ML20132G929
| ML20132G929 | |
| Person / Time | |
|---|---|
| Issue date: | 07/24/1985 |
| From: | NRC COMMISSION (OCM) |
| To: | |
| References | |
| REF-10CFR9.7 NUDOCS 8508050184 | |
| Download: ML20132G929 (99) | |
Text
E ORIGINAL dl N-k,
[
UNITED STATES OF AMERICA 7
(
NUCLEAR REGULATORY COMMISSION
.g ti!
3
~ ~ ':.'q
....a
.t In the matter of:
-a COMMISSION MEETING Briefing on Accident Source
~
Term Reassessment
'I?,)
,i (Public Meeting) j Docket No.
s.
~,;
(
s Location: Washington, D. C.
Date: Wednesday, July 24, 1985 Pages: 1 - 78 ANN RILEY & ASSOCIATES
(
Court Reporters 1625 I St.,
N.W.
Suite 921 Washington, D.C.
20006 8500050184 850724 PDR 10CFR (202) 293-3950 PT9.7 PDR
J g
1 D l SC L A I M EP 2
3 4
5 6
This is an unofficial transcript of a meeting of the Wednesday, 7
United States Nuclear Regulatory Commission held on 8
July 24, 1985 in the Commission's office at 1717 H Street, 9
N.W.,
Washington, D.C.
The meeting was open to public 10 attendance and observation.
This transcript has not been 11 reviewed, corrected, or edited, and it may contain 12 inaccuracies.
13 The transcript is intended solely for general
_ 14 informational purposes.
As provided by 10 CFR 9.103, it is 15 not part of the formal or informal record of decision of the 16 matters discussed.
Expressions of opinion in this transcript 17 do not necessarily reflect final determination or beliefs.
No 18 pleading or other paper may be filed with the Commission in 19 any proceeding as the result of or addressed to any statement 20 or argument contained herein, except as the Commission may 1
21 authorire.
22 23 24 25
1 s
1 UNITED STATES OF AMERICA 2
NUCLEAR REGULATORY COMMISSION 3
4 BRIEFING ON ACCIDENT SOURCE TERM REASSESSMENT 5
6 PUBLIC MEETING 7
Room 1130 8
1717 H Street, N.W.
9 Washington, D.C.
10 Wednesday, July 24, 1985 11 12 The Commission met, pursuant to notice, at 13 10:05 a.m.
14 COMMISSIONERS PRESENT:
15 NUNZIO J.
PALLADINO, Chairman of the Commission 16 THOMAS M.
ROBERTS, Commissioner 17 FREDERICK M.
BERNTHAL, Commissioner 18 JAMES K.
ASSELSTINE, Commissioner 19 LANDO W.
- ZECH, Jr.,
Commissioner 20 STAFF AND PRESENTERS SEATED AT COMMISSION TABLE:
21 H.
PLAINE, General Counsel 22 S.
CHILK, Secretary 23 W.
DIRCKS, EDO 24 H.
DENTON, NRR 25 D.
ROSS, OR
2 m
1 J.
MITCHELL, ASTPO 2
M.
SILBERBERG, ASTPO 3
AUDIENCE SPEAKERS:
4 None 5
6 7
8 9
10 11 13 14 15 16 17 18 19 20 21 22 23 24 25
D g
3
~
1 P RO C EE D I NG S 2
CHAIRMAN PALLADINO:
Good morning, ladies and 3
gentlemen.
We have with us today representatives from the 4
Staff to discuss the present status of the Staff source term 5
assessment, and anticipated future source term program 6
actions.
Work on source terms has been a long-range and 7
multi-pronged effort, We have had previous public discussions 8
with representatives of the Atomic Industrial Forum, the 9
American Nuclear Society, the American Physical Society, 10 interested groups and NRC Staff, the most recent being the 11 discussions with industry and non-industry groups on April 3,
12 1985.
13 Today's discussion with the Staff will focus on the 14 present status and anticipated future actions with respect to 15 the Staff's forthcoming Source Term Technology Report, 16 NUREG-0956, which is to assess the extensive body of new 17 information with respect to the val,idity and potential 18 applicability of this information to the regulatory process.
l 19 We had originally hoped that NUREG-0956, together 20 with a companion paper, would be issued for public comment by 21 June of this year, but the work has been delayed and we look 22 forward to getting a better feel of what the schedule is now.
23 Of particular interest to the Commission in today's 24 meetings are the advances which have been made in the 25 improvement of the source term, computer model and the present
4 1
Staff thoughts on the 4evelopment of a methodology for 2
applicttion of current source term data to specific cases.
3 If my fellow commissioners agree, I would like to 4
spend the last 10 minutes or so of today's meeting discussing 5
the amount of work remaining to be done and the Staff's 6
schedule.
A clear understanding of critical issues and the 7
prioritization being afforded to them would be very i
8 worthwhile.
9 I understand that Region II has requested to listen 10 by telephone, and I understand they are connected.
11 Do any of my fellow commissioners have any 12 additional opening remarks?
I should point out that
(
13 Commissioner Asselstine will be joining us but he was delayed 14 and asked us to start without him.
15 Any other comments,?
16 COMMISSIONER ROBERTS:
No.
17 COMMISSIONER ZECH:
No.
18 COMMISSIONER BERNTHAL:
No.
i I
CHAIRMAN PALLADINO:
Well then, let me turn the 19 20 meeting over to Mr. Dircks.
21 MR. DIRCKS:
I would point out, Mr. Chairman, this i
22 is a discussion of where we are in source term reassessment; j
23 it is not a reassessment of source terms today.
24 Denny Ross is going to do most of the talking.
l 25 Harold Denton is here to answer questions, but we are not
5 1
ready to talk about the regulatory implications of this work, 2
and that won't be until sometime probably in the spring of 3
next year.
4 Denny?
5 CSlide.3 6
MR. ROSS:
We have slide number 1.
The material 7
that we'11' discuss today is a work product of the Accident 8
Source Term Program Office which was set up a couple of years 9
ago for this purpose.
10 I also will note that since Region II is listening 11 in, I am taking this more detailed presentation to Region II 12 and Region IV next week.
I have been to Region I about six 13 weeks ago discussing source term, and I will be visiting 14 Regions III and V in the near future so they will all be fully 15 informed of where we are, and more importantly, where we are 16 going to go.
17 CCommissioner Asselstine, enters the meeting.3 18 CSlide.3 19 MR. ROSS:
We showed this same slide to the 20 Commission in our last briefing.
Since then, we have made 21 some progress that I'd like to discuss briefly.
22 On the slide there are sort of rectangles that 23 discuss various work products, either of our contractors or 04 special works formed for this purpose.
The last item in the 25 lefthand column is, of course, the APS Review.
All six of
6 1
these items are now complete.
The last of the documents are 2
either published or going to the Public Document Room and 3
printer this month.
4 Where we are today is the circle in the middle of 5
the chart labeled NUREG-0956.
We expect to send that to the 6
printer next week, probably about Tuesday or Wednesday, and 7
there should be many hundreds of copies available about seven 8
days later.
9 We are, as Mr. Dircks said, starting the work on 10 regulatory applications.
A very important ingredient is the 11 circle at the bottom of the chart marked " Severe Accident Risk 12 Reduction Program," where we apply the methods of the source 13 term code suite, as we call it, to five reference plants.
And 14 we will be recomputing the accident frequencies as well as the 15 accident source terms for these frequencies, and a convolution 16 of these items would produce the rebaselining risk.
17 That information then fee,ds across the dotted line 18 to the righthand. side of the graph, which is in general the 19 1986 work and it is the applications in the regulatory arena.
20 That work will be discussed in the future.
21 We regard what we have here today, NUREG-0956, as 22 basically a science and engineering report.
There's some 23 brief discussion of risk insights but it's fundamentally a 24 description of the new source term methodology.
25 We have labeled the " Severe Accident Risk
7 1
Rebaselining Report" as NUREG-1150, although we haven't yet 2
written the first world.
We use that as a catch-all so that 3
we can describe our work.
And I expect we will be prepared to 4
discuss progress on that in probably three or four months.
5 COMMISSIONER BERNTHAL:
Denny, can you give us a 6
milestone date for these two?
Maybe you're planning to do tha 7
at some later point here, but exactly when do you expect to 8
have NUREG-0956 out?
I know that that's very close, I guess, 9
to being out, but then what about these further milestones?
10 MR. ROSS:
Okay.
August 7th is the nominal coming 11 out of the printer's date for NUREG-0956.
And two points on 12 that.
First, we have ordered a large number of copies and we 13 expect to have an extensive mailing list.
And we will be 14 emphasizing, hopefully, or re-emphasizing this is being issued we'll publish a Federal 15 for comment, and we are envisioning 16 Register notice, and of course, we have an extensive mailing 17 list already.
18 We would envision a 90-day comment period, and 19 then after analysis of the comments, then we would probably 20 reissue that in final form, say, the first of next year.
21 The bottom circle, Severe Accident Risk Reduction it's 22 Program, which I call NUREG-1150, is scheduled to be 23 an essential ingredient to NRR and the technical results will 24 be given to them as they accrue in a linear fashion, starting 25 probably September and October.
The report itself in draft
8 1
form would also go out for comment, I'd estimate on the order 2
of March or April 1986.
However, NRR will have the material 3
for its auditing purposes earlier.
4 Now we have a very detailed milestone schedule for 5
that work, but that is small increments of work.
We do plants 6
one or two at a time, and then the work comes out one or two 7
plants at a time.
8 COMMISSIONER BERNTHAL:
I guess my concern is that I it sounds like we're already a year 9
think originally, 10 behind from where we wanted to be at this point on the 1150 11 effort, and I'm sure there are good reasons for that.
But
(
12 your slide makes it graphically evident that the bottleneck is 13 going to be these two documents before we can move ahead with 14 things that we need to do on the regulatory front.
And that's 15 why I'm concerned that we be able to meet some kind of 16 deadline here for 1150.
17 MR. ROSS:
Well, that's n,o t quite the case, 18 Commissioner Bernthal The regulatory intent -- and this is 19 fully consistent with the Severe Accident Policy Statement 20 is to have available what we call the individual plant 21 methodology.
These five plant are nothing more than an audit.
22 The industry, through IDCOR, is developing its 23 individual plant methodology.
They also are doing several in fact, 24 reference plants, and they are just at the point of trying to 25 they're meeting this very day on this topic
o 9
1 approve a proposal for how to analyze specifically each plant
~
2 in a systems role, how you can extrapolate or do additional 3
calculations or whatever is needed to justify the generic 4
extension.
5 We have' looked at the work in a very brief fashion 6
last week.
It looks very promising.
They have yet to agree 7
that that's the right formula, they have yet to test it.
They 8
don't expect to tender it to Licensing until early next year 9
as their proposal for individual plant analysis.
10 What Harold needs is our five calculations for audit 11 purposes, and woven throughout this is working with IDCOR to
.{
12 get consensus or agreement or, if necessary, imposition of a 13 regulatory solution on each of the technical issues.
14 So it's a close race.
But I don't regard our work 15 s pacing.
I think both the,NRC and IDCOR are working very 16 hard for the start in the spring of 1986 of doing what was 17 called out in the last few pages, pages 15 and 16 of the 18 Severe Accident Policy Statement.
That is,.doing each 19 specific plant for vulnerabilities that might be peculiar to
' '2 0 that plant.
I don't think we're pacing, but it's close.
21 MR. DIRCKS:
Well, you are right in a way.
Last 22 year there was heady optimism about this whole thing, that we 23 were going to move through this like a hot knife through i
24 butter, and think that was conveyed by many people on the l
25
- outside, l
. ~
o a
10
~
1 We have reached the Promised Lanc -- there are going 2
to be factors of 10 reductions, the source term was going to 3
be reduced.
4 COMMISSIONER ASSELSTINE:
In fact, we got a letter 5
from Warren Owens urging that very point not very long ago.
l.
6 So the ardor doesn't seem to have worn off.
7 MR. DIRCKS:
Well, I think as we've moved into it, 8
we're not alone.
I think the industry group, the IDCOR, has 9
confirmed much of the complexity that people were predicting 10 sometime ago.
I think we've seen in almost every major 11 nuclear country in the world the same cautious progress.
I've 12 seen work coming out of the OECD, CS&I work, and almost you 13 can see this curve decreasing as they get into this science.
14 We saw the indication of it with the American 15 Physical Society report that said wait a minute, you've got 16 some uncertainties here, you've got to go back and look at 17 plant-specific data.
And I think i,f anything, that's been a 18 major factor here.
And the containment performance is the-19 other major factor.
20 So we can move on the methodology, but when you get 21 back into getting the results out, you're going to go almost 22 plant by plant.
And I think this is where we are today.
23 COMMISSIONER BERNTHAL:
Well, I don't want to leave i
24 the impression here that I want to rush into something that 25 doesn't have supporting data.
My concern is that we have a
11 1
structured, coherent, milestoned program that gets us where we 2
want to go and allows us to keep our eye on the ball 3
I guess a good point in relation to that to be
/
4 mentioned as well -- and I hope we'll discuss this a bit more i
is how many of these specific issues 5
as we go along here 1
6 that APS and others have pinpointed as being fundamental 7
questions that are still attached to source term research do 8
we have some sort of reasonable assessment on?
How long is it 1
9 going to take now to carry out the recommendations of the 10 study group, of the distinguished study group, that studied 11 the steam explosion issue some months ago and came to a 12 conclusion?
Do we have a program laid out that will get at 13 the answers for that?
14 A similar question can be asked about core-concrete 15 interaction, and a similar question can be asked about the 16 so-called direct heating effect, I guess -- you guys know more 17 about this than I do, the atomizati,on of core and all that 18 stuff.
And how many other things like that are there.
And 19 have we got some sort of timetable laid out, have we got a 20 program to review so that we know where we're going with this 21 stuff and it gets done?
22 MR. DIRCKS:
Okay.
I think you've laid out the 23 basic questions I think as we move through the briefing we 24 can hit most of them, and then if we don't hit them let's go 25 back and take them up one at a time.
1 12 1
CHAIRMAN PALLADINO:
As long as we have interrupted, 2
in considering the relationship between NUREG-0956 and 3
NUREG-1150, one can't help but notice that you have a program 4
of ongoing research indicated in 0956 that has milestones out 5
to 1987.
Now, this would imply that there is important 6
research work to be done before you can really make some of i
7 the assessments that you're looking for.
8 I guess my question is:
is it basic to making your 9
assessments, or is much of this confirmatory work?
10 MR. ROSS:
We regarded much of it as confirmatory, 11 and I think it will show up in '86 and '87 as we can better 12 quantify the uncertainty.
13 However, the NUREG-1150, the Risk Rebaselining 14 Program, will use today's source term code package, not next 15 year's research.
16 CHAIRMAN PALLADINO:
But when you say today's 17 profile, what do you mean?
18 MR. ROSS:
The methodology, in fact, is depicted on 19 the next slide.
20 CHAIRMAN PALLADINO:
All right.
Well, maybe we'll 21 pick these up.
22 COMMISSIONER ASSELSTINE:
Just before we go to that, 23 though, Bill, let me make sure I understand your position.
I 24 think that upper righthand box, which is Changes in Source 25 Term-Based Regulation, I take it the Staff's position is still
4 13 1
that at this point, we've got some more things to do.
We have 2
to go through this fairly careful process of analysis that 3
you've described here, and that now we still aren't able to 4
rush to judgment in terms of rolling back some of those 5
regulations.
6 MR. DIRCKS:
Well, I think what we're saying is yes, 7
we do have analysis to do and we've got work to do, and we're 8
not only doing it ourselves; we're relying very heavily on the 9
work of IDCOR.
And I think our programs and IDCOR's are very 10 heavily intertwined.
So we're doing it in pace with them.
11 And it's essential I think also to put a plug in 12 here for the continuation of the IDCOR effort, because that 13 work is basic and needed to any progress we're going to make 14 on our front 15 COMMISSIONER BERNTHAL:
Are we going to talk about 16 these three or four areas that I mentioned, and then also 17 about any others of which you are a, ware that I am not, and get 18 some sense of exactly what your program is?
And starting with 19 the steam explosion thing where you are a little ahead of the 20 others, I should think, at least in planning?
21 MR. ROSS:
Yes.
What we had intended to do is we 22 have looked at the different phenomena that, taken 23 collectively, represent the source term method, just to point you mentioned milestones.
24 out some work.
And we do have 25 We have a backup slide where we take the eight measure areas
14 1
of research -- and of course, our report has more details --
2 and we show little triangles which are milestones, stretching 3
from here through the end of fiscal
'87.
So we have done the 4
work that says what are the key milestones, when will they 5
occur, and more importantly to us, how much is going to cost 6
and how are we going to get the money.
Because there is 7
obviously an interlocking between getting the work done and 8
getting the budget to fund it.
9 And we have details available if w'e have the time 10 and the necessity to go into that.
11 MR. DIRCKS:
Yes.
We will hit the details because I
(
12 think it's important to try to get through that.
13 MR. ROSS:
Let's go to slide number 3.
14
[ Slide.]
15 We also showed t h i,s slide at the last Commission 16 meeting.
This represents a collection of models.
Each model 17 has its own name.
And taken togeth,er, this collection really 18 is the source term methodology.
Some of these codes have been 19 used for a number of years and have been modified, some of 20 them were created in the last two or three years especially 21 for this purpose.
22 And the flow is starting with the upper lefthand 23 corner where you see the word ORIGEN.
The general flow is to 24 calculate the fission product inventory in the fuel; how is it 25 released from the fuel following an inadequate core cooling
15 1
situation; if it's released from the fuel how does it behave 2
in the reactor coolant system before it gets out; if the 3
vessel is failed due to molten core debris penetrating the 4
vessel and the core ex-vessel debris winds up on the bottom 5
basemat, how does it interact with the concrete; and what 6
fission products are released in this ex-vessel situation.
7 And then, for all of the fission products that are released to 8
the containment, how do they behave before the containment 9
fails.
10 And there's special applications if you have a 11 suppression pool or if you have any ice condenser.
(
I have some what I call tutorial slides 12 So these 13 on several of these models and we'll go through them.
- Again, 14 this is the methodology.
Manuals for these codes are 15 available, fully available.
16 We believe that collectively, these methods 17 represent a major advance over the, Reactor Safety Study l
18 methods and we'll discuss why.
I would point out that there 19 are uncertainties that remain, and these are the subject of 20 the continuing research and we'll discuss those areas.
21 So with that introduction, what I want to do is go 22 through about six tutorial slides.
23 COMMISSIONER ASSELSTINE:
But Denny, before you 24 leave that one, just a couple of general questions on the 25 codes as a group.
16 1
When I look at information and some of the reports 2
we have been getting, itke the April 1985 report on the 3
research program from Sandia and a couple of the other 4
information notices referring to some of the British work on 5
radiation effects on the release of iodine, it appears to me 6
that some of these things are bringing to light phenomena that 7
are or may not be accounted for properly in the codes.
8 1 guess l'am wondering what you are doing to correct 9
those deficiencies and when, if you have got a timetable and a 10 program for reaching a decision that in fact these codes have 11 now been validated to take into account all the phenomena that 12 we are now aware of.
13 MR. ROSS:
Well, the best illustration, I think, on if you look at the code marked 14 a particular point is to be 15 VANESA, which is one of the models that the APS had particular 16 difficulty with because it wasn't well described in a manual in fact, the manual has only rec,ently been released for 17 18 comment, this code is well-founded, and in fact, we believe it 19 has the right chemistry in it.
20 I think there are two things that one would look at 21 to get the good, warm feeling that indeed the VANESA model is 22 the right chemistry.
What it does is provide the source term 23 from the ex-vessel pool of corium up into the containment, and 24 there are all sorts of high temperature chemistry reactions 25 going on.
17 1
In order to get, say, the broad input from people 2
around the world, one thing we do is we widely circulate the 3
manual that describes the code, which has a lot of 4
information.
It's a very detailed manual on the basic 5
chemistry.
6 Now, within'the NRC family we also have an 7
experimental program where we take, for example, molten UO2 8
doped with stable isotopes that could be representative of 9
fission product species,and heat it up to high temperature and 10 dump it on concrete and measure what happens, 11 I think a coupled analytical experimental program is
._(.,
12 what is really needed, and if you take that with a wide 13 circulation of the manual and its physics and chemistry, you 14 get the confirmation you are looking for.
15 Now, to the extent we widely circulate all the 16 manuals and to the extent we finish the experimental 17 confirmation of some of these things, I think we shall have 18 done our work.
19 COMMISSIONER ASSELSTINE:
Okay.
But in terms of, 20 say, the April 1985 report from Sandia, though, they talk 21 about carbon monoxide and hydrogen generation as a result of 22 one of those experiments, and they say this is a significant 23 observation in light of current codes that fatted to account 24 for reducing the nature of the oxidic core debris, 25 Does that mean that in those cases, you have gone
1 18 1
back and you have now revised the code to account for that 2
phenomenon as well as some of these others?
3 I guess what I am wondering is are we in the process 4
of a continually evolving set of codes or are we to the point 5
where we can say with a fair degree of assurance that for all 6
of the natura'l phenomena that we are aware of now, as a result 7
of all the experimental programs under way, we have gone back 8
and corrected the codes wherever that was needed and we are 9
truly in a process now of simply just doing a few more 10 experiments to validate or provide assurance that we have 11 covered everything, or are we still in the learning process 12 with these things?
13 MR. ROSS:
I think I will get Mel Silberberg to 14 answer your first point.
15 MR. SILBERBERO:
Commissioner Asselstine, we have at 16 this point corrected in what we call the Source Term Code 17 Package those principal items that,have been pointed out 18 during the peer review process, and the ones that we 19 immediately have the capability to do in terms of data or 20 improved models.
21 We are on the alert as we proceed to the research 22 program and get what I would call a truly landmark result or a 23 result that would dictate a need to make such a change in what 24 we have, or at least to take the insides from that information l
25 and judge the results that we have and take that into account
19 1
in the utilization of the codes.
But if it's an important 2
effect that comes from the research program, then we would 3
take steps to change that 4
We feel that at this point, we have enough stability 5
with the information we have to proceed forward, but again, 6
with a watchful eye as closely coupled to the research 7
program.
8 COMMISSIONER ASSELSTINE:
How about the example that 9
I gave? _Have they been able to go back and make the changes 10 in the codes so that you account for that phenomena?
11 MR. SILBERBERG:
I believe you were referring to 12 chemical changes within the core-concrete reaction.
13 COMMISSIONER ASSELSTINE:
Yes.
14 MR. SILBERBERO:
Which, in fact, are now taken into 15 account in the VANESA Code. Jn other words, the detailed 16 chemistry of hydrogen and carbon dioxide coming from the 17 core-concrete interaction, in fact,,is treated in an 18 equilibrium fashion within the VANESA Code.
19 COMMISSIONER ASSELSTINE:
Did Sandia just not 20 realize that, or was that a change that was made in the VANESA 21 Code after April 1985 when they issued their report?
22 MR. SILBERBERG:
That's a little puzzling because I 23 believe that that particular accountability had already been j
24 in the code.
25 MR. ROSS:
We can check.
The text that you read ma l
20 1
have been referring to codes in other places other than 2
VANESA. They didn't call that, as you read it, a deficiency.
3 The other countries 4
COMMISSIONER ASSELSTINE:
Right, but they say in 5
light of current codes --
6 MR. SILBERBERO:
Yes, we will have to check that.
7 COMMISSIONER ASSELSTINE:
Well, I mean there is a 8
collection of these that I have sort of collected over the 9
past few months.
There is another one that was highlighted in 10 a weekly information report for the week of December 14, which 11 talks about chemical modification of cesium iodide containing 12 serosols during a hydrogen burn, and that one again concludes 13 that demonstration of copious generation of free iodine in 14 this manner may have an impact on current source term 15 modeling.
16 Is that another one where you have gone back and 17 changed the models or the codes to, reflect the phenomena that 18 was described there?
19 MR. SILBERBERG:
That is a case where --
20 COMMISSIONER ASSELSTINE:
Again, the Sandia 21 experiments.
22 MR. SILBERBERG:
Yes.
The information has come 23 in.
It is being reviewed, obviously, within our community and 24 discussed between the staff and the contractors.
It is one of 25 those things that one wouldn't necessarily make a change the
21 1
next month, but at some point as the evidence builds to be 2
convincing and having people check, then we would have to take 3
that into consideration.
4 COMMISSIONER ASSELSTINE:
Fred mentioned earlier 5
some of the APS concerns, and one of them that I remember that 6
stood out in my mind when Professor Wilson made his 7
presentation was the need that he noted to assure internal 8
consistency within the codes.
I remember the example he gave 9
about doing basic conservation of energy calculations, that it 10 didn't appear that those codes used consistent bases 11 throughout them.
12 Is that something that you have now corrected?
13 MR. SILBERBERG:
Yes.
For example, the MARCH Code 14 had been using a subroutine called INTER for the core-concrete 9
15 interaction, and that inconsistency we were obviously all 16 aware of over the past several years, and certainly it has 17 been pointed out during our peer r e.v i e w s.
l 18 What we have, now done in the Source Term Code in effect, taken INTER 19 Package is gone and made that change and have now coupled MARCH with CORCON, so that is an 20 out 21 example of consistency.
22 COMMISSIONER ASSELSTINE:
You were aware of that 23 one.
Have you done a review of all the codes now so that you 24 can say with confidence that you are satisfied that all these 25 codes conform to fundamental scientific principles and that e
22 1
they all use consistent internal assuptions?
2 MR. SILBERBERG:
Pretty much so, right, at this 3
juncture.
Absent any new data or building up of new data, in I
4 terms of consistency, we feel we have now done it.
For 5
example, having coupled the thermal hydraulics with fission 6
product deposition, MARCH and TRAPMELT.
They are closely 7
coupled because of the thermal hydraulics.
We have made that 8
change.
9 COMMISSIONER ASSELSTINE:
I guess what I'm trying to 10 get a sense for is, both from the standpoint of the concerns 11 that APS had raised and from the standpoint of the 12 experimental work that has been going on for the past several 13 months and that is likely to continue for the next couple of 14 years, how confident are you that these codes now really are 15 validated, that new information is not going to result an 16 significant changes to the codes, so that I can understand'the 17 basis for your later statements in,the presentation that l
18 really go to the heart of the extent to which the codes can I,
19 now be used?
i 20 MR. ROSS:
I have two comments.
We had assembled, 21 shortly after the APS report, a list of internal checks that 22 were made on the various codes for consistency.
I think quite 23 a bit was done in a manner internal to the individual 24 laboratories.
It was not reported and perhaps not appreciated 25 outside.
23 1
I am just looking at a partial list here from the 2
MARCH Code, and we have other checks, doing an energy balance 3
to check the in-vessel steam generation rate, and doing an 4
energy balance for the containment pressure and temperature, S
verifying the hydrogen production against the fraction of the 6
zircalloy oxidised, and is a relatively long list of internal 7
checks.
8 The APS did comment that there had been enough 9
people look at it that they felt that no major phenmenon 10 overlooked, and that's our position.
11 Now, the way that we will get more detailed
[
12 knowledge is when we start applying our more detailed 13 codes. The computer code MARCH models the meltdown progression 14 of the core.
We have a much more detailed core thermal model 15 under development that can do a portion of what MARCH 16 does. This code will model a melt progression up to a point.
17 It's developed at I d a'h o.
18 We have a more detailed melt progresssion code that 19 would take over from there and go all the way through the 20 vessel. It has got more detail and more sophistication than 21 MARCH does.
22 We don't think it would modify the basic 23 scenario. It could modify the timing.
For example, we have a j
(
24 very detailed melt progression code called MELPROG being 25 d=veloped at Sandia.
It might shed further light on the
24 1
timing and mode of bottom head failure.
MARCH, I think, 2
releases it all at once.
3 There are theories and contentions that you burn a 4
small hole through which rapidly progresses to a larger hole, 5
and you have a time-dependent release.
6 When we are comfortable with this code and have 7
started applying it in a more detailed fashion, I think we 8
will get some different answers.
We don't think it will make 9
a major difference, but the key to more detailed understanding 10 is to pursue these more detailed codes, and that's why we are 11 doing it.
(
12 CHAIRMAN PALLADINO:
But in response to Commissioner 13 Asselstine's point, you are keeping abreast of the research 14 and you are giving attention to information that ought to be 15 reflected.
16 MR. ROSS:
Right. I think, for example, one of the 17 big issues, I think, that everybody,has focused on is the 18 ex-vessel situation.
We have a long list of experiments we 19 are running, mostly at Sandia, but some are in cooperation 20 with the Federal Republic of Germany.
21 As the data develop and are compared with the code 22 predictions, then undoubtedly new models will be put in and 23 old ones will be taken out, and what is left will be modified 24 or revised.
So we will be continually updating the code. That 25 we have.
25 1
I think what we have to be sensitive to is something 2
so startling coming up in the future as to alter the 3
fundamental severe accident scenario, and that is where we are 4
alert to and it is part of our program.
We don't expect it, 5
but we are certainly looking.
6 CHAIRMAN PALLADINO:
Let me suggest for the moment 7
that we go on, and we can come back to any further questions 8
on the codes, 9
COMMISSIONER ASSELSTINE:
Yes, that's fine, although to at some point I would like to hear also what the Staff has 11 done to answer or respond to the APS comment about doing 12 additional peer review work on the codes so that there could
(-
13 be some independent review and confirmation of the codes.
14 MR. ROSS:
That is built into the discussion.
15 Okay, now let's go to slide 4.
16
[ Slide.3 17 What we are going te have,is a number of slides 18 where we illustrate one of the meltdown sequences.
The 19 sequence that we have chosen for illustration is sometimes 20 called the Station Blackout.
The letters TMLB stand for a 21 transient loss of main feedwater, loss of auxiliary feedwater, 22 and at least loss of enough AC power that the accident 23 progresses.
So it's a few hours loss of AC power.
24 The Greek letter delta follows that, which stands 25 for the containment failure by overpressure.
Depending on how
26 1
you input the code, you can either get an earlier or late 2
overpressure, and we have conditional probabilities for both.
3 The example we are going to show here is early 4
overpressure.
The containment failure process, of course, 5
would have to be done separately.
The Source Term Code 6
Package does not calculate the expected failure pressure for a 7
containment.
That is done elsewhere.
8 Now, this sequence happens to be one of the dominant 9
sequences for some.PWRs, but the sequences taken as a whole in 10 0956 were not selected to be the risk-dominant sequences for it the five plants but were a broad range of sequences, high
(
12 pressure and low pressure, containment bypass.
The purpose 13 was to challenge the methods.
14 So we have for this illustsation a complete loss of 15 feedwater, and if you will turn to the next page, it 16 111ustrates the type of information that comes from MARCH.
17 MARCH is sort of the driver code f o,r the whole suite, and 18 4mv sg other things, it computes such things as shown on this 19 table.
So you see, as the accident progresses, your steam 20 generator dries out.
21 The reactor system is at high pressure because there-22 has been no failure yet, and the coolant is being lost through 23 the relief and safety valves, and after you uncover the core, 24 it will melt, slump, collapse onto the bottom head.
25 Eventually the bottom head fails, in this calculation about
27 1
2-1/2 hours.
In this illustration, the containment fails 2
shortly thereafter.
3 As a user input, you can select a containment 4
failure pressure, which MARCH calculates the containment 5
pressure. You can get an early or late failure because you 6
want to be able with the rest of the code family to calculate 7
the fission product source term.
8 Now, in this I think the failure pressure was down 9
around 85 psia, whereas the Surry failure pressure is somewhat 10 higher.
11 The actual containment failure probabilities will 12 come from NUREG-1150, s.
13 So this illustrates some of the information that 14 MARCH would provide, and i t. is essential to the rest of the 15 scenario.
16 COMMISSIONER ASSELSTINE:
How would this change, 17 say, for a B&W plant, if you used D, avis-Besse instead of 18 Surry?
19 MR. ROSS:
Well, we can talk about this up to a 20 point Obviously, the steam generator will dry up much 21 sooner, and then up until bottom head failure, then you could 22 just accelerate everything.
By the way, we have not modeled a 23 B&W plant in our reference plants.
Once the bottom head 24 fails, then everything else that follows is plant specific.
25 The cavity design for B&W and Westinghouse is
t 28 1
different.
The containment failure pressure obviously is 2
plant specific.
The mode of concrete attacked is cavity 3
dependent, so it's hard to speculate.
4 COMMISSIONER ASSELSTINE:
But you haven't run one of 5
these for a B&W?
6 MR. ROSS:
We have not.
7 CHAIRMAN PALLADINO:
When you talk about bottom head 8
containment failure, are you talking about gross failure?
9 MR. ROSS:
Gross failure, yes; that's right It's a 10 gross failure.
11 CHAIRMAN PALLADINO:
Do you consider it to be a
(-
12 complete loss of the containment?
13 MR. ROSS:
Yes.
We considered at one time whether 1
14 there might be a 15 CHAIRMAN PALLADINO:
Is that the way these 16 containments fail?
17 MR. ROSS:
Yes.
That is,our current belief 18 As you know, there is what we call sort of 19 colloquially an Appendix J 1eakage.
Appendix J,
which is the 20 small, few tenths of a percent leakage, we think the leakage 21 would go up some, because you are higher than the design 22 pressure.
But we don't think at present that there would be a 23 leak before break where you sort of self-relieve the 24 containment.
5 25 Our current belief is that you would go up to some
s i
29 1
1 pressure and then fail in a dramatic fashion.
j l
2 Now we have one test series that confirms this.
We 3
have had a series of scale model experiments.
4 COMMISSIONER ROBERTS:
Oh, the one at Sandia?
5 MR. ROSS:
Yes.
We have not yet done the reinforced 6
concrete model, and we hope to.
7 CHAIRMAN PALLADINO:
That's the one I was thinking 8
of.
9 MR. ROSS:
Right.
We have a special containment 10 performance working group that looked at this very issue, and 11 they more or less conclude that you get relatively small i
12 leakage until you have some gross deformation, and then you 13 get relatively large leakage.
14 COMMISSIONER ASSELSTINE:
Did you also get 15 containment failure by perso,nnel error or maintenance error in 16 terms of leaving valves open?
17 MR. ROSS:
Yes.
And when,we look at what we call 18 our containment event tree, this is the so-called beta bypass 19 failure mode.
We modeled that as a conditional probability.
20 I think our current modeling is 2 times 10 to the
-3 21 as the so-called beta mode failure probability.
There has 22 been a lot of work on that, and there have been a few 23 instances around the world where it has occurred.
Pacific 24 Northwest Laboratory did an assessment, and it turns out, the 25 values used today and in the reactor safety study are about
4 t
30 1
the same.
2
[ Slide.]
3 Page 6 illustrates some ingredients of two of the 4
codes, ORAGEN and CORSAR.
We just illustrated some of the 5
radioactive species.
I have six of them listed.
6 And some important points on this, the ORAGEN code 7
calculates the species inventory.
For each of the species, 8
when you go to do some kind of risk or health effects 9
calculation, you have to then use the right radial toxicity 10 for each of the species, which varies very widely.
11 The computer code, CORSAR, then, models the release 12 of each of these species.
It is temperature-dependent.
13 I would note here that the cesium -- there is sort 14 of what you might call a relative to iodine, a surplus of 15 cesium, so the iodine is shown as being bound with the cesium, 16 and the rest of the cesium would be cesium hydroxide.
And if 17 you look at the boiling points, you,can sort of classify the 18 cesium iodide, cesium hydroxide, and tolerium as relatively 19 high volatile, relative to the last two items, the strontium 20 and lanthanum, which have very high boiling points, and they 21 would be relatively low volatile.
22 COMMISSIONER BERNTHAL:
When you say inventory and, 23 in fact, when you list species, I assume you are referring to 24 the usual, canonical isotopic species.
Or are you referring 25 to9 all of the xenon isotopes produced in the core?
4 a
31 1
MS. MITCHELL:
These are the sum total of all of the 2
xenon -- for instance, both the stable and unstable.
3 MR. ROSS:
Since the fission product release from 4
fuels is temperature-dependent, this also illustrates the fact 5
that you need a core thermal model from MARCH to get the 6
temperature, so you can drive the fission product release.from 7
the core.
8 CS11de.]
9 Page 7 illustrates the path of the fission products 10 once they are released from the core.
We'll use the TRAPMELT 11 code to model this deposition reactor coolant system.
This 12 illustration is again for the TMLB', and the flowpath before 13 the bottom head fails is up through the pressurizer, through 14 the exist pipe, and down to the quench tank, and that's the 15 heavy black line on your figure.
16 And the various physical phenomena that are modeled vapor, condensation, chemical 17 are shown on the left side 18 reaction, aerosol formation, agglomeration and growth, 19 settling and deposition on walls.
20 None of this collection of phenomena was modeled in 21 the reactor safety study.
22 CHAIRMAN PALLADINO:
You say none of them?
23 MR. ROSS:
None of this has been modeled.
There was 24 no treatment in the reactor safety study.
25 Now, of course, this sequence-dependent as well as
32 1
plant-dependent.
In other sequences, the flowpath would be 2
different, and again, this is for the deposition prior to 3
failure of the bottom head.
4 COMMISSIONER BLRNTHAL:
It seems to me, the clear this is all very interesting, but the key 5
question here 6
question is not so much whether these programs are going to 7
change as time goes on, because they will -- we will learn new and it's maybe not even so much whether there are 8
things 9
specific things that have been fixed that people have pointed to out, but rather whether the overall concern that APS and 11 others have raised about benchmarking and verifying these 12 codes, whether that systematic deficiency, in your judgment, 13 has been remedied over the last year.
14 It may be that these are all great, but if I ordered well, not a disinterested 15 this suite of codes now from 16 observer, but a very interested observer, let's say, in Europe 17 somewhere, and you sent the package,to me, how would ! verify 18 that it had been adequately reviewed and benchmarked and that 10 I could believe what was in it?
20 That, in fact, was the problem at the APS.
It 21 wasn't so much that they questioned that things were right or 22 wrong.
It's just that they couldn't find out whether they 23 were right or wrong.
24 Has that deficiency been adequately addressed now?
25 MR. ROSS:
I'm going to give it what I call a
t 33 1
conditional yes, because your question had a lot of facets to 2
it.
3 One thing that we think is inadequately addressed is 4
the coupling.
In other words, they had a comment that there 5
were bits and pieces.
The source term code package couples 6
these such that there is not artificial linkage going back and 7
forth.
They remarked that some of the routines had to be done 8
differently and separately from some of the other routines.,
9 So I think that part is done.
10 We did have an independent laboratory do a QA effort it was before 11 and see if they could take the code package
(
and reproduce the work.
And they 12 it was a code package 13 did.
They could.
14 COMMISSIONER ASSELSTINE:
Who did thatt 15 MR. ROSS:
Brookhaven.
16 MR. SILBERBERO:
Let me just add that in the putting 17 together of what we call now the s o,u r c e term code package, 16 which takes into account changes that are appropriate and 19 identified over the last several years, we have, in parallet, 20 an ongoing QA and benchmarking effort at Brookhaven now, 21 running lockstep with the Battelle/ Columbus people as the get 22 this code out on the street and into the 1150 process, which 23 is, in fact, happening right now.
24 MR. ROSS:
Now the individual manuals discuss model 25 verifloation, and in addition, on Slide No.
2, one of the
j 34 1
balloons that was completed was Oak Ridge's report, which is 2
the assessment of the codes.
I think you have to take all of i
3 these together.
4 Tl e reason 1 gave you a conditional yes is that i
S there is more verification yet to be done.
There are more j
6 experiments to be done.
And so I think the verification will i
7 get better.
i j
6 The reason that we are advocating use at this time j
9 is, we believe that there is a plateau that %#e have reached i
1 10 that is much higher than the reactor safety study.
And as the l
i 11 APS said, we don't think we have neglected any major i(
12 phenomena, it's just relatively marked improvement over the I
13 reactor safety study.
We think it makes it advisable that i
14 these codes can be used at this time.
t j
15 COMMISSIONER BERNT,HAL:
Have you taken steps to 16 remedy the defiotency in forums for publication of such codes 4
17 and data and developments?
j 18 We talked about this a year ago, and, of course, the 19 APS raised that point as well There are two kinds of peer 1
20 review, after all.
There is one that is sort of a closed-shcp I
21 kind, where you ask somebody of your choosing to take a look 22 at it.
And the other is where you have a legitimate 23 community, all with their own special interests and points of 24 view, that review these things in the open literature.
And so 25 far, it appears there may not even be an appropriate forum for
35 1
that to be done.
2 Have we thought about that at all?
3 MR. ROSS; Let me comment on that.
In our 4
Conclusion IV, which we will be coming to, which relates to 5
this topic, I think first let's look at the publicly 6
available.
All of the manuals will be publicly available.
7 We hope that by circulating 0956 broadly, we can at 8
least make people aware of a good portion of the science.
9 They would need these manuals if they wanted to take it any 10 further.
11 COMMISSIONER ASSELSTINE:
Are the codes themselves
(
12 publicly available7 13 MR. ROSS:
Yes.
14 COMMISSIONER ASSELSTINE:
Okay.
So the codes and 15 the manuals.
16 MR. ROSS:
Yes.
17 COMMISSIONER BERNTHAL:
Byt you know, pubito 18 available 19 MR. ROSS:
Well, I'm not through.
20 COMMISSIONER BERNTHAL:
Okay.
21 MR. ROSS:
The thing that the APS focused on is 22 wider publication and archival peer review journals.
And we 23 certainly -- it's been recommended, and we certainly intend to 24 encourage it, and I've had some initial discussion with 25 Professor Theofanous, who edits such a journal, as to whether
l 36 i
his journal would be available.
I think it's " Nuclear 2
Engineering and Design."
And he indicated that
- a. e thought it 3
could be.
you know, the 4
Now whether, in fact, these 5
journ.', which has its own referee process, would accept these 6
or not, I don't know.
I think the best we could do is 7
encourage our contractors to submit them, which we are doing 8
and will continue to do, and we will try to find appropriate 9
journals for this purpose.
10 CHAIRMAN PALLADINO:
1 should point out, archival 11 journals are interested in making sure that what is presented (l
12 is worthwhile, and the process takes quite a bit of time.
You 1
i 13 don't get results immediately published in an archival 14 journal s
15 MR. ROSS:
Yes.
2 16 CHAIRMAN PALLADINO:
I'm not saying one shouldn't do i
17 it, but I think that's more a confi,rmatory or check process 4
to than it is a complete peer review.
19 MR. ROSS:
Yes.
I think we made it clear that 20 sometimes it would be two or three years before it would show 1
21 up.
22 COMMISSIONER BERNTHAL:
You're right.
But that 23 shouldn't deter you from doing it.
I mean, there's a time lag 24 of a year or two, to be sure, and then the preprint is l
l 25 available publicly, but we're doing that already.
The point 1
I
i l
i i
I 37 1
is that there is a special effort and attention that goes into 2
the actual publication, and things change sometimes between 3
preprint and publication.
4 MR. ROSS:
Well, I believe we're doing this.
I made S
what I think is a special effort to solicit from Professor 6
Theofanous a proposal, and we have had discussions with our 7
contractors.
We are encouraging them.
I think this will just 8
have to proceed.
I think to the extent that nothing shows up, 9
we may just have to encourage a bit harder or change the verb.
10 COMMISSIONER BERNTHAL:
Well, we have mentioned this 11 before, but if there is a deficiency here, and if there simply
(
12 is not an appropriate forum -- and I don't know whether the maybe the NRC 13 Theofanous journal is appropriate or not 14 should, as a matter of policy, seek to get an appropriate 15 forum estabitshed and take the lead.
16 MR. ROSS:
Well, I submit that we have already done 17
- that, We call it the Water Reactor, Safety Meeting.
I don't 18 know I could be any broader than that.
And it has the 19 additional advantage of being timely.
The work that is 20 reported there may not be more than two or three months old.
21 COMMISSIONER BERNTHAL:
How can it get peer reviewed 22 in two or three monthst l
23 MR. ROSS:
Well, if you want it to be timely and you 24 want to get, let's say, what ! call a frank and open opinion 25 of the researchers, then you get what you get at the Water
~ -... -.
38 1
Reactor Safety Meeting.
A person comes in, he runs a new 2
experiment, he may have run it three months ago, his work is it is unfettered.
He reports his stuff.
It's not the 3
not 4
NRC giving a speech, except at the opening a'n d closing.
The 5
rest of it is the researchers.
you know, I 6
COMMISSIONER BERNTHAL:
But then 7
don't want to belabor the point -- but then, we are still 8
lacking a step.
Whether it's that or something else, I simply 9
don't know, Denny, whether there is an appropriate forum or 10 not.
But if there's not and if there is a void in the 11 literature that needs to be filled and encouraged by this 12 agency, even to the point of supplying some funding perhaps, !
13 think we ought to do it, so that the process is carried 14 through the way it ought to be.
15 MR. ROSS:
So do I.
That's why I even offered to 16 this " Nuclear Engineering and Design" Journal, if they needed 17 a grant to help them with the admin,istrative expenses, wo 18 would do it.
19 I am not wedded to that journal either.
I am just, 20 you know, trying anything we can find.
21 COMMISSIONER BERNTHAL:
Sure.
22 CHA!RMAN PALLADINO:
I think there is a fundamental 23 problem here.
If we don't find an appropriate forum, the 24 implication is, well, then, maybe we ought to start one.
But 25
! am not sure that those are the things that people read.
The
39 1
people that we are hoping will read don't necessarily read 2
those kinds of journals.
So it is important to get them into 3
the journals where the specialists have an interest.
4 Otherwise, our own journal would be not much different from 5
the reports that we put out.
6 COMMISSIONER BERNTHAL:
No.
You misunderstand me.
7 I'm not saying that the NRC should start publishing the 8
equivalent of a physical review or something.
I'm saying we 9
might foster whoever it would be.
Maybe North Holland or 10 somebody.
Most of these journals tend to be supported by page 11 charges indirectly by funding agencies anyway.
(
I if there's a need 12 So as long as you can 13 don't know if there's a need, but if there is a need and a 14 void in the literature, people will read it if there's a void 15 and it's established as a legitimate scholarly journal 16 I just don't know.
I'm asking the question.
17 MR. ROSS:
Well, we're ju,st going to have to watch 18 and see how successful we are with the journals we have.
If 19 it doesn't work, we'll have to do something else.
20 MR. DIRCKS:
But just to remind you, you know that 21 it was at the initiative of this agency that we got the 22 American Physical Society involved in this.
If we carried on 23 with our normal procedures, we might have just published NUREO 24 documents and gotten comments and gotten the normal laboratory 25 circulation.
40 1
But I think it was to get this extra effort that we 2
went to the American Physical Society.
I think that gave us a 3
great number of benefits, and it does show that when you get 4
one of these learned societies interested, you can get good 5
results out of it.
6 I think we have gotten the feeling that we've got to 7
do more of this and get more of these societies interested in 8
the work.
That's one reason why we've gone to the National to get them interested in getting 9
Academy of Science, too 10 their community interested in our program.
11 CHAIRMAN PALLADINO:
I think we've said enough on 12 this point for the moment 13 COMMISSIONER ASSELSTINE:
Denny, have you been able 14 to go back to some of the people that served on the APS review and talked to them about the Professor Wilson and others 15 16 steps that you have taken to try and address some of the 17 concerns that were raised and get a, sense for the degree to 18 which they are sattsfled that those efforts do address the 19 concerns?
20 Reproduceability is an example.
You have described 21 what you have done.
22 MR. ROSS:
No, we haven't done that I think by 23 sending them 0956 for comment, that will be a good forum to do 24 that.
But I have had no contact with them.
25 COMMISSIONER ASSELSTINE:
It might be worth doing
41 I mean, for us to havs a sense, if 1
that specifically 2
Professor Wilson were sitting here today, whether he would 3
- say, "Yes, the efforts that have been undertaken since 4
February have really addressed the concerns that we had, 5
concerns about validation of the codes, concerns about seeking 6
greater independent peer review, the concerns about 7
reproduceability, and those things,"
8 CHAIRMAN PALLADINO:
Do'you want only Wilson's 9
COMMISSIONER ASSELSTINE:.
No.
What I'm saying is, 10 some of the people 11 COMMISSIONER ASSELSTINE:
I think it's the
(,
12 committee, because I have a feeling that you'll get different 13 opinions from the various members of the committee.
14 MR, DIRCKS:
I think what we could do is make sure 15 we get the document to them and then perhaps organise a 16 meeting with them where we could sit down with them and go 17 over the 0956 after they've had a o, hance to go over it and go 18 through some sort of a question-and-answer.
19 COMMISSIONER ASSELSTINE:
I think that would be 20 useful.
And I also think, on the point that we were just l
21 discussing, our concerns about seeking this kind of archival 22 review and following through on the APS eifort shouldn't be 23 viewed as a criticism of what you've been doing.
It sounds to 24 me ikke you've really been trying to make an effort in that 25 area.
1 l
t l
42 l
1 i
The problem may not be lack of effort.
The problem l
2 just may be finding journals and organizations who are willing 3
to help us.
l 4
MR. DIRCKS:
Well, we could do this reassembling of l
5 the peer group.
l 6
CHAIRMAN PALLADINO:
I believe in the interest of 7
time it would be better if we allowed the next twenty minutes 8
or so for you to complete the presentation and then come back 9
to some of these issues, because that is exactly what I wanted l
10 to do, as ! indicated in my opening remarks, l
11 MR. ROSS:
Let's go to Slide 8.
l 12 CSlide.3 13 This shows in a tutorial fashion the features that l
14 are modeled by the CORCON-VANESA pair.
At this point, the i
15 presumption is that the molten core, together with steel from 16 the bottom head of the vessel, has come to rest on the basemat 1
17 of the containment.
i 18 We show in this picture the reactur oavity wa!!,
19 which is the term you would use !! it were a PWR.
For a 20 boiler, you'd use the term pedestal This is just symbollo of 21 the code.
i
(
22 The molten core debris interacts with the concrete f
23 and produces various gases and attacks and goes on through.
24 I might note, for the PWR basemats that are on the i
25 order of ten or twelve feet thick, there is now a good belief
i l
43 r
I 1
that the more likely event is that you would not get basemat 1
2 penetration.
This is probably expressly useful in terms of L
i 3
the China Syndrome.
With the treaty signed last night, maybe 3
i c
4 this will improve relations there also, I don't know, 5
CLaughter.1 I
I
}
6 The temperature of the molten oore debris and 7
concrete is all-important because it affects the computer 8
code, VANESA, and its modeling of fission product releases.
l l
]
9 The gas flow is equally important.
l 10 So this just illustrates the next step in the t
I 11 fission product path.
We've modeled various concretes, and we i
(
12 think we've made a major advance in modeling with respect to l
13 the reactor safety study.
4 14 CSlide.1 i
15 Now if we go to Slide 9,
which is the last of the l
16 tutorial slides, we are now -- what's modeled here is a FWR l
f r
]
17 containment, and you have the fissign products that, before I
18 vessel fatture, came out whatever penetration was opening, 19 Itke the relief and safety valve, and after vessel fatture, i
i 20 you have fission products coming as predicted by the VANESA I
I 21 routine.
And up untti the point of ostastrophic containment f
l 22 failure, you have processes going on that would tend to reduce 1
1 23 the containment source term, such as settling and deposition l
24 on walls.
And if you have AC power, you'd have performance of 1
25 engineering safety features, such as sprays and filters,
i i
P t
l 5
44 3,
1 If it happened to be a plant with a suppression pool l
2 or an ice condenser, then you'd have some retention there l
3
- also, i
t 4
Now this illustrates the main code now, and for j
5 botters, the SPARX is the suppression pool code, and you have i
6 a special routine for ice condensors.
7 (511de.3 j
j 4
Now taken together on Slide 10, I think we've i,
l 9
111ustrated the major advances in the source term technology, l
10 listed by these seven items.
These are advances since the i
11 reactor safety study.
I think we've discussed these in some f
(-
12 detail already, except possible the fourth item, aerosol j
13 behavtor in containment.
I 14 Ten years ago, the empirical model, CORRAL, was 15 used.
Now we have what we think is a very strong mechanistto j
I i
j 16 model with much more data, i
f 17 CHAIRMAN PALLADINO:
You, call these major advances I
i 18 in source term technology.
Does this mean it's all donet i
19 MR. ROSS; No.
The next slide, there are still some i
I 4
20 uncertainties.
We don't consider that we're through.
l i
i i
21 CHAIRMAN PALLADINO:
Okay.
[
1 22 MR. ROSS:
It's just that it's such an advance, we i
I j
23 think it's a good replacement because it's so much better.
i I
24 I think I've talked individually about these j
i i
25 advances, and I think in the interest of time, we ought to i
45 t
point out on Slide it, there are still some areas of 2
uncertainty.
3 CS11de.3 4
Now these eight areas are discussed in more detail 5
in Chapter 7 of our report, in that we have a research program 4
that we think addresses all of the remaining areas of 7
uncertainty.
4 The general feeling is that as this research is 9
completed, these areas of uncertainty will narrow.
We do have 10 a special program to continue the uncertainty study, and wo 11 discussed it somewhat with you last time, an uncertainty study
(
12 done at Sandia.
Taking the existing code package, we are 13 starting a program at Brookhaven to extend and refine and 14 enhance and enlarge the uncertainty, and we hope to come up 15 with a work product with an uncertainty bound on the source 16 term.
17 About a year from now, we,will have done two plants, 18 several sequences, one boiler and one PWR, and we hope to be 19 able to issue a new quantitative estimate of uncertainty then.
! shouldn't 20 COMMISSIONER BERNTHAL:
I hope that 21 take this too literally.
These are major areas of 22 uncertainty.
I trust that they are areas where we still have 23 more to learn, because it looks like it covers almost 24 o v e r y t h i.ig.
25 MR. ROSS:
Well, yes, we fussed a little bit with
45 1
the word major, and we thought that it's relative to the other 2
work we do.
There is some other work going on in the seve**
3 accident decision.
These are the most important areas that 4
remain.
5 CHAIRMAN PALLADINO:
If we're going to use the word <
6
" uncertainty" -- we use the word " uncertainty" to imply a 7
number of things.
6 Are these areas where more knowledge is ne e d e r: 7 Are 9
they areas of ignorance?
10 MR. ROSS:
Well, let me give you an example We had it a graph that, I believe, Mr. Chairman, ycu referred to as the
(
12 so-called mosquito graph, because the first time we showid it.
13 it looked Itke a mosquito, and as we looked at it, sbre enough 14 it did.
15 It's the containme,nt inventory as a function of 16 time, and when we looked at the uncertainty on this, if we 17 make some conservative assumptions,or some optimistic 16 assumptions or best estimate, we can see a big range a
19 large range, as a funetion of time, between the top and the 20 bottom of the curve.
in this 21 Now that means, if there's a large range that the uncertainty 22 case, it could be two or three decades 23 range is large, and research, if properly done, can narrow the 24 uncertainty range.
That's what we meant here.
! see.
Well, I don't know how
47 1
you reduce uncertainty, unless you have a lot of various 2
experiments and you have enough data to discern what brings 3
about differences in consequences.
4 MR. ROSS:
But we think, as we do the remaining S
work, which we can call confirmatory, like the illustration !
6 just had for the containment inventory of fission products, we 7
think we're too pessimistic, and we would take out some 8
pessimistic models and replace them with what we think is 9
more realistic, and that narrows the uncertainty range, and 10 that's what we're trying to do.
11 CHAIRMAN PALLADINO:
One of the problems, I guess,
(
12 we're all having is trying to figure out what it is that still 13 needs to be done, and you look at the two lists, the major 14 advances and the areas of uncertainty, and there's a great 15 deal of overlap.
And maybe at the end, you could somewhere 16 along the line indicate, either by example or a couple of 17 examples 18 COMMISSIONER ASSELSTINE:
It does sound Itke what 19 the staff is saying, though, is that we've made a lot of 20 progress in these areas, but we still have a long way to go.
21 CHAIRMAN PALLADINo; Well, but no, that's my 22 problem.
I'd like to know what do we have to do to get there, 23 and how far away are we?
24 MR. ROSS; We think that the work that we need to 25 finish is well-defined in our "Further Research" chapter,
48 1
which is Chapter 7.
2 All of these areas of uncertainty are not onl / in 3
the reactor safety study, but they're much worse.
Some of the 4
items, Item 4 on this slide, retention and revaporization of 5
the reactor coolant system, was culled out by the APS, We 6
have a program to model that, make some experiments, and we're 7
running the experiments.
8 Based on that, we'll have to decide whether and how 9
or even if we should put it in one of our future predictive 10 codes.
11 COMMISSIONER BERNTHAL:
Now I can understand why 12 that particular item is a major area of uncertainty, because 13 it's extraordinarily complicated and difficult business.
14 Why is natural circulation in the vessel a major 15 area of uncertainty?
16 MR. ROSS:
- Well, th-thing that's important here is as the internals heat 17 the temperature of the internals, a,n d
\\
18 up, the TRAPMELT code is less effective.
So right now there's 19 a feeling that you'll get some flow in the vessel that would 20 be up from the center of the core and down through the edge of 21 the bundles, and this will promote some circulation and give 22 you a different thermal history than we are now modeling.
23 COMMISSIONER BERNTHAL:
Is the uncertainty in the 24 thermal progression itself, or is the uncertainty just in the 25 ability of the codes to model it?
I mean, that stuff is
49 1
pretty easy compared to the other point, the retention and 2
revaporization business, it would seem to me at least.
3 MR. ROSS:
Well, when we look at the computer codes 4
that we would tend to use for this, like the loss of coolant 5
codes, we hadn't applied to it.
Sometimes the 6
high-temperature properties are just not even built into the 7
code, because we're talking -- they were designed for 8
COMMISSIONER BERNTHAL:
So the codes are inadequate; 9
is that what you're saying?
10 MR. ROSS:
Well, we're in the process of changing, 11 but they were -- ~the codes that we started with weren't 12 designed for this, and we had to -- like they wouldn't have --
13 they wouldn't handle metal surfaces at 2000 degrees, because 14 they didn't ever expect that to happen.
15 COMMISSIONER ZECH:
It seems to me what you're 16 saying in this slide is that there are areas in all of these 17 listed items of some uncertainties,,but certainly in other I
18 areas listed, you have a certain degree of confidence about 19 properties and characteristics and interactions.
20 But I think ysu're trying to tell us that in this l
21 list, there are areas of uncertainties, but not the whole, not 22 everything is uncertain that you're giving us.
23 MR. ROSS:
That's correct.
That's right.
They 24 contribute to the width of the uncertainty band.
25 A good example is the last one, containment failure L.
50 1
modes.
A lot of people will glibly say that the failure 2
pressure for containment is two to four times design.
- Okay, 3
we've run a series of tests on steel, and we got up to five 4
times design.
5 What's the right number?
It could make a 6
difference.
7 COMMISSIONER ZECH:
Right.
8 MR. ROSS:
So we'd like to -- saying two to four or 9
two to five times design, it seems too large.
We ought to be 10 able to characterize it better than that.
11 COMMISSIONER ZECH:
Right.
12 COMMISSIONER ASSELSTINE:
It seems to me we need to 13 understand not only what you're doing to address or reduce 14 these uncertainties, but also what you are proposing in terms 15 of how the codes address those uncertainties, take them into 16 account now, and how you propose to make decisions with the 17 codes, given those uncertainties.
18 You mentioned areas where there are variations 19 between a low and a high range that may cross several orders 20 of magnitude.
21 I think at some point we need to understand what you 22 do.
Do you just pick the middle?
23 MR. ROSS:
I think we have a very good program on 24 that.
The severe accident risk reduction program has built 25 into it, and has had for several years, the requirement to
4 51 1
characterize the cost / benefit ranges, so the point value will 2
be shown, as well as the extreme.
So if we look at any 3
prevention or mitigation feature that could be applied to an 4
existing reactor, and we compute the cost and the benefit of 5
that feature, we will show now a point, but a complete 6
uncertainty range, and as this range starts to come up into 7
the cost / benefit ratio of 1,
then I think the regulator will 8
have to pay a lot more attention to it.
I'm talking about the
~
9 extreme of the range, not the point value.
10 So that is part of the severe accident risk 11 rebaselining program, and it should be coming out.
It's part 12 of the NUREG-1150, and it all ought to be available in six 13 months or so.
14 COMMISSIONER ASSELSTINE:
Okay.
15 MR. ROSS:
- Okay, 1,et's go to Slide 12 now.
16
[ Slide.]
17 And this is one point wheye we might save a bit of 18 time.
We have a number of conclusions.
The first five are 19 lumped together under the title, " Source Term Science."
The 20 next four are lumped under the title, " Source Term Insights."
21 We have two that are lumped under the title, " Risk Insights;"
22 two for " Continuing Research."
We have thirteen conc *usions 23 and three recommendations.
24 I don't propose to read the slide.
On Slide 12, we have talked about Conclusion 4 already 25 they emphasize
52 1
for publication.
2 I wanted to emphasize on Conclusion 5 that a lot of 3
detailed plant knowledge is required to apply these codes, and is not 4
generic extension is something that we just haven't 5
yet available.
It's a part of the IDCOR effort.
6 There are a couple of things I would like to point 7
out on the source term insights on Slide 13.
8
[ Slide.3 9
On Conclusion 6,
we used the term -- we calculated 10 source term for selected accident sequences.
Just to remind 11 you, we have not done these yet for all of the risk-dominant C..
12 sequences for the referenced plants.
That work is in process.
13 And on Conclusion 7,
we've discussed the containment 14 behavior.
Equally important, perhaps, under some sequences, 15 is reactor coolant system retention.
16 On Conclusion 8,
I give some examples on plant 17 design.
There is a sequence called,the V sequence, where you 18 have the interfacing LOCA, where the reactor would melt, and 19 there is a leakage path bypassing containment.
At the end of 20 that leakage path, the outlet may be underwater or it may be 21 abovewater, and the difference is plant-dependent, and it 22 makes a big difference on the source term.
The concrete 23 cavity details make a big difference.
The engineered safety 24 features make a big difference.
25 So I would emphasize in Conclusion 8,
more detailed
53 1
models are needed, more time en plant design and construction.
2 COMMISSIONER ASSELSTINE:
Is thrs a significant 3
factor in terms of deciding how detailed a review you have to 4
do of individual plants 5
MR. ROSS:
Yes.
in order to resolve 6
COMMISSIONER ASSELSTINE:
I 7
these issues?
Indeed, whether the more detailed reviews are sufficiently detailed 8
guess the PRAs that have bee done 9
to reach those kinds of conclusions.
10 Would they have picked up some of the 11 vulnerabilities that we see from time to time when they crop C..
12 up?
13 MR. ROSS:
From what I've seen of the IDCOR 14 proposal, their methodology would pick up vulnerabilities.
15 They're very serious.
The work of the outline is very 16 detailed.
I think they equally share the concern about q,uite expensive, 17 generic extension.
It would be a 18 time-consuming job.
19 CHAIRMAN PALLADINO:
How do you factor in the human 20 intervention, for example, at Three Mile Island.
If they had 21 let the safety features work, you might have had a different 22 result.
They intervened, and things got worse.
23 At Davis-Besse, they intervened, and things got 24 better.
25 COMMISSIONER ASSELSTINE-They got worse and then
4 54 1
better.
2 CHAIRMAN PALLADINO:
Yes, worse and then better, but 3
nevertheless, the intervention of the human being is an 4
important factor.
5 What I guess I'm getting at is, do you make 6
assumptions about the failure of the engineered safety 7
features.
8 MR. ROSS:
Yes.
A'n d in fact, on Conclusion 9,
there 9
will be some operator actions that have to be considered.
10 There is a belief that in the boiling water reactor that some 11 of the important sequences can be mitigated by proper operator 12 action, and they are developing what they call emergency 13 procedure guidelines for this purpose, which would then be 14 converted into plant operating emergency procedures.
15 What we will have to do -- in the first place, the 16 guidelines and procedures have to be reviewed, and then in 17 some way we have to model how likely we think it is to produce 18 a success, and if we think it works, then we'll model it and 19 give a certain conditional probability to it And more 20 importantly the NRR and the Region will have to make sure the 21 procedure is in place and that it works.
22 But operator action, it has to be modeled.
23 COMMISSIONER BERNTHAL:
I guess I think that 24 Conclusion 6 ought to be reworded somehow.
I'm not sure that 25 calculating new source terms provides a sufficient test of the
55 1
codes.
It might be a needed test, but I don't think you can 2
conclude from that that they are sufficiently tested, can you?
3 COMMISSIONER ASSELSTINE:
Or at the minimum, you 4
have to explain what justifies the conclusion.
5 MR. ROSS:
Okay, then I will explain.
Sufficiency, 6
in our terms, was high-pressure, low-pressure containment 7
bypass.
It was a broad range of thermohydraulic challenges.
the thing is, if we could come up 8
Perhaps a necessary test 9
with a different sequence that had new phenomena that we 10 haven't tested, then Conclusion 6 would be wrong.
11 COMMISSIONER BERNTHAL:
As a matter of curiosity on
(
12 this suite of codes or the codes that line up there, are names 13 of individuals ever associated with those or is it just do people put their 14 organizations?
I mean is it recognized 15 names on these and take responsiblity for them, or is it just 16 a laboratory or a group or an organization?
17 MR. ROSS:
I think in gen,eral the manuals have 18 names.
There is a name on the cover, in gene ral The 19 principal investigator. I think in many cases the NRC research 20 engineer is listed also as the project monitor.
21 On Slide 14, Jocelyn Mitchell, who has labored long 22 and hard over 0956 for the last few months, is going to have a 23 discussion.
24 MS. MITCHELL:
We wanted to provide a perspective of 25 how the analytical procedure that we have talked about so far
1 4
56 1
could produce changes in what we might perceive as the risk 2
of a plant.
3 CSlide.]
4 At this point it is an illustration only.
We had a 5
available for one' plant of the five that were studied two 6
things.
This was for the Surry plant.
We had a good coverage 7
of the risk-dominant sequences for this plant because we did 8
in BMI-2104 four sequences and investigated several different 9
failure modes for the containment.
So it was possible to put 10 those together with a preliminary evaluation of a containment 11 event tree from the Severe Accident Research Program and to 12 try to make a prospective of what this new procedure would 13 mean.
14 We have no reevaluation today of the sequence 15 frequencies for Surry.
That,will come.
It is being done now 16 under the Severe Accident Research Program, and it will be 17 absorbed into NUREG-1150, which you,have heard about, with new 18 frequencies.
So we took as given for this comparative study 19 the individual sequence frequencies from the Reactor Safety 20 Study for Surry.
21 We did change the site model to make it a 22 Surry-specific site and not a set of sites, as was used in the 23 safety study itself, and therefore we had a three-step 24 process, which is illustrated on Slide 14, where we started 25 off using the WASH-1400 source terms as they were given in the
57 1
document, along with its view of the containment event 2
tree.
3 This is to provide a base case to compare using our 4
new site model 5
We then changed the source term analytical procedure 6
but kept the Reactor Safety Study view of how the containment 7
event tree would go, and we provided the second step in that 8
table.
'J And finally, we combined the same sequence 10 frequencies with the same source term evaluation but we 11 absorbed the new containment event tree from this preliminary 12 SARP program, and we produced the third line on the slide.
13 We compared the base case with the final slide, and 14 you will find for the two measures of the many possible 15 offsite consequences that we chose to look at, early and 16 latent fatalities, that there is a reduction factor of 10 for 17 early fatalities and a factor of 4,for latent fatalities.
18 When you see our third recommendation in a little 19 while, you will see that we suggest that for any application, 20 you certainly oughtn't to leave without a discussion of the 21 uncertainties.
We nave to leave you with a promise for the l
l 22 future.
23 We have talked about our programs that are in place 24 for NUREG-1150 to do a good job on the uncertainties, but I 25 did want to tell you that for any risk uncertainty, we have
C 58 1
four areas that will contribute: from the event frequencies, 2
from the calculation of the source terms themselves, from the 3
containment behavior and from the calculation of the 4
consequences.
5 For this particular comparative study, since we have 6
used a comparable site model and we have used identical 7
sequence frequencies, those two particular areas are not as 8
germane to this particular comparison as the other two.are.
9 If we can go to Slide 15 10 COMMISSIONER ASSELSTINE:
On the uncertainties, it before you leave that one, have you been talking with the ACRS 12 about how you intend to treat uncertainties in NUREG-1150? And 13 I am mindful of the discussion we had with the ACRS not too 14 long ago on their concerns about our ability to treat 15 uncertainties or to explain how we treat uncertainties in 16 these kind of analyses.
17 MS. MITCHELL:
We are def,initely talking with the 18 ACRS about all parts of this.
We have had one meeting with 19 them and another is upcoming in the next week, next week.
20 That will definitely be a part of it.
So far, we have been 21 working on the analytical procedure, the technical issues at 22 the moment.
23 COMMISSIONER ASSELSTINE:
Given what the ACRS said 24 to us at our last meeting with them, I think that's a very 25 important part, to explain this as clearly as we can, how
59 1
uncertainties are being considered and treated in making these 2
kinds of judgments so that they can get a sense and we can get 3
a sense here, as well, for the adequacy of our treatment of 4
uncertainties because right now I don't think we have done a 5
very good job in explaining how we do it.
6 COMMISSIONER BERNTHAL:
Are you willing to give a 7
wild estimate, at least, at this point, 95 percent confidence 8
level, the certainties that would attach to these numbers?
9 MS. MITCHELL:
No.
10 CLaughter.)
11 COMMISSIONER BERNTHAL:
That is succinct.
12 COMMISSIONER ASSELSTINE:
Or could you give us 13 numbers that you would attach a 95 percent confidence level 14 to?
Would they be these numbers?
15 CHAIRMAN PALLADINO:
You don't attach confidence 16 level to lack of data, and I hope we always lack data on some 17 of these things because we are talk.ing about improbable 18 events. If we get a lot of data, then they become very 19 probable events, and then you can do a great job of assessing 20 the confidence limit.
21 COMMISSIONER BERNTHAL:
Let me try again.
Without 22 being so specific, would you make a guess that nobody could 23 hold you to, 95 percent confidence level Is the difference we 24 are seeing between WASH-1400 and BMI 2104 -- this is a in your judgment, is that 25 qualitative answer now
9 60 1
difference, which is something less than an order of or let's 2
magnitude, likely to be a difference considerably 3
just say a difference larger than the envelope of uncertainty, 4
or is the envelope of uncertainties attached to each likely to 5
envelop that difference pretty well?
6 MS. MITCHELL:
As long as you promise not to read 7
the transcript after this is over 8
COMMISSIONER BERNTHAL:
A very qualitative 9
statement.
I would say the uncertainties are 10 MS. MITCHELL:
11 large, and I think before you attach a great deal of 12 confidence to whether it is a factor of 3 here and 4 there, 13 you ought to be very careful.
14 I do want to point out one other thing before we do 15 leave this slide, that from the three-step process for this 16 plant, there seems to be as much change due to the source 17 term analytical procedure as to the containment behavior.
18 That might not be the case for other plants.
It happens to be 19 the case for this one, but other plants may be very different.
20 COMMISSIONER BERNTHAL:
Fair enough.
Thanks.
21
[ Slide]
22 MS. MITCHELL:
Moving right along.
I have told you 25 about conclusion 10, which concerns the fact that the Surry 24 risk reduction shows about equally from source term and from 25 containment behavior.
I did want to say again that the source
61 1
term coverage for the other plants that were looked at in 2
BMI-2104 was not comparable to that from Surry.
In particular 3
for Peach Bottom we looked at one containment failure mode 4
rather than several containment failure modes, and that 5
particular failure mode was for release directly to the 6
environment and not for any release that might go from the 7
containment into the reactor building.
8 We are redoing the Peach Bottom analyses taking this 9
into account, and also making some changes in the plant model, 10 but for none of the other plants is it possible to give any 11 analyses before you can draw any conclusions about changes in 12 estimated risk.
13 However, I would note that for some of the plants, 14 for Peach Bottom, the analyses that have been done so far do
,15 not show a large reduction opmpared to the reactor safety 16 study results, and that would indicate that caution should be 17 used before trying to extend what we have done for Surry to 18 any other plant.
19 COMMISSIONER ASSELSTINE:
Does Conclusion 10 for 20 Surry mean that everything went down?
How about the 21 telluriums and the actinides?
22 MS. MITCHELL:
I think there are some chemical 23 element groups for which the number is numerically higher for 24 Surry for some sequences, however, it is not significant 25 COMMISSIONER ASSELSTINE:
How do you compare the
i 62 1
containment reevaluation, your ability to do that as compared 2
with, I gather, the difficulty in coming t. p with, at least at 3
the present time, with containment performance criteria, the 4
fact that we are still in the process of being a few years 5
away?
How much confidence can you have in the containment 6
reevaluation given our inability, so far, at least, to say 7
here are criteria for evaluating containment performance?
8 MS. MITCHELL:
I think they try to look at what is 9
the status of the particular plant.
This effort in relooking 10 at the containment event analysis is a very much more 11 structured kind of an analysis.
It is laid out where other 12 people could now go and see exactly how it was done and agree 13 or disagree.
14 Right now it is a judgment They ask, I think, for 15 Surry 59 questions, how Itkely is it that the pressure vessel 16 will fail at high pressure versus low pressure, and they go 17 around and get from many handfuls o,f different sources 18 insights that they can find from other PRAs, from the 19 containment loads working group studies, from the containment 20 performance working group studies, and get an answer to that 21 question.
22 They have made three separate estimates, one which 23 is a conservative estimate that maximizes the probability of 24 the earlier containment failure modes, one that is an 25 optimistic estimate and minimises the earlier containment
63 s
1 failure modes, and one which is the central The one we have 2
used here is the central estimate.
3 COMMISSIONER ASSELSTINE:
Okay.
That's what you 4
used in calculating the numbers on the preliminary slide.
You 5
picked the middle.
6 MS. MITCHELL:
That's right. They evaluated what 7
they called a central estimate and we used that.
But that is 8
undergoing peer review now. The evaluation of the tree needs 9
to go along with how the accident is perceived to go, so the 10 whole analysis should be interlocked starting with the people 11 who determine the probabilities, how likely is it that this
( --
12 paricular component will fail.
13 Everybody ought to go together, and when you get 14 finished, they all ought to be modeling the same exact 15 sequence or series of sequences.
16 MR. ROSS:
I think you need NUREG-1150.
I think the 17 ACRS needs it before we pick up the, subject of the containment 18 criteria.
Turning to page 16, there are a couple of 19 conclusions on continuing research.
In our report we are 20 addressing Chapter 7.
21 Conclusion 12 is we think we have a program 22 addressing the areas, and Conclusion 13 picks up on the major 23 conclusion of the American Physical Society that work ought to 24 go on.
25 Unless we decide to come back to this topic for more
o 64 1
details, that's all I wanted to say on this slide.
2 COMMISSIONER ZECH:
In the research area, have you I
3 been able to get anything conclusive at all yet from that 4
experiment that was done in Idaho on the plant that was 5
purposely melted down*
6 MR. ROSS:
The test was run in LOFT fission product 7
2 about a month ago, and the initial judgment was it seemed to 8
be a very successful test in that the center bundle went up to 9
very high temperatures, above 4000 degrees Fahrenheit, and 10 released a good quantity of fission products.
It will be 11 useful in verifying some of our codes, but the data just
(
12 aren't cranked out yet.
13 COMMISGIONER ZECH:
I see.
But you are attempting 14 to get ahold of that.
You are communicating with them and you 15 will factor that in.
16 MR. ROSS:
Yes.
It appears to be a very useful 17 test.
18 CHAIRMAN PALLADINO:
But are you going to try to 19 apply your codes to see if you get the same results they gott 20 MR. ROSS:
Oh, they were applied as a pre-test 21 prediction. So that's even harder.
It would be interesting to 22 see how that comes out.
23 CS11de.3 24 COMMISSICNER ZECH:
Thank you.
25 MR. ROSS:
On Slide 17 there are some
65 i
recommendations.
Again, since this report is going out for 2
comment, we will be asking for comment on these 3
recommendations, and in particular we would expect people like 4
the ACRS to comment, and from today's discussion, maybe the 5
APS would have some comment 6
CHAIRMAN PALLADINO:
APS?
7 MR. ROSS:
American Physical Society.
8 CHAIRMAN PALLADINO:
Well, are you going to send 9
this report out for comment by the APS and any other program 10 plans that you have to see whether or not we have heard them 11 correctly in terms of their recommendations?
(
12 MR. ROSS:
Yes, I am now.
I think it's a good idea.
13 CHAIRMAN PALLADINO:
Yes.
And to follow up the 14 point that Commissioner Asselstine made, I think that we 15 should overtly ask them.
16 MR. ROSS:
Yes, we will 17 MR. DIRCKS:
I think we will not only ask them, but 18 if we can organize some sort of a seminar with them, we would 19 like to sit down and exchange information.
20 CHAIRMAN PALLADINO:
Yes.
21 MR. ROSS:
Recommendation 2 is to get more detailed 22.
insight.
You have to run the more mechanistic codes. In some 23 cases these are recommendations to ourselves.
We did note 24 that we had not intentionally biased the code package one way 25 or the other towards conservative or optimistic, and we do
66
^
1 need very good coupling to get the uncertainties.
2 It appears likely that the uncertainty allowance in 3
the regulatory arena will depend on the regulatory 4
purpose. So, depending on whether it is rulemaking for Part 5
100 or for emergency planning or equipment qualification or 6
backfits, the uncertainty allowances will probably be f
7 different for different applications.
8 COMMISSIONER ASSELSTINE; Does that mean, Denny, i
9 that in those areas where you have this rang'e of uncertainty, 10 a pessimistic or an optimistic bound, that your approach has 1
2 11 consistently been throughout the codes to select the middle?
~.
12 MR. ROSS:
Yes, and carrying along with it whatever 13 knowledge you would have on the range.
i 14 On Slide 18, this is what we envision doing in the 15 next few months.
i 16
[ Slide 3 17 Issue for comment.
Conti,nue meeting with the ACRS.
18 We think having this document bound together where it looks 19 like it's reasonably coherent will help our ACRS review and 20 then we will continue upgrading our models to reduce 21 uncertainties.
22 CHAIRMAN PALLADINO:
When you talking about using 23 means or median, are you recognising that when you use factors 24 rather than plus or minus certain increments, that you have a l
25 very distorted distribution?
And if you have a plus or minus
67 1
of a factor of 10 of something, let's say measured 100 miles, 2
that means on one end it's 10 miles and on the other end it's 3
1000 miles.
So there is quite a distortion.
4 And yet I think you are using symmetric 5
distributions.
At least most of the formulations that I have 6
seen.
7 COMMISSIONER ASSELSTINE:
Is that right?
8 MR. ROSS:
I don't think so.
9 CHAIRMAN PALLADINO:
I guess I'm coming ba;k to 10 whether or not you should be talking in terms of factors of 11 10.
I. haven't thought about it too much.
For example, the 12 MR. ROSS; We are talking about 13 chart on fatalities, two points.
Acute radiation injury is a 14 threshold phenomenon, so if you put a sensitivity factor, in 15 one case it may go up very much and in an,ther case it may 16 disappear.
So that maken it difficult.
17 The distribution function,, which will be reported in 18 Appendix D of our report for fatalities, or for latent 19 fatalities, we will have the distribution in the report, but I 20 would say in general they are not symmetric The table 21 Jocelyn showed you was mean values, but the way, not median.
22 That is our direct presentation, so we are ready for 23 questions.
24 CHAIRMAN PALLADINO:
Well, I had said early that we 25 want to talk about where you are going, but I guess that's
68 1
what you have been telling us.
You have answered the question 2
when do you expect 0956 out, at least sometime early August 3
MR. ROSS:
Yes.
4 CHAIRMAN PALLADINO:
And you have indicated 1150 5
probably possibly March of April of
'86.
6 MR. ROSS:
Yes.
7 CHAIRMAN PALLADINO:
We didn't touch on, and maybe 8
you are not prepared to touch on when we would be getting into 9
the right side of that box you showed, t.". e boundary that you 10 showed earlier.
This is trying to apply source term-11 information to particular applications.
(
12 MR. DENTON:
Our plan, Mr. Chairman, is to wait 13 until closure is achieved on the methods and models that have 14 been discussed this morning.
We have been reluctant to get 15 out front of Research in this area because there still 16 appear to be uncertanties in some areas, and when we have 17 meetings on these topics, there are,still individual 18 laboratory researchers who have views on one side or another 19 of the approach being taken.
20 So what I feel we need in NRR is a technical basis 21 which is acceptable to the Commission for us to begin applying 22 the changed regulations, but it seems to me I shouldn't go 23 proposing changed regulations until you are satisfied that 24 there is an adequate technical basis for understanding the 25 difference.
I
69 1
Going back to what Commissioner Bernthal was talking 2
about, we need, in effect, a consensus that this information 3
is now the proper set of information to use to reevaluate.
4 So we will be moving in parallel with Research.
I 5
think the present' plan is they would give us next spring the 6
results for the five or six plants that they recalculated, and 7
if everyone is in agreement by that time that that's the way 8
to go, we would then pick those up and begin to apply that 9
approach over the next year or two with IDCOR to all the 10 individual plants on a per-plant basis, and also look at what 11 regulations perhaps need to be changed.
(
12 It does appear the results are going to be 13 plant-dependent, containment design-dependent, and we are 14 waiting for a consensus.
15 CHAIRMAN PALLADINO:
Well, I can't argue with 16 you. You have to reach a threshold of knowledge before you try 17 to apply it, and I recognize also that there will always be 18 irgrovements in the knowledge, but one could look ahead to the 19 extent of saying, well, it is going to take certain kinds of 20 very important information that we ought to start l
21 accumulating, basically preparing to get a running start when j
22 the report comes out 23 MR. DENTON:
We are doing that, and I think we and 24 IDCOR are working along with a great deal of coordination, 25 waiting for final closure on some of these areas of
~.- -, -
--W
70 1
uncertainty that have been mentioned.
There are cases, !
2 think, still, Denny, where two or three researchers in a lab 3
might have diff'erent views.
I have ca11ed the lab director 4
and said: Which one of the three do you recommend we use?
And 5
he says: That's your problem.
You know, wait five or ten i
6 years and science will settle down.
i 7
So we have not had a great deal of success, I think, 8
in some of these a)eas about getting closure.
I guess there 9
we would tend to take whatever the consensus Staff view is, 10 how we interpret those areas where there is not uniform 11 agreement.
12 I think what happens in the licensing case, there 13 are mechanisms to get closure.
We have hearings.
We have 14 boards and so forth.
What we have got here is a mechanism 15 moving toward closure but closure hasn't been achieved.
You 16 talked about peer journals and that type of thing.
It's very 17 difficult because of the number of, specialties involved in 18 this to get closure that's acceptable to everyone.
l 19 MR. DIRCXS:
That was our original intent when we 20 had this controversy two or three years ago, is to try to get 21 closure by going to the scientific society.
I think it 22 helped.
It identified where we needed to do work, where there I
23 were certainties and where we should close some gaps.
24 I think we have taken those factors into 25 consideration, and I think on the next round we are going to
t 71 1
have to go out and get some more input from the scientific 2
community.
The problem is as long as you have individual 3
researchers, we are going to have the debate going on.
4 CHAIRMAN PALLADINO:
Let me make a word of caution.
5 We are not in a scientific field even though we are using 6
scientific approaches and scientific information.
In the end, 7
it is an engineering field, where you have to accept 8
uncertainties and you provide for them in your design.
9 Now, I do agree you can't go and undertake an 10 engineering project until you have enough knowledge to begin 11 to deal with it, but we are never going to have the perfect 12 science because we are talking about modes of failure, and 13 failures take place in a lot of different ways.
14 MR. DENTON:
I think the way that the research 15 results are coming in, there,is more agreement on large, dry 16 containments and what I call simpler designs to analyse than 17 there are of a more complex design,,and it's possible we can 18 make a lot more progress on a plant similar to Surry than we 19 might on, say, MARX II containments, for example, which seem 20 to require a lot more attention.
21 So we might make some staged progress, depending on 22 the type of plants.
23 COMMISSIONER ZECH:
To me it is very important. It 24 seems to me that we ought to take a cautious and prudent 25 approach, which I think we are taking, and certainly we want
4 72 1
to wait for the data and be as confident as we possibly can; 2
but it does seem to me that we ought to give priority to this 3
program, and I think we are doing that I hope we are doing 4
that.
I think that's what you are telling us.
5 And also, though, as we develop something that is 6
fairly conclusive that we can make a judgment on, it seems to 7
me that perhaps we should really seriously consider doing it 8
in staged approaches and not waiting for the final, complete 9
program to satisfy everyone.
I don't think that will probably 10 ever happen, but if we can be confident ourselves, if the 11 Staff can be confident that the general conclusion is that
(
12 certain plants perhaps we can make decisions on, then it's my 13 view that that is the approach we should take.
14 MR. DENTON:
From talking to industry, I think their 15 main interest is getting closure on the severe accident 16 issue.
They are not that interested in reevaluating the 17 adequacy of existing engineered saf,ety systems, for example, 18 whether they should take a filter out or leave it in kind of 19 question.
They sort of assume they have got the plants, 20 plants are built, and the question is, is their plant safe 21 enough to operate in view of all we have learned?
22 So their main interest with IDCOR is working with us 23 as this information comes out on individual plants or classes 24 of plants.
Changes in our regulations for future plants, !
25 don't think IDCOR is that interested in as the industry.
73 1
CHAIRMAN PALLADINO:
But I think the source term is 2
so fundamental to our mission that we need to get as good a 3
handle on it, good or bad or indifferent.
4 COMMISSIONER BERNTHAL:
Could you answer a question 5
for me that I referred to early on?
Answer it briefly, I 6
hope.
If we pick the steam explosion issue, as an example, 7
what are we doing right now based on this rather careful 8
review that was done on the issue, and at what date would you 9
predict that we will have answers on that and at what cost?
10 And can you make similar appraisals or are you oh, I
11 proposing to make similar appraisals of core-concrete I guess it's their 12 always forget what the right words are 13 direct heating effect and whatever else there is out there 14 that is fundamental and generic to nearly every plant?
15 MR. ROSS:
Yes.
If you look at what we are doing in 16 particular on the steam explosion, the main recommendation was 17 that carefully controlled experimen,ts should continue, and 18 there was uniform consensus on that, and lack of consensus on 19 the need for a very big experiment you know, given availability of 20 We do intend to do just that, and I think by paying more attention-21 funds 22 to a confined versus an open experiment, better 23 instrumentation, which should be done in fiscal
'86 24 COMMISSIONER BERNTHAL:
Should be finished?
25 MR. ROSS:
Should be finished in
'86.
The
74 1
obligation we have at that time, which is roughly a year from 2
today, is to reexamine the issue: What do we know better now 3
and do we want to consider the matter closed, or do we really 4
have to run a much bigger experiment?
That issue is open.
5 COMMISSIONER BERNTHAL:
But right now you have set a 6
course and have in mind a set of experiments that you hope are 7
designed, at least, to achieve closure on that issue by the 8
end of fiscal year 1986.
9 MR. ROSS:
Yes, and at moderate scale. There will 10 still be some people a year from today who will have to
' 11 believe that much l a.' g e r scale experiments are necessary.
(s..
12 COMMISSIONER BERNTHAL:
Okay.
The next question is 13 what about the other areas, at least those two, and there may 14 be others that I mentioned?
Are we going to have a similar 15 review there?
Will there be a similar timetable and cost 16 attached?
we could have B-16 back 17 MR. ROSS:
I would like i,f 18 up.
This is just sort of a road map to what will be in 19 Chapter 7.
20
[ Slide.3 21 For each of the eight issues we have a timeline, and s
22 we have a little triangle as far as A through H.
And each of 23 these little triangles is just illustrative.
What we do in 24 Chapter 7 is we spell out what each of these mean.
For 25 example, if we were on line 7,
Milestone C says, " Complete
4 75 1
tests on direct heating of containment atmosphere," and 7C 2
shows up in early
'87, and these are calendar years.
3 So the complete report will define all of this in 4
detail 5
CHAIRMAN PALLADINO:
In 0956?
6 MR. ROSS:
0956, yes.
Did we put the roney in 7
09567 We have the money.
It's just not in the report.
8 For each of these triangles, we have assessed the 9
amount of money required to do it.
10 COMMISSIONER BERNTHAL:
So basically what you are 11 saying is that every one of those eight issues, which sound to 12 be broader issues than the ones I mentioned, but nevertheless, 13 those eight you feel should be closed and can be closed based 14 on your current projections by the end of
'87.
15 MR. ROSS:
Yes.
And my recollection is the integral 16 o; the money for '86 and '87 put together is about $60 17 million.
18 Now, there is one thing that we didn't answer.
As 19 we come to what appears to be technical resolution, would we 20 convene a review group like the Steam Explosion Review Group?
21 We haven't decided to do that yet.
What we have decided to do 22 is to revitalize the Research Review Groups, and the 23 revitalized Steam Explosion Review Group meets in about 30 24 days at Madison, Wisconsin.
We are adding people of a more 25 diverse nature, revitalizing, and I think that will be what
76 1
will provide the technical review.
2 COMMISSIONER BERNTHAL:
Well, where, for example, is 3
core-concrete?
Which category is that in here?
4 MR. ROSS:
That would be 5.
5 MR. SILBERBERG:
Let me just note, Commissioner For 6
Bernthal, that the reason why these may seem broad 7
example, item 7, pressure loads, would include within it steam 8
explosion, the ' hydrogen question, as well as direct heating.
9 So it's coupled.
In other words, a number of phenomena are 10 coupled under each category.
11 COMMISSIONER BERNTHAL:
And are every one of those
(
12 eight pacing items in one way or another for you to be able to 13 reach your final conclusions on 1150 or whatever the report 14 number is?
15 MR. ROSS:
Oh, no., These are not coupled to 1150 at 16
'all This is the final truth, i' you would, in severe 17 accidents. This is what wa
,igt
,s,t h e best we will ever do in 18 getting the uncertainty down.
19 COMMISSIONER BERNTHAL:
I don't think you will get 20 the final truth, but we hope some approximation.
21
[ Laughter.3 22 COMMISSIONER ASSELSTINE:
Following up on Fred's 23 question, is it possible under these eight categories to get a 24 listing of the key technical questions, the areas that you 25 think need to be answered and questions that need to be
77
-3, 1
answered in order to reach the broad conclusions in each of 2
these eight categories?
3 MR. ROSS:
Yes.
4 CHAIRMAN PALLADINO:
Are these questions in 0956?
5 MR. ROSS:
Yes.
And further, we have broken them 6
down into what we call the fin level, which is the contract 7
level, and for each contract, you start with what the question 8
is, the regulatory objective and the research objective.
The 9
answer is yes.
ic COMMISSIONER ASSELSTINE:
So we will be able to see 11 that in 0956.
12 MR. ROSS:
Yes.
If we don't have all the detail, we 13 might have to provide it as a supplement, 14 CHAIRMAN PALLADINO:
Well, we will be looking 15 forward to receiving copies, and I am sure you will welcome 16 any comments that we may have.
We are not the technical 17 experts, but sometimes we see things from a different point of 18 view.
19 I think this has been a very valuable session.
It j
20 demonstrates the Commission's interest in this subject, and we 21 may be looking to the fall to have a repeat, and we will bring 22 that matter up at agenda planning.
23 COMMISSIONER BERNTHAL:
Yes.
This has been very 24 useful I appreciate all the work effort that has gone into 25 preparing it
t 78 1
I wouldn't mind doing this quarterly, Joe?
I think 2
it's terribly important that we keep abreast of what is going 3
on.
4 CHAIRMAN PALLADINO:
Yes.
That's what I have in 5
mind.
6 COMMISSIONER ZECH:
I kind of agree with that.
7 COMMISSIONER ASSELSTINE:
Well, I certainly think 8
after 0956 comes out, after the ACRS has had a chance to look 9.
at it, that we might to hear both from them and get back 10 together with the Staff again and talk about IDCOR.
11 CHAIRMAN PALLADINO:
All right, Well, thank you
(
12 very much, folks.
Unless there is more to come before the 13 Ccmmission, we will stand adiourned.
14
[Whereupon, at 11:55 a.m.
the Commission meeting was 15 concluded.]
16 17 18 19 20 j
21 22 23 24 25
4
' s, 1
CERTIFICATE OF OFFICIAL REPORTER 2
3 4
5 This is to certify that the attached proceedings 6
before the United States Nuclear Regulatory Commission in the 7
matter of Commission Meeting O
9 Name of Proceeding: Briefing on Accident Source Term Reassessment (Public Meeting) 10 11 Docket No.
12 PIace:
Washington, D. C.
ts cate:
Wednesday, July 24, 1985 14 15 were held as herein appears and that this is the original 16 transcript thereof for'the file of the United States Nuclear 17 Regulatory Commission.
18
/
(Signature)
(TypedName'ofReprter) /Suzar(fe B. G ng 20 21 22 23 Ann Riley & Associates, Ltd.
24 25
n b
)
i l
Source Term Reassessment Study l
j Accident Source Term Program Office 4
i (A8REGyg Ps
+
W[
$ c-l's o
/- j
,s
)
oe c
M
(%
S
'4,,,#
i i
i l
}
i i
l V
i l
l l
m n
')
Program Relationships Reseamh Source Term Analysis l
Regulation
)
B MI-2104 l
l l
I Uncertainty Estimates SAND 84-0410 l
i I
I Status of Validation A
l ORNLITM-8842 l
A Source Term Changes in Reassessment Study L
Source Term-Based Containment Loads l
Regulations Working Group l
j NUREG 1079 l
l (Containment Performance l
Working Group 1f k
I Severe Accident l
APS Review Risk Rebaselining/
l Severe Accident Risk Reduction j
Regulatory w
implementation program (SARRP) l l
1 i
I i
l J
Sourco Torm Codo Suito i
Fission Product Thermal Hydraulic Transport Behavior j
__ _._ _ q ORIGEN MARCH l
Fission Product Overall l
Inventory in Fuel Behavior of El Reactor Coolant System, "1
==
yI l
, Molten Core, and u
Containment l
gl CORSoR l
3ll g
l c
Release l
Retained in Fuel l
U from Fuel I
I I
i i
l i
E TRAPMELT
]i MERGE l
e I
e I
Detailed Temperature, l
l Reactor Coolant System Pressure, and Flow in l
\\
Transport and o
l Retention j
l I
i l_______________a l
I i
I l
CORCON VANESA l
-l r
E 8
Release from
===='m Detailed Core-
-s Core-Concrete Melt Concrete Temperature l
and interactions a
E l
l 6l l
tl I
5 j
NAUA, SPARC, ICEDF l
l Containment Transport l
and Retention l
l l____
Release of fission products to the environment: Source Term
)
i Example Core Meltdown Sequence 3
l TMLB' d Sequence-Surry Plant Transient initiated Sequence l
Loss of All Off-Site and On-Site AC Power Failure to Remove Heat Through the Steam Generators I
i I
l l
i
?
~
)
e e
i i
Y l
)
Accident Event Times i
Event Time, Minutes j
Steam Generator Dry 67.5 l
Core Uncovery 95.5 i
Start Melt 118.3
{
Core Slump 146.3 j
Core Collapse 148.0
. Bottom Head Failure 152.8 Containment Failure 152.9 Reactor Cavity Dry 177.2 i
Start Concrete Attack 254.2 i
i o
)
I l
Inventories and Properties of important
)
Fission Product Species inventory Melting Point Boiling Point 4
Species (kg)
( C)
( C)
Xe 260
-112
-107 l
Csl 25 626 1280 l
CsOH 133 315 990 l
Te 25 450 988 f
SrO 56 2430 3249
> 4000 La2O 73 2315 i
3 l
4
Fission Product Deposition in the RCS i
i l
Fission Product l
Deposition Processes g
i Vapor Condensation on Surfaces l
Chemical Fleaction l
with Surfaces
, c
===,
Aerosol Formation 91 r
,\\
3
=
m f
Aerosol Agglomeration
?!?
and Growth i
j Gravitational Settling I
I 11 5
Deposition on Walls l
?
+
1
_N j
f l
r
8.
0.Fce.
rmes.qw.rc
?
- hk #$ I o-
.lxl
- 9 ?.
a<
O y.,.
k!
~.
.m a
8'.'
C
['
.'h
' " g' j
g LL. %
22
% %. A s
- j $ >;i E
.,$q..c
.o.o y
a... :
m g gg o j y, e,
,7. 2
% O
.4p.; r.;
oc o-
$ z.o u, Oe
.. N.fg** * ' ~"s e
4c s
cg O
G Q mcEa e o y
. Ec2..c
.;$ 9f.90 %'.
W 2ooa e c
- 6 E
eOO 3: 0 f,$).j g(
$g 0O b
E c.g
$ c) 2m gi p
.4'.:. ur oO o
o e $ 3.6 hMa Q% 8 k
- ' ;'.5fo'.'pM,l<
=C ZM gQ V C)
Om pgj
,gg e
gQ
@ *b $ $
d.
5 e,$
E bj o.5 g
&E=
-!5 lP:i5.:s: jib,z'u,p c
I
'5Ei e.
o a
.y n
m i
i'd 1
>v pa o.
O.
3
.. r u m m$ h I 'r.
O
+
k
- u. Efd
!f aa een !A -
a, a
u
.ha f
me x
CC l
,,,,,-n...
--n-.--
-,,,-.-----..,-,.,--w-
- c.,
a Containment Deposition Processes
~
a 7
Fission Product Deposition Aerosol Agglomeration and Growth Gravitational Settling Deposition on Walls
't ESF Performance 1
(
m
,m i.
c.
_w p4 l,
i..' REACTOR
~
.r
.' t d.
g.
.n
........*,s., *, o.
... ~
- I l
i t
i l
Major Advances in Source Term Technology 1
Fission Product Chemistry j
Retention in Reactor Coolant System (RCS)
Core Melt Progression l
l Aerosol Behavior in Containment Fission Product Release from Fuel Core-Concrete Interactions Containment Pressure Loads j
N i
l i
1 i
Major freas of Uncertainty
- Natural Circulation in the Vessel l
- Core Melt and Hydrogen Generation i
- In-Vessel Releases from Fuel
- Retention and Revaporization in RCS 4
l
- Core-Concrete Interactions l
- Scrubbing by Pools and Ice Compartments
- Containment Pressure Loads I
- Containment Failure Modss I
t
Source Term Science l
Conclusion 1. The BMI-2104 Suite of Computer Codes Represents a Major Advance and Can Be Used to Replace the Reactor Safety Study Methods.
l Conclusion 2. Principal Omissions and Oversimplifications in j
i the Reactor Safety Study Methods Have Been l
l Corrected.
Conclusion 3. Remaining Areas of Uncertainty Have Been i
Identified and Indicate Areas of Research that l
Should Be Pursued.
Conclusion 4. The New Analytical Procedures Have Been Extensively Reviewed (including an APS Review) and Documented.
Conclusion 5. The Analytical Procedure is Complex, and l
Successful Application Requires a Thorough Understanding of the Problem.
l p
l.
b l
l Source Term insights Conclusion 6. New Source Terms Have Been Calculated for Selected Accident Sequences, and These l
)
Sequences Have Provided a Sufficient Test of l
the Computer Codes.
I
}
Conclusion 7. For Most Accident Sequences, the Largest l
l Single Factor Affecting Source Terms is l
Containment Behavior.
Conclusion 8. Source Terms Were Found to Depend Strong-i l
ly on Plant Design and Construction Details,
}
Thus Making Development of Useful Generic
}
Source Terms Difficult.
Conclusion 9. New Source Terms for Many Accident l
Sequences Were Found to Be Lower than I
Those in the Reactor Safety Study, but Some i
Were Larger. Therefore, Generalizations Are Inappropriate.
1 i
I
4 Updated Risk Estimates for the Surry Plant Using WASH-1400 Accident Frequencies Early Fatalities Latent Fatalities Analytical Method (per reactor year)
(per reactor year)
WASH-1400 Source Terms 4.0 x 10-5 1.6 x 10-2
~
WASH-1400 Containment Evaluation BMI-2104 Source Terms 1.1 x 10-5 6.7 x 10-3 WASH-1400 Containment Evaluation BMI-2104 Source Terms 3.1 x 10-6 3.4 x 10-3 Containment Reevaluation
- Sources of Uncertainty Event Frequencies Source Term Analytical Procedures Containment Behavior Consequence Calculations
- Uncertainties Will Be Taken into Account in NUREG-1150 e
1 l
-i i
f Risk insights i
Conclusion 10. A Comparative Risk Appraisal for the Surry Plant Using BMI-2104 Source Terms and a Containment Reevaluation Shows a Reduction in Estimated Risk Compared with i
the Reactor Safety Study.
Conclusion 11. For the Other Plants, Further Analyses Need to Be Made Before Any Conclusions Can Be Drawn About Changes in Estimated Risk, and Significant Reductions May Not Be Found in all Cases.
[
f i
[
l i
t I
/
Continuing Research Conclusion 12. Research Programs that Address the Remaining Major Areas of Uncertainty in the Source Term Technology are Currently in Place and Being Pursued by the NRC.
~
Conclusion 13. A Major Conclusion of the American Physical Society Study Group Confirms the NRC Staff Position that Source Term Research Must Be Continued in Order to Complete the Regulatory Actions Being Considered.
e e
j
' ~
f-
}
)
i l
Recommendations
~
Recommendation 1. The New Source Term Analytical Methods Should Be Used to Reevaluate Regulatory Practices While Additional Confirmatory Research is Being Completed.
l Recommendation 2. The Source Term Code Package is the l
Recommended Tool for NRC Analyses.
Additional insights Can Be Obtained with the NRC's Detailed Mechanistic Codes and Their Experimental Data Bases.
Recommendation 3. The Source Term Code Package Pro-l vides Best-Estimate Results (i.e.,
Without Intentional Bias). Close Coupling Between the Research Effort and the Regulatory Effort Will Be Required in Assessing Uncertainties and Evaluating Technical issues.
9
9
\\
l l
l Near-Term Staff Actions on Source Term Science i
i lssue NUNEG-0956 for Public Comment on Entire Report I
l Including Conclusions and Recommendations.
4 Continue Meetings with ACRS on Source Term Science.
4 I
improve Analytical Models and Reduce Uncertainties
=
l Through Further Research.
i l
4 1
N i
\\
kkkkkk9kkkiGkt0kQ0hf0GQhkhk((g0th((G0g(kghghg(q(g(g(q(qhghkq(p(q(q( g(q(
q TRANSMITIAL 'IO:
Docment Cbntrol Desk, 016 Phillips l
j ADVANCED COPY 'IO: /
/
%e Public Document Rxan cc: OPS File j
OM N C&R (Natalie)
Attached are copies of a Comnission meeting transcript (s) ard related meeting i
docment(s). Wey are being forwarded for entry on the Daily Accession List and placment in the Public Docuent Rx2n. No other distribution is requested or required. Existing DCS identification nebers are listed on the individual j
doc ments wherever known.
f Heeting
Title:
b t u h.en J r.)
(I e i c ji A
. &i s.r c de n-A
(
/C Ln m on n
- v ~t:
Meeting Date:
7/; 4 /f f Open Closed 3
DCS Copies 3,
(1 of each checked)
[
Item
Description:
Copies Advanced original May Duplicate jl To PDR Document be Dup
- Copy
- 1.
TRANSCRIPT 1
1 i
When checked, DCS should send a j
copy of this transcript to the g
LPDR fort id /Aui a i e, < s,a ls ]
t d
i 2.
i i
[
j 3.
i i
?P
}
l lL i
E
.i D
l
(PDR is advanced one copy of each document,
' Verify if in DCS, and P-Change to "PDR Available."
h
[
two of each StrY paper.)
i hh hh h$
h h
$$h h$
kI$$b$$
bI l
lib $$$$
.