on when there 16 was very little data to develop models and the computer 17 capability wasn't as great, it was much, much easier to put in 18 conservatisms or conservative models which were simpler than to 19 try to work out the details which, in a lot of cases, would be 20 very hard to do. So, those are the three sources of 21 conservatism.
22 Well, because of those three sources of conservatism, 23 most licensees would end up calculating peak cladding 24 temperatures very near 2200 degrees and some plants now have 25 restricted power peaking factors. A few of them are even O Heritage Reporting Corporation (202) 628-4888
266 g)
( 1 derated a little bit to meet the 2200 degree limit.
2 Also, during this time fuel improvements and research 3 has resulted in reduction in other limits,-DNB limits which are 4 not associated with LOCA, making LOCA limits the most 5 restrictive.
6 Also, we have found that pressurized thermal shock,-
7 fuel management, power maneuvering, power upgrades require 8 increased peaking factors, while the error-corrections that 9 have been made over the past 15 years and mcdel changes over 10 the years have resulted in even tighter LOCA limits.
11 The net result is that LOCA limits are limiting plant 12- operation for some vendor designs, as pointed out.
13 The next thing is the interpretation of documentation 14 requirements have resulted in a requirement to report and 15 immediately resolve any errors in evaluation models over 20 16 degrees Fahrenheit. And, in fact, there are other parts --
17 there are other rules, such as when the vice president of the 18 utility submits and says, "This is true and correct to the best 19 of my ability," has caused lawyers sometimes to say: "Any 20 error, you must come in and fix it right away, otherwise, you
!. have lied in your affidavit saying that everything is true and 21 i
22 correct."
23 Well, this has made us all jump through hoops that 24 none of us really want to do and has nothing to do with safety.
l 25 And this usually means immediate plant restrictions Heritage Reporting Corporation (202) 628-4888 l
l
267 o
d-) .I and long hours for the NRC Staff and licensee staff until a
'2 possibly trivial. problem can be solved. I'll say vendor staff, 3 too.
4 DR. KERR: That first bullet. I don't see that there 5 is anything inherently wrong with that. If something is going 6 to limit plant operation, the fact that it is LOCA is not 7 necessarily --
8 MR. LAUBEN: Well, it limits it -- okay, fine. That 9 is true. Something is going to do it. But it limits it in 10 ways that are not -- what do I want to say?
11 They are not helpful to safety or good plant 12 operation. And, in other words, if you are spending all your 13 time monitoring axial flux shakes and axial flux shakes really O 14 don't have a lot to do with safety. And it restricts you in 15 your ability to maneuver in what you would consider normal 16 ways, then it is probably too restrictive.
17 And really what we are talking about is we have to 18 have such very tight limits on peaking factors, very flat 19 cores, that makes it difficult to operate the reactor. And it 20 really doesn't have that much to do with safety.
21 MR. SHOTKIN: Norm, you are reading it, I think, not 22 in the way it was intended. When we started *.his, we were 23 asked: Why are you doing this? Is it going to benefit anyone?
24 So, we had to look at the plants and see: Is anyone going to 25 use this revised rule? And we found out that some vendor O Heritage Reporting Corporation (202) 628-4888
268
/
k) 1 designs, they probably will use it. That is the reason for 2 that statement.
3 MR. LAUBEN: Well, I will get into it a little bit 4 later about the safety benefits, because that is a hard thing 5 to look at: safety benefits.
6 But let me just say this part of it, anyway. If 7 pressurized thermal shock considerations would want you to have 8 lower fluxes at the vessel walls, then higher peak core should 9 be a safety benefit in that sense. It reduces the flux to 10 wall, reduces your ponential for thermal shock.
11 However, I weuldn't want to overplay that too much 12 because in a certain sense, all that does is delay it until a 7
13 longer time and somebody decides they can get a life extension k' 14 or something like that, instead of reaching the PTS limit in 30 15 years, well, maybe you will reach it in 40 years, but you still 16 might get there. So, you know, in a sense it is a safety 17 benefit, but it would be very hard to quantify that it would be 18 that great a safety benefit. But it is in the direction of the 19 safer operation as far as pressurized thermal shock to allow 20 more peak cores.
21 MR. WARD: Of course, it is those several maybe kind 22 of soft but still probably real safety benefits that led the f 23 ACRS to make its comment about grandfathering.
24 MR. LAUBEN: Right. Okay, and we will get into that.
l l 25 Okay, ECCS research and more recent rulemaking l
() Heritage Reporting Corporation (202) 628-4888 1
l
269 I history. ECCS research on large break LOCA dominated'the NRC l(' )
2 research from '75 to '79 and when the TMI accident happened.
3 And one major ECCS research program, 2D/3D, is still underway.
4 And, now, of course, ROSA program is also underway.
5- NRC has spent over $700'million on ECCS research and 6 the total cost is well over $1 billion, maybe a billion and a 7 half, when DOE, industry and foreign research is included.
8 And, as early as '76, it was apparent from the 9 research that was done up to that time, that Appendix K methods 10 were highly conservative and did not accurately represent 11 actual plant behavior during a LOCA.
12 (Continued on next page.)
13
() 14 15 16 17 18 l
i i 19 20 21 22 23 l 24 25 Heritage Reporting Corporation (202) 628-4888 l
270
( 1 And advance notice of public rulemaking was, in 1978, 2 proposed a multiphase rule change, and a limited Phase I was 3 followed by a more comprehensive rule at some later date.
4 'The Phase I i:ule was just to take care of some of the 5 administrative things. It wasn't to look at the technical-6 issues, but rather to look at things like reporting 7 requirements and clean up some of the things that obviously had 8 caused some difficulty in Appendix K. However, that didn't get 9 -- actually, and that was proposed in '78. I know that the 10 actual work on that began in '77, so this rulemaking activity -
11 - that makes this rulemaking activity 11 years old.
12 The TMI accident in '79 resulted in delay in the rule 13 revision, and the redirection of a lot of the experimental
% 14 work, semiscale, became more heavily involved in small breaks 15 and transients, and that part was added to the LOFT. LOFT was 16 redirected in that direction as well.
17 So a lot of things were done to look at the more 18 probable small breaks and transients that were spurred on by 19 the TMI accident.
20 In response to a request by General Electric for 21 relief from local restrictions on BWRs, the staff adopted a 22 method to make use of improved ECCS knowledge was in the 23 context of the existing rule. The staff would allow an l 24 evaluation levol that was totally best estimate except for l
25 those features specifically required by Appendix K.
x (d Heritage Reporting Corporation l (202) 628-4888 l
l L ._
t 271 L r k 1 Remember in one of my earlier slides I mentioned 2 three sources of conservatism, those which the vendors put in 3 because it was easier to -- the information wasn't there and it 4 was easier to put in simplified conservative models, plus the 5 fact that the staff would traditionally make sure that all the 6 models were conservative. And then there were those things 7 that were specifically required by Appendix K.
8 So that was -- the first step was that calculation 9 which just had those specific requirements of Appendix K in.
10 The rest of it could be all best estimate. And our feeling was 11 that would certainly meet the letter of the law. But as a 12 condition for this evaluation model approval, a second 13 calculation would be required, and that is a totally best
'- estimate calculation along with an evaluation of the 14 15 uncertainty of the best estimate calculation. And the 16 uncertainty must show that the required Appendix K conservatism j
17 was sufficient to cover the uncertainty in the best estimate i
l 18 calculation.
1 19 In other words, for instance, the most popular and 23 frequently delimited -- really, the most real, of course, is 21 the 2200-degree limit -- and that is that the calculation, the l
i 22 licensing -- the new licensing calculation, which is the 23 evaluation model, best estimate plus Appendix K -- this one l
24 right here -- would be below 2200 degrees, and that the best l 25 estimate calculation with uncertainty would be less than ample, l
O Heritage Reporting Corporation (202) 628-4888
272
(~)
(m/ 1 Now,. jumping ahead a little bit, the point is that
'2 the new rule is very much like this except you don't have to do 3 this line out. It's just that the best estimate with the 4 uncertainty has to be less than 2200.
5 That's -- so in essence 83-472 was an important step 6 towards this new rule, and the experience gained in doing this 7 is directly applicable to submittals under the new rule.
8 MR. SCHROCK: Norm --
9 MR. LAUBEN: . Beg pardon?
10 MR. SCHROCK: -- you've avoided commenting on the 11 meaning of the estimated uncertainty, and I gather from what 12 Lou said earlier that there is a controversial point here which 1
13 involves previous statements that the intent of that is that 14 it's a two sigma meaning.
15 MR. LAUBEN: I could get to -- I'm going to talk 16 about that later. This is just for the historical context right 17 now, and -- !
i 18 MR. SCHROCK: Fine, if we'll talk about it --
l l
19 MR. LAUBEN: Yes, we -- I definitely will cover it 20 later. <
21 MR. SCHROCK: Okay.
l l 22 MR. LAUBEN: I'm not sure it'll be to my 23 satisfaction, but we'll cover it later.
I i 24 MR. WARD: Norm?
25 MR. LAUBEN: Yes.
Heritage Reporting Corovration ,
(202) 628-4888
e -- -
273
)- l' MR. WARD: On the bottom line there, weren't there 2- some cases that were calculated where the -- turned out to be a 3 little embarrassing? I guess I --
4 MR. LAUBENs Okay, let -- as a matter of fact, I
- 5. think in two slides I'll get to those embarrassing cases.
6 MR. WARD: Okay 7 MR. LAUBEN: They may not be stated quite as 8 embarrassingly as they really were, but they're -- but I'll 9 point them out to you.
10 The staff used the 83-472 approach favorably, since 11 more realistic calculations are considered beneficial to safety 12 as well as beneficial to the plant. Like I say, I don't want i 13 to overstress the safety benefit, but anyway, let me just say 14 that the disadvantages are two calculations must be performed 13 and reviewed, the one with the Appendix K requirements and then 16 the one with the uncertainty. And without a basis in 17 rulemaking, the 83-472 approach is subject to legal challenge.
18 And now I'll get to that embarrassing questions here.
19 MR. WARD: Well, are ycu going to say something for -
20 - was there legal challenge, or are you going to go on --
21 MR. LAUBEN: No, no, the lawyers were concerned that 22 there would be, because, you know, I -- hey, lawyers like 23 rules, and if you have rules, your, you know, your case is 24 always a lot tighter, a lot better. So since 83-472 is kind of 25 a, you know, a fairly major departure from previous O Heritage Reporting Corporation l (202) 628-4888 t
274
(~h s ,j 1 interpretations, they were concerned that this could be subject 2 to challenge in the rule, because one of the things that.the 3 rule, the Appendix K, did say was that the degree of 4 conservatism should be roughly the same in the models.
5 So if they could jump on that and say, hey, this 6 really violates that precept, that was a concern that that 7 might be a -- something that would be subject to challenge.
8 In 1934, a comparison of best estimate results, 9 TRAC-P and TRAC-B, for typical plants from the four vendor 10 evaluation models and evaluation models with modifications to 11 remove Appendix K canservatism showed the following:
12 conservatisms in current evaluation models varied greatly but 13 was generally larger than 1000 degrees, and I -- and although 14 it doesn't say it, I think some of these comparisons were the 15 ones that caused a little bit of embarrassi,ient, because it 16 turned out that when you took -- well, that the -- that some of
, 17 the vendor models didn't necessarily show that they were more
. 18 conservative than best estimate models which is the Appendix K.
19 But anyway, we didn't really ever totally resolve 20 that question because we found out that, except in the one case l '21 where we did work very closely with the vendor, which was i
22 Westinghouse, we did understand the differences in the 23 calculations. We never really -- I don't think we ever really 24 decided whether we were comparing apples and oranges with some 25 of those, quote, "embarrassing" cases.
O Heritage Reporting Corporation (202) 628-4888
275 k./. 1 MR.' WARD: I thought there were cases where there N .2 were indications that the best estimate plus uncertainty, l3 actually gave you a higher-calculated peak clad temperature 4 then.the best estimate plus EM.' In other words, that was an 5 indication that the rule was not sufficiently conservative.
6 MR. LAUBEN: Lou, maybe your memory -- I thought it 7 was the way I'd said i.t, but --
8 MR. SHOTKIN: That was on the GE.
9 MR. LAUBEN: Well, we can go back and look.
10 In any case, it wasn't the.way we had thought it 11 should be, and I thought it was that the --
12 MR. SHOTKIN: I think one of the vendors, but it was
- 13 just -- it was the blowdown peak, but it was well less than 14 2200.
15 MR. WARD: Okay, there was a difference, but we're 16 talking about way down at some level far below the limit.
17 MR. SHOTKIN: Three or four hundred degrees below 18 2200, at least.
19 MR. WARD: Okay.
20 MR. LAUBEN: And anyway, it did show that the decay 21 heat was certainly the major conservatism in the evaluation 22 models, worth about three or four hundred degrees. And 23 removing the decay heat conservatism usually resulted in 24 shifting the time of the peak clad temperature from the reflood 25 to the blowdown phase of a large break, and I'll have to say 2
Heritage Reporting Corporation (202) 628-4888 i
276
~
1 usually because that wasn't always the case, either.
.('.)Y 2 Most Appendix K conservatism, that is, from the 3 standpoint of thermohydraulic things especially, is during 4 reflood. That is,.the prescriptions on less than one inch per 5 second on using the kinds of reflood models that were 6 predicated on carryover rate for accident, things like that.
7 And that also that there wasn't a lot of best estimate data 8 available. So most of the thermohydraulic conservatisms were 9 during reflood.
10 And once the time of the peak clad temperature is 11 shifted into the blowdown phase, the evaluation models may not l 12 have sufficient conservatism to cover the uncertainty in the 13 calculation. Thus our interest in uncertainty.
(') 14 MR. WARD: What do you mean the shifting of the time 15- of peak clad from the reflood to the blowdown?
16 MR. LAUBEN: Yes, it --
17 MR. WARD: I mean, I get that there's some trend in 18 there.
19 MR. LAUBEN: Well, yes. When you go to more and more 20 best estimate calculations like I think Novak will tell you i 21 about the CSAU work, the peaks appear to be principally l
l 22 blowdown phase.
! 23 MR. WARD: Okay. But that's in going from EM to best 24 estimate calculations?
25 MR. LAUBEN: Yes, as you remove the conservatisms.
I O lieritage Reporting Corporation (202) 628-4888 t
277
(~)
\~/- 1 MR. WARD: That's the trend you're talking about?
2 MR. LAUBEN: Yes, right.
3 MR. SCHROCK: But it is surprising to me if that's 4 the only thing that you remove that you achieve that result, 5- because you're saying that now with a lower decay power you get 6 the peak clad temperature earlier.
7 MR. LAUBEN: Yes, and that's because you're dominated 8 so much by stored energy. The peak occurs within four to eight 9 seconds. It's before any ECCS has a chance to get the --
10 MR. SCHROCK: Well, changing the decay power doesn't 11 change the stored energy contribution.
12 MR. LAUBEN: Yes, that's right. And what we're s 13 saying is when you remove the decay heat, you're going to drop 14 the reflood peak, which is very much dependent on the decay 15 heat. And it has very little effect on the blowdown.
16 MR. SCHROCK: That's not my point.
17 MR. LAUBEN: Oh.
18 MR. SCHROCK: My point is this, that if you reduce 19 the decay power, that contribution to the total power coming 20 from decay power, then you'd have a lower total energy input.
21 And you're saying that if that's the only change you make that 22 somehow the core is going to achieve a lower peak temperature 23 than it would when it had that higher power in this 24 computation.
25 MR. LAUBEN: Let me see, maybe I'm not quite O Heritage Reporting Corporation l
(202) 628-4888
278 1 . understanding you. What I'm saying is that since decay heat so 2 dominates reflood, reducing decay heat's going to lower the
'3 reflood.
4 MR. SCHROCK: I guess a better way of interpreting it 5 would be to ask do you get a quench during blowdown or don't 6 you get a quench during blowdown. If you had a quench during 7 blowdown with the excessive decay power, and now you've changed 8 the magnitude of that peak in relationship to a peak during 9 reflood, it's suspicious that you would get --
10 MR. LAUBEN: You do get a quench, but it's 11 principally.due to blowdown dynamics, the fact that you have 12 the enormous --
13 MR. SCHROCK: You get a quench in both cases with the O 14 old, abnormally high decay power, and also --
15 MR. LAUBEN: Oh, I see. Okay, what -- okay --
16 MR. SCHROCK: It's a question of which of two peaks 17 is the higher of the two, isn't it?
18 MR. LAUBEN: Sure, okay.
19 MR. SCHROCK: Okay.
20 MR. LAUBEN: And in blowdown, although it may not be 21 a quench because the quality may be very high, you get very 22 good heat transfer just the same after you go through the core 23 flow reversal and your stored energy begins to be removed. You 24 know, you go up and then down and then up again, and the down 25 is due to the fact that you're removing the stored energy just O Heritage Reporting Corporation (202) 628-4888 i
279
( 1 due to the blowdown heat transfer, heat removal due to the high 2 velocities accompanied by the blowdown, and the fact that l3 you're removing -- the stored energy is being removed. '
4 Now, whether that's a -- it's not, in the evaluation 5 models, it's -- we restricted -- that was a conservatism in 6 evaluation models -- we restricted any return to nuclear 7 boiling that you would call a quench. It's not restricted 8 under the new rule, and I don't know specifically whether you
'T 9 have a quench, a requench, during blowdown or not.
10 Do you know of any --
11 MR. ZUBER: We analyze this in several ways, but you 12 can -- you obtain a turnaround, I mean, reach a peak even 13 without a quench, because of the good cooling. Then if you 14 don't obtain a quench you start again increasing. But reaching 15 the peak can be just without a quench, just because you cannot 16 -- you remove the energy before you can conduct it to the 17 cladding. I think that's that balance, and we did quite a btt
(
(
18 of senq1tivity analysis -- quote, unquote "did" -- and this was 19 our conclusion.
20 But then if you don't obtain quench, then the 21 temperature increases again. But at first we can be governed 22 completely just by the flow of the vapor and cooling effects.
23 MR. LAUBEN: Yes, we always receive a sharp drop even i 24 in evaluation model calculations, and even when quench wasn't 25 allowed, and that was strictly due to the -- what Novak says, i
O Heritage Reporting Corporation i (202) 628-4888
280 cm k-) I the high velocity and the good cooling even without quench.
2 MR. KERR You get this both in the PWR and BWR. Is 3 this the effect you're talking about? ,
4 MR. LAUBEN: Yes. Of course, the PWR takes out a 5 little bit more stored energy, because it's really just a one-6 sided break, and it's a bit slower. So there -- a bit more 7 stored energy gets taken out there.
8 Okay. Well, based upon the above findings, simply 9 refining Appandix K based on current knowledge without further 10 consideration of uncertainty was not viewed as acceptable. In 11 other words, when we saw GE's peak clad temperatures drop like 12 a rock when they submitted their safer analysis, we decided, 13 well, it isn't just enough to accept new models. We need to O 14 consider -- further consideration of uncertainty would have to 15 be done.
16 So this, combined with a favorable experience with 17 83-472, resulted in an approach to rulemaking consistent with 18 83-472.
19 So we have another notice of proposed rulemaking 20 almost nine years exactly later from the first one, which was l 21 March of last year, which included the proposed rule, the draft
(
22 regulatory guide, and draft compendium of ECCS research. The 23 rule itself references the regulatory guide, so for that reason 24 we do need to have a regulatory guide that goes along with this 25 rule, and a draft compendium of ECCS research, and the O Heritage Reporting Corporation (202) 628-4888
281
[ 1 regulatory guide references the draft compendium frequently.
2 So the three are a compatible group. Where the --
-3 well, let's see, I'll get into the details of what the package 4 looks like and what parts have to go out, but we're scheduled 5 to -- well, I'll get to that, you know.
6 Public comment period ended for the rule on July 1st 7 of '87. Thirty-three organizations or entities commented.
8 Generally they were favorable, as Lou pointed out, which 9 resulted in no changes to the proposed rule, but some 10 modifications to the regulatory guica, and further 11 consideration of our own resulted in other changes to the 12 regulatory guide as well.
13 The current -- it is currently scheduled to be O 14 published -- the final rule is currently scheduled to be 15 published August ist.
16 MR. WARD: Norm, we need to take a break sometime 17 during your presentation, so let's take it now.
18 MR. LAUBEN: Sure.
19 MR. WARD: You've got a --
20 MR. LAUBEN: Yes, that would be fine.
21 MR. WARD: All right. Ten minutes, please.
22 (Whereupon, a brief recess was taken.)
23
. 24 1
25 I
O Heritage Reporting Corporation l (202) 628-4888 i
282 1 (Slide.)
2- MR. LAUBEN: Continuing on, this goes into'a little 3 more detail, but talks about the same sort of things-that Lou 4 mentioned.
5 The key features of the new rule are a -- it's very 6 general rule which requires realistic calculations of the 7 system performance with a consideration of uncertainty such 8 that there is a high level of probability that the criteria 9 would not be exceeded.
10 In this case the criteria that has the most 11 vulnerable in that sense is the 2200 degree criteria.
12 The criteria itself, of 2200 degrees, is unchanged.
1 13 Alternatively, the second feature of the rule is a Appendix K
() 14 features remain acceptable as an optimal method not requiring 15 an uncertain evaluation; tnat is, that we are grandfathering 16 existing evaluation models and allowing future use of Appendix 17 K.
18 In other words, the future models do not have to be 19 according to the best estimate mode. They can be Appendix K as 20 well.
21 MR. WARD: Does that mean that they can be, this 83-22 342 or whatever it is?
24 MR. WARD: Yes.
25 MR. LAUBEN: Yes. Because that was allowed under
~
' () Ht-8.tage Reporting Corporation (202) 628-4888
283 1 Appendix K.
[ [
2 MR. WARD: Yes.
3 MR. LAUBEN: That means it's allowed under --
4 -however, I'm not sure I know why -- someone brought this.up_to 5 me before, and I tried to understand what their point was.
6 Maybe it has more to do with how 83-472 things have-been 7 reviewed. But to my mind, I look at the new rule as being 8 simpler than 83-472 because you don't have to do that one 9 calculation. But other than that, it would be the same.
10 However, if someone arguably can say that things were 11 not reviewed to the same standard when they were an 83-472 12 review, then possibly'one could arguably say that something 13 under 83-472 would be -- that might be preferable to a new
() 14 rule.
15 But in principle, I wouldn't think that that would be 16 the case.
17 MR. WARD: You haven't had -- you know, you haven't 18 had the CSAU methodology as a standard for 83 --
19 MR. LAUBEN: CSAU.
20 MR. WARD: Yes.
21 MR. LAUBEN: OKay, that is true, and that is 22 certainly -- that is one thing that is of -- but, however, we 23 are still using the work that was done from the CSAU to 24 evaluate proposals that are currently under review.
25 In other words, the smart things we learned from the
! Heritage Reporting Corporation
( (202) 628-4888
. - - - ,_. _ _ _ _ . .~ .- ,
284 1 CSAU, we are using'to evaluate the things that are currently 2 'under review, and finding, I think, that in some cases there is 3 a fairly -- the path is reasonably consistent.
4 MR. WARD: Well, let me see if I understand. Let's 5 see, have there been submittals and approvals of those under
. 6 83-4727 7 MR. LAUBEN: There have been -- well, there was a GE ,
8 safer, and that was some time ago, but that was under 83-472.
9 There are currently two submittals in-house; one by 10 Westinghouse for four of the two-loop UPI plans, and one by CE 11 for the other two Westinghouse two-loop _ plans.
12 MR. WARD: I really do mean for the GE, but have 13 there been -- you said the safer, but I mean have there been !
O 14 actual licensees, plant submittals? ,
15 MR. LAUBEN: No plant submittals yet, no. ,
16 MR. WARD: Now when you evaluate, when you review the 17 uncertainty allowance that's developed under 83-472, will you 18 use the same standard for doing that that you would use for 19 reviewing the uncertainty allowance under the new rule?
20 MR. LAUBEN: I would hope that they would be
- 21 reasonably consistent. I can't say for certain, because in the 22 first place, a lot of the review under 83-472 for the
- 23 Westinghouse model is almost -- has been done, and I think you 24 are scheduled to hear about that in June, or when? I think
! 25 it's on the --
- C:)
Heritage Reporting Corporation (202) 628-4888 i
1 M
285 Ii 4J 1 MR. WARD: UPI.
2- MR. LAUBEN: UPI, yes.
3 MR. WARD: "< n that's right.
4 MR. LAUBEN: I think it's scheduled for June, yes.
5 But my judgment is that it's reasonably consistent.
6 That's my judgment.
7 MR. WARD: Why wouldn't it be exactly the same?
8 MR. LAUBEN: Because I don't think that -- because, 9 first of all, I think CSAU is done pretty independently, okay?
10 And I don't think that the people who were reviewing the 11 Westinghouse and CE UPI models had full benefit of all the CSAU 12 work at the time. 'And my feeling is that to a large degree 13 they are pretty consistent, and I felt good about that. But n
t_)
s 14 that's also one reason why a year and a half ago we wanted the 15 CSAU to go forward, because we felt it was really important 16 that the staff had some, some experience of their own in doing 17 this.
18 Whether it's complete or not right now isn't nearly s 19 as impo.rtant as the process that's being used, and the things 20 that they are coming up with that turn out to be important.
21 MR. WARD: But the method that General Electric used 22 to develop the -- and certainly the allowance they used was 23 very, very different from CSAU.
24 MR. LAUBEN: It was different. One thing they spent 25 a lot of time on with GE and I ':hink Westinghouse is too, but Heritage Reporting Corporation (202) 628-4888
. ,, ~. - . ~ . .. .
.7 B
286 ,
() 'I which CSAU isn't because it's a code uncertainty is the overall n' ~2 calculation uncertainty which includes plant input. They spent 3 a lot of time on that on GE and Westinghour And my 4 assessment of that is that that's being done fairly
.5 conservatively. ;
i 6 However, as far as the code part, I'll;have to say, t l
7 yeah, I think -- I think that what they did for safer was a 0- .little bit different, noticeably different from CSAU, yeah. ,
9 MR. SCHROCK: Isn't the answer on this question 10 whether -- ralated more to whether the vendors are using CSAU 11 .in the arguments'that they are presenting to justify the 12 uncertainties 19,at they claim in their proposal?
13 MR. LAUBEN: Yes. They are not because they haven't ;
14 becc intimataly involved with the CSAU. It's up to the f
15' staff --
[
16 MR. SCHROCK: Their rationale is not CSAU, but it's [
- - 17 something else. f IP MR. LAUBEN
- Yes.
19 MR. SCHROCK: And the staff will use CSAU to judge i
a 20 whether it's okay.
21 MR. LAUBEN: It s up to the staff to understand what i .
l 22 is going on at CSAU and understand what's going on with the F 23 vennors so that if they feel there !s something that "STU came
.- up with that really needs to be considered with what the a vendors are doing, that they will do that.
sw Herito3e Reporting Corporation (202) 628-4888
u 287
( '1' MR. CATTON: Seems to me CE submittal being in hand 2 would be a good one to test the process.
3 MR.-LAUBEN: Yes, I thinh it will.
4 MR. CATTON: And maybe on a one-to-one basis.
5 MR. LAUBEN: Yeah, I think so, yeah. We will meet 6- with them next week, I guess. .
7 OR. KERR: You mentioned in-an earlier bullet, 8 earlier side that some of the characteristics of the review 9 process under the old rule Appendix K were what you referred to 10 as traditions developed by the staff.
11 MR. LAUBEN: Yes.
12 DR. KERR: Do you expect some of the traditions to 13 develop in connection with this revised rule that may influence O 14 the results?
15 MR. LAUBEN: Well, yeah, I tl' ink so, and I think that 16 we can see that under 83-172 we already have one tradition, and 17 that is the way GE handled their plant parameter uncertainties 18 has served as a pattern.
19 You knew, admittedly what CE and Westinghouse has 20 done is not ex.ctly what GE has deae. You wouldn't expect it 21 to be exactly. But it served as a pattern, and the staff is 22 familiar with what the -- especially NRR is familiar with what 23 they did when it came to uncertainties involving plant 24 parameters. ,
25 do I think there will be traditions developed, and
[
Heritage Reporting Corporetion I
(202) 628-4888
288
() I hopefully CSAU is -- well, no doubt CSAU will be a tradition'in 2 terms f evaluating the code uncertainty.
'3 DR. KERR But it appeared to me that you were saying 4 that that particular tradition actually introduced conservatism 5 that were not part of the rule.
6 MR. LAUBEN: The old tradition.
7 DR. KERR Yes.
8 MR. LAUBEN: That's rigitt.
9 DR. KERR: And would you expect the new traditions 10 might introduce conservatism that are not part of the rule?
11 MR. LAUBEN: Possibly, because the rule is very 12 general. And if it's a lot easier -- you know, doing a 100 13 percent bona fide, purely statistical assessment is an ,
14 extremely expensive proposition.
15- MR. WARD: Impossible. (
16 DR. KERR I simply find it occasionally useful to 17 find to view things from a licensees viewpoint.
18 MR. LAUBEN: Sure.
- 19 DR. KERR
- And if a licensee comes in and does not 20 know ahead of time how his proposal is going to be reviewed as l 21 some sort of hidden agenda which is not part of the rule, it 22 seems to me this is confusing and inefficient,
, 23 MR. LAUBENs Okay. But I think there is always that 24 kind of a risk whenever anything is being tried.
25 DR. KERR: I do, too, and I that's the reason --
- (:) Heritage Reporting Corporation i
(202) 628-4888 j
'289
. ~ N.
1 MR. LAUBEN: And I don't know that there is too much 2 we can do about it.
3 DR. KERR: That's che reason that I think one should 4 have people of wisdom and good will who attenpt to prevent that 3 sort of thing.
6 MR. LAUBEN: Gvod point. That's.why we had the CSAU 7' panel and a lot of involvement of people to help us develop 8 some traditions that would be well founded scientifically and 9 would minimize, and I think it is working out that way. It is 10 minimizing any excessive conservatism that -- any excessive 11 conservatism that are coming into their process anyway.
12 DR. KERR: And, of course, traditions can also be in 13 the other direction as well if one permits that to develop.
O 14 MR. LAUBEN: Right.
15 MR. WARD: Novak, did you have something you wantea 16 to say?
17 MR. ZUBER: I agree with this discussion. I will 18 comment during my presentation.
19 What is important here from my point of view as a 20 member of NRC and as a taxpayer and technical man, I think it 21 is important to maintain uniform standards. And I think NRC 22 and the staff of NRR, once this is implemented and one of *.he 23 results .4 the PRG, their concern I think it was, and also Mr.
24 Michelson was you have to provide some guidance, how to use it, 25 how to apply it and we shall do that.
O Heritage Paporting Corporation J'62) 628-4888
i 290 j) 1 But once this is in place, we didn't intend to do 2 this when we started this work, but this came as a need and a ;
3 results of presentation to your committee and to the PRC group, 4 we should develop guidance and standards and what is 5 acceptable.
6 Once this is in place, I think it will-then rest on 7 the' staff of NRR to apply it uniformly to the industry. And I 8 think this can be done.
9 What we are trying to do is something which is 10 doable, practicable and reasonable, and yet still I mean to 11 maintain something which is defensible in a technical 12 community, and I am convinced it can be done.
13 The problem here I can see as being an ex-professor
() 14 when somebody has to compete for students, is you can obtain a 15 larger amount of students if you obtain a larger amount of ;
16 students if you decrease the requ?.rements. I mean entrance l 17 requirement or the standards of the course, which means you 18 have a lot of students going with the least resistance.
l 19 This is like inflation degrades the value of i 20 everything. Eventually the aheap money ruins the good money, i
l 21 I think this is something which this agency, I mean l
22 the Commission and committees like that should really look for, ,
j 23 but to maintain high standards. It can be done. There is no 1
24 doubt this country has the capability of doing it.
, 25 MR. LAUBENs Okay, moving on to the additional key l
L Heritage Reporting Corporation ,
(202) 628-4888 i
,,.-,,c. -
. -,..---c, -,r 4m.m -_m , , , _
_ _ , - - - , - - r--
+. .
291
,n
(_)- 1 features in the new rule.
2 _(Slide.)
3 This slide talks about reporting requirements which 4 we spent a lot of time on early on in developing this,:a lot of 5 discussions with the lawyers, and we got a lot of comments back 6 -because our experience with Appendix K, as I mentioned in 7 earlier slides, wero sometimes excruciating when it came to 8 reporting requirements and error fixes and things like that.
9 So we felt, and this is -- you know, this is the sort 10 of thing that can jerk the industry this way and that way.- So 11 we really tried to do the best job we could with respect to 12 error change in reporting requirements.
- 13 And it's really a three tiered process. And the 14 first t.hing is that -- well, in general, what we are trying to 15 do is reporting requirements are specifically stated to relax 16 and clarify current interpretations of Appendix K documentation 17 requirements.
18 Appendix K talks about 20 degree model change, and it 19 is silent about errors. When errors, like I said, some legal 20 interpretations have been if there was an error, you have got 21 to fix it right away or else your vice presidents are lying and 22 stuff like that.
23 And what the lawyer said is if you specifically state 24 ways to handle errors, then that relaxes the vice president's 25 requirement to quickly make a fix no that he is not lying O Heritage Reporting Corporation (202) 628-4888 i
=
I fr 292
-( ) 1 anymore. So that making these things specifically clear in the 2 rule is a great benefit to us, the industry, and it makes the 3 lawyers a lot happier too.
4 So the first thing was that -- the firct provision is 5 that all error corrections and changes, and we're not making 6 any distinction ba*, ... error corrections and changes, but all 7 error corrections and changes are reported annually according 8 to one of the regulations, and I can't remember the name of it 9 and I probably should have put it in there. But there are 10 reporting requirements. I think it's the one that tells you 11 to -- you submit an annual report and it resides at the plant 12 and it's subject to NRC inspection.
13 I think it's 50.4. That's the reporting requirement O 14 tnat's part of the rule.
15 The second thing is that all errors in changes or 16 cumulative absolute values of the errors resulting in greater 17 than 50 degree PCT change shall be reported within 30 days.
10 In other words, the idea is you can't come in, here's 19 one that is plus 49 and minus 36, and plus 37 and minus -- you 20 know the net may be 5, but we are saying absolute value, and 21 they have to be added together in the absolute value so that if 22 you have lots of changes, and that total amount exceeds 23 absolutely 50 degrees, you have to report that within 30 days.
24 And then no further action is required beyond the 25 reporting.
O lieritage Reporting Corporation (202) 628-4888
~
t ki 293 c( 1 DR. KERR:- I hope the rule itself doesn't use-that 4
2 language.
3 MR. LAUBEN: What's that?
4 DR. KERR Because the absolute value of the 5 cumulative error permits one to do the observe befora one 6 selects the absolute value.
7 MR. LAUBENs Okay, let's see.
8- DR. KERR And I don't think that's what you have in 9 mind from what you just said.
10 MR. LAUBEN: We can look at the rule.
11 DR. KERR Well, you don't have to now, but you ought 12 to -- if that's what you --
L 13 MR. L..UBEN : It's in your big package. Yes, okay,
(:) 14 right. The idea -- okay, yes.
,. 15 Anyway, no further action is required beyond 16 reporting. The staff may then decide to take whatever action 17 they feel is consistent with safety. That is, whether they 18 believe that this may indicate something more than what was 19 submitted or what.
20 They then have to submit -- let's see -- let me get
! 21 the rule out now that you mentioned it, because I think there 22 is something missing.
23 In case you wanted to follow along, it's Enclosure E, 24 and -- no, that's incorrect.
25 Okay, that's right. Now I thought there was O Heritage Reporting Corporation (202) 628-4888
) 1 somothing in there, but that's fine.
d-2 So then it's up to the staff then what further action 3 may be required. .
4 For-errors or changes resulting in exceeding a 5 criteria, that is, the 2200, they are immediately reportable, ,
6 it says four hours here,-but I'm not sure four hours covers all 7 the cases that are in 50.55E, 50-72 and 50.73. But we wanted 8 to make this consistent with those reporting requirement 9 sections of the rule.
10 Also, 50.4 says that you have to have a plan for- ,
11 getting back in compliance, and there is detalis about how 12 they -- they have to submit that plan, a schedule. Then the 13 staff has 60 days to approve the schedule for that plan of 14 getting back into compliance. So that takes a lot of the knee-i 15 jerk response out of --
16 DR. KERR: The licensee must make an immediate ;
17 proposal, t 10 MR. LA"BEN: Yes. 3 19 DR. KERR: Which the staff has 60 days to consider.
20 MR. LAUDEN That's correct. ,
I' 21 DR. KERR My God, that seems inconsistent.
22 MR. LAUBEN 3eg your pardon?
23 DR. KERR That seems inconsistent.
24 MR. LAUBEN: That's what 50.4 says. ,
25 DR. KERR Well, the fact that it's nonsense doesn't O Heritage Reporting Corporation (202) G28-4888 !
F
295
[')
s-1 mean you should continue to do it.
2 MR. LAUBEN To me, I don't think it's nonsense.
3 DR. KERR: Well, look, you want to (:orrect an error.
4 MR. LAUDEN: Yes.
5 DR. KERR: And you have got to do it immediately.
6 MR. LAUBEN: Yes.
7 DR. KERR: And then the staff has 60 days to sit 8 around and consider it.
9 MR. LAUBEN: No, no.
10 DR. KERR: Huh?
11 MR. LAUBEN: You don't have to do it immediately.
12 You have to propose a schedule.
13 DR. KERRt Tell us what you are doing to do q,) 14 immediately, it says.
15 MR. LAUBEN: Right.
16 DR. KERR: With no fault.
17 Well, if you feel good about that, I guess it's okay.
18 I don't feel good about that. I would --
19 MR. LAUBEN: You would rather have the staff --
20 DR. KERR: No, I would rather have you give a 21 licensee a little bit of time to think about what he is going 22 to do to correct an error rather than come in with an 23 immediately proposal.
24 It may well be wrong jf he has to do it immediately.
25 MR. LAUBENs He's just proposing a schedule.
p)
Heritage Reporting corporation (202) 628-4888
296 g
(-) 1 DR. KERR: The rule may say that.
2 MR. LAUBEN: Yes, the rule says that.
3 MR. SHOTKIN: The rule doesn't say when you start 4 counting the four hours, by the way.
5 DR. KERR: I'm not talking about the four hours. I'm 6 talking immediate steps of being in compliance must be 7 proposed.
8 Now I don't know what that means. I assume it means 9 that you say what you are going to do.
10 Now if all you have to do is come in and say, I'm 11 going to do something, that's okay. But if what you have to 12 come in with is a proposal of what you plan to do, I don't see
_ 13 the point of an immediate.
\' MR. LAUBEN:
14 If the change or error is significant, 15 the applicant or licensee should provide this report within 30 16 days, and include with the report a proposed schedcle for 17 providing a reanalysis or taking other actions as may be needed 18 to show compliance of 50.46.
19 This scheduled may be developed using an integrated 20 scheduling system provided previously approved by the facility 21 for NRC. For those facilities not using an NRC-approved 22 integrated scheduling system, a schedule will be establish <d by 23 the NRC staff within 60 days of receipt of the proposed 24 schedule.
25 I guess that's just saying what I did thou.
g)
\
w i
Heritage Reporting Corporation (202) 628-4888
297
() 1 DR. KERR: We aren't talking about the last. bullet, 2 are we?
3 MR. LAUBEN: No, that's right. Let me see.
4 MR. WA!iD: Norm, how does the discovery of an error-5 come about? I mean, what's an example of how one might -- this 6 is an engineer deep in the bowls of an office somewhere finds 7 an error in the calculation?
8 OR. KERR Well, do you remember recently, about four _
9 or five years ago they discovered an error in the calc'0ation 10 of seismic response?
11 MR. LAUBEN: You are right. Engineers will go over 12 calculations, and I think in one example one of the vendor j 13 licensees over in Europe was looking at the calculations and
) 14 discovered what he thought was an error. They discussed it 15 with the vendor here and they decided it was an error, and 16 they --
17 MR. WARD: Is that when the clock started, the four-18 hour clock started?
19 MR. LAUBEN: No. Well, let's see.
l 20 Yes, because there are immediately reportable 21 requirements in those parts of the regulations, 50.55, .72 and 22 .73, right, i
23 When he finds the error and confirms an error, then i 24 the clock is supposed to start.
25 MR. WARD: But in practice, you know, there are some i
CE) Heritage Reporting Corporation
( (202) 628-4888 I
298 (3
s_/ 1 institutional difficulties with this sort of thing which you 2 know as well as I do, probably better. And it just seens to 3 me --
4 MR. LAUBEN: That's right. In other words, in 5 practice what does happen is that it does -- the organization 6 will take some time to confirm that it's an error. You know, 7 nobody wants them to jump right off the ground and come and 8 tell us if they are not sure yet. So they do do some 9 investigation to confirm that it's an error.
10 And I think they all have organizations, management 11 organizations that they have to, safety organizations that they 12 have to go through to -- you know, it isn't just the engineer 13 calls up the NRC on the phone. They do have a safety 14 organization where they decide that it is a safety impact, a 15 regulatory impact, and then they notify the NRC.
16 In these cases, I'm not aware of people just sitting 17 on these things after they found something out without making a 18 concerted effort to determine what the error is and some i
19 initial assessment, at least, as to what its effect is, because 20 you have to make that assessment of the effect to decide 21 whether it's going to be over 2200.
22 If you find the error, tne clock doesn't start i
23 because you haven't decided what the effect is yet. So it does 24 give them some time --
25 MR. WARD: So it's a complex sort of institutional Heritage Reporting Corporation (202) 628-4888
O 299
)) I thing.
2 MR..LAUBEN: Yes.
3- MR. WARD: But it just seems to me that the rule, the 4 words in the rule-I should somehow, you-know, more gracefully 5 acknowledge the realities of this and not just'say'as if it's 5 some event that's observable in the plant. I mean that's a 7 very different-thing.
8- DR. SHOTKIN: Try and_make it clear, let's see if I 9 have got it right.
10 This is not part of the rule at all. This is --
3 11 MR. WARD: No, it's another rule.
12 DR. SHOTKIN: Yes, it's part of the regulations that
, 13 we had to refer to if the criteria is exceeded.
() 14 MR. WARD: Okay. No, I understand that. But it 15 seems to me there might be an opportunity or a need to change 16 that although --
17 MR. LAUBEN: I think, you know, certainly 50.72 and 18 .73 are not long rules, and people are thinking about them all 19 the time, because reporting is a big part of what industry has 20 to do.
21 MR. WARD: Yes. But there is a big difference in 22 reporting --
23 MR. LAUBENs Okay.
I 24 MR. WARD: -- an event that occurs in the plant.
25 MR. LAUBEN: Versus some analytical problem.
Heritage Reporting Corporation (202) 628-4888 l
300'
(
1 MR. WARD: Right.
2 MR.' LAUBEN: Yes, but like I say, I think the fact 3 that they need to assess what the effect is -- I would say 4 this. The clock starts the minute they have made that 5 assessment of what the effect is. That's when the clock 6 starts.
7 MR. WARD: But in a real working institution, it's 8 awful hard to determine when that is. And I think you 9 unnecessarily stress the conscious of --
10 MR. LAUBEN: Conscious of the --
11 MR. WARD: -- licensees and the vendors and 12 everything. I mean it's just an unfair, unrealistic situati.on P
l 13 to put in there. There needs to be some different way of 14 defining the reporting requirements for that sort of thing.
15 MR. LAUBEN: Well, that really would -- you know, I 16 am not the lawyer, but I think that would require a change in 17 the reporting requirements section as well.
10 MR. WARD: Yes.
i 19 MR. LAUBEN: Because, you know, anytime you are 20 outside your design basis, or you are violating your technical 4
l 21 specifications, that's reportable, and that depends on l
22 calculations, not just ECCS calculations.
23 DR. KERR We aren't talking altogether about 24 reporting. We're also talking about what you are going to do 25 to come into compliance.
O Heritage Reporting Corporation (202) 628-4888 a
301
(~%
() 1 MR. LAUBEN: That's right. Okay, that's in here, 2 that's correct. And that was the point you brought up earlier 3 about letting the staff sit around for 60 days to decide 4 whether the schedule is all right.
5 I guess being a staff maybe it sounds like we made it 6 easy for the staff and hard for the members in terms of 7 schedule.
8 DR. KERR: Well, no, I don't think it's unreasonable 9 for -- I don't think it's unreasonable for someone to have some 10 time to contemplato a change.
11 MR. LAUBEN Yes.
12 DR. KERR: But it seems to me it's equally reasonable 13 for the licensee to have some time to decido what he is going
(
(J 14 to do to correct something. Otherwise, he may make a serious 15 mistake.
16 I don't want to push it any further. I don't 17 understand the logic, but this is not the first time I have 18 failed to understand logic, and I don't expect it will be the 19 last.
20 MR. LAUBEN: Well, you know, the reason we put a time 21 in here was part of the problem --
22 DR. KERR: I'm sorry. The point I am making is not 23 that you put a time, but that you put zero time. To me 24 immediate is --
25 MR. LAUBCN: You don't mind the 60 days #or the NRC Heritage Reporting Corporation (202) 628-4888 i
L
b t
'302 I 1 to decide if the schedule is okay.
2 DR. KERR No, I don't think that's unreasonable, j 3 -MR. LAUBEN: But the fact that we really didn't give 4 the licensee any time to develop the schedule. :
S. L?. KEPR Right.
6 MR. LAUBEN: Except it says he has -- if the change 7 or error is significant, the applicant or licensee shall
- 2. 8 provide this within 30 days, and include with the report of 9 proposed schedule for providing -- so he doesn't have to come 10 up with a schedule instantly. He has 30 days to come up with a [
11 schedule.
12 MR. CATTON: So this last bullet -- ,
13 MR. LAUBEN: Is wrong.
2
.O 14 MR. CATTON: -- is wrong.
L 15 MR. LAUBEN: Right. He does have to report it -- t i
16 MR. WARD: He has to report it, but that's it.
l
! 17 MR. LAUBEN: Right. ;
18 MR. WARD: Sort of like an advance --
19 MR. LAUBEN: Yes, that's right. But he's got 30 days ;
t 20 to work out his schedule. I i
21 MR. WARD: Okay. Well, that helps you. l 1 ;
22 MR. LAUBEN: We give him 30 days; we've go 60.
r i
23 MR. BOEHNERT: I don't think that's right. If he's !
24 out of compliance, it says he has to take immediate steps to f 25 propose -- to go back under compliance. The last part down ;
L 1
s Heritage Reporting Corporation (202) 628-4888 s
303
/~ I there says that.
\s) 2 MR. WYLIE: I think that's in the proposed rule.
3 What's the difference? There is a difference when you talk i
4 about proposed and final.
5 MR. LAUDEN: Which section is it?
6 MR. WYLIE: Somebody just handed me one that's 7 marked. .
8 (Continued on next page.)
9 10 11 12 13 p
() 14 15 16 j 17
- 18 19 20 21 22 23 24 25 Heritage Reporting Corporation (202) 628-4888
304
(-) 1 MR. BOEHNERT: This is Section 3 (iii), the last 2 sentence down there before Section 4, 3 DR. KERR: It says, "Errors or changes that would 4 result in the calculated plant performance exceeding any of the 5 criteria would mean that the plant is not operating within the 6 requirements of the regulations and would require immediate 7 reporting and iramediate steps to bring the plant into 8 compliance."
9 MR. LAUBEN: You're right. You're right. The 10 affected applicant or licensee shall propose immediate steps.
11 DR. KERR: His proposal for immediate steps could be 12 another plan.
13 MR. LAUBEN: Well, he could say, "I'm going to do O
k/ 14 something." That's all it would require.
15 DR. KERR: Well, I hope he says something more than 16 just "something." I'm going to fix the calculations and 17 resnalyze it and this will be done in a certain amount of time.
18 MR. BOEHNERT: I guess you haven't really defined 19 "immediate steps." Is that what you're saying?
20 MR. LAUBEN: Yes. But I think that is also 21 consistent with other parts of the regulations. If you are out 22 of compliance, you are supposed to propose immediate steps.
23 Gee. I thought this was the good part.
24 (Laughter.)
25 This next one is a self-serving statement about how O
\~J Heritage Reporting Corporation (202) 628-4888
305
() 1 well we did with Appendix K. After.$700 million of research, 2 the only non-conservative feature of Appendix K found is the 3 Dougall-Rohsenow correlation. Thus, this previously acceptable 4 correlation was removed from Appendix K.
5 Evaluation models approved prior to the rule revision 6 that make usage'of Dougall-Rohsenow now would continue to be 7 acceptable as long as the current level of overall conservatism 8 in the evaluation model is not significantly changed. A 9 significant change in overall conservatism is defined as 10 changes or error corrections of which the net effect plus or 11 minus would result in a 50 degree decrease in peak cladding 12 temperature from that which would have been previously 13 calculated.
14- In other words, the next time something goes amuck 15 with somebody's model that uses Dougall-Rohsenow, he has got to 16 fix it. And that change has made the peak clad temperature go 17 down by at least 50 degrees. So that's the Sword of Damocles 18 over those folks who are using Dougall-Rohsenow.
19 I think some people say we ought to just take 20 Dougall-Rohsenow out over a period of time or quickly. But 21 then you run into the conflict of grandfathering existing 22 models. So, we are saying, "Here is a limited grandfather.
23 You grandfather as long as you are still reasonably in the same 24 order of conservatism, but the minute you find and make a 25 change or an error and your clad temperature goes down and you O Heritage Reporting Corporation (202) 528-4888
306 f3
(_/ 1 use the Dougall-Rohsenow, it is now time to reassess because we 2 know the Dougall-Rohsenow is non-conservative."
3 The regulatory guido defines the high level of 4 probability stated in the rule as a 95 percent probability. It 5 provides general features expected in best estimate 6 calculations and on certain evaluations, but does not specify 7 or prescribe any specific methods in the way that Appendix K 8 does.
9 It lists a lengthy number of acceptable best-estimate 10 models, such as the 1979 ANS/ ANSI decay heat and CATHCART-PAWEL 11 metal-water reaction above 1900 degrees.
12 DR. KERR I must have misunderstood Mr. Shotkin this s 13 morning because I would have thought what he said would mean s
14 that the 95 percent is no longer in the regulatory data.
15 MR. LAUBEN: It is in the regulatory guides. It is 16 not la the rule.
17 DR. KERR: So, it is okay to have something that is 18 not understandable in the regulatory guide, as long sa it is 19 not in the rule.
20 MR. LAUBEN: Yes, because it is a lot easier to 21 change regulatory guides as you get smarter than it is rules.
22 DR. KERR: But it is desirable to have at least one 23 non-understandable statement in each regulatory guide.
24 MR. LAUBEN: At least. You have to keep people busy.
25 But, hopefully, we have -- I'll tell you what: This O Heritage Reporting Corporation (202) 628-4888
307
() 'l probability stuf f is a real test for engineers like me. 'And I 2 am -- we talked about this with people who have a good l 3 statistical background very frequently. And we have changed the- !
4 wording in the regulatory guide and I hope it is better.
5 HHR . WARD: What is the wording? I think it may just
. 6 be your bullet here is the problem.
7 MR. SHOTKIN: Well, the misunderstanding is that a 8 probability level at 95 percent. That's mentioned everywhere.
9 Then th6re is something called a confidence integral that we ;
i 10 are not requiring. !
11 DR. KERR But a probability level implies that you !
12 have a probability distribution that can be identified and i
13 verified. And you don't have such a distribution in this >
14 situation. You have a lot of uncertaintios, but they aren't ;
15 statistical uncertainties. Many of them are modeling ;
16 uncertainties and other kinds of uncertainties. And you don't 17 really know what they are.
! 18 MR. LAUBEN: I think that -- what I would hope that -t
- 19 when Novak talks later, he will show you that we know that 20 stuff better than I think that you are seeing. At least in i 21 terms of the work that has been done at CSAU. f 22 DR. KERR I will wait and have llovak convince me.
23 MR. WARDt Well, that may be so, but still there i 24 isn't a nice clean probability distribution that is known or
! 25 knowable.
l i
(:) lieritage Reporting Corporation L (202) 628-4888 ,
i
,, ,, __. , , , _ . - - - - , . . _ - . . , . . - , . , _ _ , _ - , ~ . _ , . - -. . . ~ . . . . _. _ -, , - - - - - - . -
308 (3
(-) 1 MR. LAUBEN: And by golly, it sure would be an 2 overstatement to call it normal. That's right.
3 MR. SCHROCK: But there is likely a generally 4 proposed rule that says it is recognized that the probability 5 cannot be determined using totally rigorous mathematical 6 methods due to the complexity of the calculations. So, you are 7 not going to back away from that statement. Are you?
8 MR. LAUBEN: No, no.
9 MR. SCHROCK: So, what Dr. Kerr is saying is right.
10 MR. LAUBEN: What we attempted to do more recently in il the regulatory guide is -- I can get into it, now. Some 12 commentor said: "By golly, the NRC should provide confidence
,_ 13 limits." And we thought, "Boy, I guess maybe there is a little J 14 bit of confusion. There certainly is in my mind about what we 15 should be doing and shouldn't be doing."
16 And we talked to our statistician. He said, "If you 17 can define the probability curve, then you don't need to L
18 consider confidence limits in the kinds of methods we are 19 using." And that is the kind of language I think we reflected 20 most recently in the regulatory guide, which is Enclosure F on 21 page 40.
22 DR. KERR That is interesting because it would be my 23 view that unless you can define the probability distribution, 24 you can't specify confidence limits.
25 MR. WARD: You need to talk to another statistician.
O o Heritage Reporting Corporation (202) 628-4888
309 1 DR. KERR Go ahead, Norm. What-page was that?
2 MR. LAUBENs Page 40, actually it goes to the top of 3 page 41 which ends up: "These techniques which may. require the f
-4 use of confidence level," on the bottom of page 40, we are 5 talking about using other techniques.
6 We say, "Other techniques which account for the 7 uncertainty in a more detailed manner may be used. These i 8 techniques may require the use of confidence levels which are t
9 not required by the above approach." Which is what we define 10 in the middle of page 40.
11 DR. KERR You are going to drive me to agreeing with 12 Harold Lewis if you keep this up; but go ahead.
4 13 MR. SCHROCK: F.
14 MR. LAUBEN: Enclosure F.
15 DR. KERR: You've heard the story of the two ladies
! 16 who lived across the alley from each other in Brooklyn and they 17 never could agree on anything because they were arguing from 18 different premises.
l 19 MR. LAUBEN: Yes. Okay. Let's see. The last bullet i
! 20 here is that in some cases, data appropriate to model
- 21 assessment is provided. Before Novak had the TPG and the CSAU, l
4 22 he convened a group of experts to look at some data and quality
- 23 some of the data to be put in a regulatory guide as acceptable 1
- 24 data. So, there is -- not every provision of the regulatory 25 guide specifies or lists applicable data, but there is Heritage Reporting Corporation
! (202) 628-4888
. ..- . , ~ , . -,,-,__-,..-_.-..__.-,._.._,,,-.-._.m.o _
-o -
3 3 '
.l. fig
nb i'
+
i, y _..L 310
- y. sh N '(,) .1 reference'to.the compendium which lists lots andilots of data.
2 , MR. CATTON: Is there a reference to the CSAU effort 3 of Novak in the regulatory guide?
4 MR. LAUBEN: I don't know if the regulatory guide has
'S that in it.
6 MR. SHOTKIN: There is a reference to the compendium.
7 MR. CATTON: Is there any reference-to Novak's work?
8 MR. LAUBEN: Yes. The actual example that he is 9 going through now.
10 MR. SHOTKIN: Let me tell you the ground rules we 11 worked with. We are not allowed to put anything in the
- 12. regulatory guide that is not published. So, we can't reference 13 something to be published. So,- the only roterence to the CSAU 14 is whatever is in the compendium, which will be published by 15 the time the regulatory guide comes out.
16 MR. LAUBENs There will be a volume 2. I think we 17 mentioned this. There will be a Volume 2 of the compendium 18 which will have the CSAU stuff in it.
19 MR. SHOTKIN: I can state that with high confidence, i'
20 MR. LAUBEN: But I don't know about the probability.
21 The draft compendium, NUREG 1230, was a major summary 22 and bibliography, 1200 pages of extensive ECCS research and as 4 23 it says later, contains a description - "Will later contain
]
24 the results of NRC's method of evaluating uncertainty." That 25 .will be Volume 2. This is provided a proof of principle and a
O Heritage Reporting Corporation (202) 628-4898
e 311 1 audit tool and example of how to perform an uncertainty 2 evaluation, but_is not a requirement or an acceptable method in 3 the regulatory guide sense. That's why we can put it in -
4 Volume 2.
5 Schedule I as far as the compendium is concerned --
6 DR. KERR: Ir this compendium supposed to be a 7 complete description of reports of all NRC ECCS research?
8 MR. LAUBEN: Well, hopefully, it references just 9 about everything. And I can probably say with a high degree of 10 confidence that those things which aren't referenced are 11 probably referenced in the references. .
12 DR. KERR Would it make any sense to also put 13 together one that references research that has some 14 significance as far as this rule is concerned? Fairly. good 15 research and applicable research?- Or is it assumed that all of 4
16 this research falls into that category? ;
17 MR. LAUBEN: No. I think in some of those casec, -
}
18 like I said, the regulatory guide, itself, points out some of 19 the things that we feel are well qualified. But there is a lot i
20 more in the compendium than just those things that are 21 referenced in the regulatory guide.
22 NUREG 1230 has had wide foreign and domestic 23 distribution, however, we haven't received many comments to 24 date and we are not inclined to make major edits. AlthougP 25 after I read that and after I read some of the compendium, I l Heritage Reporting Corporation
- (202) 628-4888
l g
312 1 think I personally am going to try to make an effort to be more !
'2 comprehensive in my own review of the compendium, 3 ,
-DR. KERR How many pages? I 4 MR. LAUBEN: 1200 pages. ;
5 DR. KERR Don't you suppose the lack of comments may 6 indicate that very few people have looked at it?
7 MR. LAUBEN: That's right. That's a hell of a lot to 8 read and when you start looking at it in detail, you realite it 9 is a big, big job to do. That's right. You can take one .,
10 subject -- I've looked at it in some subjects and realized 11 that, "Hey, you could really spend an awful lot of time 12 reviewing this." It is an enormous document and a lot of 13 people did a lot of work. And to review it thoroughly would O
t-14 have taken an enormous amount of time. That's true.
15 Let's see. It would be helpful to receive comments s
16 from ACRS consultants. The alternative is that we would have l 1 17 to dig through transcripts on the CSAU and we will refer to ;
18 results in a separate report planning distribution for July l r
19 '88. Distribution of the main report with appendices to be ;
20 issued as Volume 2 in September about the CSAU. l I
21 Now, we get to the ECCS rule package, itself. As Lou
, r 22 pointed out, the letter -- there was a transmittal letter in
- 25 the ECCS packet I think that you have which is a transmittal 24 letter to NRR, OGC and ARM, who are the three major offices
- 25 that have to concur in the package.
)
Heritage Reporting Corporation (202) 628-4888 l
l
313
(_/) 1 The transmittal letter went to CRGR. There is the 2 Commission paper, itself. And I wanted to point out here that 3 the original Commission paper that we sent to you all last week 4 by Express Mail had a lot of comment summaries in it. That has 5 been deleted because that is also in enclosure D. So, we 6 decided that we didn't need to duplicate those comments in the 7 Commission paper. It just made the Commission paper ,
8 unnecessarily long.
9 Other than that, there is not too much significantly 10 different. Well, there isn't anything significantly different 11 in the Commission paper compared to the version you got last e
12 week.
13 There is the proposed notice of rulemaking which is 14 historical, since that was last year. The memorandum from 15 Joe Stello transmitting that -- transmitting the Commission 16 saying, 'Go ahead and do it." ACRS in the peer review 17 summar.les which you are all pretty familiar with. Enclosure D 18 ts a summary of public comments on the rule. Enclosure E is a 19 notice of final rulemaking which includes the rule. F is the 20 regulatory guide. G is the summary of comments on the 21 regulatory guido and the staff response to those comments. H 22 is environmental analysis which has to go along with any rule 23 change. I is a regulatory analysis which is the cost benefit 24 that has tc go along with it. And J is one which is new and 25 added to discuss the margin inherent in 2200 degree Fahrenheit O Heritage Reporting Corporation (202) 628-4888
T-Os , ,
j
/ 314-n is) 1 limits. That was a question-that_ arose in the Staff. And I
- 2 will talk about that a little bit later, too.
3 I'm going to talk about all of these, some of them i 4 very briefly. Very briefly.
S MR. SHOTKIN: Norm, could I make one point based on a-
/
6 remark on the compendium.
.7 MR. LAUBEN: Sure. i 8 MR. SHOTKIN: Gary reminded me. The compendium was !
9 worked on by many people. The final version was put together 10 here at Idaho. And there was a peer review of that by several 11 peoplo that Idaho did arrange of the compendium.
(
12 MR. WARD: A peer review?
13 MR. SHOTKIN: Yes. In other words, had people read ,
k- 14 it and comment on it that didn't have anything to do with 15 writing it. I think the impression was made that nobody has 16 looked at it. And I think we just wanted to make that point: i 17 That it has been looked at independently.
18 MR. SCHROCK: So, this is a rather liberal 19 interpretation of peer review.
20 MR. SHOTKIN: We didn't have a peer review list. This 21 was not anything that required a group to look at it. It was 22 just: Did anyone read it who didn't write it?
23 DR. KERR: What is its use expected to be?
i 24 MR. SHOTKIN: This is supposed to be a road map to 25 whoever is interested in finding out what research has been 4
l Hetitage Reporting Corporation (202) 628-4888
.o -
lii
. .[
?
315
( ). I done over the past 15 years that provides the technical basis 2 for this rulemaking. This road map is included in the 3 compendium. It includes -- I think a fairly reasonably, maybe
'4 excellent summary, in some places of the research that has been 5 done. But it certainly refers to reports where you can find 6 out more if you are interested.
7 HR. LAUBEN: I think part of this was motivated -- I
'i 8 can appreciate this very much. Denny Ross wanted this done
[ 9 because anyone who was through the first battle on the firs'.
10 rule knows that the most severe comments, criticisms of during
- 11 the first ECCS hearing was that nobody knew where the ;
12 information was. And there was no information. And how could 13 somebody who wasn't close to this know where to find anything O- 14 or become informed? So, this is to answer that criticism that 15 I was -- I know we all.got burned on a lot. And I think it is j 16 a good idea. !
17 MR. WARD: Well, I recall, you know, we had some sort 18 of a brief review a year ago. I thought it was a fairly 19 impressive documentation.
{
4 20 MR. LAUBEN: It is. Absolutely, f 21 MR. SCHROCK: In an earlier meeting here in Idahe j 22 Palls, we had some discussion about the status of this thing 23 with regard to the rule change. And I think at that time, Lou 24 was saying that the Staff viewed this as having no official l
- 25 status and I think you were getting advice from at least }
l- C:) :
- Heritage Reporting Corporation !
(202) 628-4888 l
)
f 6
l 1
3?6 tx
( 1 consultants on this committee, that maybe it should be 2 considered to have some official status.
3 MR. LAUBEN: Well, in the sense that -- I think Lou 4 also mentioned that you can't reference anything that is not 5 published. And it is a very convenient thing to reference in 6 the regulatory guide. And, so, it is. And those parts that 7 are referenced in the regulatory guide will be published in 8 Volume 1 at the same time or before. In that sense, it is 9 official.
10 This reiterates the status of the rule package as Lou 11 mentioned before. The Office of Research has concurred in the 12 package, transmitted it to NRC, OGC and ARM for concurrence.
13 The package was transmitted to CRGR for information. CRGR can O 14 decide whether they need to look at this, again. Our ,
15 suggestici. to them is that they don't, since the rule hasn't 16 changed from the last time they reviewed it.
17 There are no substantive changes to the rule. I f
18 think ARM said we should change -- I can never remember if it 19 is "must" to "shall" or "shall" to "must." But something that 20 had some tradition, I gueen. as to whether you are talking 21 about individuals or entities.
22 There are modifications to the regulatory guide which 23 we will talk about. The uncertainty methodology has been 24 reviewed by the ACRS and the external peer group chaired by 25 Neal Todreas of MIT. And I hope I am reflecting this O Heritage Reporting Corporation (202) 628-4888 n - ,, - -
'* 317 ,
1 cobrectly, the'ACRS reserves the right to final approval until [t 2 the uncertain methodology demonstration is complete. !
, i 3 -Like I say, Enclosure A and B, we will just discard
'4 those things. .They are notice of rulemaking and transmittal i
5 letter from the Commission, j 6 Enclosure C is the uncertainty methodology with the f 7 ACRS comments and then the peer review report and individual-8 panel member reports. And I wasn't going to go through 9 anything on that, just to mention what it was. Since that is l 10 also really information that is there and nothing - there-is a
11 nothing in there that the Staff did. That's just transmitting j 12 that information. !
13 MR. CATTON: Is there some reservation on the part of i 14 the peer review group with respect to capability within NRR to 15- exercise the uncertainty methodology? Are you going to respond j i
16 to that? !
17 MR. LAUBEN: No. I hadn't planned to. Novak can --
18 we can discuss -- I think maybe a better time would be when he i 19 gets up and talks about them. I i
20 Here is a summary of the public comments Generally, 21 they supported the ECCS evaluation approach. They supported i
22 grandfathering of the conservative Appendix K. That was the i
23 ACRS portion. i 24 I have another slide that I can put up about that if i 25 you would like to see a little more about that. Okay? I just t
1
() heritage Reporting Corporation j
, (202) 628-4888 l
, N 318 ,
( l' thought I would summarize it here, first. ,
i 2 Let me just go through a summary of the.public 3 comments and then I.can go back to that. Okay? The commentore !
3 t
4 did not feel that an explicit' degree of conservatism need to be- t 5 specified, which one of the Commissioner-Asselstine questions.
- 6 In other words, did we need to specify how much conservatism
- 7 there should be. How close to 2200 should it be?
8 DR. .KERR Do you know what Commissioner Asselstine j
.. 9 meant by that question? Did anybody ask him? When he said, 10 "explicit degree of conservatism."
11 MR. LAUBEN: I think there was -- it was felt that 12 DR. KERR Nobody asked him other than just that the 13 document that came that said, "Should an explicit degree of 4
O 14 conservatism be specified?" It is just a matter of curiosity 15- because I would not have known what he meant.
16 MR. LAUBEN: I'm afraid I don't remember either. At
- 17 one time, I think I did, but I don't remember now.
4 18 MR. CATTON: What are you going to do if somebody i 19 comes in and cranks it up with the 2200 was the best estimate.
j 20 MR. LAUDEN: Plus uncertainty? That's okay. They
[ 21 can do it. It's going to be a little bit harder because the l
22 closer you get to -- and we talk about this in the regulatory 23 guide. The closer you get to the limit out in the tall, the 24 more you have to justify it.
- 25 MR. CATTON
- I think the present calculations, at
(
Heritage Reporting Corporation (202) 628-4888
s 319
(~T 1 least for blowdown peak, are what? 1300, 1400 degrees?
LJ 2 MR. LAUBEN: Yes.
'3 MR. CATTON: A lot of margin.
4 MR. LAUBEN: Yes, but that also does not include 5 plant parameter uncertainty either.
6 MR. CATTON: That's true.
7 MR. LAUBEN: Another Asselstine question was the 8 comments -- felt they did not negate the Commission's 9 contention that increased fission products with power operating 10 is negligible when compared to an uncertainty fission products 11 release.
12 In other words, Asselstine was afraid that if we 13 increased power, it would increase fission products and,
() 14 therefore, that would increase risk. But nobody felt that that 15 was a concern either. And I think that --
16 DR. KERRt I'm sorry. Would or nobody felt that that 17 would be a concern?
18 MR. LAUBEN: What would be a concern. Because the 19 raaximum power increase you could get out anything like this 20 would probably be in the neighborhood of 5 percent for any 21 plant and in terms of risk, that is well within the uncertainty 22 of any risk assessment. So, I think, in general, people 23 thought that that small increase in fission product inventory 24 was noi. a risk.
25 Also, should there be a separate independent review?
0)
(_ Heritage Reporting Corporation (201) 628-4888
}
.. .. . _ .. ~ .. .. . . _
320 ,
1 That was another Asselstina question, such us by the American 2 Physical Society or some other entity. And the most commentors 3 felt that the ACRS review would be the appropriate review-4 specifically because ACRS had the most background on what was 5 going on. And, of course, we do know that as far as the CSAU. <
6 there was an independent review by the Todreas group.
7 Grandfathering Dougall-Rohsenow was generally 8 supported. There was some variability in comments in there.
9 Generally supported the use of uncertainty evaluations and 10 there were many comments on the reporting requirements.
11 Interestingly enough, some thought they should be more 12 restrictive. Some thought that they should be less, but, I
( 13 clearly the overwhelming majority supported what we had
(:) 14 suggested on reporting requirements. ,
15 Support exists, maintenance of the existing 16 acceptance criteria, that is the 2200 and 17 percent, 17 et cetera. One comment suggested broadening the scope of the 18 rule to include zirconium-based alloys. I think I'll have some 19 comment's on that later on in the regulatory guide. Maybe not.
j 20 Anyway, we felt that since there were other rules, 21 lika 50.44, that discussed -- that specifically mentioned 1 22 zircaloy, the rule mentions zircaloy specifically. It doesn't l 23 say zirconium alloys. It says, "zircaloy." And zircaloy is 24 something that has a known composition: Zircaloy 2 and I
, 25 Zircaloy 4.
4 I
O Heritage Reporting Corporation l (202) >28-4888 l
l l
321
() 1 There have been proposals for other zirconium-based
. 2 alloys, so it was thought that we should use language that is a -
3 little less restrictive than just zircaloy. But we felt that 4 since other rules mentioned zircaloy, that that would be -- ,
s 5 those rules would need to be changed, too, if somebody came in 6 and was to use a different cladding material.
7 And, in addition to that, that would. require, since 8 we can't really say what someone may propose to do in the 9 future, we thought that it would be safest to leave it this ,
10 way. ,
11 MR. WARD: Safest to use what?
~
12 MR. LAUBEN: Zircaloy. Leave zircaloy in the rule.
~
13 MR. WARD: And you would leave it up to Staff
() 14 interpretation.
i 15 MR. LAUBEN: Well, it probably would require a rule 16 change for someone to come in with a different -- [
17 MR. HARD: I wrs going to ask you about the stainless 18 steel, but leave that aside for now. i 19 MR. LAUBEN: Stainless steel doesn't fall under J
20 Appendix K or the 50.46. And there are a couple of stainless 21 steel reactors.
4 22 MR. WARD: Yes.
23 MR. LAUBEN: One or two. One operating, I think. ,
24 MR. WARD: No, there's two. Adam and Rowe, I think.
25 MR. LAUBEN: I think one of them is.
I O. Heritage Reporting Corporation (202) 628-4888
- \
322 1 MR. WARD: Okay. I don't understand..-The rule now -!
2 says zircaloy?' [
3 MR. LAUBEN: Yes. The old rule said "zircaloy". The 4 new rule says "zircaloy." We didn't change that.
5 MR. WARD: Are there some cores that use a different 6 zirconium alloy?
7 MR. LAUBEN: Not yet.
8 MR. WARD: Not yet. So, you are just going to face 9 that against all the rules when the time comes.
10 MR. LAUBEN: My suggestion was that if someone has a 11 different alloy, there are two paths they may be able to 12 follow. One is to propose a rulemaking, which would change all 13 those rules that talk about it. The other is to demonstrate 14 that their alloy is not that different from zircaloy.
15 MR. WARD: Actually, zircaloy is just a name, in a .
16 certain sense.
17 MR. LAUBENs Although it does have certain specific 18 concentrations of materials in it.
19 These are changes that we made in the Federal 20 Register notice that includes the final rule. Now, none of 21 these changes are in the rule, itself. ;
22 MR. WARD: You didn't go over all the rest of the !
23 points, j 24 MR. LAUBEN: I'm sorry. I was going to talk about the j 25 grandfathering. Show you another slide.
l
() Heritage Reporting Corporation l (202) 628-4888 i
cr 323
(~~
(- 1 MR. WARD: Yes. And you didn't go over all the 2 points on that previous slide.
3 MR. LAUBEN: I missed something on the previous 4 slide? Okay.
5 (Continued on next page.)
6 7 i 8
9 10 11 12 13 14 15 16 17 18 e
19 20 21 22 23 24 25 lieritage Reporting Corporation ,
(202) 628-4888 !
i I
e - - - ---- - - . _ . . . . . ,_ _
324 k._) 1 MR. LAUBENs Oh, I am sorry, you are right. I beg 2 your pardon. One commenter wanted to broaden the scope of the 3 rules to include STGR's consequent to LOCA.
4 DR. KERR Does consequent mean caused by or 5 following?
6 MR. LAUBEN: Caused by. And Lou mentioned, that was 7 prioritized as generic issue 141. And it is probably coming to 8 come out a drop or a low, indicating that it will not need to 9 be considered. Part of our rationalo for that if there are 10 problems with steam generator tube integrity, that they should 11 be addressed in the steam generator tube integrit.y inspection 12 program. And that the inspection program should assure that 13 steam generator tube integrity is maintained to a high O'- 14 probability during the transients and accidents. So that was 15 our position on that.
16 Oh, yes. Another commenter said that we should 17 concern ourselves with long-term decay heat removal, but that 18 is being addressed elsewhere. And nobody felt that that needed 19 to be addressed in ECCS. Besidas long-term cooling is one of 20 the criteria.
21 Another commenter said that the rule is not legal 22 because it is based solely on cost saving considerations. ThTt 23 is not quite true. Also, I thin that cost saving does not make 24 it illegal anyway.
25 That the experimental basis for the rule is q
LJ Corporation Heritage Reporting (202) 628-4888
s' s
- . 325
- 1 inadequate,-because full-scale bypass data is not available.
2 Well, it is available now. I do not know how generally 3= -available it is to the industry as a whole.
4 MR. ZUBER: It is available to all of the industry.
5 -MR.-LAUBEN: It is, okay.
6 Before I go on to the next part, I was going 7 to -- you do not have copies of these slides,-so we will make 8 sure that you get them.
1 10 rule be grandfathered indefinitely. There were 20 of the 33 11 comment letters that addressed the question. There were 18 12 commenters who recommended the grandfathering. Many cited eY.13 ting Conservatism to Appendix J.
13 Several said that it 14 would be very burdensome to those who determined that the rule 15 does not provide an economic benefit.
16 In other words, this is not a cheap process to t-17 develop a best estimate model in an uncertain methodology. So 18 that was another comment, j 19 MR. SCHROCK: It seems that the intent there was in 20 the word indefinitely, not in the word grandfathering, and it
- 21 is only grandfathering that is addressed.
l-22 MR. LAUBEN: Well, what you are saying is that there l 23 should have been some timo period specified before you have to p
24 change?
l 25 MR. SCHROCK: The question was should this be
! C:)
i Heritage Reporting Corporation (202) 628-4888 i
i
-v,------- m.,,,-w.-- my,,._ . . . _ _ , _ , ,
1 326 1 grandfathered with some time limit.
2 MR. LAUBEN: I think that is the question that they
.3 got.
'4 MR. SCHROCK: And the answer is that it ought to be 5 grandfathered.
6 MR. WARD: That is exactly right. I think that is 7 the problem. We thought that it was appropriate to 8 grandfather, but we questioned whether indefinite 9 grandfathering was appropriate.
10 DR. KERR Well, in a previous slide, it indicates 11 that somebody said or that the NRC staff said the risk was 12 neg.\igible.
13 MR. LAUBEN: Right.
14 DR. KERR Did the NRC staff say that?
15 MR. LAUBEN: In our regulatory analysis, we sty that 16 the risk --
17 DR. KERR Of grandfathering?
18 MR. LAUBENs Okay. We say that whether you choose 19 the Appendix K method or whether you choose the best estimate 20 method with uncertainty that the difference in risk is almost 21 impossible to evaluate the difference in risk.
22 DR. KERR: When you say that it is negligible, you 23 say that it is impossible to evaluate?
24 MR. LAUBEN: Well --
25 DR. KERR: There is a significant difference between O Heritage Reporting Corporation (202) 628-4888
327
[ ). 1- those two statements.
'( 2 MR. LAUBEN: If you cannot evaluate it, it must be 3 pretty low.
4 DR. KERR: Oh, oh.
5 MR. LAUBEN: Let me get out the. regulatory analysis 6 and we can read what it says, but I think that it says that it 7 is low.
~
O MR. WARD: It does not say that.
9 MR. LAUBEN: You are right. I sure do not want to 10 misspeak. The regulatory analysis is Enclosure I. The effect
~ 11 of what they are discussing here is what would be the effect of 12 changing the rule. And the effects that they are looking at
!- 13 here, the safety and risk effects on page five, i t, that you
) 14- might be able to increase the power by five percent.
l And what they are saying is that the increased risk 15 16 represented by this decrease in margin and increase in fission 17 product inventory is negligible, and falls within the
, 18 uncertainties of probabilistic risk assessment.
19 MR. WARD: Well, that is not the same question.
20 DR. KERR: That is a different question. What I am 21 looking at is your slide on P-4 that says, "Discussion of l 22 grandfathering doss not mention safety benefits af new rule and 23 states that the effect on public risk is negligible." Now I am 24 not sure what effect was referred to. I assume tnat it meant i
25 the effect of grandfathering.
Heritage Reporting Corporation (202) 628-4888
328 s
(~'/
N_, 1 MR. LAUBEN: The way that I look at grandfathering is 2 that grandfathering has the offect that you can either use a 3 new calculation or an old one.
4 DR. KERR: My question is did the staff say this, 5 that the effect of grandfathering on risk was negligible? That 6 was the impression that I got from the slido. It is the one 7 labeled changes from 12/87, your slide.
8 (Pause.)
9 MR. LAUDEN: I guess it is sentence, "As described in 10 the regulatory analysis, Enclosure I, the probability of a 11 large break is so low that the choice of best estimate versus 12 Appendix K has little effect on public risk."
13 DR. KERR So you think that it is a va.11d statement g
O 14 that the staff concluded that the risk before and after the 15 change is negligible?
16 MR. LAUBEN: Right, little offect.
17 DR. KERR Now is this based on any analysis, or just 18 on engineering judgment, did the staff do any risk analysis to 19 conclude --
20 MR. LAUBEN: What we find out with a lot of these 21 things that have to do with large breaks --
22 DR. KERR: Let me complete my question. It may not 23 be a good question. But what I am trying to find out is if 24 there was any ef. fort made to analyze risk before and after. It 25 is possible, it seems to me, that this change could increase pb Heritage Reporting Corporation (202) 628-4888
329
{) 1 risk, particularly if nobody looked at the whole thing.
2 My question is has there been any systematic 3 investigation to determine whether the risk associated with 4 plants analyzed and operated according to the new rule is more 5 or less sound or about the same as the risk of plants operated 6 under Appendix K?
7 MR. LAUBEN: No more significant investigation than 8 just to look at the risk from large break LOCA which is very 4
9 low. And any of the risk assessments that are done on large i
10 break LOCA have very little to do with any Appendix K 11 calculations. And the probability is so low that almost 12 anything that you tack on to large break LOCA risk turns out to 13 be very low, whatever it is.
hh 14 DR._KERR Somebody has looked at that systematically 15 and determined that?
16 MR. LAUBEN: Well, in the sense that there have been 4
17 lots of PRAs that have looked at the probability of large break 18 LOCA.
19 DR. KERR: I am simply asking whether anybody took 20 this new rule and systematically investigated to determine that 21 there is no significant change in risk. I think that the 22 answer is no.
23 MR. LAUBEN: No.
24 DR. KERR: Okay, f
25 MR. WARD: Do you think that the risk in large break O Heritage Reporting Corporation (202) 628-4888 l
330-1 LOCA is because of or in spite of the fact -- your observation 2 is that the current wisdom that risk from large break LOCA is 3 very, very low.
4 'MR. LAUBEN: Right.
5 MR. WARD: Is that because of or in spite of the fact 6 that the NRC has spent $700 million researching it?
7 MR. LAUBEN: Probably in spite of. It is due to the 8- fact that probability of the accident is so low, i 9 DR. KERR: It occurs to me that the only operation
. 10 with which we have had experience up to now is that under 11 Appendix K. And while we do not have enough data yet to 12 validate in any sense the predictions of PRA, at least all of 13 our experience is based on that.
14 Now we may be changing operating conditions.
15 Significantly or not, I am not sure. And we have absolutely no 16 experience on which to base the risk that may be associated
. 17 with the new operation. Nor do we have apparently any l
18 systematic reevaluation.
19 MR. LAUBEN: Well, let me see, i
20 DR. KERR: I am trying to think this through. I am
! 21 not being critical.
22 MR. LAUDEN: First of all, we know that most systems, 23 most NSSS systems, are certainly not -- if you took all of the 24 ECCS rules away, that they could only increase their power by 25 five or ten percent at most, they would try to operate under an
! !!eritage Reporting Corporation l (202) 628-4888
331 1 envelope of peaking factors that would be in most cases not 2 enormously different from the envelope that they have attempted 3 to do before they became-ECCS limited.
4 In other wordt, it does not look to me that the 5 changes that could accrue because of going to a best estimate 6 analysis would be that significant in terms of operating the 7 plant.
8 DR. KERR: We have a history in this country of 9 having learned a lot from accidents. Prior to TMI-2, we 10 thought that the large break LOCA was the thing that should be 11 studied. After TMI-2, we concluded that there small break 12 LOCAs were fairly important.
13 I do not know some of the things that we may learn 14 from the next accident. And I do not think that anybody does.
15 And I cannot help but believe that we are decreasing 16 conservatism some by this new rule.
17 It appears logical and plausible to do that from what 18 you have said. But I do not know what effect that this may 19 have on operational characteristics, or what people will do 20 with plants. And I am not confident that we know enough about 21 operating in a now regime that we can predict.
22 I was simply asking whether any group of people had t
23 sat down and tried to systematically look for things that 24 nobody has thought of that may result from the new operation 25 that is possible under this new rule, which I think will permit
)
, IIeritage Reporting Corporation (202) 628-4888
332 As 1 less conservatism tnan Appendix K arid the old rule permitted.
2 MR. LAUDEN: In the sense of higher power levels, and i
3 more peaks, and less surveillance of those power shapes. Those 4 are the only three things that I am aware that anyone has come 5 up with.
6 DR. KERR The excuse for doing this, I am told, is 7 that one is removing conservatism.
8 MR. LAUBEN: Unnecessary conservatism.
9 DR. KERR: Whether unnecessary or not, they are being 10 removed.
11 MR. LA') BEN : Yes.
12 DR. KERR: And one is thereby permitted to do certain 13 things operationally than were not permitted.
C) k- 14 MR. LAUBEN: Possibly raise power.
15 DR. KERR Whether these are good or bad, I do not 16 know. A plausible case has been made that they will not have 17 any effect on risk.
18 MR. LAUBEN Because removing some of these may have 19 some good effects and some bad effects, as the regulatory 20 analysis points out. Some of the goods things that it does is 21 that the more highly peaked core helps you with the PTS 22 considerations.
23 Another advantage is that you have less restrictive 24 testing on ycur diesel generators, because you do not have to 25 have such rapid start times. And without having to frequently O Heritage Reporting Corporation (202) 628-4888
i 333 1 test diosol generators to ensure that they have these rapid 2 start times, your diesel generator reliability is increased.
3 DR. KERR:- I am puzzled by that. Because I would 4 have thought that the principal impact there would be on 5 whether you talk about instantaneous double end pipe break.
6 And this has not_ changed that, has it?
7 MR. LAUBEN: It has not changed that. But as Lou 8 pointed out, the leak before break as applied to ECCS is out 9 for public comment now.
10 DR. KERR: But this rule change will not have much to 11 do with when you need the diesel generators?
12 MR. LAUBEN: Well, yes, it does. Because in a best 13 estimate sense, you may find that you do not need the ECCS as 14 quickly.
15 DR. KERR By a few seconds?
16 MR. LAUBEN: It is more than that.
17 DR. KERR Well, okay. By my impression was that a 18 complete blowdown has occurred and reflood has begun in 19 something like a minute or so in these situations.
20 MR. LAUBEN: I must say that I am not sure about 21 that.
22 DR. KERR: But my principal point was, and I think 23 that you have answerud it there, is that there has not been a 24 systematic look in a risk analysis sense to determine what the 25 risk change might be.
O Heritage Reporting Corporation (202) 628-4808
334 1 MR. LAUBEN: I think that part of the reason is that 2 the best that they can tell, except in increase in power where 3 you know what effect of risk that has, is that these other 4 things are within the uncertainty.
5 DR. KERR I would'just point out the number of 6 things that were discovered when the first risk analysis was 7 carried out. A number of things were discovered that nobody 8 thought of, because they had not looked at the total picture on 9 a systematic basis. And it may be that nothing would be 10 discovered by looking at this on a systematic basis, but at 11 least it has not been done.
12 MR. LAUBEN: It has been looked at, but maybe 13 systematic, I guess that I am having trouble deciding about 14 systematic. I guess systematic is sort of a subjective word, 15 too. In terms of a PRA type thing, it has been looked at. And 16 what they said is that these kinds of things usually do not 17 enter into PRAs. So we feel that the uncertainty is very 18 small.
19 MR. WARD: Maybe that is not the right systematic way 20 to look at it.
21 MR. LAUBEN: It could be.
22 MR. WARD: But there might be another systematic.
23 MR. LAUBEN: There might be another systematic way, 24 and it has not been looked at that way.
25 MR. SCHROCK: If I may comment. It seems that there O Heritage Reporting Corporation (202) 628-4888
335
- O- 1 might be another aspect of why the rule change is being made, j 2 And that is that it takes us from a situation where the basis f i
3 of the plant regulation is a calculation which is total. fiction 4 to a basis where the calculation has some reasonable i
5 engineering basis, as we understand it today. And that maybe j 6 should be discussed a little more.
7 MR. LAUBEN: Yes. -However, to a certain degree, that 0 was accommodated for the more probable small breaks as part of 9 the TMI action plan under TMI Item 2K-330 where we stress the 10 importance of vendor models which did a better estimate job of 11 calculating the more probable small break. And still within !
i 12 the context of Appendix K.
13 But most of the things that are specified in j
. 14 Appendix K are large break type specifications anyway. Decay l l 15 heat, of course, has a heavy implication in both. But we did !
J 16 emphasize either upgrading small break models during that ;
i' i 17 2K-330 review, or demonstrating the applicability of the
! 18 conservative models by comparing them to better estimate l 19 models.
20 And my feeling is that since the small breaks are the
)
- 21 more probable ones than the ones that can lead you into more 22 difficulty in terms of operator actions and stuff like that,
)
23 that it was more appropriate to look at best estimate type 24 analysis in a small break regime, and we did that.
\
25 MR. WARD: I think that you are skirting the point
!O Ileritage Reporting Corporation (202) 628-4888 i
i 336
( s 1 though. And I guess that I agree with what Virg said. I am i 2 sort of disappointed that you have not emphasized this more.
3 That it is just a healthier thing to describe accidents from a ;
4 realistic perspective than from some artificial perspective. I t
5 And that was really the thrust of the comment on 6 grandfathering.
7 And as an example related to large break LOCA, I 8 think that there are instances where pieces of equipment in a !
9 plant are adjusted, so that the calculated response of the 10 plant is optimized. In other words, minimizing heat 11 temperature, by means of an evaluation model.
12 And what I remember in particular were ice condenser 13 plants. I forget what yo. call them, secondary.
14 MR. LAUBEN: Well, the UHI accumulators.
15 MR. WARD: Yes. And they were actually adjusted 16 based on an evaluation model calculation. ,
j 17 MR. LAUBEN: Yes, that is right.
18 MR. WARD: I do not know how to show that it is 19 better that you are reducing risk by adjusting them based on a l 20 realistic calculation. But in my heart of hearts, I know that 21 is right.
22 MR. LAUDEN: I think two things. The probability of 23 a large break is low, so it probably does not make much i 24 difference. And some of the UHI plants, if they take the UHI !
25 out, and they are thinking of doing that anyway, so it would j l
Heritage Reporting Corporation l (202) 628-4888 I
337 1 not appear to me that UHI has a large benefit in safety.
2 MR. WARD: As an example, and maybo it is the only 3 example, and maybe there is not anything else. But you 4 mentioned the diesels, for example. I do not know what else t 5 there is.
6 MR. LAUBEN: I guess maybe what you and Professor 7 Kerr ato saying that- t' -- and yes, we looked at all of these 8 things. I do not t -
that anyone quantified the pluses and 9 minuses of all of t: ..e things. I think that is probably true. l l
10 I think that in the traditional way tnat we evaluate risk, that <
11 these things do not fit very comfortably into that. And they 12 are not something that someone would 100, t in a traditional 13 risk analysis and say I can put a number on this. It would not ;
14 change it very much which is what i.t would not do. !
15 MR. CATTON: What about Chapter 15, do they use the l 16 EM modsla to calculate those?
17 MR. LAUBENs Oh, yes. This is a Chapter 15 event.
i l 18 MR. CATTON: What about all of the others, are,there i 19 not a whole string of things?
20 MR. LAUBEN There are a lots of accidents in Chapter 21 15, yes.
- 22 MR. CATTON
- And they are used to develop set points i
23 and everything else for operations. [
i 24 MR. LAUBEN: Yes.
25 MR. CATTON1 It seems to me that things cannot be Heritage Reporting Corporation t
(202) 628-4888 i
r
338
'- - (%> 1 anything but better if you do it with best estimato.
2 .s MR. LAUBEN: You have a better knowledge of the 3 accident, s that what you are' driving at?
4 MR. CATTON: Well, you are going to operate'your 5 plant based on the results of your Chapter 15 calculations.
6 MR. LAUBEN: Yes.
7 MR. CATTON: And the Chapter 15 calculations are EM
+
8 calculations. If you go to best estimate calculations, should 9 you not be a lot better off?
-10 MR. LAUBEN: Well, the other Chapter 15 accidents ao 11 not have to abide by these.
12 MR. CATTON: Oh , they do not, they are not EM?
13 MR. LAUBEN: Right. They do not have to abide by the
(~)
k/ The LOCA is unique. There 14 same rules. ECCS is almost unique.
15 are conservatistas, but they are not codified in the same way 16 that Appendix K codifies the conservatisms for LOCA.
17 Okay. Back to this. One commenter was not opposed 18 to the rule, but thinks that it should be considered when 19 significant experience is to be gained. We thought that we had 20 the experience now, that we did not have to wait. But he makes 21 no recommendation as to what policy the Commission should
[
l 22 pursue in the meantime.
23 And one last commenter's views were not clear
! 24 regarding grandfathering. And as I said, the NRC staff thinks 25 that grandfathering is appropriate based on cost benefit. In
! Heritage Reporting Corporation 8
(202) 628-4888 l
lL
3'39
<~
(_) 1 other words, wo would imagine that if anyone were to -- I have 2 to say that we do not have the cost benefit in our pockets that-
^3 we ought to have. We asked'someone to'do a quick one, and 9e 4 do not have it.
5 But I would just look at the probability of a large 6 break itself. I think that'we know how it would probably come 7 out. I just opened it up again, did I not?
8 MR. WARD. If you think in that perspective, you find 9 that there is no point in having the rule at all.
10 MR. LAUBEN: Well, I do not think that is quite true.
11 MR. WARD: If you evaluated this, you may indeed find 12 that you would not have to have things like accumulators. If 13 you did this strictly on the basis of risk, you may eliminate
) 14 accumulators.
15 MR. CATTON: Even containment?
16 MR. LAUBEN: Maybe even containment.
17 MR. WARD: Well, I think that we better move along.
18 MR. LAUBENs Okay. These were the changes in the 19 Federal Register notice that has the rule. And as I said, none 20 of these are changes to the rule itself. They are only changes 21 in the statement of considerations and things like that. And 22 the discussion of grandfathering on page four does r.ot now 23 mention the safety benefits. And it states that the effect on 24 public risk is negligible, and that was in referenco to the 25 regulatory analysis that was done. Small reference is added to O lieritage Reporting Corporation (202) 628-4988
J f
'- 340 1 the compendium.
2' MR. WARD: What do yoa mean by the first' item?
3 MR. LAUBEN: That is what we were just discussing.
4 MR. WARD: No, Ao, what has been~ changed?
MR. LAUBEN: We used to talk about some safety 5
6 benefits to the new rule.
7 MR. WARD: And you are not talking about them 8 anymore?
9 MR. LAUBEN2 They are not in there anymore.
10 ,
DR. KERR: He says that they are negligible.
11 MR. LAUBEN: This was part of the problem. You get 12 yourself between a rock and a hard place. And the risk l
! 13 assessment that was done, albeit it was limited, not the one on l 4=/ 14 my right, it said that the effect on public risk was i
15 negligible, whether you use Appendix K or the new rule.
16 So on that basis, we would not want to stress the 17 safety benefits or the safety debits either way. There are 18 some advantages and disadvantages in terms of safety, and they 19 are small.
l 20 So we deleted. There was just a small mention of l
21 public risk in the original Federal Register notice. I dc not j 22- even think that you probably saw the original Federal Register 23 notice that we had proposed on this, this Federal Register 24 notice. And I am just saying that we had this one around, and 2" we have made those changas since December.
(h
%) Corporation Heritage Reporting (202) 628-4888
_ _ ~ . - -,_ ,,
t . 341-(3
's) MR. WARD: I.thi~nk that you have gotten through'this.
2 We have talkei about.it.
'3' MR. LAUBEN: ~ Yes, we talked-about those before.
4 The regulatory guide, which is Enclosure F. As Uou 5 mentioned, we merged Appendices A and.E. We structured a new 6 appendix. It does not now have'a letter designation,'eince 7 there is only one appendix. It highlights the model 8 correlation and data where applicable. In other words, we just-9 outlined it a little more clearly, I think.
10 We added in the regulatory guide on page 18, 11 Section 1.2.4, to say that we should account for physical L 12 chemical changes in the core, but i the ex-fuel materials.
13 That is if there are physical chemical changes in the control 14 rod material or something like that. For instance, control 15 rods when they'get up to 1500 or 1600 degrees, they can begin 16 to melt. And if there are changes that need to be accounted 17 for, they should be. Also, we made modifications based on the 18 comments which the next several slides will address.
19 MR. CATTON: The peer review of CSPU said that=it 20 needed to be much more prescriptive.
21 Where would the prescriptions go? It seems to me 22 that they would go in the reg guide.
23 MR. LAUDEN: Yes. That is where we talk about the 24 uncertainty now, in the reg guide.
25 MR. CATTON: They were concerned that it wns not O Heritage Reporting Corporation (202) 628-4888
i s
~'
342 k,p /
1_ prescriptive enough. Not that it was doable or undoable, but 2 that it was not prescriptive enough. If that is the case, it 3 seems to me that you have got to put it in the reg guide, if
-. 4 you are going te meet their comments or respond to their 5 comments..
6 MR. LAUBEN: Well, I think that we tried to define it 7 a little better, and I will get to that a little-bit later.
8 MR. CATTON: But you know what their concern is, it 9 is the lack of the -- one of them put it very bluntly and I 10 will not -- but that there may be a lack of sufficient talent 11 within licensing to do these sorts of things without it being 12 far more prescriptive.
'13 MR. LAUBEN: But you know, we have also had the
+. 14 experience with Appendix K that when you are very prescriptive, 15 that you also run into trouble.
16 MR. CATTON: That in why I asked earlier if you were 17 going to reference the document, and that could solve the whole 18 problem. If you had the document available, you could 19 reference it, and the prescription would then be in the 20 document, or the document could be as prescriptive as you 21 wished.
22 I do not know what the timing is for the 23 documentation, but maybe it might be soon enough. Mr. Zuber l
l
!' 24 has his hand up to tell you when it will be available.
l l 25 MR. ZUBER: With ever'rthing possible, we should O Heritage Reporting Corporation l
(202) 628-4888
p:
343
- r3 k-) 1 finish the calculation in August. And we intend by all means 2 to have a draft by September.
3 And let me address what you are bringing, and many 4 people' brought this up in the comments, and various members.
5 This came up many times. This was a concern. And what we are 6- trying to address is that we shall provide that. That is 7 important in what we address. Now whether this is going to be 8 good enough or not, I do not know. But a document should be 9 available as a draft in September. But being available as a 10 draft and being available as a final document maybe six months 11 later in a paper to review, to comment, and change this and
. 12 that.
. ,_ 13 But we intend definitely to finish the calculation in k) 14 August and finish the document in September. And the reason 15 for this delay is just in order to address the questions and 16 concerns that the PRG group had. And there will be a report 17 from our management to address this concern that we are 18- discussing today. But this produced a delay in our producing 19 the final results and calculations.
20 But we can say today that we can do them. I am 21 confident that the methodology will demonstrate it. And this 22 will provide a document that will address some of these 23 concerns. But we cannot have them before September.
24 MR. LAUBEN: Guidance does not needed to be provided 25 in the regulatory guide. We often provide guidance in other VO Heritage Reporting Corporation -
(202) 628-4888 t
m
{'
k; 4 344
,m
(,) 1 ways. But'it might be appropriate to consider modifying _the 2 regulatory guide later. I would not want to do it now.
3 MR. CATTON: Well, it looks like you cannot.
4 MR. LAUBEN: But I think that may be an appropriate S. vehicle.
6 MR. CATTON: Now there are also two QA documents 7 floating around in some kind of form that are examples of how 8 you ought to describe your code, or at least they will be when 9 modified appropriately.
10 Do you plan to reference them? They should be NUREGs 11 very soon, I would think.
12 MR. LAUBEN: No. We do talk about appropriate 13 documentation.
14 MR. CATTON: There is an example of appropriate 15 documentation that you could reference. And that document, you 16 know, was a key to this whole process. So you have a n' ice 17 example that you could reference in your reg guide.
18 MR. LAUBEN: It is possible. It is a possibility 19 that it could be included, in the future just to be added.
20 MR. CATTON: It could be added to the one that you 21 have now with just a sentence somewhere when you are talking 22 about documentation.
23 MR. LAUBEN: Yes, it could. They are not published i 24 yet, but if they are imminently going to be published.
25 DR. SHOTKIN: If we follow all of the suggestions l
l- Heritage Reporting Corporation I (202) 628-4888 l
l i
(
, . . ~ . - . - _ .
345 1 that we have heard at these meetings, it is unclear when we 2 would'have them published.
3' MR. ZUBER: It has a-NUREG number. Your document has 1 a NUREG number.
5 MR. LAUBEN: It;has a NUREG number, but it has not 6 been published.
7 MR. ZUBER: What is the status then?
8 MR. LAUBEN: We are working on the final. We are 9 working on the final editing and the final incorporation of 10 comments and stuff like that.
11 MR. ZUBER: When are you going to get it ready?
12 MR. LAUBEN: The way,that we are going, I do not 13 know.
\J 14 MR. ZUBER: It should be available in June. It 15 already has a NUREG number, and it should be available in June.
16 DR. SHOTKIN: We heard certain suggestions yesterday 17 primarily from the consultants cf new information that they l
- 18 would like to see in these QA documents.
19 MR. CATTON: In the TRAC QA document, my recollection 20 is that all it really needed was the description of the l
i 21 numerical. By the same token, as a modified reg guide, you l
22 could also talk about a revised QA document that includes that 23 chapter. And the document that is published could say that a 24 revision is planned with it. Then you are in business.
25 MR. LAUBEN: I think that we will take that under O
V Heritago Reporting Corporation (202) 628-4888
1 346 3
~J- 1 advisement.
. s 2 MR. CATTON: If you remember ~ that.was a key piece of s
3 information thet was needed.- v MR. LAUBEN: I think that there are advantages to 4
5 doing something-like-that, i
6 DR. KERR: 'You could-put in a footnote that'said if- i 7 you use freedom on information that you can obtain a document.
8 that might be very useful here.
9 (Continued on next page.)
10 ,
11 i 12
- 13 1 14 15 16 i
17 l
18 19 20 21 22 1 23 24 [
25 O Heritage Reporting Corporation
- (202) 628-4888
347
[h \/
~
1 MR. LAUBEN: NRC received nine responses. We
/
2 solicited comments on.should the regulatory guide list models,-
3 data and model evaluations procedures that the:NRC considers to 4 be acceptable for realistic calculations of ECCS performance.
5 Are the models, data, a.nd model evaluation procedures listed in
'6 Appendix A to the draft regulatory guide appropriate? Does 7- Appendix B,of the draft regulatory guide appropriately discuss 8 the features of the best-estimate codes and the estimation of 9 the overall calculational uncertainty?
10 MR. CATTON: What were the answers?
,- 11 -
MR. LAUBEN: Ah-ha, we are going to get to that.
12 DR. KERR: One answer was that the eliminate Appendix
- 13 B.
h As/ 14 MR. LAUBEN: We merged Appendix B with Appendix A, 15 and we now just call it "The Appendix".
16 DR. KERR: What other comments did you receive?
17 MR. LAUBEN: Oh, Comment one was extent the comment 18 . period. Well, in effect we did that by considering late 19 comments, but we didn't formally extend the comment period.
20 Then we got comments addressing Question 1.
21 Commentors, that's the plural. Commentors said the reg 22 guide -- one commentor said the reg guide was overly 23 -descriptive.
24 Our brief answer was that we didn't think so.
25 (Laughter.)
i O Heritage Reporting Corporation (202) 628-4888
y ';
- i' .
~,*
348'
-1 MR. LAUBEN: I am going to_be brief with some of
}
2 thesa answers.
3 Another commentor said the best estimate methodology-4 is burdensome. Our answer was, that is why Append!x K is 5 grandfathered.
6 DR. KERR: If I were you, I would not use'that as a 7 response, because it seems to me you --
8 MR. LA'JBEN : _I am being very brief here.
-9 DR. KERR: No, but the real reason is because you 10 felt that you didn't necessarily interfere with safety, because 11 the fact that something is burdensome if it really alves an 12 added safety --
13 MR. WARD: That'. one reason why it's grandfathered,
) 14 okay.
15 MR. LAUBEN: Right. I was under some bravity here.
16 I you are right.
17 NRC should do more for industry in quantifying 18 uncertainties. We said, no that's really the industry's job.
19 They wanted us to do more than just the CSAU. They wanted us 20 to do a uncertainty study for all the basic plant types, ,and I 21 say that is their job, not ours.
(
22 So we are doing it to demonstrate, and we are doing 23 it to educate ourselves so we know how to review these things.
l 24 DR. KERR: If the rule requires a plant-specific l 25 analysis --
l
() Heritage Reporting Corporation (202) 628-4888
m 349 i 's
'w/
1 MR. LAUBEN: Oh, yes, sure.
2 So we didn't think we should do all that work for 3 everybody.
4 MR. WARD: Unless you have really got something to 5 say, we can read these as you go along.
6 MR. LAUBEN: OKay. I won't even read them.
7 Oh, do you want me to just go -- well, let me see.
8 MR. WARD: If you have something you want singled 9 out.
10 MR. LAUBEN: Yes, right.
11 About this No. 7, I think we even covered No. 7 about 12 zirconium, l
7.,
13 MR. WARD: It is in here, i \
l 14 MR. LAUBEN: We originally --
15 MR. WARD: Okay, any comments.
16 MR. LAUDEN: Take a course in speed reading. Reg 17 guide insufficiently developed regarding uncertainty. Boy, 18 that just sounds like the Peer Review group, doesn't it?
19 And our response is guidance purposely general so 20 that applicability to a wide range if computer codes is 21 retained.
. 22 MR. CATTON: The Peer Review Group, I think, was 23 referring to having guidance for NRC in carrying out a review.
24 So that may be the rationale for putting it into the reg guide.
25 MR. WARD: Yes. You don't think that would be in the rm Heritage Reporting Corporation (202) 628-4888
350 l' reg guide, would you?
2 MR. CATTON:' Well, I guess not. It probably.would be 3 more internal in your organization where you have --
4 DR. KERR: -- technical position.
5 MR. WARD: Manual chapter.
6 MR. CATTON: Is that what they call that?
7 MR. WARD: No, no, no. That would be a manual 8 chapter.
9 MR. LAUBEN: I agree. And that certainly is true.
10 That is why I think the CSAU is so important. It is for our 11 benefit.
12 MR. CATTON: If it is to be beneficial for you, I l .
13 guess they are saying it has to be description. This sounds
\
(%-) 14 like something that Novak wanted to do with.
15 MR. LAUBEN: Well, no, not just Novak.
16 Okay, I think I do want to point out No. 11. I 17 discussed it briefly already when I made some Page 40 and 41 of 18 the reg guide, and that is that one commentor said that we 19 should specify confidence levels, and we tried to explain how 20 confidence levels aren't really needed, depending on the method 21 you use, and I don't know if I want to discuss that anymore.
22 DR. SHOTKIN: We thought should we get up in front 23 and try to explain it, and probably not, although what I would
- 24 like to do is just individually discuss some crib sheets with l
l 25 whoever wants to talk about the difference between confidence I
/~
(_T )
lieritage Reporting Corporation (202) 628-4888 l
L
351 I think we can discuss
~
.1 and the probability level uncertainty.
2 it individually.
3 MR. WARD: That's fine.
4 MR. SCHROCK: It seems to me though you kind of hedge -
5 the issue on this. What you say is you want a 95 percent 6 probability level, but you are unwilling to say anything about 7 how you interpret that in practical sense.
8 The common engineering approach is to assume a priori l 9 a normal distribution. In lots of cases where a normal 10 distribution is quite an exact.
l 11 MR. LAUBE2: Inexact.
12 MR. SCHROCK: Inexact. And it's widely done. And so 13 you have a relationship between your statement of 95 percent 14 probability level and what is otherwise frequently stated as 15 the one sigma standard deviation, and that does have a 16 confidence significance.
17 It means that 95 percent of the errors are contained 18 within the two sigma spread.
19 It seems to me that's not all that difficult to 20 explain, and I think you are letting your statisticians get you 21 hung up on language which shouldn't hang you up. It seems to 22 me that you want to explain it in a sense that the more 23 practically inclined people can underetand it instead of in 24 terms of statisticians who may be critics of what you have 25 done, pointing out that it's not pure statistics.
O Heritage Reporting Corporation (202) 628-4888
. . ~. .
352 1 You have acknowledged in your proposed rule that'it 2 -is not pure statistics. And so why make such a big fetish of 3 this?
4 MR. LAUBEN: That's fine with rae.
5 MR. WARD: Any comraittee questions on any of these?
6 (No response.)
7~ MR. WARD: -Any point you particularly want to make?
8 MR. LAUBEN: No, there is nothing on these.
9 MR. WARD: Okay, let's go ahead.
10 DR. KERR: I don't see in 17 how the applicant can 11 show that explicit modeling is not required without considering 12 it.
I _
13 MR. LAUBEN: Okay, it's simple.
5J 14 If you know that swelling and rupture doesn't occur 15 until 1600 degree Fahrenheit and your clad temperature is below +
16 it, then you didn't have to consider it, as a for instance.
17 MR. WARD: Anything else?
18 MR. LAUBEN: I just want to mention in 91, the leak 19 before break, the original reg guide said instantaneous break.
20 At one time we said, you know, you can consider break opening 21 time a further consideration. Since this is being circulated 22 for public comment in another rulemaking, we decided to be 23 silent in the current version of the reg guide so that we would 24 be adaptable to whatever the current proposed -- if they decide 25 to change it by rulemaking, we could account for it either way.
Heritage Reporting Corporation (202) 628-4888
e +
~
-353 s
(/ 'l (Pause.)
2 MR. WARD: Twonty-one looks kind of familiar doesn't 3 it?
4 MR. LAUBEN: Got that one again, huh?
5 .MR. WARD: Yes.
6 MR. LAUBEN: Okay, this'is my last slide, huh? How 7 about that.
8 Okay, in the regulatory analysis,,we deleted the 9 discussion of the 400 degree margin. In the original notice of 10 public rulemaking, we said that the 2200 degree limit and the 11 400 degree margin, and this is the margin that's partly there 12 because of the auto-catalytic nature of the water reaction i
e
-. 13 making clad temperature take off.
i 14 We deleted that because we -- and this also'now l 15 relates to the next part which is Appendix J -- not Appendix J, 16 Enclosure J where we talk about the margin inherent in 2200 17 degrees.
18 The new enclosures discusses the recent information
. 19 regarding fuel cladding and control rod behavior in the 2200 20 degree range, and we were not willing to say this inherent 400 21 degrees exists anymore, but we're still willing to say that 22 2200 is still a good limit, and that the margin above 2200 is 23 not significant to the rule as long as you are calculating to 24 the 2200 degree margin and you account for uncertainties.
25 DR. KERR: What does margin about 2200 mean?
Heritage Reporting Corporation (202) 628-4888
354 (qj 1 MR. LAUBENs Okay. It means, well, two things.
2 There were some PBF experiments that seemed to indicate that 3 for those experiments the clad temperature didn't take off 4 until you got to 2600 degrees F.
5 But what was really true is if you looked at those 6 experiments, they were really quite different from LOCA. First-
.7 of all, they were experiments in which there was virtually no 8 cooling, because they were trying to reach a severe core damage 9 condition.
10 Secondly, the time scale on the plots were so that 11 what looked like a fast temperature rise was really no worse 12 than the temperature rise you experience during reflood of the l
l 13 LOCA. So what looked to be a rapid temperature rise really
() 14 wasn't as rapid as we're used to anyway during LOCA.
15 So we concluded that there are too many 16 dissimilarities between these severe accident experiments and 17 calculations, and anything that looks like a design basisHor 18 even a best estimate evaluation of LOCA -- so we did some 19 assessment of -- so that's what that is.
20 Some additional PBF LOFT-FP2 calculations and 21 analysis came in. It showed, hey, this temperature takes off 22 below 2600. It starts to take off at 2200 in some casee. And 23 so it looked like what we thought was a 400 degree margin 24 really wasn't there.
25 But like I said, this all -- these "rapid temperature Heritage Reporting Corporation l
i (202) 628-4888
355 cm
()
_ I rises" really weren't that rapid, and they were in an 2 atmosphere of virtually no forced cooling anyway like you had
.3 during reflood. We decided these things weren't applicable, 4 even the ones that showed the margin or the ones-that didn't 5 show the margin.
6 And we went back and looked at other calculations and 7 decided that, by and large, that was an overstatement of any 8 margin for rapid temperature take off, and really rapid 9 temperature take off is probably closer to 2300 degrees.
10 And then if you look at the work that Piklesimer_did 11 that was also in the Notice of Proposed Rulemaking back last 12 year, it showed that brittlement criteria -- if you combined 13 the 2200 -- if you look at 2200 and the 17 percent, and look at
- 14 the margin to clad failure along the 17 percent line, it 15 doesn't look like its 400 degrees there. It may be more like 16 200 degrees before you would get brittlement failure. So 17 albeit, it would be at much higher to.nperatures.
18 So we decided that we had best just not talk about 19 200 degree margins because it probably isn't there.
20 DR. KERR: Since it's not there, it's not 21 significant.
l 22 MR. LAUBENs Oh, I see. It's not significant, but 23 it's a different reason. It's not significant to the rule 24 because the rule is looking at the margins of 2200, not the 25 margin above 2200.
O Heritage Reporting Corporation i (202) 628-4888
356
. 'l DR. KERR:- 'Thank you.
2 MR. WARD: Ckay, thanks, Norm..
3 I propose instead of going to Dr. Zuber's hour and a 4 half presentation that we break for lunch now, and let's come-5 back in an hour which would be 1:35.
6 (Whereupon, at 12:35 p.m., the meeting was recessed, 7 to resume at 1:35 p.m., this same day, Wednesday, April 20, 8 1988.)
9 10 11 12 13 14 15 16 i
17 18
- 19 20 l 21 22 l
23 i
24 25 Heritage Reporting Corporation (202) 628-4888 I
357 1 AFTERNOON SESS ION 2 1:40 p.m.
3 MR. WARD: Mr. Zuber, are you prepared?. Ready? I
-4 don't know if you're prepared or not, but are you ready?~
5 MR. ZUBER: My name is Novak Zuber. I am from the 6 Nuclear Regulatory Commission. And I will discuss the CSAU
, 7 methodology status. One advantage of being able to make a-8 presentation after lunch, you hope that the wolves are satiatri 9 and you won't be thrown as a sacrificial lamb or a sacrificial 10 buck in my case.
11 I will discuss the CSAU methodology and these will be 12 divided into three parts. In particular, I-will discuss the 13 results of the peer review group and industry response. Then I 14 will discuss the current status and end up discussing the 15 future schedule.
16 And my presentation on the results of the peer review 17 group will be divided into three parts. First, I will discuss 18 the response of the PRG to two questions which were raised and 19 asked from them by NRC. Then I will discuss some comments and 20 recommendations made by PRG and, finally, I will address most 21 of the concerns or the main concerns that PRG had and our 22 response to these comments. And as a result of these concerns, 23 we decided to address them, we had to get our management to 24 give us the money to do it. But there was kind of a delay. And 25 we had approximately a three-month's delay because of this.
O Heritage Reporting Corporation (202) 628-4888 i
358
. (. *1 And lt will discuss what these concerns were, what our responses 2 and'that the delays occurred.
3 But we thought that since this concerned our wallet.
4 and the review group did an excellent job, I think we should 5 address them and try to. resolve them before we go out with a 6 final document.
7 The first question that NRC asked from PRG to 8 consider and to respond was: Is the methodology systematic, 9 logical and practical?
. 10 And the answer came in two parts. You have all these 11 letters you have here. They judged the methodology to be 12 systematic and logical. On the question of practicality, they 13 were divided and the main reason was after talking to them was
()_
14 that some of them didn't think that we finished the job. At i 15 that time, we only had information on the blowdown and they i
16 didn't have any information.on the reflood. And some of them 17 thought that we should have done more calculation that we 18 presented.
19 On the other, the conclusion was considering for the 20 large break LOCA blowdown calculations, they thought that the 21 results completed to date appeared to be practical.
22 Our comments on their response is shown on this ,
23 viewgraph. We are presently applying the methodology to the 24 reflood PCT and we are really confident that this application 25 to the reflood PCT will be as practical to the reflood as it 7
\ Heritage Reporting Corporation (202) 628-4888 4
L 359 2
1 was shown to be to the blowdown. Based on the results to date, 2 I think we can make this statement. We shall prove it in a few 13 months.
4 As far as practicality in general, it will always be 5- depend on the cost-benefit basis. It will depend on who is 6 applying what is the advantage. You cannot set up a standard.
7 This will have to be decided by the applicant how he decides to 8 go. What this methodology provides, it provides a guidance on 9 what is important, what is reesonable and what is effective in 10 this methodology. I think this is a guidance which we can 21 provide to NRR when they review the applicant's submittals.
12 The second question that NRC asked from the PRG was:
13 Did we demonstrate the methodology?
14 And, again, they divided the answer in two parts.
15 The firct was addressed to the blowdown, PCT. The other on to 16 the reflood PCT. They did not identify any difficulty which 17 would preclude us to demonstrate the method is applicable and 18 further demonstration for the blowdown PCT. Again, these are 19 the quotes from the Dr. Asselstine's letter. However, since we 20 didn't have any results at that time for large break LOCA, they 21 made these statements. However, this does not assure that the 22 demonstration of the methodology for a large break LOCA reflood 23 phase and SDLOCA's are achievable.
24 Our comments to this response is shown on this 25 viewgraph. Again, we are confident that the work which in O Heritage Reporting Corporation (202) 628-4888 i
360
. n
-} Il progress now on the reflood PCT, we should assure that the
'2 practicable end should demonstrate that we can do it. There is 3 really nothing which would preclude us to reach this goal 4 witt_n a few months.
5 We are also optimistic that this methodology can be 6 applied to other cases of interest to NRC, not only to the 7 large break LOCA, but to other scenarios and in fact I will 8 discuss it if I have time and your; permission that we want to 9 apply this methodology also to severe accident initiated by the 10 large break LOCA.
11 MR. WARD: Small break LOCA.
12 MR. ZUBER: Small break LOCA. Pardon me. You are ,
13 right. So, in this sense, we would address some of the 4
f]'
\- 14 concerns which the PRG had whether we can do it for a small 15 break LOCA and the answer: We are confident we can do it.
16 MR. WARD: Let me ask. You know, in a sense, from a 17 regulatory standpoint, you probably need this methodology under 18 the new rule only for the large break LOCA.
19 MR. ZUBER: Yes. In a strict sense for the large 20 break. But I think we would like to apply those also to small 6
21 break to show how we can use these, but especially we would i
22 like to do it for the severe accident and for the accident 23 management in order to use more effectively our resources.
24 This was one of the comments I will discuss in a moment that l 25 the peer review group made, that one of the benefit of this O Heritage Reporting Corporation (202) 628-4888
361 1 methodology is to use more efficiently resources. Research 1 2 resources to organize code development and experiments. And-3 this is what we want to do.
4 But as far as this rule, I think we are only 5 addressing large break LOCA because this is what it addresses, 6 anyway.
7 MR. WARD: I guess it is a little different approach.
8 For large break LOCA, you are making an assessment so that NRR 9 will be able to judge licensee's assessments.
10 MR. ZUBER: That's correct.
11 MR. WARD: But for other transients or small break 12 LOCA, you will make sort of a one-tinae assessment.
13 MR. ZUBER: What we intend to do is that we want to Q 14 do it in two stages. First, we want to do a first pass. To 15 dig very fast to see -- we want to bring, I think I'll go back ,
16 into history. Going through this process for the last eight 17 months, it is not my conviction, but the entire TPG that we I
18 could have done a much better job, more efficient and with 19 better results had we had a unified approach to experiments, to 20 scaling, to core development and core assessment early in the 21 game.
22 MR. WARD: All along?
23 MR. ZUBER: All along. Had we had this seven years 24 ago, ten years ago, we would have been in a much better 25 position. We would not hear, for example, these kind of heated O Heritage Reporting Corporation .
(202) 628-4888 ,
s- ,
362
~
) 1 discussion or animated discussion we had yesterday because they 2 would have addressed this all along. Having had this 3- experience and we want to write the report: "Lessons. Learned."
4 'I think we would like to apply this.to the next stage, how we 5 can do it when we want to manage an accident.
6 And, again, in an accident, it cannot have a reactor 7 in a serious accident status. What we need is analyze ,
8 management options: How to do when and what? For this, we use 9 codes. Then we have to be sure what is the uncertainty in the 10 code calculations parameters of interest to the management on 11 this accident. And this is where we want to apply this 9
12 methodology. To put experiments and scaling experiments, 13 analysis, score development together. Not to do it piecemeal,
( 14 but do it in an integrated way.
15 The PRG made a very 1nteresting comment which I will 16 talk about and that is the reason.I put it in here. The 17 professor gave a good definition. I think this was alco put ,
i
- 18 back in the letter from Professor Todeas.
19 The give a definition of the methodology in the 20 following way. CSAU is a formal method for combining
- 21 quantitative analysis and expert opinion in a controlled way to i
22 minimize the subjectivity of the experts in arriving at 4 23 computed values of uncertainty.
! 24 And this is really true. What we are trying to do is 25 use the analysis with expert judgment in addressable way, a i
Heritage Reporting Corporation (202) 628-4888 i
363 1\ 1 knowledgeable way so that we can really come to some valvae 2 which we can stand behind.
3 And this is the comment of TPG: They providei y
4 better definition than we did and we shall be using it.
5 MR. SHOTKIN: Is that the whole TPG or a small 6 fraction of it?
7 MR. ZUBER: No. This is the TPG because we discussed i
9 it last ,nonth.
9 They made many recommendations, but anyway one of 10 them is here and this is again in the letter. And the 11 statement is similar to what the ACRS made in their letter.
12 "The peer revieu strongly encourages the work on the 13 development and demonstration of the methodology."
-O 14 I think this is almost identical as you did.
15 "The pursuit of such a program on comprehensive 16 utilization of research results, not only has the potential 17 benefits for operating reactors, but offers a focus for 18 planners and researchers in the conduct of future 19 investigations."
20 And I am thinking that what I would like to discuss 21 now is this is where we want to use this method for severe 22 accidents.
l 23 And another benefit identified, these reports, the QA 24 documents which came out of, as a by-product of this activity.
25 DR. KERR Are you confident that you can use it in i
i () Heritage Reporting Corporation (202) 628-4888
364
() 1 : calculating severe. accidents?
22 MR.-ZUBER: Yes.
3 DR. KERR Or you just think it is a good' idea to' 4 try?
5 MR. ZUBER: Well, I think it is a good idea to put in 6 trust. But I am confident we can do it. Tho' reason is what we 7 would like -- I will describe it in a moment -- to put 8 experiments and analysis together as we develop it step by 9 step. The problem we had we were testing for the last eight 10 months was we had done some experiments, repeated experiments 11 in a similar scale. We didn't really look at what scale we 12 should worry. We didn't look what improvements to do. You
- 13 know, I can do many improvements. It's like a wish list. -I
() 14 would like to have Cadillac, a Jaguar, a Porche. You know, you 15 have to put priorities. To have priorities, you have to get a 16 measure. And what is the payoff? Payoff function? The 17 success function?
18 And what we really would like to develop in this 19 approach is to obtain a quantifiable numbers so we can develop 20 this. We made this improvement; the payoff is this. If we 21 make this improvement, the payoff -- so that management can 22 say, "Yes, we should follow this or that."
23 But what this means You have to take experiments 24 and analysis together. If you don't do it -- we do a job 25 piecemeal, we may do 50 experiments on one scale and miss the i .
Heritage Reporting Corporation (202) 628-4888
365 (O> 1 rest. No. It can be done. Of this, I am confident.
=2 MR. WARD: Novak, what about applying this general 3 approach to other things? I mean I have seen -- we always 4 complain about the uncertainties and probablistic risk 5 estimates.
6 MR. ZUBER: I think we should deal with that 7 directly. You see, you have to say what -- on what basis this 8 probablistic risk analysis has been made? What information 9 they got, you see. Was it biased? Was it representative or 10 not. And what we would really like to do, this would be like a i
11 front-end. This is what is important. Now, determine the risk 12 of that.
13 MR. WARD: Are you familiar with -- what was it 14 called? The Quest Program? I guess it was that attempt to 15 quantify uncertainties related to risk estimates?
16 MR. ZUBER: No. Of that, I am not familiar. There 17 are several approaches one could really follow here.
18 Okay, let me discuss -- do you want to listen to this 19 or shall I go to -- I have three slides on this, if you want to 20 look.
. 21 MR. WARD: No, go ahead.
22 MR. ZUBER: What we intend to do is to address one of 23 the concerns that the peer review group hadt Can you apply it 24 to other transients, et cetera, large break LOCA. And we say we 25 can do it. And we shall do it in the context of a severo i Heritage Reporting Corporation l (202) 628-4888
~%
t 366 1 accident. And your overall objective would be to bring all 1 t
-2 these experiments done to date and analytical codes within the 3 CSAU framework. Then to provido -- experiments as cods 4 development and to provide a comprehensive evaluation'for what 5- is important. Not to do it in a fragmented way as was donc to 6 date.
7 And the way we -- the fourth objective would be to do j 8 a first pass evaluation of code uncertainties to calculate i.
i 9 parameters of interest to in-vnssel accident management.
10 MR. CATTON: Novak?
11 MR. ZUBER: Yes.
12 MR. CATTON: Does that mean you are going to need QA
_ 13 documents of coces like MELPRG and SCADFA --
14 MR. ZUBER: Absolutely right. Absolutely right. We 15 have to -- again, I am speaking --
16 MR. CATTON: I have one word that describes that.
l 17 "Fat Chance."
18 MR. ZUBER: Look, look. It is up to NRC management.
- 19. You know, if you want to keep up, you're better off in court, 20 which are not accountable -- by people who are not accountable.
21 I mean they are wasting taxpayers' money.
22 MR. CATTON: That's right.
i 23 MR. ZUBER: I think there should be some discipline 24 in code development, discipline in experiments, especially wLen f
i 25 our funds are getting smaller and smaller.
O Heritage Reporting Corporation (202) 628-4888
367 pf s_ 1 Yes, we need it. We need it definitely.
.2 We want to identify.the cost-effective basi:s, what 3 kind of code improvements would be needed to reduce code 4 uncertainties and to provide NRC with an auditable, traceable, 5 and quantifiable methodology for in-vessel accident management.
6 Now, let me address the concerns that PRG had and our.
7 response and our actions. If you read those lecters and the 8 records -- the letter that they wrote, I think you get one 9~ thing: We didn't provide them with any criteria for 10 establishing what is sufficient for a PIRT, what is sufficient 11 for a data base and what is the complete methodology.
12 To tell you the truth, when we started this matter 13 eight months ago or so, we didn't think of doing it. We wanted 14 just to demonstrate that methodology. But as a result of our 15 presentation to the ACRS and to the PRG, we realized this has 16 to be done. And this we shall be doing.
- 17 And now TPG's response is
- We recognize the concern 18 as very valid and they are in process of developing these 19 criteria. And, as I said, our management, their response to do i
- 20 this kind of work, and we shall provide this in our report.
21 MR. WARD: What is PIRT?
l 22 MR. ZUBER: This is the process identification --
23 tables. This what we identified. This is important. We shall 24 then focus our attention on that.
I 25 MR. WARD: I remember.
Heritage Reporting Corporation (202) 628-4888 l
\
368 l' MR. ZUBER: The second concern or one of the 2 concerns, the second here, is PRG was not really convinced.that
~
3 we really addressed in sufficient data the problem of scale.
4 Whether the codes kept to. scaling capability. There were 5 several reasons for this. We felt optimistically that we can 6 do it. Maybe we oversold our case.
7 'Another reason was the BNL had this in their 8 presentation, but they were cut short because the chairman of 9 the group wanted to save time, so, they didn't present this 10 .information which would have had effect on their thinking. But 11 w9 recognize this is a very important question and I think we 12 want to address it. And, in fact, we are doing it. And let me 13 just describe where we stand. And we shall provide more O 14 information in our final report.
15 MR. CATTON: That is going to be much tougher for the 1
16 severe accident?
17 MR. ZUBER: You bet. You bet 't will. But it just l 18 will be done.
19 We are approaching this program in several ways.
20 One, we have developed a general approach and we want to apply 1
l 21 it to the large break LOCA and to severe accident. And we want
- 22 to do this, first apply it to the large break LOCA because >s 23 have a data base. We have been working on this problem /
l l 24 may not have the final results before we finish our analysis 25 here. But we shall be applying this method first to large break Heritage Reporting Corporation (202) 628-4808 i
369
,7-~
's / 1 LOCA to show how it works and then to severc accidents as a 2 follow-up.
3 The approach is based on the hierarchical structure 4 of scaling. Ano what we have to date, we have developed this 5 top level scaling criteria or functional and we are preparing a 6 NUREG report and, hopefully, it should be out sometime in June 7 or July, depending on other work. But we will have this 8 document definitely this summer before the final report of the 9 TPG coes out.
10 Now, the question 18: Why do we need this method and 11 why would we need this method either for the large break LOCA 12 or for severe accidents?
13 First, we have to determine whether the code is
- 14 applicable to the scenario and that it can scale-up. Then we 15 have to identify what kind of experiments do we need, 16 especially for severe accidents. At what scale we need this, 17 otherwise, we may waste running experiments which are not 18 portinent. And, finally, what is important, we have to have 19 this index of success or failure to quantify the effect of 20 scale distortion.
21 Unless we can say, "If I change the experiments at 22 this scale, I am gaining so much," you don't give a measurable 23 way to say, "This is really cost effective."
24 I think this is a good management tool for NRC or 25 anyone else. If you are able to develop such a quantitative A
U Heritage Reporting Corporation (202) 628-4888
370 l
- 1 approach to this problem.
2 Now, what ve would do in this case, this is the CSAU 3 methodology as it is applied to the large break LOCA. And what 4 we would intend to do to modify, to have another activity here 5 connected with experiments and scaling and create a 6 quantitative index of success from experiments and see what 7 kind of experiment to uEe in establishing an matrix 8 mathematics. So, this would Lo a modificetion of this CSAU 9 methodology. So, obtain an index of coaling success for the 10 entire facility and also establish the methods here. And this 11 shall be described in this report. But this is how they would 12 apply this methodalogy to small break LOCA and severe accident.
_s 13 But to demonstrate it, we shall first develop this for large 14 break LOCA.
15 Now, the PRG had another concern. Some members of the 16 group -- not all of them -- some of them got really concerned 17 with nodalization and the comments they mada is You didn't 18 vary the nodalizaticn to see the effect on nodalization on your 19 uncertainty.
20 We had a rational, we still like it, to fix the 21 nodalization and do everything, all our calculations based on 22 the particular nodalization selected and then we have an 23 uncertainty for this particular nodalization.
24 We shall address these questions day after tomorrow.
25 We have some members in our group, TPG, particuler Dr. Wolfe, Heritage Reporting Corporation (202) 628-4888 i
I 4
'371
() I who wrote something they want to discuss in great detail on 2 Friday. We are aware of these questions. We are going to 3 examine it in detail on Friday. I cannot give you the answer,
- 4 but we shall discuss it and we shall address it one way or the 5 other.
6 Another comment that the PRG group had was: The
- 7. methodology considered uncertainty of primary safety criteria; 8 the uncertainty of other parameters is not considered.
9 They were a little bit -- not all of them, some of 10 them -- concerned that we addressed only the PCT and we didn't 11 take the uncertainty of other parameters.
12 Our comment down here is two comments. Our response 13 to this concerning the following: As far as the uncertainty, I
() 14 think it is cost-effective only to address the primary 15 licensing criteria. This is essentially PCT. It is not cost 16 effective to address all the criteria.
17 MR. CATTON: One of the earlier steps is to make some 18 sort of a statement about the goodness of the codes. At that 4
19 time, all of the variables are considered. So, your answer is l 20 really not complete.
l 21 MR. WILSON: Let him go to the next part of the l
22 answer.
t t
23 MR. CATTON: Okay, I didn't know there were two i
24 parts.
25 HR. WILSON: There are two parts, r
f)v Heritage Reporting Corporation (202) 628-4888 i
W f
372
=1 MR. 7UBER: We recognize in order to defend a' method, 2 really depends on looking at other parameters. And I think 3 this is provided by assessment reports. This is where you will ;
-4 get the feel: Is there a physical basis? Is the code good 5 enough? Can the code be applied with some confidence to this 6 particular case?
7 So, what we need -- this is where the assessment
- _ 8 reports come, but I' don't have to run uncertainty on any one of 9 the secondary parameters. This will give me the goodness of 10 the code in a general way. The numbers on the uncertainty, I do 11 it on PCT or whatever else I may consider.
12 Another concern the group had was that we didn't-13 consider the thermal-hydraulic effect or channel effect. And r
L. 14 this was a good comment.
15 We responded we realized it. We responded immediately our management, again, wanted us to address this. L 16 17 We did it. We performed the calculations and got the
-18 results --
19 MR. CATTON: Shouldn't Westinghouse get the credit 20 for that concern? I believe they should.
21 MR. ZUBER: Yes, Dr. Hochricher (ph) made this 22 comment and I want now to say here that we are not only 23 responsive to the peer review group, but also to the inputs 24 from industry.
25 MR. WILSON: If I could offer Dr. Hochricher was an O Heritage Reporting Corporation (202) 628-4888
, a a
f 373 1 ad hoc member of-the PRG at that particular time.
2 MR. ZUBER: Thank~you. This will be in the public 3 record.
4' Another concern that the PRG had was that we didn't 5~ only look at the cross products, the effects of a
6 ' thermal-hydraulics and the fuel. And we recognized this also, 7' .but we didn't think at the time that this was important. It 8 was our feeling that what we showed at that time was 9 conservative and that we dion't have to consider this cross 10 product effects, the thermal-hydraulics and the fuel.
U 11 Nevertneless, in order to respond to the concerns of 12' .the PRG, we decided to do these calculations. Also to see how r
13 far off we were from our original judgment. Again, we get 14 money from our management to do this.
15 The results show that we were right in the first l 10 place. But, anyway, we did it and you will see the results.
17 Another concern that PRG had which we addressed is ,
18 they were concerned that we used in our work an uncertainty of 19 results which were only used in code assessment and 20 development. In essence, we are using data which we already 21 uaed before like LCFT, particularly LOFT. And they made the 22 comment: In order to show that this methodology, our numbers 23 on uncertainty are good, we should use some "virgin" data, 24 which we didn't use before. l 25 And we realized this. This is very desireable, I
O Heritage Reporting Corporation >
(202) 628-4888
374 s
O 1 except there are very few virgins left today as far as 2 facilities are' concerned and we don't have a large pick of 3 them, so,-you have to work with what's available. l 4 Nevertheless, we intend to respond to this concern as 5 good, as well as we can. We got data from PKL which we didn't 6 use before and we shall include it in our analysis.
7 Mr. Fell from U.K., he wrote a several page comments 8 and recommendations. It is in your package. He suggested that 9 we should also include OECD calculations of a LOFT test. We 10 already did this and got it. We will show you the results.
11 So, in a sense, we tried to be as responsive to the 12 concerns and address the most important concerns that the'PRG 13 had. And they are summarized on these next two slides. The f.
r
% 14 concern of the hot channel is addressed. We did address it.
15 We are well aware of the cross products. We included the 16 independent experimental data from other facilities. The PKL 17 will be done in the near future. ;
18 And we are still in the process of evaluating or l 19 addressing other concerns and these are shown on this slide.
20 Here is the summary. We shall develop the success criteria and 21 the sufficiency criteria. We shall provide the rationale and 22 address the question of nodalization in greater detail. We are 23 addressing the scaling in much more detail that probably they 24 would have expected us. We intend to do that.
25 We shall provide the sufficiency criteria and in ;
Heritage Reporting Corporation (202) 628-4888
375 c
km) 1 addition to the PCT. This will require a much more expanded 2 documentation than we really thought originally because this 3 would require some additional effort. And they are especially 4 included in PKL. So, we won't be accused that we didn't 5 consider "virgins."
6 DR. KERR: Excuse me. What is NPP in number "b."
7 MR. ZUBER: Nuclear products.
8 DR. KERR: Thank you.
9 MR. ZUBER: It is similar to semiscale. This really 10 summarize -- I mean the first part of the presentation deals 11 witt. our response and comments from the PRG. Gary Wilson will 12 now tell you where we stand today and the results we have. I 13 will summarize then the future work.
O
\/ 14 DR. KERR I was impressed with your comments on 15 dealing with the severe accident issue. And it seems to me what 16 you propose is very appropriate, It does strike me that there 17 may be a significant difference between that effort and the 18 current one in that you -- and you mention this when you said 19 it makes sense to concentrate on peak clad temperature because 20 it is a principal method for regulation.
21 In the severe accident arena, you don't have those 22 landmarks and one of the things that it seems to me you have to 23 do is to decide what it is one needs to know.
24 MR. ZUBER: I didn't put it here because this was not 25 a topic. What I really wanted to show that we shall address Heritage Reporting Corporation (202) 628-4888
~
1 L1 u.
376 1 scaling and_with this context. And we went to apply this 2 methodology to other accidents. It is not one which only 3 applies to large break LOCA. If that is the case, it would not 4 be worth it for NRC to spend the money.
5 But this thing, I think if you have another G opportunity,-I will go into this of how we want to do it.
7 There are other things. This has to be also combined, just to 8 give you a flavor, with an event. You see a manager has to 9 know what is the success rate. What we have to try this-10 methodology which is -- represented by experiments on the code.
11 To management actions: What does he do at a 12 particular point in time?_ Doer he take the left step or the 13 right step? One is success spot, one drives him into numbness.
14 And then you have to develop these event trees and then do the 15 uncertainty. Once you get the uncertainty, you relate it to 1
16' the event which is then related to management actions. I think 17 this is how we would like to put our arms around the problem.
l 18 Gary, it's all yours.
19 (Continued on next page.)
20 21 I
- 22 i
i 23 24
- 25 l
i
() Heritage Reporting Corporation (202) 628-4888 l
l
{f 377 1 MR. PLESSET: What code or codes do you have in i
2 mind? !
3 MR. ZUBER: For a severe accident?
n 4 MR. PLESSET: We are talking about codes. ,
5 MR. ZUBER: Right. For the severe accident, you have .
6 combinations. We did decide that once we embark on this 7 program, then we should identify the code, look at the 8 experiment, and then proceed. But that we should have a 9 prerequisite that any code that we shall be using the 10 documentation and the QA document. Otherwise, it a game. You 11 always have a new version of a code that is not assessed. And 12 once you want to assess that code, they will tell you I have a 13 better one on the horizon. It is a game well played.
- / 14 MR. WILSON: I am Gary Wilson with the INEL. And as 15 Novak indicated, I am going to be showing you results of those 16 items from the PRG that we have addressed to this point in 17 time. Those topics are indicated here. We will be looking at 18 the additional data that we have used that were not used in 19 code development and assessment. The effects of additional 20' hydrodynamic cross products, and the effects of the hot channel 21 thermal hydraulics on PCT.
22 The comment that came from the PRG in this particular 23 instance related to that we had riot used virgin data in our l
24 analysis. The data that we had used had actually been used in 25 the code development and assessment program to some degree.
)
Heritage Reporting Corporation (202) 628-4888 i
378 A)
(_ 1 Some to a high degree, and some to a very low degree. And they 2 suggested that we try to look for virgin data. And we have 3 done that. We still have some stuff in progress.
4 What I am going to show you here relates that we did 5 go in and pick a LOFT OECD test.
6 (Slides being shown.)
7 As indicated here, we used 37 additional blowdown PCT 8 points. I will give you the bottom line, and then I will show 9 you, and then I will tell you again what the bottom line is.
10 The change in the experimental data availability at the 95 11 percent confidence limit was less than one percent, three 12 degrees, when we added this semi-virgin data to the data base.
13 MR. WARD: Is that because this was just a very small 14 population, I mean 37 points were small compared with what you 15 are adding it to?
16 MR. WILSON: The 37 points was a 16 percent increase 17 in the data base, and I will show that to you. It resulted in 18 less than a one percent change in our perception of the j 19 variability of the data. Now I am told that that is in a 20 statistical sense significant.
, 21 Now in the process of doing this, we have also added l
22 some other data points from the data bases that we had prior to 23 that prant in time. I just wanted to show you what the effect 24 of that was.
25 Just to refresh your memory, this is the data base
(
l Heritage Reporting Corporation l (202) 628-4888
379
() 1 and its analysis that we took into the program review group 2 meeting. The most important or the highlighted point here is 3- what this data told us. It gave us perceptions of the 4 variability of experimental data from this data base which 5 represents three decades of volume scale, most of which were 6 scaled on the basis of the power volume scaling.
7 And what we see is roughly a 361 degree variability 8 in the experimental data base at a 95-percent confidence limit.
9 That was the information that the program review group looked 10 at and commented on.
11 MR. SCHROCK: I am surprised to see PBP included in 12 these. That is not really an integral simulation. !
. 13 MR. WILSON: It is not an integral facility in the !
14 same sense that the other facilities that are shown here are..
15 However, it is one out of the two facilities shown in this data .
16 base that are nuclear facilities. So we thought that it was 17 enlightening to include it in the data base. The point shown 18 here in green --
i 19 DR. KERR: Did it turn out to be enlightening to ;
20 include? f 21 MR. WILSON: Well, what it did to me, in my own !
22 personal opinion, is that it extended the scaling range 23 significantly. It did not seem to be inappropriate to include 24 Jt. So we now have a much wider range with respect to scale.
25 MR. CATTON: But it is short. It is even shorter
( !
Heritage Reporting Corporation (202) 628-4888 j
380
) I than the LOFT.
2 MR. WILSON: Some of the semi-scale that is in here 3 is short. The LOFT data is short.
4 MR. SCHROCK: But the system is altogether different 5- from the' system that you are.trying to simulate and calculate 6 heat clad temperatures, so'what is the meaning of that 7 comparison? That escapes me completely.
4 8 MR. ZUBER: If I may answer. The way this is that 9 everything is built on the core really. And it is almost like 10 it has been decoupled from the rest. Essentially, it is a core 11 phenomenon. If that is well-scaled, if.that is well-scaled,
- 12 you would expect then that it should fall in this range here.
13 But if the thing was not well-scaled, then I think that it l) 14 would be out of the range. But to us, it was an indication of 15 two things. It la the core, and the second thing it is
- 16 well-scaled according to volume. ,
- 17 And when the report comes out, you will see another i 18 rationale why this is so powerful, some additional information, 19 But this gave us two things. It is nuclear light LOFT, and it 20 is governed by the core. And this really confirmed our 21 assessment.
22 MR. WILSON: In any case, by the additional of these 23 additional LOFT OECD points, we see the variability change from 24 361 degrees to 364 degrees, a three degree change. That is 25 something less than one percent. It came about through a 16 O Heritage Reporting Corporation (202) 628-4888 L
381 A
(-) i percent increase in the number of points in the data base. And 2 as I indicated before, in a statistical sense, that has 3 significance.
4 Now we have done that, we have in effect gone in and 5 added additional data points. The prior data that I showed you 6 was I believe 253 data points. And we now has 301 data points.
7 We have gone in and added semi-scale, LOBI, and LOFT points.
8 Not LOFT OECD, but prior LOFT testa. And we now see that by 9 increasing the data base by those some 48 points, that we see 10 that the variability has changed roughly 14 degrees from what 11 it was before, fortunately in the right direction. This is 12 approaching what we can get our hands on.
_s 13 MR. ZUBER: I was going to say that this is probably 14 as much data as you '<ill find anywhere in the world. They 15 scooped everything that they could put their hands on.
16 MR. CATTON: It does not seem to make much
. 17 difference.
18 MR. ZUBER: Unfortunately, it seems so.
19 MR. WILSON: he think that we are approaching a point 20 of diminishing returns, with respect to blowdown now.
21 MR. SCHROCK: I wonder if Lou would prefer you call 22 that a 95 percent probability line.
23 MR. WILSON: In this cr'e, this is a 95 percent 24 confidence limit in the way that I understand it, in that this 25 band is based on tolerance limits. And it is a statement of O Heri.tage Reporting Corporation (202) 628-4888
(
382 1 how confident you are in that band with relation to the number 2 of points in there. In this instance because we have 301 3 points, the probability and confidence limits are probably not 4 that much different. Let us not launch into that discussion 5 again in detail.
6 Now turning to the reflood concerns, talking about I
7 reflood PCT. We thought that the PRG comment related to 8 reflood as well as it did to blowdown. So we are in the 9 process of doing that. We have added eight data points that 10 came from this same LOFT OECD test.
11 The results to this point in time were that we saw no 12 discernable change in the temperature rise during the reflood, i 13 or at least that is our understanding of that. And I just want 14 to point out that we have just received some additional data 15 from PKL. I think that we have perhaps located another case 16 from the Netherlands. IT came from the ECCN in the 17 Netherlands.
18 So the next plot that I am going to show you is an l
19 interim plot. We are still working on that. But with respect 20 to the data that we had in hand, here is what we saw.
21 This is the rise rate as a function of reflood rate.
(
22 Before we added the OECD stuff, we had these points, and we had 23 these bands. This represents eight new points. It does not 24 appear to change our picture or our current understanding of 25 the reflood situation.
Heritage Reporting Corporation (202) 628-4888 -
l
383
/'_T
"( ) 1 And I would just note to you that in our NPP 2 calculations in which we are characterizing the uncertainty, 3 what we have seen is that the NPP reflood range is in that 4 range right there.
5 MR. WARD: Those two graphs, does that mean that one 6 could get by without doing any calculations?
7 MR. WILSON: If one wants to say that the 8 experimental world that we know represents everything that we 9 need to know in the future, then I guess the answer is yes. ,
10 Now in turning to the second subject that we are 11 addressing with respect to the program review comments. Well, 12 I will tell you what the bottom line is, and then I will go 13 through how we ended at that bottom line.
O kJ 14 The use of additional hydrodynamic cross products 15 that I am going to show you here resulted in a lower blowdown 16 PCT at the 95 percent probability, but the change was small.
17 It was roughly 16 degrees F, which is about one percent.
18 Now to lead you thrcugh how we arrived at that bottom 19 line. The program review group's concern was that we had 20 looked at single, double, and triple cross products in 21 developing the response circuits that we used to generate 22 uncertainties.
23 What we had not done was look at double variations in .
24 hydrodynamics. We had double variations between a single 25 hydrodynamic parameter and a fuel parameter. We had doubles, i
Heritage Reporting Corporation (202) 628-4888 l
l
384
(_/ 1 and we had triples. But what wo did not have was double 2 hydrodynamic variations at the same time.
3 So we made three additional runs with the nuclear 4 plant model that are indicated here. We chose a case where we 5 buried both the pump and the critical flow in three different 6 ways. The pump with its two sigma variation coupled with the 7 critical flow, and the one sigma variation, et cetera, as you 0 see on the slide here. It will become evident why we picked 9 those three in the next slide.
10 Nhat we have here is an indication of the relative 11 change in the PCT between the prior condition, i.e. using 12 single hydrodynamic variations as separated from the current .
13 condition now where we had done the double hydrodynamic k-} 14 variation.
15 And what we see here is this represented what the PRG 16 saw. When we had exercised critical flow through the one sigma 17 and two sigma variations, we saw as an individual effect the 18 change in PCT from the nominal condition as shown here. And we 19 see that the critical flow one sigma is the highest of those, 20 although we see other contributors down here.
21 Now when we go in and do a double variation where we 22 combine. By the way, these were among the most significant 23 that we looked at. This was the most significant individual 24 parameter. That is why we chose this set here to look at l 25 double hydrodynamic variations at the same time. And these are lieritage Reporting Corporation i
(202) 628-4888
385
() I the changes from the nominal case with these kinds of 2 combinations.
3 MR. SCHROCK: Are these variations taken to produce a 4 positive heat clad temperature change?
5 MR. WILSON: They were chosen to produce the worst 6 change, i.e. a higher PCT.
7 MR. SCHROCK: It seems strange that the one sigma 8 critical flow would produce a 106.. increase.
9 MR. WILSON: Virgil, what happens is that when you go 10 from the one sigma to the two sigma, you get a difference in 11 pattern in there. And roughly, what that looked like in terms 12 of temperature versus time was that one of them would look like 13 this, and the other one would look like, I believe that it was
) 14 like this, where this was the one sigma case and this was the 15 two sigma case.
16 The real variation was in the timing. But what we 17 believe after looking at the runs and everything is that we 18 went through a different set of -- when the code goes down 19 through the logic, we went through a different set of logic, 20 and we saw this kind of behavior.
21 MR. LAUBEN Are you sure that it could not be 22 simpler than that? It could be that when you look at the break 23 calculations, that it is not always the largest break that 24 gives you the worst heat clad temperature.
25 MR. WILSON: This fully supports that.
O Heritage Reporting Corporation (202) 628-4888
386 m
(-) 1 MR. LAUBEN: Yes. And it is not necessarily the 2 logic, but just the fact that sometimes the very heavy 3 down-flow that you get with the biggest break proves to be 4 somewhat of an 2dvantage. You blow down quicker and you start 5 your -- well, I do not want to say that. I guess that I am not 6 sure.
7 MR. WILSON: I am sure that I could ask Jim to stand 8 up, and he could give you a 45 minute discussion of exactly why 9 this happened. We know why it happened, or at least we know 10 what the code said happened. I do not believe that it is 11 significant to the objective that we are trying to achieve 12 here. It is your pleasure.
13 Now looking at adding these three cases to the seven kJ 14 cases that we had prior to this and going through the process 15 that we used to characterize uncertainty in its total sense and 16 in its ccmbined sense where we go through response surface and 17 sampling, and create a probability density function, this is 18 what we saw.
19 The old value was based on only the single 20 hydrodynamic variations that we had in the original seven cases 21 including the nominal case. Then the new data is the data 22 where we combined the three new runs that contained double 23 hydrodynamic variations wiuh the prior seven runs.
24 What we saw was n change in the mean value of the 25 probability density funct'.on. The new data base dropped that O Heritage Reporting Corporation (202) 628-4888
387
() 1 by about five degrees. The standard deviation, it showed a 2 tighter density function by about 23 degrees. And the 3 temperature on a one tail test with a 95 percent probability 4 level changed from 1379 down to 1363, a decrease of 16 degrees.
5 And this data supports the prior statement that we 6 thought that we were conservative in the original analysis.
7 And as it turned out, that was not the case.
8 Now as was indicated, Dr. Hoitreighter raised the 9 question of hot channsi hydrodynamic effects on the original 10 analysis. So we have examined that. These two statements give 11 you a little bit of background of why the original analysis was 12 done in the way that it was.
13 Going into the original analysis, we had experimental O-s 44 evidence that tended to indicate that the average fuel rods 15 drove the core global hydrodynamics. So we thought that the 16 hot channel hydrodynamics on the hot tin, the effects of that, 17 were second or third order, and we did not quantify that.
18 Now we have quantified it, and I am going to show 19 that to you. Preliminary results as of today and the analyses 20 that we have run to date indicate that on a combined 21 uncertainty basis that there may be a bias as high as 68 22 degrees on PCT if you introduce local hot channel hydrodynamic 23 effects on the hot tin. And then I am going to show you how we 24 arrived at that.
25 We revised the NPP model to contain four hot bonds.
O Heritage Reporting Corporation (202) 628-4888
388
() 1 We made.two runs, and we reran the nominal case with what I 2 call the new model. We also ran the critical flow one sigma 3 case through blowdown. The nominal case, we ran through 4 reflow.
5 To this point in time, we have done two types of
- 6 comparisons to try to quantify the effect. We have looked at a 7 run by run comparison, and I am going to show you some results 8 of that. And then we also recreated PDFs and looked at it on a 9 PDF basis, and I am going to show that to you.
10 On a run by run basis, this is the kind of effect F
. 11 that we see from the hot channel. What is shown here is clad 12 temperature as a function of time. Case one was the original 13 model. By the way, both of these are for the nominal set of j
() 14 ' conditions. Case was the base case for the original model.
15 Curve two is the hot channel model.
16 We soo no significant change in behavior except that 17 we do see an increase in the PCT. That is the kind of general i 18 effect that we saw on a run by run basis, i 19 with respect to the runs that we have in hand, this i
- 20 summarizes the information on a run by run basis. Here is the 21 nominal PCT for the old model during blowdown. Here it is with I' 22 the hot channel model. And the difference between these two is 23 24 degrees. It was an increase of 24 degrees.
24 During reflood, it went from 820 degrees for the 25 original model to 835 degrees, an increase of 15 degrees. With
! C:) Heritage Reporting Corporation (202) 628-4888 4
389
(
O) 1 respect to the critical flow run which we ran out through 2 blowdown PCT, we saw a difference from 938 originally to 960.
3 That is a 22 degree increase. So on a run by run basis, we saw 4 differences as much as 24 degrees in that curve of blowdown in 5 this instance. <
6 Now turning to a PDF type comparison, looking at the 7 total uncertainty. I will throw this up here. I thought that 8 because of this morning's discussion that we ought to say 9 something about PDFs.
10 This is a typical example of the probability density 11 functions that we are getting for PCT. This is not the final 12 one, but this is a typical one.
13 In this instance, we see that it is not normal, 14 although it approaches normal. But we are quantifying the 15 probability of the density function in our analysis. So when I 16 talk in the next slide about comparing PDFs, this kind of 17 information generated from runs where we have single, double, 18 triple, and even quadruple cross products.
19 DR. KERR: What is the significance of the statement 20 of all distributions uniform?
21 MR. WILSON: Oh, okay. This result is based on the 22 fact that for the dominant phenomenon that drives PCT 23 variations, that what we have done is that we have ranged the 24 variations in those phenomena like peaking factor and gap 25 conductors.
O Heritage Reporting Corporation (202) 628-4888
O f
I 390 1 We have ranged those from experimental data. In.some
.2 , instances, we know what the distribution looks like of each of !
3 those individual contributors.. In some instances, we do not 4 know what that distribution looks like.- In that case when we ,
f I are confronted with that, we assume a uniform distribution 5 l i
6 which in statistical terms or a statistical sense indicates the !
l 7 most lack of knowledge of the distribution of that particular 8 parameter. Any point within the range is equally likely. l'
. 9 DR. KERR: And all distributions have.been treated !'
1
! 10 that way? !
, i
- . 11 MR. WILSON
- Every dominant phenomena or parameter !
12 that went into creating the individual PCTs that eventually end ,
i :
! 13 up in this combined form for the pluses here were assumed that
! 14 we did not know what their distribution was. We assumed a 15 uniform distribution. [
16 You will note that there are some circles here. At i
- i 17 each gap, we had an idea that that was a normal distribution. l l t
( 18 The peaking factor was a normal distribution. And the fuel !
l 19 thermal conductivity was a normal distribution. You can see in i
! 20 that instance that there is not much difference in the answer. ;
i j 21 We have tried to establish checks in everything that 22 we do. This is an example of one of those checks that we tried I 23 to run at each stage. ,
t 24 How going back to the main theme. Looking at the l 25 possible or the potential offeet of a hot channel or hot
! O Heritage Reporting Corporation
(
l (202) 628-4888 l
! i I
391
(/
s-T 1 channel hydrodynamics on a total combined basis, i.e. looking 2 at the comparison of probabiljty to density functions, this is ;
3 what we saw.
4 The original model, not hot channel. The hot channel 5 model. There are two PDFs that were generated from full second 6 order response surfaces and constructed from the nominal and 7 critical flow of one signal NPP runs that we had, and gave 8 these kinds of results.
9 The mean PCT changed from 1093 up to 1176, a change ,
10 of 83 degrees increase. The standard deviation. The standard 11 deviation changed by 25 degraes. PCT at the 95 percent 12 probability, again a one tail test, went from 1511 to 1579, a 13 68 degree change. And that is where I made my original 14 statement that it would appear that 68 degrees, a bias of 68 15 degrees, may capture the effect of the hot channels.
16 Now we are going to address this subject at the TPG.
17 To dn the hot channel analysis in every case costs about 30 or 18 40 percent more in terms of computers, and even more than that 19 when you throw in labor and that sort of thing.
20 I em convinced that we could do that if we wanted to 21 spend those resources and do hot channel analysis for every 22 run. And when we ended up with the final PDF, we would see 23 something tighter than that. The hot channel would really not 24 have that much influence.
25 One of the decisions that we are facing and that we Heritage Reporting Corporation (202) 628-4888
392 1 are going to address is the cost to do ti.is refinement worth
~
2 making that value somewhat smaller. And I mention this, 3 because we talked about cost effectively and practicality. I 4 think.that these are the kind of practical cost efficiency 5 decisions that anyono who gets into this ball game of 6 uncertainty-is going to have to-face in the future.
7 And you judge whether a method is practical only if 8 it helps you, and it is the cheapest way that you-can reduce 9 risk to the public, or the capital cost up front to do the 10 process will pay you in terms of increased revenues, for 11 exemple, from a utility standpoint.
12 That completes what I had to say. I will be glad to
, 13 address any questions.
14 MR. WARD: Your intent for the hot channels, it is 15 just to use the 68 degrees?-
16 MR. WILSON: We are considering that option. And as 17 I said, we will discuss it in the next two days.
18 MR. WARD: Are there any other questions?
19 (No response.)
20 MR. WARD: Thank you, Gary.
21 (Continued on next page.)
22 23 24 25 O lieritage Reporting Corporation (202) 628-4888
i 393 1 MR. ZUBER: In a sense, we are the users of our j, 2 codes we developed, and now we have to face up to the fact how
! 3 much it costs.us to run them, and-then we can do it within a i j 4 given amount of time and a given amount of money. And these ;
4 5 are-the decisions'we have to face in the meeting with our !
l 6 group.
7 Let me then summarize the results of where we stand !
i 8 today and where we are going tomorrow and from now on. .
9 We discuss address most of the concerns that PRG had.
10 This introduce some delay, but in particular it introduced more i
11 work on the final document, because it might have more !
i 12 implications and impact than we originally thought a year ago l t
13 or a month ago. :
() 14 For the blowdown uncertainty, the original thought -- l 15 we did finish it in March of 1988. We did these additional
, t 16 calculations and this put the delay to June of '88, i
- 17 The refill /reflood we thought we would finish in 18 April. Because of the delay, we will finish in July. }
19 We intended to finish the documentation, the document l 1 !
j 20 uncertainty in May, and they finished reporting in June. This !
r h
will be delayed in August and you shall get a report in !
21
- i j 22 September. Unless something really drastic happens, we shall j r
l 23 plan to give this late. We want to finish this this fiscal i 24 year. I mean we are going to bust our whatever to finish our ,
- i 25 work by September of this year. }
3
() Heritage Reporting Corporation f
j
- (202) 628-4888 i
i
- i j l t
e 394
'1' DR. KERR: That's almost a disgracefully small 2 slippage.
3 MR. ZUBER: Well, we thought we --- well, anyway thank 4 'you ab'out that.
5 We don't practice this thing. We thought we would 6 finish, and especially we wanted to_ finish this document before 7 the reg guide goes and it will just happen in parallel. But'we 8 thought it -- the PRG did such a good job. Really, they-did a 9 very job made by these comments. We thought we really should 10 address them. It's a_ technical community. We would like to 11 -say, yes, we looked into these things, we addressed them and p 12 these are the results, and these are the consequences, but we 13 should have it finished i: September. We must finish it in O 14 September.
15 If you have any question.
16 MR. WARD: No, thank you.
17 MR 2UBER: Thank you.
18 MR. WARD: Let me address the subcommittee. I 19 propose that we have some time at the May full committee 20 meeting to review what the full committee, this final version 21 of the rule. That is, if the subcommittee recommend us going 4
22 forward with it.
23 I guess one wrinkle is that the -- they aren't nearly 24 quite as far along with the CSAU examples as we had thought 25 some months ago. It looks to me this may not be a serious gap.
O
~
Heritage Reporting Corporation (202) 628-4888
395 (s- 1 The potential for finishing up satisfactorily looks pretty 2- good.
~
3 Any comments on that?
4 MR. CATTON: The only concern is whether or not they 5 are going toflose interest in completing this work once the 6 rule is out. I don't think there is anyl problem.with them 7 completing. Actually in many respects,-at least from my own 8 experience, the reflood would be easier than the blowdown.
-9 That's my view.
10 MR. WARD: Virgil, do you have anything?
11 MR. SCHROCK: No, I don't disagree what Ivan just 12 said. I am inclined to think that there is every expectation 13 that they can finish it up reasonably, casting any doubt on the O 14 root change.
1.5 MR. WARD: Milt.
L16 MR. PLESSET. It sounds all right to me.
17 DR. KERR: I have no reason to disagree with such an 18 eminent expert.
19 MR. WARD: Okay. Well, do you think -- I don't'know 20 have we been alloted some time.
21 MR. BOEHNERT: Yes, yes, we have.
22 MR. WARD: How rauch?
23 MR. BOEHNERT: We have an hour and a half for 24 discussion of the rule, plus the letter-writing.
25 MR. WARD: Sounds a little short, I would have said a t%
U Heritage Reporting Corporation (202) 628-4888
y
. 396
- 1 couple of hours if we could get it.
2 How much do you think we need to.have the full 3 committee hear the CSAU story again? Maybe just some sort of a 4 brief, a briefer sum of the status. I mean' they have heard it.
5 They know what it is.
6 DR. KERR: I certainly found.what Mr. Zuber just 7 presented quite interesting. Maybe I had heard it before, but 8 in its latest version it seems to me it's rather significant.
9 MR. WARD: You begin to understand it better.
10 Well, why don't we plan on two hours and a half an 11 hour for the CSAU -- I think you really need to clean up for 12 the full committee, Norm, your presentation.
13 MR. LAUBEN: Don't need to put a lot of changes in 14 and stuff like that, and I don't know --
15 MR. WARD: You had a lot of good stuff there, but you 16 kind of rambled through it.
17 MR. LAUBEN: Probably nod a lot of history in it 18 either.
L -
19 MR. WARD: Well, I think the history, if you can go 20 through it briskly, is probably a good idea. I would try to go l 21 through it. That was good. But I think you just need to get a l
22 little more concise presentation on the last three-quarters of 23 it.
l 24 DR. KERR: Not more than five minutes on history.
l l 2b MR. WARD: Not more than five minutes on history?
i I
Heritage Reporting Corporation
! (202) 628-4888 l
t I
-- ,_ - _. . . _ . _ _ . _ _ _ _ _ _ ._. ... . _ ._ _ _. ~,
T 397
- r~ -
(s 11 No, I wouldn't short the history. I would have to 2 disagree with my eminent --
3 DR. KERR I thought maybe by saying five, I could 4 get-it down to 10.
5 MR. WARD: Okay. You know, don't' lovingly relate 6 every moment of the history.
7 DR. KERR: Talk twice as fast when you are talking 8 about history.
9 DR. SHOTKIN: Do you know when the full committee 10 meeting is?
11 MR. WARD: Do you just want to know whether it's 12 Thursday or Friday. Okay, it's Itay 5th and 6th, but I don't l 13 know which day you have. We will let you know.
14 DR. SHOTKIN: Dr. Zuber will be at 2D3 meseting in 15 Japan on those dates.
16 MR. WARD: Well, could you or someone just give a 17 status report?
18 DR. SHOTKIN: Yes, Gary.
19 MR. WARD: Gary? Okay, that would be fine.
20 MR. ZUBER: I should be in Washington on the 5th.
21 That's a Thursday. Friday I won't be here. I am flying.
22 MR. WARD: Oh.
23 MR. ZUBER: But then if the schedule is such that we i 24 can make the presentation on the 5th, I will --
- 25 MR. WARD
- Well, our first attempt will be to make it l
Heritage Reporting Corporation (202) 628-4888 l
l m- ar - >r---ry *
.l
.l
j- \
398 l i 1 .on the 5th. I-don't know-if you-have set the agenda for the
.2 ,-full meeting yet.
3 ~DR. KERR: Well, one always setsLit at this. point, j
'4 'but it is -- it becomes less flexible as one gets closer.
5 MR. WARD: Okay.
^
6 DR. KERR Paul can find out whether it's still --
7 MR. WARD: OKay, we will try to make it on Thursday 8 then.
9 Okay, well, gentlemen, thank you very much for your 10 time.
11 'Yes, a question?
12 Mr. WILSON: We presumed we were going to hear some 13 summary comments with respect to the RELAP5 methods and 14 correlations, Mr. Ward.
15 MR. WARD: Yes, that's right. We didn't do that 16 yesterday. Thank you, Gary, I appreciate that.
17 Yes, let's do that now then and keep the record.
18 Ivan, why don't we let you start.
19 MR. CATTON: I have a few comments to make.
20 First, I found that RELAP5 in many respects is 21 similar in its characteristics to TRAC. The QA document that 22 we went through is in need of additional information.
23- We heard several times that it was only 50 or 60 24 percent done. It clearly needs to be augmented. In 25 particular, both the QA document for TRAC and RELAP need to O Heritage Reporting Corporation (202) 628-4888
f 399 O)
- k. 1 include a chapter on numerics, and I think what happened here 2 yesterday is an example of why.
3 We heard a presentation and got one opinion. Then I 4 talked to others in the audience later, and was told that the 5 opinion I gathered from the presentation was incorrect.
6 Well, if we can't understand the person when he's up 7 presenting it, they have to write it down in a clear way. And 8 this means the descriptions need to be such that somebody who
'9 is a little better than a layperson can at.least understand 10 what they are doing.
11 The other part is the reflood. This was one of the 12 things that was left out. It needs to be included, because it l 13 is part of the code. I think all of what's in the code must be 14 in the QA document, or else take it out of the code or do 15 something.
16 The other thing I would like to see something on is 17 the impact of using a constant critical Weber number. I can 18 think of circumstances where this has an impact on things like 19 the steam binding. If you don't get the drop sizes right, you 20 are not going to get the entrainment right. If you don't get 21 the entrainment right, you are not going to get steam binding j
! 22 right.
1 23 Now you can argue that it really doesn't matter 24 because it's a low probability, but you have to do something.
25 And I think there ought to be a discussion on both TRAC and Heritage Reporting Corporation (202) 628-4888 l
l
400 T',
t <
'ss' 1 RELAP5 on this particular aspect.
2 I think in the spirit of peer review fever, the codes 3 should receive some peer review. The code development process 4 I think has been way too inbred, and outsiders who do have 5 anything to do with the code tend to be users only. And I have 6 heard people from Bechtel talk about how great the code is for 7 water hammer, and we all know condensation in the code is no 8 good, or is poorly represented. It's not a code that should be 9 used for water hammer.
10 So I think peer review, published peer review will go 11 a long way to helping more clearly delineate where the codes 12 are of value.
., 13 The QA documents should be published as soon as
()
14 possible as NUREGs, and I mentioned that earlier. I would like 15 to see them referenced in the reg guide so that the industry 16 has some idea as to what is expected.
17 I would also like to see the CSAU lessons learn 18 document produced as a NUREG as soon as possible. In 19 particular, if there is going to be something done in the 20 accident management arena, I think some of the lessons learned 21 clearly point to how a code should be structured.
22 Of course, this is relative to the troubles with 23 TRAC. I don't know if one would have the same troubles with 24 RELAP. But, nevertheless, there are lessons on code structure 25 that I think will do us a lot of good, f3 O Heritage Reporting Corporation (202) 628-4888 l
l
401
(~1 s/ 1 During some of the discussion yesterday, we were told 2 how.RELAP5 was a small break LOCA code, and not for large break 3 LOCA. This came as a surprise to me, and it is somewhat 4 inconsistent with the presentations that we. heard from the 5 RELAPS people.
6 As far as I know, RELAPS is the most used code by the 7 industry. It's going to be used by the Europeans, and in my 8 view, it certainly has been treated in the past as a large 9 break LOCA code. I don't think it should be limited to small 10 break LOCA.
11 MR. WARD: Could I ask, when you said that you 12 thought the code should be subject to more peer review, what do 13 you mean? How would you propose that it be done?
(:) 14 MR. CATTON: Well, there are a lot of people who work 15 in the thernal hydraulics area, and the present modus operandi 16 of NRC seems to be gather together some experts to give an 17 opinion on what you have at hand. And we certainly do that 10 here in that we have given our opinion. But somehow that's l^
19 different than when an outside group gets together and says, 20 gee, the code is good, bad or indifferent. It could be used l
l 21 for this, that or the other. And that needs to get out into 1
22 the user community. I'm not sure that what we do here does, l
l 23 MR. WARD: Milt.
l 24 MR. PLESSET: Well, I have long felt that RELAP was a 25 preferable code to TRAC, and sometimes I wondered why. But I l Heritage Reporting Corporation (202) 628-4888 l
l
r ,
l I
402 k/ 1 learned something yesterday that helped me in that direction, 1 l
2 because it was clear from Ransom's history that they had the 3 need of the user in mind. I think that's-kept in the forefront 4 of their attitude, because it is very clear that it's the most 5 widely used code of-any of them, and I think it deserves 6 further support. I would like to see the MOD 3 completed, and I 7 don't think that should take a long time or be a major 8 expenditure.
9 I was also very glad to hear Lou talk about putting 10 some of these codes on the shelf, like TRAC. I was thinking of 11 putting them in a vault last time I heard about it. A shelf 12 and a vault are almost the same.
13 I must admit that the more we heard about their use 14 of data, the more it .vounded like TRAC all day. But I 15 recovered from that overnight and felt a little better about 16 them this morning, and I still feel pretty good about RELAP.
17 It is a good code, and with relatively modest 18 extension, it could be very good.
19 That's about all I have to say, Dave.
20 MR. WARD: Virgil.
21 MR. SCHROCK: Well, like Ivan, I am concerned about 22 the missing elements in the QA document. First of all, I 23 noticed there is a title discrepancy between this and the TRAC 24 one. This is called models and something else. I don't know 25 the reason for that. Maybe there is a simple explanation, but Heritage Reporting Corporation (202) 628-4888
403 p)
\~ 1 it seems to.me that if the CSAU methodology depends upon the 2 . existence of' good, solid QA documents, then there ought to be a 3 uniform method of identifying what-is a good, solid QA 4 -document.
5 And it does seem to me that that has to be a complete 6 description of all of the models that are contained in the 7 code, otherwise there will forever remain the suspicion that 8 there is something wrong with the result that comes out of 9 something that hasn't even been identified in the QA document.
10 Ivan mentioned a couple of things. Yesterday I.
11 mentioned the missing flooding models. We had quite a lot of 12, discussion about the flooding models in TRAC, and there was 13 quite a bit of concern about them. I think we need to know O 14 what the flooding models are in,RELAP5.
15 I am not saying that because I suspect they weren't 16 identified because of some deficiency in them, bis t if it is 17 true that there is something like 40 percent of the models not 18 accounted for in the present document, it's really not that 19 adequate.
20 Another item is the one of the part of the critical 21 flow model that has to do with horizontal stratification 22 upstream. I have dono some work on that which I had understood 23 was going to be introduced into these codes. And Gary Johnsen 24 told me something about an effort in the U.K. that had to do 25 with that. He then gave me a copy of a report entitled O Heritago Reporting Corporation (202) 628-4888
404
/~N
(/ 1 "Assessment of Horizontal Stratification Entrainment Model in-2 RELAPS/ MOD 2". And Gary suggested that I review this and give 3 him a critique on it.
4 I can't do that if I don't know what the-model is in 5 RELAPS. I don't have the model in RELAPS, and it comes as a 6 surprise to me that it's already in RELAPS if it's different' 7 from what we proposed in the NUREG report.
8 So that one, in my mind, needs a considerable amount 9 of clarification. But that's the only other specific item I 10 guess that I could cite at the moment.
11 Perhaps a listing of what all is included in this 12 additional 40 percent of the code might be enlightening for us, i 13 I don't think we have identified all the things, but certainly 14 we might want to know what they all are.
15 In any case, I have the firm view that the QA 16 document should be complete. It ought to have all of these 17 models.
18 That's all I have.
19 MR. WARD: Thanks, Virgil.
20 Charlie and Bill, give you the opportunity to say 21 anything you want?
22 MR. WYLIE: There was -- on what Ivan said about the 23 large break LOCA. I heard quite a bit yesterday about what 24 else was needed to make -- additional work to make it 25 applicable to the large break LOCA, and that's something that O Heritage Reporting Corporation (202) 628-4888
405 1 needs to.be identified.
2- MR. CATTON: We also heard that there was a 3 consortium that was going to take that, so that should be a' big 4 help to NRC.
5 MR. WARD: For what, RELTP57 t .
6 MR. CATTON: To essentially turn RELAPS into a large 7 break LOCA. I guess it was the German, Swiss, there was 8 another company, Northeast Utilities and also --
9 MR. ZUBER: U.K.
10 MR. CATTON: U.K. and Sweden. They are all very 11 interested in RELAPS, and I guess they are willing to put some 12 money where their mouth is.
13 So many of the needed modifications will be paid for 14 by these other people.
15 MR. SCHROCK: Well, Dave, I did have one other point, 16 end that is the idea of modifying correlations that are found 17 J.n the literature.
18 When we discussed this is Los A'amos, I expressed 19 seme rather strong views on that, and I haven't really changed 20 my mind about this process. I think it's presumptuous and 21 dangerous for code developers to have this philosophy that this 22 is a perfectly acceptable engineering approach to a problem.
23 It seems to me that it relates also to this peer 24 review aspect that Ivan discussed earlier. If in fact there'is 25 a reason for making a change in a correlation which can be b
a Heritage Reporting Corporation (202) 628-4888
406
()- 1 clearly justified as a result of the code development process, 2 it seems to me that documentation of the validity of that then 3 deserves to go into the literature where it can be again peer.
4 reviewed by people who are completely unconcerned with this 5 code, and who have an interest in the phenomenon which is 6 represented by the original correlations.
7 But when you end up with something more than 50 8 porcent of the constitutive relations, which are empirical 9 correlations taken out of the literature, and then modified in 10 some way in order to tune the code, not to sensitively test its 11 ability to represent the result that the experimenter that 12 produced that empirical correlation had, but instead the sense 13 the response of an integral code to some more global result in O
14 tnat integral code.
15 The gist is not a sound engineering practice, and 16 it's a practice that came out of these two decades of code 17 development in the NRC that I don't think is shared in other 18 fields. And I think it's not a good thing for the nuclear 19 industry to have this kind of a situation in existence.
20 I, for one, would like to see this aspect really 21 tightened up, and I don't know how it's going to get tightened 22 up at this stage, but it is not a good thing on the record of 23 nuclear power to have this kind of sloppy engineering approach 24 to problems that's really engrained in the system to the point 25 where people believe, honestly believe, I am not accusing O Heritage Reperting Corporation
- (202) 628-4888 i
407
) 1 anybody of bad faith here. I am saying people believe
~
~2 something that they ought not to believe, and that is that this 3 is an adequate engineering approach. It is not an adequate 4 engineering approach.
5 MR. WARD: Bill, did you have something you wanted to 6 say?
7 DR. KERR: I have nothing to add.
8 MR. WARD: Lou, I will give you a chance to make any 9 closing comment you want to make.
10 DR. SHOTKIN: I don't have anything to say except we 11 enjoyed the opportunity to present our programs and the QA 12 documents to the committee, and we appreciate all the comments 13 we heard.
14 MR. WARD: Okay, thank you, Lou.
15 That will end the reported part of the raeeting.
16 (Whereupou, at 3:10 p.m., the meeting was 1 17 concluded.)
18 19 20 21 22 23 24 25 l
Heritage Reporting Corporation ,
(202) 628-4888 l
, l REPORTER'S CERTIFICATE .
I I i i
2 i r i
3 DOCKET NUMBER: l l
4 . CASE TITLE: THERMAL HYDRAULIC PHENOfiENA SUBCO!1MITTEE 5 ']n HEARING DATE: April 20, 1988 6 , LOCATION: Idaho Falls, Idaho ;
!i 7 g i
! I hereby certify that the proceedings and evidence 8 q are contained fully and accurately on the tapes and notes r 9 i l reported by me at the hearing in the above case before the 10 l j UNITED STATES NUCLEAR REGULATORY CO!!!!ISSION.
11
. 12 l, 13 Date: April 20, 1988 1 ,
14 i .
i 15 i
- i IO 'oon, u4 _
Official Reporter j- ,7 HERITAGE REPORTING CORPORATION 18 r 1220 L Street, N.W.
- - Washington, D.C. 20005 i ,
['
19 i
i 20 i i
21 l
l 22 !
t
?
! 23 1 i
o "
25 l
s ,
l Heritoge Reporting Corporation l
- =
O O *'O PRESENTATION TO THE ACRS SUBCOMMITTEE ON THERMOHYDRAULIC PHENOMENA ON REVISION OF ECCS RULE CONTAINED IN APPENDIX K AND SECTION 50.46 OF 10 CFR PART 50
"""a"A^""
APRIL 20,1988
O O O ECCS RULE HISTORY
- CONCERN OVER THE ABILITY OF ECCS TO PROTECT THE REACTOR DURING A LARGE BREAK LOCA PEAKED AFTER A TEST IN THE SEMISCALE FACILITY RESULTED IN ALL INJECTED ECC LIQUID BEING EXPELLED FROM THE VESSEL.
- INTERIM ACCEPTANCE CRITERIA IN 1971
- LACK OF A TECHNICAL BASIS FOR INTERIM. ACCEPTANCE CRITERIA RESULTED IN A CHALLENGE AND LENGTHY PUBLIC HEARINGS.
l
- FINAL ACCEPTANCE CRITERIA (1974) OF 10 CFR 50.46 AND l APPENDIX K (ECCS RULE) l l
1
- - - aa__.- _ _ _ =- - _ _ _ - _ _ _ _
l ..
.C) e O .
! PRESENT ECCS RULE .
i~
i j
- 10 CFR 50.46 REQUIRES CALCULATIONS OF ECCS PERFORMANCE FOR j A SPECTRUM OF BREAK SIZES USING THE METHODS OF APPENDIX K.
BOTH APPENDIX A (GDC) AND APPENDIX K REQUIRE CONSIDERATION OF A SINGLE FAILURE, WHILE APPENDIX A ALSO REQUIRES CONSIDERATION OF LOSS OF EITHER ONSITE OR OFFSITE POWER. '
l
1 i
HYDROGEN < 1%
i -
COOLABLE GEOMETRY LONG-TERM COOLING
- APPENDIX K CONTAINS REQUIRED FEATURES OF EVALUATION MODELS
- (E.G., 1.2 TIMES 1973 ANS DECAY HEAT) AND ACCEPTABLE l FEATURES (E.G., MOST HEAT TRANSFER CORRELATIONS)..
l -
- APPENDIX K REQUIRES DOCUMENTATION OF EVALUATION MODELS OF l SUFFICIENT DETAIL SO AS TO REQUIRE AN AMENDMENT FOR SIGNIFICANT CHANGES (DEFINED AS A CHANGE THAT WOULD RESULT l IN A 200F CHANGE IN PCT). _ _ _ _ _ _ _ _ _ _ _ _ __ __ __
o o
~
. . . O APPLI_ CATI _QN OF ECCS RULE
- CURRENT EVALUATION MODELS CONTAIN LARGE CONSERVATISMS DUE TO:
APPENDIX K REQUIRED FEATURES (E.G., 1.2 X ANS DECAY -
HEAT)
STAFF REQUIREMENTS (E.G., CORRELATIONS THAT BOUND ALL OR MOST DATA)
VENDOR CONSERVATISMS (E.G., SIMPLIFIED MODELS)
- MOST LICENSEES CALCULATE PCT VERY NEAR 22000F, AND SOME PLANTS HAVE RESTRICTIVE POWER PEAKING FACTORS (OR HAVE EVEN BEEN DERATED) TO MEET THE 22000F LIMIT.
- FUEL IMPROVEMENT AND RESEARCH HAS RESULTED IN REDUCTION IN OTHER LIMITS (E.G., DNB) , MAKING LOCA LIMITS THE MOST RESTRICTIVE.
- PTS, FUEL MANAGEMENT, MANEUVERING, AND POWER UPGRADES REQUIRE INCREASED PEAKING FACTORS, WHILE ERROR CORRECTIONS AND MODEL CHANGES OVER THE YEARS HAVE RESULTED IN TIGFITER LOCA LIMITS.
! O O O i
l APPLICATION OF ECCS RULE -- (CONT'D) l .
l i
THE NET RESULT IS THAT LOCA LIMITS ARE LIMITING PLANT OPERATION FOR SOME VENDOR DESIGNS. -
INTERPRETATION OF DOCUMENTATION REQUIREMENTS HAVE RESULTED IN A REQUIREMENT TO REPORT AND IMMEDIATELY RESOLVE ANY l ERRORS IN EVALUATION MODELS OVER 200F. THIS USUALLY MEANS IMMEDIATE PLANT RESTRICTIONS AND LONG HOURS FOR NRC i
AND LICENSEE STAFF UNTIL A POSSIBLY TRIVIAL PROBLEM CAN BE SOLVED.
l l
1 I
l . .
I O O o i ECCS RESEARCH AND MORE RECENT RULEMAKING HISTORY i
i l
l FROM 1975 UNTIL THE 1979 TMI-2 ACCIDENT. ONE MAJOR ECCS -
l RESEARCH PROGRAM (2D/3D) IS STILL UNDERWAY.
- NRC HAS SPENT OVER $700M ON ECCS RESEARCH AND THE TOTAL ,
l COST IS WELL OVER $1 BILLION WHEN DOE, INDUSTRY AND FOREIGN :
RESEARCH IS INCLUDED.
- AS EARLY AS 1976, IT WAS APPARENT FROM RESEARCH THAT APPENDIX K METHODS WERE HIGHLY CONSERVATIVE AND DID NOT ACCURATELY REPRESENT ACTUAL PLANT BEHAVIOR DURING A LOCA.
l i
l _ . _ . _ _ - _ - . _ _ _ . _ _
O O O ECCS RESEARCH AND MORE RECENT RULEMAKING HISTORY
-(CONT'D1
- ANPR 1978 PROPOSED LIMITED PHASE I, FOLLOWED BY MORE COMPREHENSIVE RULE REVISION AT SOME LATER DATE. <
- THE TMI-2 ACCIDENT IN 1979 RESULTED IN A DELAY IN THE ECCS RULE REVISION, AND A MAJOR REDIRECTION OF ALMOST-COMPLETED LARGE BREAK LOCA RESEARCH.
- ECCS RULE REVISION AGAIN REVIVED IN 1982 IN CONTEXT OF STREAMLINING THE REGULATORY PROCESS.
- SECY-83-472 t m_ w --r- _ _ _ ___ _ _ _ _ _ _ _ _
[ O SECY-O-472 O IN RESPONSE TO REQUESTS BY GE FOR RELIEF FROM LOCA RESTRICTIONS ON BWRS, THE STAFF ADOPTED A METHOD TO MAKE USE OF IMPROVED ECCS KNOWLEDGE WITHIN THE CONTEXT OF THE EXISTING ECCS Rule.
l THE STAFF WOULD ALLOW AN EVALUATION MODEL THAT WAS TOTALLY BEST ESTIMATE EXCEPT FOR FEATURES SPECIFICALLY REQUIRED BY APPENDIX K (E.G., ALL STAFF-IMPOSED CONSERVATISMS REMOVED).
j
- AS A CONDITION FOR THIS EVALUATION MODEL APPROVAL, A SECOND l TOTALLY BEST ESTIMATE CALCULATION WAS REQUIRED, ALONG WITH j AN EVALUATION OF THE UNCERTAINTY OF THE BEST ESTIMATE I
CALCULATION. THE UNCERTAINTY EVALUATION MUST SHOW THAT REQUIRED APPENDIX K CONSERVATISM WAS SUFFICIENT TO COVER 3 THE UNCERTAINTY IN THE BEST ESTIMATE CALCULATIONS.
}
! 22000F > EM (BE + APPENDIX K) > BE+ ESTIMATED UNCERTAINTY i
l _______ _ - . .. .. . _ _ _ - _ _ _ _ - . _ ._.
1 l
O SECY-83-472 O(CONT'D) i i
1 STAFF VIEWED THIS APPROACH FAVORABLY SINCE MORE REALISTIC -
{ CALCULATIONS ARE CONSIDERED BENEFICIAL TO SAFETY AS WELL AS l BENEFICIAL TO THE PLANT. ~
}
I
- DISADVANTAGES ARE:
l 4
I TWO CALCULATIONS MUST BE PERFORMED AND REVIEWED i
l l -
WITHOUT A BASIS IN RULEMAKING, THE SECY-83-472 APPROACH .
IS SUBJECT TO LEGAL CHALLENGE.
\ .
i
- 1. - . .. . -- - . - -.-. . .
O e O
- CURRENT RULEMAKING l
- 1984 COMPARISON OF BEST ESTIMATE RESULTS (TRAC-P AND l
TRAC-B) FOR TYPICAL PLANTS FROM THE FOUR VENDOR'S
, EVALUATION MODELS AND EVALUATION MODELS WITH MODIFICATIONS TO REMOVE APPENDIX K CONSERVATISMS SHOWED THE FOLLOWING:
CONSERVATISM IN CURRENT EMS VARIED GREATLY BUT WAS GENERALLY LARGE (10000F)
DECAY HEAT WAS THE MAJOR CONSERVATISM IN TP= EMS (300-4000F)
REMOVING THE DECAY HEAT CONSERVATISM USUALLY RESULTED IN SHIFTING THE TIME OF PCT FROM THE REFLOOD PHASE TO THE BLOWDOWN PHASE OF THE LARGE BREAK LOCA.
MOST APPENDIX K CONSERVATISM IS DURING REFLOOD. ONCE THE TIME OF PCT IS SHIFTED TO THE BLOWDOWN PHASE, EMS MAY NOT HAVE SUFFICIENT CONSERVATISM TO COVER THE UNCERTAINTY IN.THE CALCULATION.
u . ._ _ . .
O CURRENTRU$ MAKING (CONT'D)
I
- Bt. SED UPON TFiE ABOVE FINDINGS, SIMPLY REFINING APPENDIX K BASED ON CURRENT KNOWLEDGE (REVISE DECAY HEAT) WITHOUT FURTHER CONSIDERATION OF UNCERTAINTY WAS NOT VIEWED AS 4 ACCEPTABLE. THIS, COMBINED WITH FAVORABLE EXPERIENCE WITH l SECY-83-472, RESULTED IN AN APPROACH TO RULEMAKING CONSISTENT WITH THE SECY-83-472 METHOD.
i i
l i
L_
i NOTICE OF PROPbED RULEMAKING O i
l l
NPRM ISSUED MARCH 3,1987 PROPSED RULE DRAFT REGULATORY GUIDE DRAFT COMPENDIUM OF ECCS RESEARCH (NUREG-1230) l l
PUBLIC COMMENT PERIOD ENDED JULY 1,1987
- 33 COMMENTS RECEIVED - GENERALLY FAVORABLE, RESULTED IN NO CHANGES TO THE PROPOSED RULE BUT SOME MODIFICATIONS TO THE REGULATORY GUIDE.
i
)
CURRENTLY SCHEDULED TO PUBLISH FINAL RULE AUGUST 1,1988.
i l
l l
\ .- - - -
O KEYFEATURESOF$ROPOSEDRULE -
- VERY GENERAL RULE REQUIRES "REALISTIC" CALCULATIONS OF ECCS PERFORMANCE WITH CONSIDERATION OF UNCERTAINTY SUCH THAT THERE IS A "HIGH LEVEL OF PROBABILITY" THAT CRITERIA WOULD NOT BE EXCEEDED. CRITERIA (E.G., 22000F) UNCHANGED.
- ALTERNATIVELY, APPENDIX K FEATURES REMAIN ACCEPTABLE AS AN OPTIMAL METHOD NOT REQUIRING AN UNCERTAINTY EVALUATION (GRANDFATHERING EXISTING EMS AND ALLOWING FUTURE USE OF APPENDIX K). -
- l. o o o i KEY FEATURES OF PROPOSED RULE (CONT)
- REPORTING REQUIREMENTS SPECIFICALLY STATED TO RELAX AND CLARIFY
,1 CURRENT INTERPRETATIONS OF APPENDIX K DOCUMENTATION REQUIREMENTS.
i
, e
' ALL ERROR CORRECTIONS AND CHANGES REPORTED ANNUALLY ALL ERRORS AND CHANGES OR CUMMULATIVE (ABSOLUTE VALUE) ERRORS RESULTING IN GREATER THAN 50 F PCT CHANGE REPORTED WITHIN 30 DAYS.
NO FURTHER ACTION REQUIRED BEYOND REPORTING. STAFF MAY TAKE ACTION CONSISTENT WITH SAFETY S!GNIFICANCE. j f
- ERRORS OR CHANGES RESULTING IN EXCEEDING CRITERIA (E.G.,2200 F) ,
i i ARE IMMEDIATELY REPORTABLE (4 HOURS) AND IMMEDIATE STEPS OF
! BEING IN COMPLIANCE MUST BE PROPSED (CONSISTANT WITH 10 CFR50.55(e),
50.72 AND 50.73).
1 i -
- O KEYFEATURESOFPROP$EDRULE (CQNT'D) i l
- AFTER $700M OF RESEARCH, THE ONLY NONCONSERVATIVE FEATURE OF APPENDIX K FOUND IS THE DOUGALL-ROHSENOW CORRELATION.
THUS, THIS PREVIOUSLY ACCEPTABLE CORRELATION WAS REMOVED FROM APPENDIX K.
I EVALUATION MODELS APPROVED PRIOR TO THE RULE REVISION THAT !
j MAKE USE OF DOUGALL-ROHSENOW WOULD CONTINUE TO BE ACCEPTABLE AS LONG AS THE CURRENT LEVEL OF OVERALL CONSERVATISM IN THE EM IS NOT SIGNIFICANTLY CHANGED. A SIGNIFICANT CHANGE IN OVERALL CONSERVATISM IS DEFINED AS CHANGES AND/OR ERROR CORRECTIONS OF WHICH Ti1E "NET EFFECT"
- (PLUS AND MINUS) WOULD RESULT IN A 500F DECREASE IN PCT FROM THAT WHICH WOULD HAVE BEEN PREVIOUSLY CALCULATED. ,
}
l
O O O REGULATORY GUIDE
- DEFINES HIGH LEVEL OF PROBABILITY STATED IN THE RULE AS 95%
PROBA BILITY
. PROVIDES GENERAL FEATURES EXPECTED IN BEST ESTIMATE CALCULATION AND THE UNCERTAINTY EVALUATION, BUT DOES NOT SPECIFY OR PRESCRIBE ANY SPECIFIC METHODS.
. LISTS A LIMITED NUMBER OF ACCEPTABLE BEST ESTIMATE MODELS (E.G.,1979 ANS/ ANSI DECAY HEAT, CATHCART-PAWEL METAL-WATER REACTION RATE).
. IN SOME CASES, DATA APPROPRIATE TO MODEL ASSESSMENT IS PROVIDED.
KRAFT COMPENDIUM 0F ECCf RESEARCH (NUREG 12301
- MAJOR
SUMMARY
/ BIBLIOGRAPHY (1200 PAGES) OF EXTENSIVE ECCS RESEARCH
- CONTAINS A DESCRIPTION (AND WILL LATER CONTAIN RESULTS) OF NRC'S METHODS OF EVALUATING THE UNCERTAINTY OF NRC'S CODES. THIS IS PROVIDED AS A PROOF OF PRINCIPLE, AN AUDIT TOOL, AND AN EXAMPLE OF HOW TO PERFORM AN UNCERTAINTY EVALUATION, BUT IS NOT A REQUIREMENT OR AN ACCEPTABLE METHOD IN THE REGULATORY GUIDE SENSE.
l :
O o o j :
i ECCS COMPENDIUM i
l ISSUED AS DRAFT NUREG-1230 (APR 87)
WIDE FOREIGN & DOMESTIC DISTRIBUTION l ONLY A FEW COMMENT LETTERS RECEIVED i
i NOT PLANNING TO MAKE MAJOR EDITS WOULD BE HELPFUL TO RECEIVE COMMENT FROM ACRS CONSULTANTS, ALTERNA- -
l TIVE IS TO DIG THRU TRANSCRIPTS CSAU SECTION (CHAPTER 4) WILL REFER j TO RESULTS IN A SEPARATE REPORT l PLANNING FOR JULY 1988 DISTRIBUTION i
l OF MAIN REPORT, WITH APPENDICES ;
! 1
. TO BE ISSUED AS VOL 2 IN SEPT l
. _ =. .
(
1 1
t O o o I
ECCS RULE PACKAGE
! - TRANSMITTAL LETTER TO NRR, OGD, ARM l TRANSMITTAL LETTER TO CRGR i COMMISSION PAPER (DELETED COMMENT
SUMMARY
)
- ENCLOSURES A. NOTICE OF PROPOSED RULEMAKING :
B. MEMORANDUM CHILK TO STELLO, DTD 1/9/87 C. ACRS AND PEER REVIEW SUMMARIES D.
SUMMARY
OF PUBLIC COMMENTS ON RULE i E. NOTICE OF FINAL RULEMAKING ...
F. DRAFT REGULATORY GUIDE . . .
G.
SUMMARY
OF PUBLIC COMMENTS ON GUIDE H. ENVIRONMENTAL ASSESSMENT .
- 1. REGULATORY ANALYSIS . ..
i J. MARGIN INHERENT IN 2200 F LIMIT i
l . - _ _ _ _ _ _ _ _ .. . . . . . - . . - . . _ - - - _ _ _ . _ _ _ _ - . _
O o o ECCS RULE PACKAGE STATUS RES CONCURRED
- PACKAGE TRANSMITTED TO NRR, OGC, ARM FOR CONCURRENCE
- PACKAGE TRANSMITTED TO CRGR FOR INFORMATION SUGGESTED NO FURTHER CRGR REVIEW REQUIRED
- NO SUBSTANTIVE CHANGE TO RULE
- MODIFICATIONS TO REG GUIDE
- RES UNCERTAINTY METHODOLGY HAS BEEN REVIEWED BY ACRS AND EXTERNAL PEER REVIEW GROUP CHAIRED BY NEAL TODREAS OF M.I.T. ACRS RESERVES RIGHT TO FINAL APPROVAL UNTIL UNCERTAINTY METHODOLGY DEMON-STRATION IS COMPLETE
O O O i
i l
ENCLOSURE C-UNCERTAINTY METHODOLOGY
.l ACRS COMMENTS '
i PEER REVIEW CHAIRMAN (TODREAS) REPORT l
INDIVIDUAL PANEL REPORTS l
l
j
~
O o o .
! ENCLOSURE D !
SUMMARY
OF PUBLIC COMMENTS GENERALLY SUPPORTED NEW ECCS EVALUATION APPROACH i
l SUPPORTED GRANDFATHERING OF CONSERVATIVE APPENDIX K METHOD l l (ACRS QUESTION) l j "
DID NOT FEEL THAT EXPLICIT DEGREE OF CONSERVATISM NEED BE SPECl-FIED ( ASSELSTINE QUESTION)
DID NOT NEGATE COMMISSION'S CONTENTION THAT INCREASED FISSION PRODUCTS WITH POWER OPERATING IS NEGLIGIBLE WHEN COMPARED TO UNCERTAINTY OF FISSION PRODUCT RELEASE (ASSELSTINE QUESTION)
INDEPENDENT REVIEW OF TECHNICAL BASIS SHOULD BE BY ACRS (ASSELSTINE QUESTION) :
I GRANDFATHERING OF DOUGALL-ROHSENOW GENERALLY SUPPORTED i
SUPPORTED USE OF UNCERTAINITY EVALUATION MANY COMMENTS ON REPORTING REQUIREMENTS ,
l u
O o o j EU_MMARY OF PUB _LI_C C_0_MMENTS (cpNT'D1 l
l
- SUPPORTED MAINTENANCE OF EXISTING ACCEPTANCE CRITERIA i
- SUGGESTED BROADENING OF RULE TO INCLUDE ZIRCONIUM-BASED ALLOYS J
t
- BROADENING OF RULE SCOPE j SGTR CONSEQUENT TO LBLOCA LONG TERM DECAY HEAT REMOVAL SHOULD BE ADDRESSED i
- RULE IS NOT LEGAL BECAUSE IT BASED SOLELY ON COST-SAVINGS CONSIDERATIONS i
- EXPERIMENTAL BASIS FOR RULE IS -INADEQUATE BECAUSE FULL SCALE ECCS BYPASS DATA IS NOT YET AVAILABLE i
O o o ENCLOSURE E-FEDERAL REGISTER NOTICE WITH FINAL RULE CHANGES SINCE 12-87 P.4 DISCUSSION OF GRANDFATHERING APPENDIX K DOES NOT MENTION SAFETY BENEFITS OF NEW RULE, AND STATES THAT THE EFFECT ON PUBLIC RISK IS NEGLIGIBLE P.5 ADDED REFERENCE TO COMPENDIUM
! P.11 (ITEM 10) CLARIFIED THAT RULE ONLY APPLIES TO ZlRCALOY CLAD FUEL REACTORS P.11 STATE THAT ANY ACTION ON CONCURRENT SGTR WILL AWAIT PRIORITIZATION OF GI-141, AND SHOULD BE ADDRESSED THROUGH THE S.G. INSPECTION PROGRAM 43
l
! O o o
) ENCLOSURE F-ECCS REGULATORY GUIDE i
CHANGES i
1
) MERGED APPENDICES A & B. RESTRUCTURED NEW APPENDIX TO HIGHLIGHT MODELS, CORRE-l LATIONS, AND DATA WHERE APPLICABLE 1 !
l '
l P.18 1.2.4.- ADDED SENTENCE TO ACCOUNT FOR i PHYSICAL / CHEMICAL CHANGES IN IN-CORE, EX-FUEL MATERIALS l
MADE MODIFICATIONS BASED ON COMMENTS I
l i
i O O O l' ENCLOSURE G-
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE
- NRC RECEIVED NINE RESPONSES I
- SOLICITED COMMENTS ON- j
- 1. SHOULD THE REGULATORY GUIDE LIST MODELS, DATA, I l AND MODEL EVALUATION PROCEDURES THAT THE NRC CONSIDERS TO BE ACCEPTABLE FOR REALISTIC CALCULATIONS OF ECCS PERFORMANCE?
! 2. ARE THE MODELS, DATA, AND MODEL EVALUATION PROCEDURES LISTED IN APPENDIX A TO THE DRAFT REGULATORY GUIDE APPROPRIATE?
DOES APPENDIX B OF THE DRAFT REGULATORY GUIDE i 3.
APPROPRATELY DISCUSS THE FEATURES OF THE BEST-ESTIMATE CODES AND THE ESTIMATION OF i OVERALL CALCULATIONAL UNCERTAINTY?
l
kNCLOSURE G-SUMMARf0F PUBLIC COMMENTS O
) ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUE)
- 1. COMMENT- EXTEND COMMENT PERIOD RESPONSE: LATE COMMENTS CONSIDERED i
Q.1 COMMENTS
) 2. COMMENT- IN ANSWER TO Q 1., COMMENTER SAID REG GUIDE
! OVERLY PRESCRIPTIVE l RESPONSE: NOT SO
- 3. COMMENT- BE METHODOLGY BURDENSOME RESPONSE: THATS WHY APPENDIX K IS GRANDFATHERED l
l 4. COMMENT- NRC SHOULD DO MORE FOR INDUSTRY IN QUANTIFYING I UNCERTAINTY
! RESPONSE: NO i
~
I ENCi_OSURE G-SUMMAO OF PUBLIC COMMENTSO ON REGULATORY GUIDE AND NRC RESPONSE l (CONTINUE)
! 5. COMMENT- EXPLICITLY STATE THAT ALTERNATE MODELS, DATA AND !
CORRELATIONS ARE ACCEPTABLE RESPONSE: IT IS CLEARLY STATED IN APPENDIX l
Q.2 COMMENTS r
- 6. COMMENT- DISAGREED WITH LAMINAR FLOW H.T. RECOMMENDATION RESPONSE: DELETED REFERENCE IN REG GUIDE TO AVOID CONFUSION
- 7. COMMENT- RULE AND REG GUIDE SHOULD SPECIFY ZlRCONIUM . ,
BASED ALLOYS NOT ZlRCALOY .
RESPONSE: SUBJECT OF ZlRCALOY VS. OTHER ALLOYS !
SHOULD BE THE SUBJECT OF SEPARATE RULE MAKING
{ .
l O O O
{
l ENCLOSURE G -
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUE)
- 8. COMMENT- REGARDING APPLICABILITY OF CATHCART-PAWEL RESPONSE: AFTER CONSULTING WITH LABORATORIES, MODIFIED REG GUIDE SO CATHCART-PAWEL ONLY PRE-AP PROVED ABOVE 1900 .
i Q. 3 COMMENTS '
- 9. COMMENT- REG GUIDE INSUFFICIENTLY DEVELOPED REGARDING UNCERTAINTY I
RESPONSE: GUIDANCE PURPOSELY GENERAL SO THAT APPLICABILITY TO A WIDE RANGE IF COMPUTER CODES IS RETAINED l
l I i l
j .
l O o o h
l ENCLOSURE G -
SUMMARY
OF PUBLIC COMMENTS ,
ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUE)
- 10. COMMENT- CONSERVATIVE COMBINATIONS OF INPUT PARAMETERS i I SHOULD BE SPECIFICALLY ALLOWED RESPONSE: THIS IS ALLOWED AS LONG AS CONSERVATISMS ARE NOT MASKING COMPENSATING ERRORS ,
- 11. COMMENT - CONFIDENCE LEVELS NOT DISCUSSED OR SPECIFIED...
i RESPONSE: MODIFIED REG GUIDE TO EXPLAIN THAT CONFIDENCE .
LEVELS NOT NECESSARILY REQUIRED l
\
- 12. COMMENT- QUANTIFYING UNCERTAINTY SHOULD NOT BE REQUIRED RESPONSE: DISAGREED, BUT NOTED THAT QUANTIFICATION NEED NOT ,
CONFORM TO RIGOROUS STATISTICAL PROCEDURES .
l l
O o o ENCLOSURE G -
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUE)
- 13. COMMENT -
UNCOVERED BUNDLE HEAT TRANSFER SHOULD ACKNOWLEDGE PRESENCE OF LIQUID DROPS.
RESPONSE: ENTRAINMENT ADDRESSED ELSEWHERE. GUIDE MODIFIED TO EXPLAIN CONDITIONS OF UNCOVERED HEAT TRANSFER
- 14. COMMENT -
DO NOT NEED TO ACCOUNT FOR BURNUP DURING ACCIDENT RESPONSE: EXPLAINED THAT THIS APPLIES TO STORED ENERGY
- 15. COMMENT -
ASSYMETRIC CLADDING SWELLING TOO DIFFICULT AND BURDENSOME RESPONSE: RECOGNIZED DIFFICULTY, ONLY REQUIRE CONSIDERATION AS IT AFFECTS UNCERTAINTY
- 16. COMMENT -
NODING STUDIES NOT NEEDED IN AREAS OF THERMAL NON-EQUILIBRIUM RESPONSE: REG GUIDE MODIFIED TO CLARIFY INTENT REGARDING BREAK AND INJECTION LOCATIONS
O o o ENCLOSURE G-
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE j (CONTINUED) l
- 17. COMMENT - FLOW BLOCKAGE NEED NOT BE CONSIDERED l
l RESPONSE: IF APPLICANT CAN SHOW THAT EXPLICIT MODELING NOT REQUIRED, THAT IS ACCEPTABLE
- 18. COMMENT - DATA LIMITATION SHOULD BE STATED SO THAT THIS l
IS NOT DOUBLY ACCOUNTED FOR IN UNCERTAINITY f
RESPONSE: DOUBLE ACCOUNTING NOT REQUIRED IF SUCH CAN BE DEMONSTRATED
- 19. COMMENT - FINITE BREAK OPENING SHOULD BE ALLOWED RESPONSE: "LEAK BEFORE BREAK" IS NOW BEING CIRCULATED FOR PUBLIC COMMENT AS APPLIED TO LOCA :
ANALYSIS AND OTHER ISSUES. THEREFORE, CONSIDERATION HERE IS PREMATURE.
- 20. COMMENT - BYPASS CONSIDERATIONS SHOULD BE EXPLICITY NOT REQUIRED FOR SMALL BREAKS RESPONSE: DEMONSTRATION OF INAPPLICABILITY IS APPLICANTS ,
RESPONSIBILITY .
O O O ENCLOSURE G-
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUED)
- 21. COMMENT - DEMONSTRATION OF "BEST ESTIMATE" CAPABILITY IS TOO BURDENSOME RESPONSE: THATS WHY APPENDIX K IS GRANDFATHERED .
- 22. COMMENT - MORE SPECIFICITY NEEDED REGARDING EQUIPMENT AVAILABILITY IN ANALYSES RESPONSE - APPLICANT BEST QUALIFIED TO DO THIS
1 O o o ENCLOSURE H - ENVIRONMENTAL ASSESSMENT NO CHANGES l
)
i ENCLOSURE I - REGULATORY ANALYSIS l
! CHANGE SINCE 12-87 l P. 6 DELETED DISCUSSION OF 400*F MARGIN. NOTE THAT ;
EFFECT ON PUBLIC SAFETY IS DEEMED NEGLIGIBLE i
' ENCLOSURE J - 2200*F INHERENT MARGIN NEW ENCLOSURE DISCUSSES RECENT INFORMATIN REGARDING FUEL, I CLADDING, CONTROL ROD BEHAVIOR IN THE 2200*F RANGE. CONCLUDES THAT MARGIN ABOVE 2200*F NOT SIGNIFICANT TO RULE 4
i t
I. ___- - , , . . . _ _ . . - . _ _ _ . _ _ _
.9 S A cct # z O
4 t
REVIEW 0F RmjRE PLANS FOR l PES T/H RESEARCH LOUIS M. SH0TKIN, USNRC O
ACRS T/H NBTEMA SLECCf'EITTEE IDAHO FALLS, APRIL 19, loES O
/ ,
O DELINE
- 1. PLAffT PERFORPAt!CE PLANS: FESEARCH AND APPLICATICfS
- 2. SOE NEW APPLICATICt!S PROJECTS
- 3. FES/NRR/AEOD TASK GFOUP CN FUTUPE T/H FESEARQ1
- 4. FY90 BUDGET EXEPCISE O
O
~_ . __ __ _ _ _ __ __. _ _ _ .
1, PLAtH PERFORMANCE PLANS RESEARCH AND APPLICATIONS I !
i 88 89 90 91 92 i ROSA-IV 'mmM"Nk"dEWEE0b"m' 2D/3D l
~ h
< , with industry ,
Q ,
c ICAP and .
O
, , l Code Consortia ,
c: ,
w C TRAC-B (MOD 2) 'm " m " ' ;
DICTATED M
W g
RELAP5 (MOD 3) l"mmu BY REGULATORY 8 TRAC-P (MOD 2) y, lmmuu, CSAU , l O 8^ sic siuo us ECCS RULE !
In-Vessel Acc.Mgmt.
(Software Expert mtg) s i""
(
(
CONTAINMENT .
m '
z
,g GENERIC ISSUES =
mummmmmmmu,
< priorityissues o TECH SUPPORT g CENTER o.
< NUCLEAR PLANT AEOD(TCC) & NRR ANALYZER AUDIT INDUSTRY SUBMITTALS ADV. REACTORS l
/
O I
- 2. SC E NEW APFtlCATIONS FFCJECTS CChTAltmT ISSUES FECUEST FRCM EPR
- GENERIC ISSUES FECUESTS FPm PES:
SGTR TUFBit:E TRIP W/0 FEACTCR TRIP
- LASALLE ANALYSIS FOR AECD
- NPA AT c TECFAIC/1 TFAINIt.'G CEhTER g
IN-\ESSEL ACCIDENT FANACEFHiT PLAN 1
l l
9
O V
- 3. Background and Charter for RES/NRR/AEOD Task Group on .
Future T/H Thermal Hydraulic Research I. BACKGROUND A. Declining Emphasis on T/H Research 7 -) All major thermal-hydraulic research programs are planned to be k> completed by tha end of CY 1991. There will be a long-term declining emphasis and resources on T/H code development and assessment. A solid base program effort will be maintained in this area to: make necessary code improvements based on new information; retain a cadre of experts; and address new issues as necessary. A T/H Technical Support Center (TSC) has been formed to efficiently accomplish both immediate and long-term goals.
This solid base program will include codes, basic studies, the T/H TSC, the Nuclear Plant Analyzer, and a significant applications effort addressing priority issues. The Task group is being asked to help formulate the "codes" portion of this base program.
B. T/H Codes and Their Uses Within NRC The Task Group should consider future agency needs for the following ,
T/H Codes: TRAC-PWR, jRaC-BWR, RELAP5, COBRA-NC (which includes
-')
COBRA-TF and COBRA / TRAC) and RAMONA-3B.
O These codes have been used to analyze both generic and specific transients and accidents that have cccurred in light water reactors (LWR's). These include loss-of-Coclant Accidents (both small-break and large-break LOCA's), ATWS and a full complement of operational transients. Enclosure 1 provides a reasonably complete list of these analyses performed under NRC funding with TRAC and RELAP.
The codes ability to calculate LWR transients has been comprehensively assessed against data from several scaled experimental facilities specifically designed to address the accident scenarios being analyzed.
Finally, these codes are used as the "front-end" of mechanistic severe accident codes (RELAP/SCDAP and TRAC-MELPROG) and also provide the boundary conditions for fuel codes such as FRAP.
O C. RES Close-out Plans for T/H Codes and Their Major Capabilities YEAR FOR CODE FINAL '.'ERSION MAJOR CAPARILITIES RELAP5 1989 1-0 PWR Accidents TRAC-PWR 1989 3-D PWR Accidents TRAC-BWR 1988 3-D BWR Accidents COBRA-NC 1987 Containment (Also,sub-channel)
RAMONA 1B 1987 BWR Accidents with Neutronics 9
i O :
i i D. RES Close-out Plans for T/H Test Facilities and Their Major l
Capabilities i-TEST YEAR FOR !
FACILITIES FINAL VERSION MAJOR CAPABILITIES i
- t LOFT 1987 PWR Accidents with Nuclear Fuel l SEMISCALE 1986 PWR Accidents FIST 1985 BWR Accidents i t
MIST 1988 B&W Accidents t 20/3D 1989 PWR LOCA (Full Scale) i r
ROSA-IV 1992 PWRAccidents(largescale) j ICAP (BETHSY, '
LOBI, GEST-GEN) 1991 PWR Accidents E. RES Current Maintenance Plans for T/H Codes f i
i MAINTENANCE jr CODE PLAN REASONS RELAPS 2-3 MY/Y Widely used TRAC-PWR 0-3 MY/Y As use dictates I TRAC-BWR Put on Shelf Little used COBRA-NC Put on Shelf Used Occasionally i RAMONA-3B Put en Shelf Used Occasionally l 110te that user conveniences will still be provided for RELAPS through its use on the Nuclear Plant Analyzer.
I i
f O !
_,_._-, _ _, _... _ _ . .. D
1 O
II. TASK GROUP CHARTER A. Discuss Close-out Plans for Final Code Capabilities
- 1. any capabilities left out (neutronics, secondary-side, containment)?
11, what criteria should be used to terminate code improvement?
iii. what level of experimental studies is appropriate to meet future needs?
B. Discuss Code Maintenance Plans
- i. which computer codes should be maintained for future agency use?
ii. what activities should code maintenance cover?
- a. for a code still under active use O
- b. for a code that is little used iii. what code user conveniences should be provided?
C. Draft recommendations on T/H code plans (including code maintenance) and submit to Reactor Systems Safety Senior Research Program Steering Group by mid-May, 1988
- 1. prioritize recomendations ii. state basis for recommendations Enclosure 1: Plant Transients Analyses Sumary, 7/30/87 i
l O
'j O ;
1
. 4. 'FY90 EUDCET EXERCISE (M$)
PROJECT FY88 $ $ _9.1 g T/H RESEARCH ROSA-IV 0.4 0,5 0,5 0,5 0,5 2D/3D 2,0 1,3 1.2 -- --
B8W TESTING 1.6 3,6 1,0 0,5 --
ICAP 0.7 0,7 0.8 0,5 --
T/H CCCES 1,8 1,7 2,0 2.0 2.0 CSAU 1,4 1,0 0,3 0,2 0,2 0,8 1,0 BASIC STUDIES 1.0 1.0 1.0 T/H RESEARCH TOTAL 8,9 9,6 6,8 4,7 3,7 REGULATORY APPLICATigLS TECH SUPPCRT CENTER 0,8 1,5 2.0 2.0 2.0 CONTAINfM 0.2 0,7 0.8 0,5 0,5 NPA/DB/SItifLATORS 1,0 0,9 1,0 1,0 0,8
[
. ADVANCED LWR'S --- 0,2 1,5 1,0 1,0 REG!'LATORY APPL, TOTAL 2.0 3,3 5,3 4,5 4,3 TOTAL T/H RES 8 REG, APPLICATION 10,9 12,9 11,3 8,4 7,2 O
,- IN-VESSEL ACC, MANAGB W 0,7 1,4 1.6 1,8 1,8
-. , , - .-_._ _ __ _- ,,_,_ _ ..... , _ _.-~,._ _ ._____ .,. _ ,_.. -_._ -.__-___.. _ _ s
O o *b ECCS RULE: UPDATE ON STATUS RULE PACKAGE e SIGNED 0FF BY RES OFFICE DIRECTOR i
e KEY NRR STAFF BRIEFED e DRAFT COMMENTS RECEIVED FROM ARM (RULES AND RECORDS) AND INCORPORATED e SCHEDULED FOR ISSUANCE IN AUGUST 1988
O O O ECCS RULE: UPDATE ON STATUS RULE i o REMAINS UNCHANGED AFTER COMMENT PERIOD REGULATORY GUIDE o EXPLICIT INCORPORATION OF LANGUAGE ABOUT
- HQM-USE OF STATISTICAL CONFIDENCE LEVEL 1
j 0 MERGED APPENDICES o OTHER MINOR CHANGES ECCS COMPENDIUM i
o MINOR COMMENTS BEING INCORPORATED INTO FINAL VERSION
I O O O ECCS RULE: UPDATE ON STATUS 1
ITEMS OF INTEREST FOR ACRS UPDATE e GRANDFATHERING EXISTING ECCS RULE (10 CFR 50-46 AND APPENDIX K) e 22000F LIMIT e SGTR (GI-141) e CSAU DEMONSTRATION OF REFLOOD PEAK BY ~8/88. SCHEDULE NOT BEING TIED TO RULE
O O o ECCS RULE: UPDATE ON STATUS CRGR o REQUESTING THAT THERE IS NO NEED FOR FURTHER REVIEW BY CRGR ACRS l 0 ACRS FULL COMMITTEE FINAL REVIEW SCHEDULED FOR MAY 1988 l
COMMISSION o RES MANAGEMENT IS CHECKING WITH COMMISSION TO SEE IF THEIR RE-REVIEW IS NECESSARY
%*y 9 9 PRESENTATION TO THE ACRS (
SUBCOMMITTEE ON THERMOHYDRAULIC PHENOMENA ON REVISION OF ECCS RULE l
CONTAINED IN l
APPENDIX K AND SECTION 50.46 i
1 OF 10 CFR PART 50 l
l .
NORMAN LAUBEN IDAHO FALLS APRIL 20,1988 I
^
EC_C_S RULE HISTORY CONCERN OVER THE ABILITY OF ECCS TO PROTECT THE REACTOR DURING A LARGE BREAK LOCA PEAKED AFTER A TEST IN THE SEMISCALE FACILITY RESULTED IN ALL INJECTED ECC LIQUID BEING EXPELLED FROM THE VESSEL.
j
- INTERIM ACCEPTANCE CRITERIA IN 1971 LACK OF A TECHNICAL BASIS FOR INTERIM. ACCEPTANCE CRITE RESULTED IN A CHALLENGE AND LENGTHY PUBLIC HEARINGS.
FINAL ACCEPTANCE CRITERIA (1974) OF 10 CFR 50.46 AND 1 APPENDIX K (ECCS RULE)
(
e e 9
i
~
~
CES RULE O
APPLICATION OF
!
- CURRENT EVALUATION MODELS CONTAIN LARGE CONSERVATISMS DUE TO:
APPENDIX K REQUIRED FEATURES (E.G., 1.2 X ANS DECAY HEAT)
STAFF REQUIREMENTS (E.G., CORRELATIONS THAT BOUND ALL OR MOST DATA)
VENDOR CONSERVATISMS (E.G., SIMPLIFIED MODELS)
!
- MOST LICENSEES CALCULATE PCT VERY NEAR 22000F, AND SOME PLANTS HAVE RESTRICTIVE POWER PEAKING FACTORS (OR HAVE EVEN f BEEN DERATED) TO MEET THE 22000F LIMIT.
l l
j'
- FUEL IMPROVEMENT AND RESEARCH HAS RESULTED IN REDUCTION IN l OTHER LIMITS (E.G., DNB), MAKING LOCA LIMITS THE MOST ~
RESTRICTIVE.
1
- PTS, FUEL MANAGEMENT, MANEUVERING, AND POWER UPGRADES i
l, REQUIRE INCREASED PEAKING FACTORS, WHILE ERROR CORRECTIONS I
AND MODEL CHANGES OVER THE YEARS HAVE RESULTED IN TIGHTER LOCA LIMITS.
i PRESENT ECCS RULE l
- 10 CFR 50.46 REQUIRES CALCULATIONS OF ECCS PERFORMANCE FOR A SPECTRUM OF BREAK SIZES USING THE METHODS OF APPENDIX K.
BOTH APPENDIX A (GDC) AND APPENDIX K REQUIRE CONSIDERATION OF A SINGLE FAILURE, WHILE APPENDIX A ALSO REQUIRES CONSIDERATION OF LOSS OF EITHER ONSITE OR OFFSITE POWER.
l .
i -
HYDROGEN < 1%
COOLABLE GEOMETRY i
LONG-TERM COOLING l
i APPENDIX K CONTAINS REQUIRED FEATURES OF EVALUATION MODELS (E.G., 1.2 TIMES 1973 ANS DECAY HEAT) AND ACCEPTABLE
- FEATURES (E.G., MOST HEAT TRANSFER CORRELATIONS).
~
~ APPENDIX K REQUIRES DOCUMENTATION OF EVALUATION MODELS OF SUFFICIENT DETAIL SO AS TO REQUIRE AN AMENDMENT FOR g GNIFICANT CHANGES (DEFINES AS A CHANGE THAT WOULD RE MLT IN A 200F CHANGE IN PCT).
~
i i
O o O
!~~ APPLICATION OF ECCS RULE (CONT'D) l-l
- THE NET RESULT IS THAT LOCA LIMITS ARE LIMITING PLANT l
OPERATION FOR SOME VENDOR DESIGNS.
l '
INTERPRETATION OF DOCUMENTATION REQUIREMENTS HAVE RESULTED IN A REQUIREMENT TO REPORT AND IMMEDIATELY RESOLVE ANY ERRORS IN EVALUATION MODELS OVER 200F. THIS USUALLY MEANS IMMEDIATE PLANT RESTRICTIONS AND LONG HOURS FOR NRC AND LICENSEE STAFF UNTIL A POSSIBLY TRIVIAL PROBLEM CAN BE SOLVED.
- b k
i ECCS RESEARCH AND MORE RECENT RULEMAKING HISTORY 1
!
- ECCS RESEARCH ON LARGE BREAK LOCA DOMINATED NRC RESEARCH FROM 1975 UNTIL THE 1979 TMI-2 ACCIDENT. ONE MAJOR ECCS
- .RESEARCH PROGRAM (2D/3D) IS STILL UNDERWAY.
- NRC HAS SPENT OVER $700M ON ECCS RESEARCH AND THE TOTAL l
! COST IS WELL OVER $1 BILLION WHEN DOE, INDUSTRY AND FOREIGN RESEARCH IS INCLUDED.
- l
- AS EARLY AS 1976, IT WAS APPARENT FROM RESEARCH THAT APPENDIX K METHODS WERE HIGHLY CONSERVATIVE AND DID NOT ACCURATELY REPRESENT ACTUAL PLANT BEHAVIOR DURING A LOCA.
e e 0
O m-g-422 o IN RESPONSE TO REQUESTS BY GE FOR RELIEF FROM LOCA RESTRICTIONS ON BWRS, THE STAFF ADOPTED A METHOD TO MAKE USE OF IMPROVED ECCS KNOWLEDGE WITHIN THE CONTEXT OF THE EXISTING ECCS RULE.
THE STAFF WOULD ALLOW AN EVALUATION MODEL THAT WAS TOTALLY BEST ESTIMATE EXCEPT FOR FEATURES SPECIFICALLY REQUIRED BY APPENDIX K (E.G., ALL STAFF-IMPOSED CONSERVATISMS REMOVED).
AS A CONDITION FOR THIS EVALUATION MODEL APPROVAL, A SECOND TOTALLY BEST ESTIMATE CALCULATIONS WAS REQUIRED, ALONG WITH
. AN EVALUATION 01~ THE UNCERTAINTY OF THE BEST ESTIMATE CALCULATION. THE UNCrnTAINTY EVALUATION MUST SHOW THAT REQUIRED APPENDIX K CONSERVATISM WAS SUFFICIENT TO COVER THE UNCERTAINTY IN THE BEST ESTIMATE CALCULATIONS.
22000F > EM (BE + APPENDIX K) > BE+ ESTIMATED UNCERTAINTY
i i
l ECCS RESEARCH AND MORE RECENT RULEMAKING HISTORY l _( CONT'D) l
- ANPR 1978 PROPOSED LIMITED PHASE I, FOLLOWED BY MORE COMPREHENSIVE RULE REVISION AT SOME LATER DATE.
w .
l
- THE TMI-2 ACCIDENT IN 1979 RESULTED IN A DELAY IN THE ECCS-f RULE REVISION, AND A MAJOR REDIRECTION OF ALMOST-COMPLETED LARGE BREAK LOCA RESEARCH.
i i
ECCS RULE REVISION AGAIN REVIVED IN 1982 IN CONTEXT OF .
STREAMLINING THE REGULATORY PROCESS.
i l
- e e e
O SECY-83-472 % CONT'D) O'
- STAFF VIEWED THIS APPROACH FAVORABLY SINCE MORE REALISTIC CALCULATIONS ARE CONSIDERED BENEFICIAL TO SAFETY AS WELL AS BENEFICIAL TO THE PLANT. .
- DISADVANTAGES ARE: .
~
TWO CALCULATIONS MUST BE PERFORMED AND REVIEWED WITHOUT A BASIS IN RULEMAKING, THE SECY-83-472 APPROACH IS SUBJECT TO LEGAL CHALLENGE.
l l
l-i I
CURRENT RULEMAKING
- 1984 COMPARISON OF BEST ESTIMATE RESULTS (TRAC-P AND TRAC-B) FOR TYPICAL PLANTS FROM THE FOUR VENDOR'S EVALUATION MODELS AND EVALUATION MODELS WITH MODIFICATIONS TO REMOVE APPENDIX K CONSERVATISMS SHOWED THE FOLLOWING:
CONSERVATISM IN CURRENT EMS VARIED GREATLY BUT WAS GENERALLY LARGE (10000F)
DECAY HEAT WAS THE MAJOR CONSERVATISM IN THE EMS (300-4000F)
REMOVING THE DECAY HEAT CONSERVATISM USUALLY RESULTED IN SHIFTING THE TIME OF PCT FROM THE REFLOOD PHASE TO THE BLOWDOWN PHASE OF THE LARGE BREAK LOCA.
MOST APPENDIX K CONSERVATISM IS DURING REFLOOD. ONCE THE TIME OF PCT IS SHIFTED TO THE BLOWDOWN PHASE, EMS
~
MAY NOT HAVE SUFFICIENT CONSERVATISM TO COVER THE UNCERTAINTY IN.THE CALCULATION.
O e e
j O CURRENT RUL91AKING (CONT'DJ O
)
I
- BASED UPON THE ABOVE FINDINGS, SIMPLY REFINING APPENDIX K
! BASED ON CURRENT KNOWLEDGE (REVISE DECAY HEAT) WITHOUT FURTHER CONSIDERATION OF UNCERTAINTY WAS NOT VIEWED AS ACCEPTABLE. THIS, COMBINED WITH FAVORABLE EXPERIENCE WITH SECY-83-472, RESULTED IN AN APPROACH TO RULEMAKING l CONSISTENT WITH THE SECY-83-472 METHOD.
i i
l l
l l
! NOTICE OF PROPOSED RULEMAKING ,
~
- NPRM ISSUED MARCH 3,1987 j -
PROPSED RULE DRAFT REGULATORY GUIDE DRAFT COMPENDIUM OF ECCS RESEARCH (NUREG-1230)
I i
- PUBLIC COMMENT PERIOD ENDED JULY 1,1987 4
l
- 33 COMMENTS RECEIVED - GENERALLY FAVORABLE, RESULTED IN NO l CHANGES TO THE PROPOSED RULE BUT SOME MODIFICATIONS TO THE i REGULATORY GUIDE.
j
- CURRENTLY SCHEDULED TO PUBLISH FINAL RULE AUGUST 1,1988.
i j .
i 9 e e
L O KEY FEATURES OF SR0 POSED RULE O l
- VERY GENERAL RULE REQUIRES "REALISTIC" CALCULATIONS OF ECCS l
PERFORMANCE WITH CONSIDERATION OF UNCERTAINTY SUCH THAT THERE IS A "HIGH LEVEL OF PROBABILITY" THAT CRITERIA WOULD NOT BE EXCEEDED. CRITERIA (E.G., 22000F) UNCHANGED.
- ALTERNATIVELY, APPENDIX K FEATURES REMAIN ACCEPTABLE AS AN OPTIMAL METHOD NOT REQUIRING AN UNCERTAINTY EVALUATION (GRANDFATHERING EXISTING ems AND ALLOWING FUTURE USE OF-APPENDIX K).
i l
KEY FEATURES OF PROPOSED RULE (CONT')
- REPORTING REQUIREMENTS SPECIFICALLY STATED TO RELAX AND CLARIFY CURRENT INTERPRI 9NS OF APPENDIX K DOCUMENTATION REQUIREMENTS.
i
- ALL ERROR CORRECTIONS AND CHANGES REPORTED ANNUALLY I
i i
i ALL ERRORS AND CHANGES OR CUMMULATIVE (ABSOLUTE VALUE) ERRORS RESULTING IN GREATER THAN 50 F PCT CHANGE REPORTED WITHIN 30 DAYS.
l NO FURTHER ACTION prQUIRED BEYOND REPORTING. STAFF MAY TAKE l ACTION CONSISTENT WITH SAFETY SIGNIFICANCE.
i i
- ERRORS OR CHANGES RESULTING IN EXCEEDING CRITERIA (E.G.,2200 F) i ARE IMMEDIATELY REPORTABLE (4 HOURS) AND IMMEDIATE STEPS OF
! BEING IN COMPLIANCE MUST BE PiiOPSED (CONSISTANT WITH 10 CFR50.55(e),
j' 50.72 AND 50.73).
j
! O KEYFEATURESOFPROP0$DRULE (CONT'D) O
- 1 1
j-
- AFTER $700M OF RESEARCH, THE ONLY NONCONSERVATIVE FEATURE OF APPENDIX K FOUND IS THE DOUGALL-ROHSENOW CORRELATION.
THUS, THIS PREVIOUSLY ACCEPTABLE CORRELATION WAS REMOVED FROM APPENDIX K. .
- EVALUATION MODELS APPROVED PRIOR TO THE RULE REVISION THAT MAKE USE OF DOUGALL-ROHSENOW WOULD CONTINUE TO BE I
ACCEPTABLE AS LONG AS THE CURRENT LEVEL OF OVERALL CONSERVATISM IN THE EM IS NOT SIGNIFICANTLY CHANGED. A SIGNIFICANT CHANGE IN OVERALL CONSERVATISM IS D5 FINED AS
, CHANGES AND/OR ERROR CORRECTIONS OF WHICH THE "NET EFFECT" (PLUS AND MINUS) WOULD RESULT IN A 500F DECREASE IN PCT FROM THAT WHICH WOULD HAVE BEEN PREVIOUSLY CALCULATED.
l*
i REGULATORY GUIDE
. D FINES HIGH LEVEL OF PROBABILITY STATED IN THE RULE AS 95%
PROBA BILITY i
- PROVIDES GENERAL FEATURES EXPECTED IN BEST ESTIMATE CALCULATION AND THE UNCERTAINTY EVALUATION, BUT DOES NOT SPECIFY OR PRESCRIBE '
ANY SPECIFIC METHODS.
l I . LISTS A LIMITED NUMBER OF ACCEPTABLE BEST ESTIMATE MODELS (E.G.,1979 ANS/ ANSI DECAY HEAT, CATHCART-PAWEL METAL-WATER REACTION RATE).
- IN SOME CASES, DATA APPROPRIATE TO MODEL ASSESSMENT IS PROVIDED.
l-l 9 e e 1
l
b FT COMPENDIUM 0F ECCSS ESEARCH(NUREG 1230) O 1
l i
I i
- MAJOR
SUMMARY
/ BIBLIOGRAPHY (1200 PAGES) OF EXTENSIVE ECCS RESEARCH i CONTAINS A DESCRIPTION -(AND WILL LATER CONTAIN RESULTS) OF
! NRC'S METHODS OF EVALUATING THE UNCERTAINTY OF NRC'S -
- CODES. THIS IS PROVIDED AS A PROOF OF PRINCIPLE, AN AUDIT TOOL, AND AN EXAMPLE OF HOW TO PERFORM AN UNCERTAINTY
, EVALUATION, BUT IS NOT A REQUIREMENT OR AN ACCEPTABLE
! METHOD IN THE REGULATORY GUIDE SENSE.
i l i l
l I. ,
i
ECCS COMPENDIUM ISSUED AS DRAFT NUREG-1230 (APR 87) -
WIDE FOREIGN & DOMESTIC DISTRIBUTION l
ONLY A FEW COMMENT LETTERS RECEIVED l
NOT PLANNING TO MAKE M.AJOR EDITS i
! WOULD BE HELPFUL TO RECEIVE COMMENT FROM ACRS CONSULTANTS, ALTERNA-TIVE IS TO DIG THRU TRANSCRIPTS CSAU SECTION (CHAPTER 4) WILL REFER l TO RESULTS IN A SEPARATE REPORT PLANNING FOR JULY 1988 DISTRIBUTION OF MAIN REPORT, WITH APPENDICES !
TO BE ISSUED AS g VOL 2 IN SEPT g g
l O O O l
l ECCS RULE PACKAGE i
- TRANSMITTAL LETTER TO NRR, OGD, ARM l
f TRANSMITTAL LETTER TO CRGR ;
1 i -
COMMISSION PAPER (DELETED COMMENT
SUMMARY
)
l -
- ENCLOSURES A. NOTICE OF PROPOSED RULEMAKING B. MEMORANDUM CHILK TO STELLO, DTD 1/9/87 C. ACRS AND PEER REVIEW SUMMARIES
SUMMARY
OF PUBLIC COMMENTS ON RULE D.
E. NOTICE OF FINAL RULEMAKING ...
F. DRAFT REGULATORY GUIDE . . .
G.
SUMMARY
OF PUBLIC COMMENTS ON GUIDE H. ENVIRONMENTAL ASSESSMENT .
- 1. REGULATORY ANALYSIS . ..
J. MARGIN INHERENT IN 2200 F LIMIT
_ . , - w.-- . , - - - - - .- - . , - - .- - _ _ _ _ _ _ _ _ _ _ _ _ _
~
l l ECCS RULE PACKAGE STATUS l
- RES CONCURRED l - PACKAGE TRANSMITTED TO NRR, OGC, ARM FOR f CONCURRENCE i - PACKAGE TRANSMITTED TO CRGR FOR INFORMATION SUGGESTED NO FURTHER CRGR REVIEW REQUIRED
! - NO SUBSTANTIVE CHANGE TO RULE
) - MODIFICATIONS TO REG GUIDE i
RES UNCERTAINTY METHODOLGY HAS BEEN REVIEWED BY i
ACRS AND EXTERNAL PEER REVIEW GROUP CHAIRED BY
[ NEAL TODREAS OF M.I.T. ACRS RESERVES RIGHT TO FINAL
] APPROVAL UNTIL UNCERTAINTY METHODOLGY DEMON-l STRATION IS COMPLETE i-
- 9 e o k-
O O O ENCLOSURE C-UNCERTAINTY METHODOLOGY
- ACRS COMMENTS PEER REVIEW CHAIRMAN (TODREAS) REPORT i
INDIVIDUAL PANEL REPORTS i
l l.
! ENCLOSURE D i
SUMMARY
OF PUBLIC COMMENTS
(
GENERALLY SUPPORTED NEW ECCS EVALUATION APPROACH i
SUPPORTED GRANDFATHERING OF CONSERVATIVE APPENDIX K METHOD l (ACRS QUESTION) l l
DID NOT FEEL THAL EXPLICIT DEGREE OF CONSERVATISM NEED BE SPECl-FIED ( ASSELSTINE QUESTION)
DID NOT NEGATE COMMISSION'S CONTENTION THAT INCREASED FISSION PRODUCTS WITH POWER OPERATING IS NEGLIGIBLE WHEN COMPARED TO UNCERTAINTY OF FISSION PRODUOT RELEASE (ASSELSTINE QUESTION)
INDEPENDENT REVIEW OF TECHNICAL BASIS SHOULD BE BY ACRS (ASSELSTINE QUESTION}
- GRANDFATHERING OF DOUGALL-ROHSENOW GENERALLY SUPPORTED SUPPORTED USE OF UNCERTAINITY EVALUATION
]
MANY COMMENTS ON REPORTING REQUIREMENTS i - - __ . . _ _ _ - . _ _ .
l
! O o o ENCLOSURE F-ECCS REGULATORY GUIDE CHANGES MERGED APPENDICES A & B. RESTRUCTURED NEW APPENDIX TO HIGHLIGHT MODELS, CORRE-LATIONS, AND DATA WHERE APPLICABLE P.18 1.2.4.- ADDED SENTENCE TO ACCOUNT FOR PHYSICAUCHEMICAL CHANGES IN IN-CORE, ,
EX-FUEL MATERIALS l
MADE MODIFICATIONS BASED ON COMMENTS 1
l
~
_ m -
ENCLOSURE G-
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE
- NRC RECEIVED NINE RESPONSES
- SOLICITED COMMENTS ON:
- 1. SHOULD THE REGUL.ATORY GUIDE LIST MODELS, DATA, AND MODEL EVALUATION PROCEDURES THAT THE NRC CONSIDERS TO BE ACCEPTABLE FOR REALISTIC CALCULATIONS OF ECCS PERFORMANCE?
- 2. ARE THE MODELS, DATA, AND MODEL EVALUATION PROCEDURES LISTED IN APPENDIX A TO THE DRAFT REGULATORY GUIDE APPROPRIATE?
- 3. DOES APPENDIX B OF THE DRAFT REGULATORY GUIDE APPROPRATELY DISCUSS THE FEATURES OF THE BEST-ESTIMATE CODES AND THE ESTIMATION OF OVERALL CALCULATIONAL UNCERTAINTY?
9 e e
SUMMARY
OFPUBLI_C_bMM_ENTS (C_ONT'D1
- SUPPORTED MAINTENANCE OF EXISTING ACCEPTANCE CRITERIA
- SUGGESTED BROADENING OF RULE TO INCLUDE ZIRCONIUM-BASED ALLOYS
- BROADENING OF RULE SCOPE SGTR CONSEQUENT TO LBLOCA LONG TERM DECAY HEAT REMOVAL SHOULD BE ADDRESSED
- RULE IS N_OI LEGAL BECAUSE IT BASED SOLELY ON COST-SAVINGS CONSIDERATIONS 1
- EXPERIMENTAL BASIS FOR RULE IS INADEQUATE BECAUSE FULL SCALE ECCS BYPASS DATA IS NOT YET AVAILABLE
~-
ENCLOSURE E-FEDERAL REGISTER NOTICE WITH FINAL RULE l
CHANGES SINCE 12-87 !
t l
P.4 DISCUSSION OF GRANDFATHERING APPENDIX K DOES NOT MENTION SAFETY BENEFITS OF NEW RULE, AND STATES THAT THE EFFECT l
! ON PUBLIC RISK IS NEGLIGIBLE 1
P.5 ADDED REFERENCE TO COMPENDIUM
! P.11 (ITEM 10) CLARIFIED THAT RULE ONLY APPLIES TO ZlRCALOY CLAD FUEL REACTORS P.11 STATE THAT ANY ACTION ON CONCURRENT SGTR WILL AWAIT PRIORITIZATION OF GI-141, AND SHOULD BE ADDRESSED l' THROUGH THE S.G. INSPECTION PROGRAM 43 9 9 e l
<. n. h g g qr y.. + .c.n
..i'... .+;. . 4f . o . 1.).; W~
QW&p{q;Q:-i
~ j.[W RM4%, ' [7_g; . ._ o.;
~ . ~
i$W
~
- ,; y9.
$~NCLOSURE G-SUMMARG OF PUBLIC COMMENTSe ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUE)
- 1. COMMENT- EXTEND COMMENT PERIOD RESPONSE: LATE COMMENTS CONSIDERED Q.1 COMMENTS
- 2. COMMENT - IN ANSWER TO Q 1., COMMENTER SAID REG GUIDE OVERLY PRESCRIPTIVE i
RESPONSE: NOT SO
- 3. COMMENT- BE METHODOLGY BURDENSOME RESPONSE: THATS WHY APPENDlX K IS GRANDFATHERED l
- 4. COMMENT- NRC SHOULD DO MORE FOR INDUSTRY IN QUANTIFYING UNCERTAINTY RESPONSE: NO 1
l l*
1
~-
j_ ,; '
4 ,
_p, ph 8s .. ',g ~_ . W ', e. 4 , .4bAis:- gy
- _ ,
}.- ...,3 ,, . . ,
1 ENCLOSURE G-
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUE)
- 5. COMMENT - EXPLICITLY STATE THAT ALTERNATE MODELS, DATA AND CORRELATIONS ARE ACCEPTABLE RESPONSE: IT IS CLEARLY STATED IN APPENDIX Q.2 COMMENTS l 6. COMMENT- DISAGREED WITH LAMINAR FLOW H.T. RECOMMENDATION l RESPONSE: DELETED REFERENCE IN REG GUIDE TO AVOID l CONFUSION
- 7. COMMENT- RULE AND REG GUIDE SHOULD SPECIFY ZlRCONIUM .
BASED ALLOYS NOT ZlRCALOY .
RESPONSE: SUBJECT OF ZlRCALOY VS. OTHER ALLOYS
- SHOULD BE THE SUBJECT OF SEPARATE RULE MAKING l 9 e e
.. . , . ~. _ , ; ,, ,, , ..n ,n, a,a. ,; y
G e o ENCLOSURE G -
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUE)
- 8. COMMENT- REGARDING APPLICABILITY OF CATHCART-PAWEL RESPONSE: AFTER CONSULTING WITH LABORATORIES, MODIFIED REG GUIDE SO CATHCART-PAWEL ONLY PRE-AP PROVED ABOVE 1900 Q. 3 COMMENTS l 9. COMMENT - REG GUIDE INSUFFICIENTLY DEVELOPED REGARDING UNCERTAINTY RESPONSE: GUIDANCE PURPOSELY GENERAL SO THAT APPLICABillTY l
TO A WIDE RANGE IF COMPUTER CODES IS RETAINED
i._
h j ENCLOSURE G -
SUMMARY
OF PUBLIC COMMENTS i ON REGULATORY GUIDE AND NRC RESPONSE I
I (CONTINUE)
! 10. COMMENT- CONSERVATIVE COMBINATIONS OF INPUT PARAMETERS l SHOULD BE SPECIFICALLY ALLOWED i
RESPONSE: THIS IS ALLOWED AS LONG AS CONSERVATISMS ARE NOT MASKING COMPENSATING ERRORS
- 11. COMMENT - CONFIDENCE LEVEL, NOT DISCUSSED OR SPECIFIED...
RESPONSE: MODIFIED REG GUIDE TO EXPLAIN THAT CONFIDENCE LEVELS NOT NECESSARILY REQUIRED
- 12. COMMENT - QUANTIFYING UNCERTAINTY SHOULD NOT BE REQUIRED l
l RESPONSE: DISAGREED, BUT NOTED THAT QUANTIFICATION NEED NOT CONFORM TO RIGOROUS STATISTICAL PROCEDURES O e e
ENCLOSURE G -
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUE) ,
- 13. COMMENT -
UNCOVERED BUNDLE HEAT TRANSFER SHOULD -
ACKNOWLEDGE PRESENCE OF LIQUID DROPS.
RESPONSE: ENTRAINMENT ADDRESSED ELSEWHERE. GUIDE MODIFIED TO EXPLAIN CONDITIONS OF UNCOVERED HEAT TRANSFER i
- 14. COMMENT -
DO NOT NEED TO ACCOUNT FOR BURNUP DURING ACCIDENT l RESPONSE: EXPLAINED THAT THIS APPLIES TO STORED ENERGY ,
- 15. COMMENT -
ASSYMETRIC CLADDING SWELLING TOO DIFFICULT AND -
BURDENSOME RESPONSE: RECOGNIZED DIFFICULTY, ONLY REQUIRE CONSIDERATION AS IT AFFECTS UNCERTAINTY ,
- 16. COMMENT -
NODING STUDIES NOT NEEDED IN AREAS OF THERMAL NON-EQUILIBRIUM l RESPONSE: REG GUIDE MODIFIED TO CLARIFY INTENT REGARDING BREAK AND INJECTION LOCATIONS I
h *
, , _ . _ . _ _ , --~ . _ . . - ,,-,__s_. .~... __ - , . - . . - , - _ . _. . , , , _ , , _ _ _ _ . _ _ , .
d -
ENCLOSURE G-
SUMMARY
OF PUBLIC COMMENTS
- ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUED) l 17. COMMENT - FLOW BLOCKAGE NEED NOT BE CONSIDERED
- RESPONSE
- IF APPLICANT CAN SHOW THAT EXPLICIT MODELING l NOT REQUIRED, THAT IS ACCEPTABLE
- 18. COMMENT - DATA LIMITATION SHOULD BE STATED SO THAT THIS i IS NOT DOUBLY ACCOUNTED FOR IN UNCERTAINITY l RESPONSE: DOUBLE ACCOUNTING NOT REQUIRED IF SUCH CAN BE DEMONSTRATED l
- 19. COMMENT - FINITE BREAK OPENING SHOULD BE ALLOWED RESPONSE: "LEAK BEFORE BREAK" IS NOW BEING CIRCULATED FOR PUBLIC COMMENT AS APPLIED TO LOCA ANALYSIS AND OTHER ISSUES. THEREFORE, ;
CONSIDERATION HERE IS PREMATURE. .
- 20. COMMENT- BYPASS CONSIDERATIONS SHOULD BE EXPLICITY NOT REQUIRED FOR SMALL BREAKS
. RESPONSE: DEMONSTRATION OF INAPPLICABILITY IS APPLICANTS
- RESPONSIBILITY .
O e e
o o O ,
ENCLOSURE G-
SUMMARY
OF PUBLIC COMMENTS ON REGULATORY GUIDE AND NRC RESPONSE (CONTINUED) l
! 21. COMMENT - DEMONSTRhTION OF "BEST ESTIMATE" CAPABILITY
! IS TOO BURDENSOME RESPONSE: THATS WHY APPENDIT K IS GRANDFATHERED .
i
- 22. COMMENT - MORE SPECIFICITY NEEDED REGARDING EQUIPMENT AVAILABILITY IN ANALYSES RESPONSE - APPLICANT BEST QUALIFIED TO DO THIS 1;
O
% o
._ . . , _ , _ _ . _ . . _ _ _ . . . .,_.m._ .. _
ENCLOSURE H - ENVIRONMENTAL ASSESSMENT NO CHANGES i
i ENCLOSURE I - REGULATORY ANALYSIS ,
CHANGE SINCE 12-87
- P. 6 DELETED DISCUSSION OF 400 F MARGIN. NOTE THAT
) EFFECT ON PUBLIC SAFETY IS DEEMED NEGLIGIBLE i
l l ENCLOSURE J - 2200 F INHERENT MARGIN
- NEW ENCLOSURE DISCUSSES RECENT INFORMATIN REGARDING FUEL, CLADDING, CONTROL ROD BEHAVIOR IN THE 2200 F RANGE. CONCLUDES THAT MARGIN ABOVE 2200 F NOT SIGNIFICANT TO RULE i
. 9 9 9
$4=
+<r o o CSAU METHC'DOLOGY Novak Zuber, USNRC Presented at the Meeting of the ACRS Subcommittee on Thermal-Hydraulic Phenomena Idaho Falls, Idaho April 19-21, 1988 f4SLOO925
Outline
- 1. Results of Peer Review Group Meeting and NRC response -
NRC
- 2. Current status -
INEL
- 3. Future schedule -
NRC g g egom
, ;l r !! ;! i!
7 2
o 9
0 p s e C L
3 u n s n
N o
r i o o t p G s s s
- e e
' u d r n
wC q a s
- eR s '
G i
vN e
C t s
n P
- R e T R dn N m d
r a o m n e t o a o eg e cs s n n Pn i s
n l
ai o r e
t o
rt o e a f e- c p
ee s nd n e en o sMs n r ge c t
l
) o s s m s s n uGp ' '
m '
Gio s G Go eRse R P
Rc Pe r Rt c P a R (P R 1 2 3 O 1 1
1 1
) ' ;l ; :
1.1 PRG's Response to NRC's Questions j Question 1: Is the methodology systematic, logical and practical?
PRG Response:
- "The methodology is judged to be systematic and logical.
i e its practicality is not uniformly accepted because of differing
- views of the breadth of demonstration calculations needed to i reach a judgement.
- When judged only in terms of the LBLOCA blowdown calculations completed to date, the methodology appears
- practical."
.- f4S 928
)
O O O
~
I j
) 1.1 PRG's Response to NRC's Questions i
(Cont'd) l l TPG's Comments:
i
)
- An application of the methodology to ihe reflood PCT is in progress.
- TPG is confident that the CSAU -methodology will be as
) practical when applied to reflood as it was shown to be for i blowdown.
l
- The practicality of any methodology will be evaluated on a I cost-benefit basis.
- CSAU can provide guidance for evaluating what is important, I reasonable and effective to the process.
f4SLOO929
'I i
1.1 PRG's Response to NRC's Questions (Cont'd) i Question 2: Has the methodology been demonstrated?
PRG's Response:
l
... no unresolvable difficulties are identified which would prevent achievement of this goal as work progresses. In a practical sense then, it can be assumed that an acceptable
{
demonstration of the methodology to the blowdown PCT is or soon could be in hand."
j
- ... however, this does not assure that demonstrations of the methodology for the LBLOCA reflood phase and SBLOCA's are achievable."
i l
l t 00930 i
- l. O O O l
1.1 PRG's Response to NRC's Questions (Cont'd)
TPG's Comments:
o The work is in progress on applying CSAU to the reflood PCT. No unresolvable difficulties have emerged to date, which would prevent its successful completion. ,
o TPG is optimistic that the methodology will be effective for other accident scenarios.
o Infact, NRC intends to extend and apply the CSAU methodology to a severe accident initiated by a SBLOCA. i NSL00947 l
.1 I
. 1.2 PRG's General Comments and
! Recommendations i
4 i
3
'~
^U Methodology:
I "A . .ai method for combining quantitative analysis and j expert opinion in a controlled way to minimize the
- suojectivity of the experts in arriving at computed
~
va!ues of uncertainty" i
i TF 3 Comment:
i i PRG provided a better definition of CSAU than TPG 1
ns 931
- . O o o 1.2 PRG's General Comments (Cont'd)
Recommendations and Benefits a "The peer review strongly encourages work on the development and demonstration of the methodology"
+ "The pursuit of such a program on comprehensive utilization of research results, not only has potential benefits for operating reactors, but offers a focus for planners and researchers in the
, conduct of future investigations" l
- "The benefits also include the comprehensive j documentation of computational codes and data
! bases" NSt_00932
1.2 PRG's General Comments (Cont'd)
NRC Comments:
- NRC intends to apply the CSAU methodology to in-vessel severe
- accident management activities 1
l
- The overall objectives of such a program will be to:
l
- 1. Bring within the CSAU framework, the research (experimental and analytical) conducted to date on in-vessel severe accident
- processes and phenomena.
- 2. Provide a basis for integrating experimental and code development activities.
! 3. Provide a basis for comprehensive and quantitative evaluation (in j contrast to a fragmented, piecemeal and qualitative approach) of the present status and future research (experimental and analytical) needs concerned with in-vessel severe accident phenomena, processes and issues.
NSLO0933 9 e e
i 1
- i. O o o l
i i
- 1.2 PRG's General Comments (Cont'd) l
- 4. Conduct a "first pass" evaluation of code j uncertainties to calculate parameters of interest to in-vessel accident management.
- 5. Identify and specify on a cost-effective basis, j what kind of code improvements would be needed
! to reduce code uncertainties.
- 6. Provide NRC with an auditable, traceable, and quantifiable methodology for in-vessel accident management.
I ,
nstoosu l
i
1.3 PRG's Concerns and TPG's Response Actions PRG's Concern:
Methodology does not offer criteria for establishing:
- The sufficiency of PIRT
- The sufficiency of the data base
- The completeness of methodology TPG's Response:
, TPG recognizes the concerns as valid and is in the l
process of developing necessary criteria and information.
t ns oss
- l. ~
O o o l
4 i
l 1.3 PRG's Concerns (Cont'd)
PRG's Concern:
The PRG was not convinced that the effects of scale l had been sufficiently characterized.
l TPG's Response:
- Part of BNL's presentation, which would have I provided more detail on scaling, was deleted by PRG's chairman, in the interest of time i
l
- TPG is addressing this concern and will provide i information on the scaling rationale l
- NSLOOS36 l
l l
l
! 1.3 PRG's Concerns (Cont'd) i
- Ongoing Scaling Activities i
i l
- NRC is in the process of developing a general j approach to scaling which will be tested and i demonstrated on LOCA and severe accident test
! facilities.
i
)
- Approach based on a hierarchical structure i
!
- Top level scaling group /functionals have been derived and will be presented in a forthcoming NUREG report together with the methodology.
l tiSLOO937 e
~
9 9
1 O O O
! 1.3 PRG's Concern (Cont'd)
- The objective is to:
Develop a method which can be used to quantify the effects of distortions and/or to provide a measure (index) of applicability of available correlations to conditions of interest to NRC applications.
- The results are needed to:
- 1. Assess the applicability of a code to the accident scenario and to evaluate its scale-up and modeling capabilities.
- 2. Identify, specify and prioritize future experiments if needed.
- 3. Quantify the effect of scale distortion.
f 4*> LOO 938
(
1 l
1.3 PRG's Concerns (Cont'd)
- PRG's Concern:
l J Some members of PRG were concerned that j the effects of nodalization on uncertainty l were not adequately addressed TPG's Response:
This question will be examined in detail at the next meeting of TPG i
ts 0939 j.
i .. O O o 1
i l 1.3 PRG's Concerns (Cont'd) i PRG's Concern:
}
j The methodology concentrates on the uncertainty of the primary safety criteria (PCT for LBLOCA). The uncertainty of l other parameters is not considered.
i l TPG's Response:
1 l
- TPG believes that it is sufficient and cost effective to determine the uncertainty only for the primary safety l criteria used in licensing / regulation decisions.
NSLOO940 i
)
l 1.3 PRG's Conce.rns (Cont'd) i l
- TPG recognizes that the defensibility of an uncertainty j quantification depends to a large extent on l demonstrations that the code simulates other important
! parameters adequately.
I i
i Such demonstrations are (or should be) provided in code j assessment reports.
~
NSLOO941 l 9 9 6 4
- . O O O l
i 1.3 PRG's Concerns (Cont'd) i .
I
! PRG's Concerns:
)
Hot channel thermal-hydraulics effects have been ignored l
l TPG's . Response:
- in response to this concern, additional calculations have
! been performed. The results will be presented at this meeting by INEL.
l _
1.3 PRG's Concerns (Cont'd)
PRG's Concern:
The effects of thermal-hydraulic / fuel cross products have not been quantified.
TPG's Response:
- TPG believed that the analysis was conservative.
- However, to respond to PRG's concern, additional calculations have been performed which show that the original results were indeed conservative.
- Results will be presented at this meeting by INEL.
il 0943
O
~
. O O 1.3 PRG's Concerns (Cont'd)
PRG's Concern:
Mucil of the experimental data used in the CSAU demonstration have been used in code development and assessment PRG's Suggestion To demonstrate the generality of the methodology, TPG should include additional data (not used heretofore) in code assessment and development.
TPG's Response:
- TPG is in complete agreement with PRG; however, data from "virgin" facilities are very scarce.
NSLOO944
l i
1.3 PRG's Concerns (Cont'd) i I
i
- Nevertheless, to respond to this concern
) ... OECD LOFT data have been included I ... PKL data will be included i
I
- The results will be shown by INEL i
l l
l .
, , ..sg.s l -
- j. O O O i
l PRG Response items Resulting in Delayed j Completion of CSAU Project .
I l 1. Items already addressed and to be presented at the i April ACRS meeting:
Hot channel and results.
- a. investigation
) b. Evaluation of the effect of double hydraulic cross l products.
- c. Inclusion of additional "independent" experimental data (OECD LOFT) in scaling study using PCT vs j LHGR.
l l
,moom j
l
PRG Response Items Resulting in Delayed Completion of SCAU Project !
l
- 2. Items that are being or will be addressed:
- a. Development of success /suff!Liency criteria for the steps in the methodology.
- b. Additional support for the standard NPP nodalization strategy.
l
- c. Additional support for the sufficiency of the scaling evaluation.
- d. Additional support for the sufficiency of quantifying uncertainty for primary safety criteria (PCT) only.
- e. Expanded final documentation covering new topics.
g , ,,ge...
.. o ,
o o CSAU Milestones .
OLD NEW o Complete transient calculations MAR 88 JUN 88 for blowdown dominate uncertainties o Complete transient calculations APR 88 JUL 88 for refill /reflood dominate uncertainties i o Complete determination of total MAY88 AUG 88 uncertainties o Complete NUREG draft JUN 88 SEP 88
! .,woms
W O' .
O O CSAU STATUS REPORT RESULTS ADDRESSING PRG CONCERNS PRESENTER: GARY E. WILSON CONTRIBUTORS: CSAU TPG &
1 STAFF AT BNL, LANL, INEL PRESENTED AT THE MEETING 0F THE l ACRS SUBCOMMITTEE ON THERMAL-HYDRAULIC PHENOMENA meap APRIL 1988 AGtDCnfAL
- - 2; , -
' LABCWAfDWY
, I
i i
~
I THE MAJOR TOPICS TO BE COVERED INCLUDE i
o ADDITIONAL DATA NOT USED IN CODE DEVELOPMENT & -
ASSESSMENT 4
o EFFECTS OF ADDITIONAL HYDRODYNAMIC CROSS PRODUCTS o EFFECTS OF HOT CHANNEL THERMAL-HYDRAULICS ON PCT l
l 9 9 9
l THE EFFECT OF ADDING ADDITIONAL OECD-LOFT BLOWDOWN PCT EXPERMENTAL DATA IS VERY SMALL o 37 BLOWDOWN PCT POINTS FROM LP-LB1 o CHANGE IN EXPERIMENTAL DATA VARIABILITY AT 95% CL WAS l LESS THAN 1% (3 0F) o CURRENT DATA BASE INCLUDES 48 ADDITIONAL DATA POINTS FROM LOFT, SEMISCALE AND LOBI WITH A SUBSEQUENT REDUCTION IN THE VARIABILITY OF 14 0F l
i 11aximum measur ed clad temperature from several scaled experimental facilities (216 data points) l during blowdown I 2300 . . . i-i i i . .
i . . . .
i ,.. . .
- APPENDIX K LIMIT -------=
l l _ ..
l -
VOLUME SCALE FACILITY SCALING CCNTERION: -
O V PDF - 1130000 POWER / VOLUME /
W '
~ O SEMISCALE
! 111700 '
i -
1800 - 0 LOoi 17700 ," -
1 m ._ A THTF 11500 ,-
' yt' y
m i O LOFT 114 8 ,-
,/ 4 /
i m
m a
_ ,- - O , av .
2 1300 ' / + y
+ 4+
m / -
i - 3 +7 1 F O
p/ 4 g++ + #
- Y4
'p SA ~
4 4 - - '
4
__3 O
+
+9 (+ ,/
JG1oF 4 4[ + /
j 4 4 ..
d "
Q 800 -
- f 4 +yY y ,%' -
~
! < ~~'
/
. . . . . 95% CONFIDENCE B AfJOS
~
2 -
/ LINEAR REGRESSION LINE -
I
+
l l l l l $ l l ! ! ! I ! !
O 5 10 15 20 i
LINEAR HEAT GENERATION R ATE (KWIFT)
I 9 e G.....
T E
=
a y
()
- w. O l
I I i N l
i i u.
i o
! v i e "O i 's O NT CO e ! '. i
- i i e
\, O M e g _ l 5
+
s\ m H
__ l %+ \ m e u
em . i. \ o e
>- ; g* o o i c= s c
- .E l \ \
- c 3 O l Ei \ ++ %o
. 8.9 c "
2
~
E' : o '
om l ':
O el, R oo '-\ o E w m - i o e i
- 2- H
-e i .: i + i. - c) 4 ODC b \, * +, D. \,
U CE cM3
- LO o 2 eo c
D +
g$
z cm .N, u 3 =e s>
\
i, 4 o o i. i to c O
3 w o o i i c) D H M 3 \ N + ++
- O<
l E$o- -
D x
o >1 o '
i
+
& \+
\
i i
" c~
w Z
3 =O m z 5 \. o e.
l o w a ++
w I o m .-' c. e i
+ + +
i O
mw a 1 s T) 3 o < \ 4 + H pc i e
i o ma o
\, 5 "+. +
+++* ,\., $
=
S N 5.!
m e i o8o88m Em-se,
\ <*
+\ c-e a :, esssss i. . g,.,
. w Ex o \
i.
z o -
l > 2 i i '
to " ,
E . ,
c l 's a D \
,E l
' \ .
+ <<< \
i x < '
m l U - \ \
l l 2 "
u_ 2 e_ ms r u.
e \ + \
- -f -
i
\
i ewoIo c. .
l 8
c.mara _.; i
$ + ++
i t
O O <G 0 + M i
' l
' I 1 O
O O O O O O O O O O M CO M co m N - e--
(d 930) 38nivuadWal OV lo WnWixvW
, . , , - . ~ , --
Maximum measured clad temperature from several scaled experimental f acilities (301 data points) during blowdown 2300 i i i
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -- - - - - - - - A P P E N D I X K L I M I T - - - - - - - - - - -- - --- - - - - - - - - - -- - - - - - - - - - - -- --- - -
E Facility Scaling Criterion:
0 '
tu o PBF ~ 1/30000 Power / Volume 9 1800 _
o SEMISCALE 1/1700 ,;,
-7$,o ,_
uJ a LOBl 1/700
++
cc ;+
- o
' +
D o TitTF 1/500 + +
+k g k + LOFT 1/48 ,-,- o
' ' 'o
, . o+ P f++,*
' ' h 3 1300 -
0g *
+
$ + /*U o 4p 8 _
F-O a +n O,+ + .
+
. g+ +o f + ipd ' 'o -
y,
'8 4 ,~
h g
++ +oo6 * + +*8 ,
~'
__; + a
+++++o ,,
- +
U
+o
.a y
++
+
++
+
+
+
350 0F 3 4
+
3 a+ g +a 800 -
p a o -
1 a $
6
+ ss , +t -7+
6 Y + fy ,{ *'+ -
4 . -
95% Confidence lines s '
Linear regression line T = 585 + 60 (LHGR) 300 O 5 10 15 20 LINEAR HEAT GENERATION RATE (KW/FT) 9 9 9 NSL388-2
l 0 o o i
l ,
i l ADDITION OF OECD-LOFT REFLOOD DATA DID NOT CHANGE OUR i CURRENT UNDERSTANDING 0F THAT PHENOMENA i
l 0 8 DATA POINTS FROM LP-LB1 i
o NO DISCERNIBLE CHANGE IN TEMPERATURE RISE DURING REFLOOD l 0 FURTHER ANALYSIS USING DATA FROM THE FRG (PKL) &
NETHERLANDS (ECN) IS IN PROGRESS l
i
Measured temperature rise data from scaled experimental f acilities (91 points).
1200 ; i i i i k, Volume Scale 1
i O Semiscale 1/1700 O FLECHT 1/400 u_ 900 - -
g ',', A LOFT 1/48 W '
- E LB1 C) "
o
- '\, O CCTF 1/21 W '
w i, + SCTF 1/21 o ,
m a w 600 -
o 'N
[ o i D o 's H g
~
's Regression line. ATrise = 355 U r o_ g 95% Confidence lines -
3 -
W '
300 -
0
,o ran _,,-
, o . , '
i o - . _ , "
i + -
's,(( 4 g _______----------_ _________
., o -
p
-j---------____________,_____________________p______t___,_______2________________
O 2 4 6 8 10 REFLOOD RATE (IN/S) 9 e e HSL8712-1
~
~
O O O THE USE OF ADDITIONAL HYDRODYNAMIC CROSS PRODUCTS RESULTED IN A LOWER BLOWDOWN PCT AT 95% PROBABILITY; THE CHANGE WAS SMALL (16 0F = 1%)
o PRG CONCERN - EFFECTS OF COMBINED HYDRODYNAMIC VARIATIONS HAD NOT BEEN EXPLORED o THREE ADDITIONAL NPP SIMULATIONS HAVE BEEN PERFORMED USING DOUBLE HYDRODYNAMIC VARIATIONS PUMP 2 SIGMA + CRITICAL FLOW 1 SIGMA PUMP 2 SIGMA + CRITICAL FLOW 2 SIGMA PUMP 1 SIGMA + CRITICAL FLOW 2 SIGMA 1
l
THE RELATIVE CHANGES IN PCT BETWEEN SINGLE AND DOUBLE HYDRODYNAMIC VARIATIONS ARE ILLUSTRATED BELOW PCT CHANGE TYPE VARIATION VARIATION FROM NOMINAL FROM NOMINAL SINGLE CRITICAL FLOW 1 SIGMA 106 OK CRITICAL FLOW 2 SIGMA 97 PUMP 1 SIGMA 14 i PUMP 2 SIGMA 35 DOUBLE PUMP 25 + CF 1S 88 PUMP 2S + CF 2S 98 PUMP 1S + CF 2S 98 e e o
. O O O THE CHANGE IN THE TOTAL BLOWDOWN PCT UNCERTAINTY BETWEEN THE DATA BASES CONTAINING SINGLE AND DOUBLE HYDRODYNAMIC VARIATIONS IS SHOWN BELOW (VALUES IN OF)
OLD NEW PARAMETER VALUE VALUE CHANGE MEAN PCT 1107 1102 -5
-23 STANDARD 172 149 DEVIATION l
T05 1379 1363 -16 OLD VALUE = DATA BASE CONTAINING ONLY SINGLE l HYDRODYNAMIC VARIATIONS l NEW VALUE = DATA BASE CONTAINING SINGLE & DOUBLE l
HYDRODYNAMIC VARIATIONS I
I 1.
DURING THE PRG MEETING THE ABSENCE OF HOT CHANNEL HYDRODYNAMIC EFFECTS IN THE ORIGINAL ANALYSIS WAS QUESTIONED o EXPERIMENTAL EVIDENCE TENDED TO INDICATE THE AVERAGE FUEL RODS DR0VE THE CORE GLOBAL HYDRODYNAMICS o THUS THE EFFECTS OF HOT CHANNEL HYDRODYNAMICS ON THE HOT PIN WERE THOUGHT TO BE 2ND OR 3RD ORDER & NOT QUANTIFIED I
~
$ @ e
. O O O PRELIMINARY RESULTS FROM A HOT CHANNEL ANALYSIS INDICATE A BIAS (PENALTY) 0F 68 0F MAY BE SUFFICIENT TO ACCOUNT FOR THE EFFECTS OF LOCAL HOT CHANNEL HYDRODYNAMICS ON THE HOT PIN o THE NPP MODEL WAS REVISED TO CONTAIN 4 HOT BUNDLES o TWO RUNS HAVE BEEN MADE:
NOMINAL CASE THROUGH REFLOOD CRITICAL FLOW 1 SIGMA CASE THROUGH BLOWDOWN o TWO TYPES OF COMPARISONS HAVE BEEN MADE:
RUN BY RUN PDF BY PDF l
l
1 ORIGINAL MODEL 2 Il0T CilANNEL MODEL 1000 , , , , i , , ,
l ; ; , , , ,
T ,
800 -
l a:
a 4
F w:
a:
w i S' m
H ll S 600 --
.a _
o i
l l
i I
400 ' ' ' ' ' ! ' ' ' '
E 'O I ? ? ? ;? 2 MAGNUH 2.0 16.56.53. 88/03/21.
< 0 50 100 150 Tlae (s) l BASE CASE AND HST CHANNEL COMPARISSN a
1
. O o o ON A RUN BY RUN BASIS OF COMPARIS0N THE HOT CHANNEL CONTRIBUTION TO PCT IS < 25 OK (45 0F), FOR THE ANALYSIS PERFORMED ~TO DATE PCT (OK)
HOT CHANNEL ORIGNINAL MODEL HOT CHNL MODEL EFFECTS CASE BD RF BD RF BD RF NOMINAL 832 820 856 835 24 15 CRITICAL 938 960 22 FLOW 1 SIGMA l
l l
1 ON A COMBINED UNCERTAINTY . BASIS (PDF) THE HOT CHANNEL EFFECT ON PCT IS 68 0F, FOR THE ANALYSIS PERFORMED TO DATE TEMPERATURE (OF)
HOT HOT PDF ORIGINAL CHANNEL CHANNEL PARAMETER MODEL MODEL EFFECT MEAN PCT 1093 1176 83 STANDARD 218 193 -25 DEVIATION l PCT AT 95% 1511 1579 68 l PROBABILITY PDF's WERE GENERATED FROM FULL 2ND ORDER RESPONSE SURFACES CONSTRUCTED FROM THE NOMINAL AND CRITICAL FLOW 1 SIGMA NPP RUNS 9 4 e
N.162 m,#
o nwmuu q --; e +m e2 y=a -. i 1 % .
M-+ :athF %=t %4= . = ~
, g.=c == . ; c 3- :- 22.a-w _1 .,
,-o v . ==d-i- -i COMPARING CALCULATED AND EXPERIMANTAL LBLOCA r,_ -
7.=_
(3 7 =E i CT3 =- PROBABILITY DISTRIBUTION FUNCTIONS N '- = =
g..
li.. - e e g _-__--2 :-r_2 r -
_ _ -- g
_ . _ _~ , .-
__.3____. ,
--:=_._ .
- p. --__._-
a: :.::_ _
.a:-
-. , - . . .__-_,----_ . ~~
4 . _ . - _
I -'
, p___ _ _ _ .;
1
.-._y_ : _ _ _
_. 4
. . . . J
-Mt.
r x
__ _ ; j l
-e r u m.
x, _
+ m1I r1TKTWIMITTTIRT TMTFilH99 -
p
.- -WA1 "_"T l- ;N -~E] MAL. I ~ _l
~~
v '
- -- T z -=, 9 ___ .___
E
- = _ ._ .2 E
o
,i l 4s Lit tmfmftM- ET5Rrtet- mt-
.'_'-.~,j~
0
. r- 1
\r L I gIjy,g ji 1,yp 1 n E
m
_-. _s - -_ -
'd g ,7 q 4
,3 .g mgg 6 9_EW4-
.._.__.__._4._____.
p _ _ . - . . _ _ .
yy I_ .
-\ ~~~~-----
i 2 +
..__LMbK KMbt3 DalD* h/f(*
i w i \ -
@y 1,0. a-i1M.- -_ = - ' s AL1: ;
-hV 4.LM ~- = = ix ,A . -- - - - - - -
a w t -- = u .1_t-.
p q .i "-&
--+*1. . .C=^1 1[ .-h u: r& 14_t-.& R- z-f- dua > -' i Q : 1 .i h-- i &J i -i =~ i .k--- j ' * ' ' ~
- E = : . . --- l = =- =i- fi vi - MY L l =n -= i =# r -- '
- b -+ i---- :- '
k W
n + " "%M ~-M -f-i_- I.3.J == =i--M- i r 4 -i- i- . . 2
- -.rr
=- _- i
.=-1_= 6__- r- ; -
fy - _ 2- = = - -+i h- : - _._.==-E-,----- -
- _\=+ ; _=1- .: -
h ,-- Mi }~~_ 23 :-$.f.;3 ' .
,(,j;_,4 ,,+ p. ; .z, p~~~~~
.--d : 5.--IE .- _ ' =_ -5 ~nTr=5 -[_Z-=5 i-h553. : "I-~ : -
I ! ~ = i 5-L 3i . ~i - - I _,
- 6 t= ==+-- = = EC ~J - -~ : :_ ZZT_ _.: =__=i iis :-la i< - a O
o ==
_._i
_ . .:'--t -3_. EG
- 7=__
_[__1_Y :. _ M-6 _ _~- __ _ __.. -Hi:F: i 7 '. _ . -.
_ :.j -
. ._1_
.T;,, g _ _ . _ _ .
LA.
g n. . ._- _ . :__ .
_ : . .- 1. -
- T . ----
U $ . .: ; ; -- - n---'. ..._t.._. .
4 - h H - ~ ~ ~
_ . . _ . ,_..~_~.C.4.
..._.J. _. !
. . A.
T, - . . . . _ ._ _ _ _ _ _ _ . . .
, __ _ ; _ _ _ _ , , __. g_. i ,
,._ _. . _. - - -4. -
i
, .}4 ,
,, , y I 11 v _.
!i -
ii! !
, , i ii i ti
!i'
!i'- i i . . i . k_ .
i ii 2 t a '
_ _,- 64 : - - - - -
1 .
___,1 2_
- _ a __... .
(m,\
y -_.
v _
4 .. -
\
E .
! ~~
in-5 n~ .hn n i, i . L M am 10bo 12co 1400 - 1600 1800
-